Proceedings of the 4Th Japan and US Workshop on Seismic Measures for Water Supply

The 4th Japan/U.S. Workshop on Seismic Measures for Water Supplies

The 4th Japan/U.S. Workshop on Seismic Measures for Water Supplies was held in Kobe, Japan in January 2005 to commemorate the 10 year anniversary of the Great Hanshin-Awaji Earthquake. The basic objectives of the workshops are:

• to create a forum for the active discussion and exchange • identify and document best available current technologies in water system seismic mitigation practices • provide a practical information source on seismic mitigation efforts for water utilities • identify technology improvements needed to practically and efficiently advance water system seismic practices.

Attendees included 14 representatives from the U.S., 41 from Japan and 3 from Taiwan. There were 24 papers presented during seven meeting sessions during the three day event. AwwaRF subscribers that attended and provided both technical papers and presentations include: Contra Costa Water District, San Francisco PUC, East Bay Municipal Utility District, Los Angeles Department of Water and Power, City of San Diego, Portland Bureau of Water and Seattle Public Utilities.

The workshop was funded by Awwa Research Foundation and the Japan Water Works Association, as well several other organizations and utilities.

Proceedings of the 4th Japan and US Workshop on Seismic Measures for Water Supply

January 26-28, 2005

JWWA, Kobe, Japan

The 4th Japan and US Workshop on Seismic Measures for Water Supply

*** Workshop Agenda ***

Jan.26, 2005

1. Registration 8:45-9:00 Tachibana Training Center

2. Opening Ceremony Greetings 9:00-9:10 Nobuhiro Matsushita (Kobe City) 9:10-9:20 Masakazu Akagawa(JWWA) 9:20-9:30 Elizabeth Kawczynski(AwwaRF) Marilyn Miller (EBMUD) Pei-Chung Hsu (Taipei Water Department)

3. Group Photograph

4. Keynote Speech 9:30-10:00 Masanori Hamada (Japan) 10:00-10:30 Masanobu Shinozuka (US)

Coffee Break 10 minutes

5. Presentation Part Ⅰ ( Time: 20 minutes including Q&A)

＊Chairpersons(10:40-12:00) Shiro Takada (Japan) & Don Ballantyne (US)

Session Ⅰ: People’s Cooperation for Seismic Measures 「水道の地震対策推進と住民の協力」

10:40-11:00 S1-1 The Evaluation of the Public Awareness and the Measures against Earthquakes with the Use of a Survey Questionnaire （アンケートに見る住民意識と地震対策） ---Takashi Kashiwamura (Hachinohe Regional Water Supply Authority) Japan

11:00-11:20 S1-2 Earthquake Countermeasures in Yokohama and Cooperative Activities with Residents （横浜市の地震対策と住民との協働について） ---Hironori Nukui (Yokohama Waterworks Bureau) Japan

11:20-11:40 S1-3 Consideration for Resident Participation in Seismic Hazard Mitigation Measures to Secure Water Supply （水道の地震対策における住民協力についての一考察） ---Masahiro Kimura (Osaka Prefectural Waterworks) Japan

Session Ⅱ: Risk Assessment「リスクアセスメントと分析」

11:40-12:00 S2-1 Verification of Effectiveness of Pipeline Renewal Based on Damage Analysis on Recent Earthquake （最近の地震被害分析にもとづく水道管路更新効果の検証） ---Toshio Toshima (Japan Ductile Iron Pipe Association )Japan

12:00-13:30 Lunch 90minutes

＊Chairpersons(13:30-14:30) Charles Scawthorn (Japan) & Tim Collins (US)

Session Ⅲ: Risk Management 「リスク管理」

13:30-13:50 S3-1 Bringing Pragmatic Engineering to Earthquake Preparedness at Contra Costa Water District (コントラコスタ地区における震災技術としての実用技術の適用) ---Stephen J. Welch (Contra Costa Water District) USA

13:50-14:10 S3-2 Seismic Upgrade of Water Facilities─An Asset Management Approach(水道施設の耐震性向上――施設管理の研究) ---William F. Heubach (Seattle Public Utilities) USA

14:10-14:30 S3-3 Assessing and Managing Risk―Planning Future Upgrades to San Diego’s Water System （リスク調査と管理―サンディエゴ水道の将来に向けた向上計画） ---Michael E. Conner (City of San Diego Water Department) USA

Coffee Break 30 minutes

＊Chairpersons(15:00-16:20) Masakatsu Miyajima (Japan) & Luke Cheng (US)

15:00-15:20 S3-4 Case Study of a Backup System for Water in Kinki Region (近畿圏における広域的なバックアップ体制の整備に関する一考察) ---Eizo Seki (Japan Water Research Center) Japan

15:20-15:40 S3-5 Uniform Confidence Hazard Approach for the Seismic Design of Pipelines (パイプラインの耐震設計に適した被害を一様に分散させる研究) ---Craig Davis (Los Angeles Department of Water and Power)USA

SessionⅣ Theme: Seismic Performance「地震の挙動」

15:40-16:00 S4-1 Measures Against Active Faults for Distribution Trunk Lines and Seismic Observations （配水幹線の活断層対策と地震観測） ---Takashi Furuya (Yokosuka City Waterworks and Sewerage Bureau)Japan

16:00-16:20 S4-2 Water System Seismic Performance, 1994 Northridge－ 1995 Kobe Earthquakes （1994 年ノースリッジ地震及び神戸地震－水道施設の挙動） ---Le Val Lund (Los Angeles Department of Water and Power) USA

Coffee Break 20 minutes

6. Welcome Party at the New Otani Kobe Harborland

18:00-20:00 Greetings Miyoji Konae （JWWA, Osaka Office） Marilyn Miller(EBMUD) Masanobu Shinozuka(UC, Irvine) Shiro Takada (Kobe University) Nobuo Ando(Kobe City)

Jan.27, 2005

7.Presentation Part Ⅱ

＊Chairpersons(9:00-10:20) Nobuhisa Suzuki (Japan) & Stephen Welch (US)

Session Ⅴ: Mitigation and Prevention of Damage「震災予防と準備」

9:00-9:20 S5-1 The Outline of Seismic Measures of Tokyo Waterworks Bureau ―From the Point of View of Cooperation with Regions― （東京都水道局における震災対策の概要～ハード＆ソフト対策、地域との連携を踏まえて～） ---Hiroshi Yamada (Bureau of Waterworks, Tokyo Metropolitan Government)Japan

9:20-9:40 S5-2 Emergency Operation and Countermeasures for the Water Supply System in Taiwan Learned from 1999 Chi-Chi Earthquake （1999 年の台湾集集地震に学ぶ水道システムの緊急戦略と対策） ---Wei-Sen Li (National Science and Technology Center for Disaster Reduction) Taiwan

9:40-10:00 S5-3 Evaluation of Seismic Upgrade Construction of Hanshin Water Supply Authority ―In the case of pipe line― （施設耐震計画の実績評価と今後の課題－管路の整備状況－） ---Keiichi Murakami (Hanshin Water Supply Authority) Japan

10:00-10:20 S5-4 Current Status and Subject of the Seismic Upgrade of Kobe Water System after Ten Years from the 1995 Hanshin-Awaji Great Earthquake （震災後１０年を迎えた神戸市水道耐震化の現状と今後の課題） ---Tetsuro Kijima (Kobe Municipal Waterworks Bureau Planning Division)Japan

Coffee Break 20minutes

＊Chairpersons（10:40-12:00） Toru Tomioka (Japan) & Craig Davis (USA)

Session Ⅵ: Seismic Proof Design of Waterworks and Other Facilities 「水道施設と関連施設の耐震設計」

10:40-11:00 S6-1 Design Guideline for Seismic Resistant Water Pipeline Installations （水道管路施設耐震設計指針） ---John M. Eidinger (G&E Engineering System Inc.) USA

11:00-11:20 S6-2 ALA Guidelines for Pipeline Analysis Methods and Appurtenance Design Methodology (米国ライフライン連合のパイプラインの解析法と付属施設の設計方法の指針） ---Bruce Maison (EBMUD) USA

11:20-11:40 S6-3 Seismic Design Issues on Water Transmission Pipelines （送水管の耐震設計上の問題） ---Luke Cheng (San Francisco PUC) USA

11:40-12:00 S6-4 Seismic Diagnosis of Extensive Water Distribution Network （大規模ネットワークの耐震診断および地震被害予測手法） ---Nobuhisa Suzuki (JFE R&D Corporation)Japan

12:00-13:00 Lunch 60minutes

＊Chairpersons(13:00-14:50) Toru Tomioka (Japan) & La Val Lund (USA)

Session Ⅵ continued,

13:00-13:20 S6-5 Seismic Upgrade of Prestressed Concrete Water Tanks (プレストレスト・コンクリート貯水槽の耐震性の向上) ---David D. Lee (EBMUD) USA

13:20-13:40 S6-6 Evaluation of Scenario Earthquakes and Examination of the Seismic Resistant Design Method of Waterworks in Hiroshima City （広島市における想定地震の評価、耐震設計手法の検討） ---Kenji Totoki (Waterworks Bureau, The City of Hiroshima)Japan

13:40-14:00 S6-7 Seismic Upgrades of Pump Station Located on Liquefiable Soils in Portland, Oregon (オレゴン州ポートランドにおける液状化地盤上にあるポンプ所の耐震性向上） ---Tim Collins (City of Portland Water Bureau)USA

Coffee Break 10 minutes

Session Ⅶ : Restoration after Earthquake「被災後の復旧」

14:10-14:30 S7-1 Introduction of the Disaster Information Management System of Osaka Municipal Waterworks （水道局災害情報システムの導入） ---Hiroaki Miyazaki (Osaka Municipal Waterworks Bureau)Japan

14:30-14:50 S7-2 Damage to Water Supply Pipelines System due to the 2004 Niigata-ken Chuetsu Earthquake and Its Restoration （2004 年新潟県中越地震による上水道管路網の被害と復旧） ---Masakatsu Miyajima (Kanazawa University) Japan

Coffee Break 20 minutes

8. Final Discussion

15:10-16:50 ＊Chairpersons Hiroyuki Kameda (Japan) & David D. Lee (USA)

Presentation by Craig Davis (Los Angeles Department of Water and Power)

9. Closing Address

16:50-17:00 Elizabeth Kawczynski (AwwaRF) Toru Tomioka (JWWA)

Jan.28, 2005

9. Technical Tour

Introduction: Kazuhiko Mizuguchi (Kobe Municipal Waterworks Bureau)

＊＊＊ Tour Itinerary ＊＊＊

9:00 am Tachibana Training Center (Guidance)

9:30 am Departure (by bus)

10:00am-11:30am Construction of Large-Capacity Transmission Main at Motoyama

12:00pm-1:00pm Lunch at Kobe Bay Sheraton

1:30pm-3:30pm Disaster Reduction and Human Renovation Institution

4:00pm Arrival at Shin-Kobe Station and at Hotel New Otani

Greetings Nobuhiro Matsushita Kobe City

Thank you for the kind introduction. I am Nobuhiro Matsushita, deputy mayor of Kobe City. I would like to welcome you all to Kobe.

As Kobe City observed the 10th anniversary of the Great Hanshin-Awaji Earthquake on January 17, 2005, it held a memorial ceremony in the presence of their Majesties the Emperor and Empress. Also, the United Nations World Conference on Disaster Reduction was held in Kobe from January 18 to 22. I would like to take this opportunity to express my sincere appreciation for the help and support extended from the rest of Japan as well as abroad, with which our waterworks facilities have finally been restored to its current state. It is significantly meaningful that the 4th Japan-US Workshop on Seismic Measures for Water Supply is being held in Kobe this memorable year. I would like to extend a warm welcome to you all on behalf of Kobe City.

I was told that this workshop is being held under the cosponsorship of the Japan Water Works Association and the American Water Works Association and aims to foster working-level officials who promote the creation of earthquake-resistant waterworks systems by sharing information on research and studies on earthquake and antiearthquake measures.

This year, Kobe City is promoting the project dubbed “Sending Messages from Kobe, Ten Years after the Quake.” In this project, we disseminate the experience of and lessons from the earthquake disaster, problems and achievements of restoration efforts, and the future direction of the development of Kobe.

Recently, there was the Niigata Chuetsu Earthquake in Japan and the Great Sumatra Earthquake abroad. Furthermore, it is predicted that large-scale earthquakes will likely to strike the Tonankai and Nankai regions within the next thirty years. There is growing concern about earthquake disaster.

Aiming to create waterworks facilities that can be restored quickly in the event of a disaster, Kobe City is currently implementing an earthquake-resistant facility project, and such measures as the installation of high-capacity water pipes are being taken. As we work to continue providing a stable, safe, and high-quality water supply, we ask for your understanding and cooperation.

I hope this workshop will produce successful results through lively discussions and the sharing of experiences and expertise. I expect that such results will be disseminated throughout Japan and contribute to future waterworks services. I would like to conclude my speech with wishes for your prosperity and happiness. Thank you. Greetings from Masakazu Akagawa JWWA

Good morning, ladies and gentlemen. My name is Masakazu Akagawa, Director General at the Japan Water Works Association. I would like to deliver a short message for the opening of the 4th Japan-US Workshop on Seismic Measures for Water Supply. Holding this fourth workshop here in Kobe was made possible thanks to the cooperation of the Water Supply Bureau of Kobe City. I appreciate the assistance provided by all parties involved. Ten years ago, Kobe City suffered substantial damage from the Great Hanshin-Awaji Earthquake. The restoration of the city, however, ended up being strong and admirable. I believe all victims would have had a lot of difficulties.

Last year, Japan was hit by many typhoons, and many areas throughout the country were flooded. In addition, a strong earthquake hit the Chuetsu Region of Niigata Prefecture at the end of October. Essential utilities, including water, electric power, and city-gas supplies, were damaged, having a great damages on communities. The Japan Water Works Association established a headquarters to take measures immediately after the earthquake, and the entire water related industry implemented full support activities. We learned to take such rapid action from the Great Hanshin-Awaji Earthquake that we experienced 10 years ago.

On the other hand, internationally, a great earthquake occurred off Sumatra, Indonesia, at the end of last year, when a tsunami hit countries bordering the Indian Ocean and many precious human lives were lost. Currently, countries throughout the world are carrying out full-scale international support; everyone must have been surprised at and feared the extensive damage on a global scale.

I believe this fourth workshop, being held in such a time, is a very meaningful international conference. I would guess that each country around the world takes various measures against earthquakes, depending on its natural condition or social status. Such measures, among others, need practical findings maintained by working-level officials from each country who have detailed knowledge of seismic measures.

Therefore, we expect each of you, who are engaged in waterworks in Japan, the United States, and Taiwan, to promote the active sharing of opinions and discussions in this workshop. I hope that the outcome of this workshop will be transmitted all over the world as a specific policy for seismic measures for waterworks and utilized in an effective way. In closing, I wish all of you who have attended the workshop further prosperity and happiness. Thank you very much.

1 Greetings Elizabeth Kawczynski AwwaRF, Denver, CO

Thank you to JWWA for sponsoring this 4th Japan/U.S. workshop on seismic practices for water utilities and thank you to the City of Kobe for hosting the workshop. Kobe is a beautiful city and it is especially interesting to be here during the 10th year commemoration of the Great Hanshin-Awaji earthquake.

After every major earthquake there is something to learn. Seismic engineers and practitioners are continuously learning from every event so that the aftermath and recovery can be managed in the shortest amount of time with the least disruption to the people they serve. At AwwaRF our mission is to advance the science of water to improve the quality of life and this is the very reason that we continue to support of the Japan/U.S. exchange on seismic issues for water utilities.

It is a pleasure to be here today as part of the workshop. We look forward to many good presentations and most importantly a good exchange of information so that we can all learn from each other and help those that we serve. Thank you.

4th Japan-U.S. Workshop on Water System Seismic Measures

Greetings From Marilyn L. Miller

Director General Akagawa, Deputy Mayor Matsushita and all our colleagues from Japan, Taiwan and the U.S.:

Welcome to the 4th Japan-U.S. Workshop on Seismic Measures for Water Systems. This is the continuation of a long-standing relationship of cooperation between the two organizations, JWWA and AWWARF, to improve the seismic safety and performance of our respective water systems. On behalf of the U.S. participants, I would like to thank the JWWA and the AWWARF for organizing this workshop. I would also like to thank Kobe Waterworks for hosting this workshop.

In the next three days, the workshop will emphasize the practical understanding and implementation of the seismic practices among the water agencies in the countries of Japan, the United States and Taiwan. We will be providing updates on our exciting seismic improvement programs in the U.S. and looking for new and improved techniques for seismic safety. We are looking forward to discussing seismic issues with our Japanese and Taiwanese colleagues.

This year is the tenth Anniversary of the Kobe Earthquake and it is very important to have this workshop held in Kobe City. The experiences of the Kobe Earthquake and the reconstruction of the City and its water system have provided water agencies around the world with greatly improved techniques for the design of water systems to resist the effects of damaging earthquakes.

The AWWA Annual Conference & Exposition will be held on June 12-16, 2005 in San Francisco. EBMUD will be doing a demonstration of our fault crossing technology utilizing flexible hoses to quickly restore water service where pipes are disrupted by earthquake effects. We would be happy to meet with and arrange tours for our Japanese, Taiwanese and U.S. friends attending the conference. We will make our experienced seismic engineers available for meetings and discussions. EBMUD will be hosting a dinner for our Japanese and Taiwanese colleagues as well as the US representatives to the Kobe workshop. You are all invited to join us.

Thank you for your tremendous efforts to make this workshop possible. I am looking forward to a very successful and informative conference. Greetings Pei-Chung Hsu Taipei Water Department

Good morning ladies and gentlemen. My name is Pei-Chung Hsu. I am from Taiwan. I work as the deputy commissioner of Taipei Water Department, and as the secretary general of Taiwan Water Works Association. Thank you for your kindness to invite us to attend the 4th Japan and US workshop on seismic measures for water supply today.

The delegation of Taiwan includes Professor Ban-Jwu Shin from National Taipei University of Technology, and Dr. Wei-Sen Li, he is from National Science and Technology Center for Disaster Reduction. I do believe, through broad-spectrum discussions, fruitful suggestions and viewpoints would be presented for us to mitigate the catastrophic disaster.

We are still remember, five years ago, an earthquake, named Chi-Chi, had shattered countless buildings and bridges with thousands of causalities, left nothing but destruction in the central Taiwan. Our friends, around the world, stretched their helping hands without hesitation at first moment to share the suffering we took. So I am confident the international collaboration would offer the most effective relief when any major disaster would take place.

I would like to express my deep gratitude for your invitation. And finally, I would like to make a humble suggestion, if possible, Taiwan Water Works Association would have the honor to hold the workshop in the coming future. We believe we will do our best to present you a suitable platform to share and cherish the wisdom and experience of humankind. Here I would like to send our sincere appreciation and wish you have a successful workshop.

Thank you. 1

MEASURESMEASURES FORFOR EARTHQUAEKEARTHQUAEK DAMAGEDAMAGE MITIGATIONSMITIGATIONS FORFOR WATERWATER SUPPLYSUPPLY SYSTEMSSYSTEMS ININ JAPANJAPAN

byby MasanoriMasanori HAMADAHAMADA

School of Science and Technology， Waseda University， Tokyo, Japan 2

CONTENTSCONTENTS

1)1) LessonsLessons LearnedLearned fromfrom TheThe 19951995 KobeKobe EarthquakeEarthquake 2)2) JSCEJSCE Recommendations,Recommendations, RevisionRevision ofof DesignDesign CodeCode andand ReinforcementReinforcement ofof ExitingExiting StructuresStructures 3)3) ThreatsThreats byby FutureFuture EarthquakesEarthquakes 4)4) BriefBrief ReportReport onon DamageDamage byby thethe 20042004 NiigataNiigata ChuetsuChuetsu EarthquakeEarthquake 3 TheThe DamageDamage toto WaterWater SupplySupply SystemSystem inin MajorMajor citiescities ofof TheThe HyogoHyogo PrefecturePrefecture byby TheThe KobeKobe EarthquakeEarthquake

Number of Total Money Days for Charateristic Features City Hoseholds of Loss million$ Complete Recovery of Damage Water Cut off ･Filtration Plants 650,000 ･Transmission Pipes Kobe 287.2 9070 (100%) ･Distribution Pipes ･Head Office Buildings 157000 ･Reservoirs Nishinomiya 41.6 70 ･Transmission Pipes (95%) ･Distribution Pipes 193,000 ･Transmission Pipes Amagasaki 2.8 14 (100%) ･Distribution Pipes ･Filtration Plants Ashiya 33,400 13.4 64 ･Raw Water Tunnel ･Ground Slide Total in Hyogo Prefecture 1,265,730m $506.90 4

DamageDamage toto WaterWater FacilitiesFacilities

(1)(1) ReservoirsReservoirs

(2)(2) RawRaw WaterWater TunnelsTunnels

(3)(3) FiltrationFiltration PlantsPlants

(4)(4) PumpingPumping StationsStations

(5)(5) DistributionDistribution BasinsBasins

(6)(6) HeadHead QuarterQuarter BuildingBuilding DamageDamage toto WaterWater SupplySupply FacilitiesFacilities CausedCaused byby 5 TheThe 19951995 KobeKobe EarthquakeEarthquake Collapse of The Niteko Dam

Damage to Raw Water Tunnel

Slope Slide at Filtration Plant

Collapse of Head Office Building DamageDamage CausedCaused byby LiquefactionLiquefaction--InducedInduced 6 LargeLarge GroundGround DisplacementDisplacement 193 cm cm 180 184cm 94 99 202cm 197cm 95 86 201cm

60 69 66

68 40 99 290cm328cm 71 91 300cm 79 Extensive Soil Liquefaction of Manmade Land Horizontal 45 71 Displacement Liquefaction-Induced Ground Displacement

Breakage of Foundation Piles 7 DamageDamage toto DistributionDistribution MainsMains andand ServiceService PipesPipes inin MajorMajor CitiesCities inin HyogoHyogo PrefecturePrefecture

Distribution Mains Service Pipes Total Damage Damage City Number of Number of Number of Length Ratio Ratio Breakages Breakages Households (km) (1/km) (%) Kobe 4,002 1757 0.44 89,584 650,000 13.8

Nisinomiya 966 1019 1.05 41,237 163,800 25.2

Amagasaki 847 130 0.15 13,324 193,300 6.9

Ashiya 184 408 2.22 3,316 33,400 9.9 Total in 8,517 3,778 0.44 180,213 1,365,600 13.2 Hyogo Prefecture 8 CausesCauses ofof DamageDamage toto BuriedBuried WaterWater PipesPipes

(1)(1)StrongStrong EarthquakeEarthquake MotionsMotions

(2)(2) LiquefactionLiquefaction andand ItsIts--InducedInduced LargeLarge GroundGround DisplacementDisplacement

(3)(3) SlopeSlope SlidingsSlidings andand FailuresFailures ofof EmbankmentsEmbankments 9 LiquefactionLiquefaction--InducedInduced GroundGround StrainStrain andand RuptureRupture ofof AA SteelSteel WaterWater PipePipe

2 43 0 100m 2 cm 5 2 3c 3 4 m 3c 36 m cm

2 36 77 5 cm 31 0 35 3 cm 8 11 3 82 cm

Damage Ground Strain Point

Liquefaction-Induced Ground Breakage of Welded Steel Pipe Displacement and Strain 10 EstimatedEstimated EarthquakeEarthquake FaultFault andand AreaArea ofof JMAJMA IntercityIntercity 77

601 Epicenter Epicenter （H） Takarazuka 305 （H） 446 Nishinomiya （V） Ashiya Awaji Is.

561 818 （H） 774 （H）（H） 332 502 379 KOBEKOBE （V）（H）（V） 283 OsakaOsaka BayBay Akashi （V） N 616 Maximum Acceleration （H） (gal) Epicenter Area of JMA Intensity 7 (MMI = 05 10km EstimatedX) Earthquake Fault GroundGround SurfaceSurface AccelerationAcceleration 11 CausedCaused byby thethe KobeKobe EarthquakeEarthquake Acceleration Record Acceleration Response Spectrum ） 2

） h=5% 500 NS 2 2000 （cm/s NS 0 （cm/s 1000 -500 Max.=820.5cm/s2 EW 500 Acceleration 30 40 50 Time（sec）） 2 500 EW Max.=619.3cm/s2 100 Acceleration Level for Design of Water Facilities （cm/s 0 before The Kobe Earthquake

Acceleration Response Spectra 50 -500 0.1 0.5 1 5

Acceleration 30 40 50 Time（sec） Natural Period（sec） 12 BasicBasic ConceptsConcepts forfor ImprovementImprovement ofof EarthquakeEarthquake ResistanceResistance ofof InfrastructuresInfrastructures ProposedProposed byby JSCE(JapanJSCE(Japan SocietySociety ofof CivilCivil Engineers)Engineers)

(1)(1)ExaminationExamination ofof earthquakeearthquake resistanceresistance againstagainst TwoTwo levelslevels ofof earthquakeearthquake groundground motionsmotions Level 1 : Conventional level of intensity with a moderate probability of occurrence Level 2 : Large intensity as observed during the Kobe earthquake with very low probability of occurrence

(2)(2) PerformancePerformance--BasedBased DesignDesign GovernmentGovernment PoliciesPolicies forfor DamageDamage MitigationMitigation 12‘ ofof WaterWater SupplySupply SystemSystem MinistryMinistry ofof Health,Health, LaborLabor andand Welfare,Welfare, JapanJapan

(1)(1) EstablishmentEstablishment ofof GuidelinesGuidelines forfor SeismicSeismic DesignDesign andand ImprovementImprovement ofof FacilitiesFacilities andand WaterWater PipesPipes

(2)(2) EstimationEstimation ofof DamageDamage toto WaterWater SupplySupply SystemsSystems againstagainst FutureFuture EarthquakesEarthquakes

(3)(3) PlanningPlanning ofof EmergencyEmergency MeasuresMeasures ImmediatelyImmediately afterafter EarthquakesEarthquakes andand RestorationRestoration andand ReconstructionReconstruction WorksWorks

(4)(4) PromotionPromotion ofof SeismicSeismic ImprovementImprovement ProgramsPrograms byby FinancialFinancial SupportSupport toto WaterWater UtilitiesUtilities GuidelinesGuidelines forfor EarthquakeEarthquake ResistantResistant DesignDesign ofof 13 WaterWater SupplySupply FacilitiesFacilities andand WaterWater PipesPipes JapanJapan WaterWater WorksWorks AssociationAssociation

Two Levels Earthquake Ground Motions Level 1 Ground Motion,which 1000 may occur once or twice Level 2 during the life time of the (0.7,100 Level 2 100 facilities and has a moderate ) magnitudes of intensity (0.7,70 Level 1 ) 10 Level 2 Ground Motion,which (0.5,20 has a low probability of ) occurrence ,but has a strong Response velocity : Sv'(cm/s) : velocity Response intensity such as those observed Response velocity:Sv’(cm/s) 1 0.1 1 10 during the Kobe earthquake NaturalNatural periodPeriod of of surface surface layer layer:T(s) : T(s) Design spectra for design of buried water pipes ObservedObserved BedBed RockRock MotionMotion andand LevelLevel 22 14 DesignDesign MotionMotion ResponseResponse VelocityVelocity 1000

Kobe Univ. Kobe Univ. Higashi Kobe (NS) (EW) (N12E) cm/s) （ 100 Higashi Kobe-33m (N78W)

Port Island-83m Non - Exceedance Probability 90% 10 (EW) Non - Exceedance Response Velocity Probability 70%

ｈ＝15％ 1 0.1 1.0 10.0 Period (sec) 15

ConsiderationConsideration ofof TheThe EffectsEffects ofof LiquefactionLiquefaction--InducedInduced LargeLarge GroundGround DeformationDeformation forfor thethe DesignDesign CodeCode

1)1) LiquefactionLiquefaction--InducedInduced PermanentPermanent GroundGround StrainsStrains forfor TheThe DesignDesign ofof BuriedBuried PipesPipes

2)2) ExternalExternal ForcesForces onon FoundationFoundation PilesPiles fromfrom FlowingFlowing LiquefiedLiquefied GroundGround 16 LiquefactionLiquefaction--InducedInduced GroundGround DisplacementDisplacement andand StrainStrain 410 0 100 200m 433 308 142 385 244 226 195 170 385 186 Horizontal 167 260 298 270 Displacement 328 108 (200cm) 282 147 87 191 2% 433 159 91 109 81 108 Tensile Strain 315 112 74 93 120 140 Comprossive 150 89 83 352 143 Strain 179 80 76 83 314 68 75 75 135 81 140 97 73 189 160 72 Damage66 81 Rate of Water Pipes (A, T, K Joints) 84 65 55 235 40 82 61 35 312 123 90 138 54 30 N ≒ 8.3 ε 101 142 25 54 76 20 98 251 59 15 57 64 10 52 Damage Rate Damage 282 98 Rate Damage 52 54

N (Number/Km) 72 N (Number/Km) 5 70 0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 Tensile Ground Strain (%) GroundGround DisplacementsDisplacements andand StrainsStrains forfor DesigningDesigning ofof 17 WaterWater SupplySupply FacilitiesFacilities Ground Behind Quaywall Ground Displacement 100m

εG : Ground Tensile Strain =1.2~2.0% QUAYWALL

Inclined Ground Ground Displacement

Distribution of Ground Displacement

Empirical Formula δG=0.77～0.96Hθ ： δG Maximum Ground Displacement in Horizontal Direction H ：Thickness of Liquefied Soil (m) θ ：Ground Surface Gradient (%) TensileTensile GroundGround StrainsStrains CausedCaused byby TheThe 19951995 KobeKobe EarthquakeEarthquake18

HUKAE- UOZAKI- HAMA MIKAGE- HAMA HAMA OASAKA BAY Mean Value 0.91% MAYA WHARF Non-Exceedance Ground ONO- HAMA Probability Strain(%) ROKKO ISLAND 50% 0.76 80 70% 1.21 PORT ISLAND 70 90% 1.88 60 03km50 Non- Exceedance Non- Exceedance Probability 70% Probability 90% 40 εG=1.21% εG=1.88% 30 20 10

Probability of Occurrence (%) 0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 Ground Tensile Strain (%) EarthquakeEarthquake ResistanceResistance DesignDesign ofof PilePile FoundationsFoundations againstagainst19 LiquefactionLiquefaction--InducedInduced GroundGround DisplacementDisplacement

Soil Spring

Non-Liquefaction Ground Soil Displacement

Liquefaction Drag Force by Ground Flow Soil (Equivalent Earth Pressure : 0.3) Flow Flow ImprovementImprovement ofof EarthquakeEarthquake ResistanceResistance 20 ofof WaterWater SupplySupply SystemsSystems

1) Replacement of Low Strength Pipes with High Earthquake- Resistant Pipes 2) Reinforcement of Water Facilities : Reservoirs, Filtration Plant Facilities 3) Establishment of Emergency Water Supply Systems ・Construction of Emergency Water Storage Facilities ・Mutual Support System among Water Utilities 4) Construction of Data Base for Water Supply System by GIS ReinforcementReinforcement ofof ExistingExisting WaterWater FacilitiesFacilities 21 --TokyoTokyo MetropolitanMetropolitan GovernmentGovernment--

Mt. Fuji

Yamaguchi Reservoir

Murayama Lower Reservoir ReinforcementReinforcement ofof EarthEarth DamsDams byby TheThe 22 TokyoTokyo MetropolitanMetropolitan GovernmentGovernment Body of Existing Dam Additional Embankment Additional Embankment

EL.86.000

Cross-Section of Earth Dam SourceSource RegionsRegions ofof Tokai,Tokai, TounankaiTounankai andand NankaiNankai EarthquakesEarthquakes23 （（ NationalNational CouncilCouncil frofro DisasterDisaster PreventionPrevention））

Tonankai earthquake(M8.1)

Nankai earthquake(M8.4) Tokai earthquake(M8.0)

◎ ◎ ◎ Kyoto Nagoya Shizuoka Osaka ◎

Nankai trough 200ｋｍ DamageDamage EstimationEstimation byby NationalNational CouncilCouncil forfor 24 DisasterDisaster PreventionPrevention

Tonanakai-Nankai Tokai Earthquake Kobe Earthquake Earthquakes Collapsed and Burnt 222,000-260,000 330,000-360.000 117,000 Houses Death 7,900-9,200 12,000-18,000 5,520 Water 1,375,000 4,000,000 1,265,730 (Households) Electricity 1,300,000 2,500,000 2,600,000 (Households)

Lifelines Gas 720,000 750,000 857,400 (Households) Loss of Properties $220 billion $430 billion $100 billion FutureFuture EarthquakesEarthquakes beneathbeneath MetropolitanMetropolitan AreaArea 25 （（ NationalNational CouncilCouncil frofro DisasterDisaster PreventionPrevention））

Predicted Source Areas Damage Estimation

・ Collapsed Houses 190，000 ・ Burnt Houses 180，000 ・ Death 12，000 ・ Refugees 6，500，000

0 50km BriefBrief ReportReport onon DamageDamage CausedCaused byby 26 TheThe 20042004 NiigataNiigata--ChuetsuChuetsu EarthquakeEarthquake

Chuetsu Area, Niigata Prefecture Outline of damages N Human : Death 40, Injured 2,641 Sea of Japan Refugee 100,000 House : Totally collapsed 431 Nagaoka Sado Island Power failure 278,000 Water supply failure 110,000 Main Shock Ojiya 2000 1000 Date : 17:56, October 23, 2004

Tokamachi 500 Magnitude : Mj 6.8 Focal Depth : 8.0 km 200 Epicenter 100 Aftershocks

Niigata Peak ACC[gal] Approximately 700 times Prefecture 50 20 6.0

Number of Water Total Length Number of Damage Cut Off of Water Pipes Water Rate Notes Households (km) Pipe Breakage 1/km Nagaoka DCIP(A,K Joints) 1084 City 70,000 341 0.31 , VP Ojiya DCIP(A,K Joints) 12,000 316 102 0.32 City , VP Kawaguchi 1,500 Town

Main Caused of Damage to Water Pipes

1) Liquefaction and Its Induced Ground Failures (Lateral Spread) 2) Slope Failures and Embankment Sliding 3) Earthquake Ground Motions GroundGround FailureFailure andand DamageDamage toto WaterWater PipesPipes byby TheThe 28 20042004 NiigataNiigata--ChuetsuChuetsu EarthquakeEarthquake

Ground Sliding Failure of Embankment

Up-Lift of Sewage Manhole Natural Slope Sliding DamageDamage toto WaterWater PlantPlant FacilitiesFacilities 29

Mul-Function of Generator by Flooding Ground Subsidence due to Liquefaction

Ground Subsidence due to Liquefaction AccelerationsAccelerations onon GroundGround SurfaceSurface byby 30 20042004 NiigataNiigata--ChuetsuChuetsu EarthquakeEarthquake Tokamachi - NS Ojiya - EW Kobe Earthquake 1500 1000 Tokamachi-NS Max=1716 Gal 2 500 5000 0 -500 After The Kobe -1000 Earthquake Acceleration cm/s -15000 10 20 30 40 50 Time(sec)

1000

1500 500 1000 Ojiya-EW Max=1314 Gal 2 500 0 -500 -1000 Before The Kobe Acceleration Response Spectrum (Gal) Acceleration cm/s -15000 10 20 30 40 50 Earthquake Time(sec) 100 0.1 0.5 1 5 Period (sec) Cover page

Title: Evaluation of Hydraulic Transients and Damage Detection in Water System under a Disaster Event

Authors: Masanobu Shinozuka (Contact person) UCI Distinguished Professor and Chair Department of Civil and Environmental Engineering University of California, Irvine E-4150 Engineering Gateway Irvine, CA 92697-2175 Phone (949)-824-9379 Office Fax: (949)-824-9446 Fax E-mail: [email protected]

Xuejiang Dong (co-author) Research Associate Department of Civil and Environmental Engineering University of California, Irvine E-4315 Engineering Gateway Irvine, CA 92697-2175 Phone (949)-824-9388 Office Fax: (949)-824-9389 E-mail: [email protected]

1 Evaluation of Hydraulic Transients and Damage Detection in Water System under Disaster Events

Masanobu Shinozuka1, and Xuejiang Dong2

ABSTRACT

This study explores the methods of rapidly detecting and locating the damage in a water delivery system taking advantage of sharply transient change in hydraulic parameters such as water head and flow rate under disaster events. In addition, we also considered the detection of equipment malfunction within the system that can cause similar and often more serious transient states. For this purpose, we used computer code HAMMER by Heasted Methods [1] in an ARC/GIS platform so that the inventory, operational, and management features can be integrated into the transient analysis all at the outset. We envision that the emerging sensor and data transmission technology will make it possible to monitor, process, and analyze the key hydraulic data in real-time, and extract the signature (such as maximum water head gradient) which can most effectively identify the severity and location of damage and malfunction. The proposed technology will serve as a next generation of the SCADA (Supervisory Control and Data Acquisition) system. Current generation of SCADA system that the utility industry deploys primarily for the purpose of system operation, not for rapid response to acute transients resulting from severe damage sustained by pipes, sudden stoppage of pump operation , and the like. We demonstrate by numerical simulation that the local water head gradient, for example, can serve as key signature, and that the source of damage/malfunction is at the joint closest to this point of maximum gradient. This research is also consistent with the national effort in enhancing the level of homeland security as described in the 2003 Academy Press publication “Making the Nation More Secure” [2] which identified future development of SCADA as one of the most critical agenda items for enhanced national security.

INTRODUCTION

Urban water delivery network systems, particularly the underground components, can be damaged due to earthquakes, severely cold weather, heavy traffic loads on the ground surface, and other causes. In all these situations, the damage cannot be detected and located easily, especially immediately after the damaging event. In recent years, real-time or near real-time damage assessment and diagnosis of buried pipelines has attracted much attention from researchers focusing on early detection of the damage severity and location. However, due to large size of network and complex nature of the physics that affect the pipe damage, particularly under seismic waves, such detection still remains difficult to achieve. As a possible solution, time histories of these and other parameters can be simulated with the aid of computer codes that are capable of transient analysis and allow creation of new nodes at the location of pipe damage. In addition, we also considered the detection of equipment

1 UCI Distinguished Professor and Chair of Department of Civil and Environmental Engineering, University of California, Irvine, E-4150 Engineering Gateway, Irvine, CA 92697-2175 2 Research Associate, Department of Civil and Environmental Engineering, University of California, Irvine, E-4315 Engineering Gateway, Irvine, CA 92697-2175

2 malfunction within the system that can cause similar and often more serious transient states. It is envisioned that, in the very near future, the current generation of SCADA system that the utility industry deploys primarily for the purpose of system operation, not for rapid response to acute transients resulting from severe damage sustained by pipes, sudden stoppage of pump operation, and the like. We demonstrate by numerical simulation that the local water head gradient, for example, can serve as key signature, and that the source of damage/malfunction is at the joint closest to this point of maximum gradient. Hence, given adequately dense sensor network data, we can reduce the problem to that of finding numerically and in real-time the point of maximum gradient from the optimal 3D surface fitted to the data. The magnitude of the gradient is expected to relate closely to the severity of the anomaly that caused the transient and this relationship will be the subject of future study. Also, it is the matter of future study to develop, by means of analytical simulation, first principle procedure of detection and associated algorithm. All these, together with field experiment and verification, lay the foundation for establishment of an optimal strategy toward cost-effective monitoring and emergency response taking into consideration, among other things, the type, number and location of sensors and nodes to be installed. Regional water utilities (e.g. Irvine Ranch Water District) will participate in this research providing technical data so that their system can be used as test-bed. The localization technology described above can also be applied to other lifeline systems such as power and transportation networks with appropriate modifications. The essence is the integration of a dense sensor network with rapid and robust data transmission capability, with the software capable of recognition of damage unique for each lifeline system. This research is also useful in enhancing the level of national security as described in “Making the Nation More Secure” (2003). This publication identified future development of SCADA as one of the most critical agenda items for enhancement of national security. In the last ten years or so, many researchers attempted to develop real-time damage and assessment diagnosis techniques for buried pipelines subjected to earthquake ground motion. Some researchers focused on establishing the relationship between damage ratio (breaks per unit length of pipe) and ground motion, taking the soil condition into consideration (e.g., Nishio [3], Tanaka[4], Yamazaki [5] ). Eguchi [6] put forward a method in which nominal damage estimated through some earthquake parameters is updated gradually based on the collection of post-earthquake observation information. In addition, Takadea and Ogawa [7] discussed seismic monitoring and real-time damage assessment, and Shinozuka et al [8] developed a methodology to detect the damage location and severity with the aid of neural network methods, and applied the method to a water network that consisted of 31 nodes and 50 pipes under the assumption of equilibrium flow. However, real water networks consist of a much larger number of nodes and links in a more complex topology and the system will produce a transient state of behavior when it suffers from damage. In 2004, Shinozuka and Dong developed a GIS-based methodology in which the correlation analysis was used to determine the damaged pipes using MLGW water system [9]. The present study demonstrates a method where the transient analysis is carried out between two equilibrium states of the water flow before and after the damage, which leads us to a rational, and cost-effective damage identification procedure for water delivery system.

3 DAMAGE DETECTION AND LOCALIZATION OF PIPE NETWORK

Hydraulic Transients A hydraulic transient represents a temporary flow, pressure, and other hydraulic conditions that control a hydraulic system, in this case, a water delivery system, between an original (first) steady state and final (second) steady state the system achieves after the effect of the disturbance that caused such a transient is absorbed into the second state. The disturbance includes such events as a valve closure or opening, a pump stopping or restarting depending on power supply, and pipe damage or break leading to water leakage. The transient can produce a significant change in water head and pipe pressure. In fact, as described in more detail in what follows, it is envisioned that the sudden change of such pressure will generate a measurable pressure wave and used for detection and localization of pipe damage due to seismic forces. If the magnitude of this transient pressure is beyond the resistant capacity of system components, their failure can induce disastrous effects on the water system [1] where it is also suggested that these effects include (1) high or low transient pressure which results in pipe burst or collapse, (2) high transient flow which can loosen the deposits and rust and thus degrade water quality, (3) high transient forces on pipe bends and other fittings, which can cause joints to move, and (4) column separation and vibration which may cause pipes to rupture, and flanged pipes and fittings (elbows and bends) to dislodge. Therefore, it is more than prudent to simulate the transient behavior of the water system under various adverse scenarios in order to assess the magnitude of their impacts. In this study, the industry-grade computer code HAMMER developed for the transient analysis of hydraulic systems by the Haesead Methods is employed to generate time histories of key hydraulic parameters (primarily water head and flow rate). The analysis is carried out for a hydraulic system as shown in Figure 1 (intact water system) which appears in HAMMER User’s Guide [1]. This water system consists of two reservoirs, one pump, one valve, thirty-eight nodes and 54 pipe links. In the ensuing analysis, we consider a case in which pipe break is assumed to occur at the mid point of link 111 in which case, a node is created and labeled node J10 as shown in Figure 2. The physical parameters of nodes and pipes are listed in Tables 1 and 2 where the following comments apply.

1. Pipe or link has two ends or nodes. One is defined as “from node” and the other as “to node”, which are labeled as FNode and TNode in Table2. 2. In Figure 1, connection node represents a junction at which two or more links meet. Consumption node means that the node from which water is supplied to the user at a specified level of flow rate and water head. 3. In Figure 2, the damaged network with pipe break location ‘+’ at J10 (in Table 1 in bold and italic) is shown. In this paper, the damage is modeled as an orifice. The original pipe P111 (in Figure 1) is divided into two pipes P11 and P12 (in Figure 2, P11 and P12 in bold and italic in Table 2). 4. The water head is measured in meter and the water flow rate is measured in cubic meter per second (cms). For flow rate, the positive value means the water flows in the direction from FNode (from node of pipe) to TNode (to node of pipe) while the negative value means the flow in opposite direction (from TNode to FNode).

4 Table 1 Nodal Parameters LABEL CATEGORY ELEVATION LABEL CATEGORY ELEVATION J1 Junction 412 J23 Junction 396 J2 Junction 395 J24 Junction 397 J3 Junction 395 J25 Junction 408 J4 Junction 386 J26 Junction 390 J5 Junction 380 J27 Consumption 395 J6 Junction 420 J28 Junction 396 J7 Junction 395 J29 Consumption 396 J8 Junction 395 J30 Junction 396 J9 Junction 395 J31 Junction 396 J10 Orifice 395 J32 Consumption 397 J11 Junction 410 J33 Junction 410 J12 Junction 420 J34 Consumption 420 J13 Consumption 390 J35 Junction 372 J14 Junction 396 J36 Junction 360 J15 Junction 397 J37 Consumption 355 J16 Junction 397 PJ1 Junction 363 J17 Junction 380 PJ2 Junction 363 J18 Junction 420 PMP1 Pump 363 J19 Junction 435 Res1 Reservoir 383 J20 Consumption 410 Res2 Reservoir 456 J21 Consumption 385 VLV1 Valve 395 J22 Junction 395

Table 2 Pipe Parameters Pipe Length Dia FNode_Head TNode_Head Flow Label FNode TNode (m) (mm) D-W (m) (m) (cms) P1 PJ2 J1 50 600 0.019 463.5 463.2 0.473 P2 J1 J2 380 600 0.019 463.2 461.5 0.473 P3 J2 J3 300 450 0.021 461.5 460.3 0.208 P4 J3 J4 250 450 0.021 460.3 459.3 0.208 P5 J4 J5 400 450 0.021 459.3 457.7 0.208 P6 J5 J6 250 450 0.021 457.7 456.7 0.208 P7 J6 Res2 175 450 0.021 456.7 456 0.208 P8 J2 J7 500 450 0.02 461.5 458.4 0.265 P9 J8 J9 300 450 0.02 458.4 456.5 0.265 P10 J9 J13 250 300 0.023 456.5 455.1 0.084 P11 J9 J10 100 300 0.022 456.5 395 0.09 P12 J10 J11 50 300 0.022 395 455.6 0.09 P13 J12 J11 200 300 0.025 455.3 455.6 -0.039 P14 J13 J17 300 300 0.024 455.1 454.4 0.054 P15 J9 J14 200 300 0.022 456.5 455.2 0.091

5 Table 2 Pipe Parameters (Cont’d) P16 J15 J14 150 300 0.027 455.1 455.2 -0.024 P17 J11 J15 200 300 0.024 455.6 455.1 0.051 P18 J15 J16 200 300 0.028 455.1 455.1 0.017 P19 J16 J12 175 300 0.025 455.1 455.3 -0.039 P20 J17 J21 300 200 0.024 454.4 451.8 0.037 P21 J17 J18 500 300 0.03 454.4 454.5 -0.013 P22 J14 J18 200 300 0.023 455.2 454.5 0.067 P23 J18 J22 300 200 0.024 454.5 452 0.037 P24 J19 J18 200 300 0.029 454.5 454.5 -0.017 P25 J15 J19 250 300 0.024 455.1 454.5 0.058 P26 J23 J19 300 200 0.024 452.1 454.5 -0.035 P27 J19 J24 300 200 0.024 454.5 452.2 0.035 P28 J20 J19 220 200 0.033 454.4 454.5 -0.005 P29 J16 J20 260 300 0.024 455.1 454.4 0.056 P30 J25 J20 200 200 0.024 452.4 454.4 -0.041 P31 J21 J26 500 200 0.026 451.8 449.8 0.025 P32 J22 J21 400 200 0.03 452 451.8 0.008 P33 J22 J27 500 200 0.024 452 448.1 0.035 P34 J23 J22 450 200 0.031 452.1 452 0.006 P35 J23 J30 300 200 0.034 452.1 448.9 0.035 P36 J23 J24 150 200 0.032 452.1 452.2 -0.006 P37 J24 J25 200 200 0.029 452.2 452.4 -0.011 P38 J27 J26 400 200 0.026 448.1 449.8 -0.025 P39 J28 J27 200 200 0.028 448.4 448.1 0.015 P40 J28 J29 150 200 0.029 448.4 448.5 -0.01 P41 J29 J30 150 200 0.026 448.5 448.9 -0.02 P42 J31 J28 150 200 0.032 448.5 448.4 0.005 P43 J30 J31 200 200 0.039 448.9 448.5 0.015 P44 J31 J32 250 200 0.029 448.5 448.3 0.01 P45 J32 J24 400 200 0.024 448.3 452.2 -0.04 P46 J25 J33 150 200 0.025 452.4 451.5 0.03 P47 J33 J34 500 200 0.025 451.5 448.6 0.03 P48 J35 J17 200 200 0.025 453.3 454.4 -0.03 P49 J35 J36 200 200 0.025 453.3 452.1 0.03 P50 J36 J37 200 200 0.025 452.1 451 0.03 PMP1D PMP1 PJ2 50 600 0 463.5 463.5 0.473 PMP1U PJ1 PMP1 50 600 0 382.8 382.8 0.473 PS1 Res1 PJ1 50 600 0.019 383 382.8 0.473 PVD VLV1 J8 0.3 425 0 458.4 458.4 0.265 PVU J7 VLV1 0.3 425 0 458.4 458.4 0.265

6 P10 Pipe #10 J9 Joint #9 VLV1 Valve #1 PMP1 Pump #1

Figure 1 Water Delivery System

P10 Pipe #10 J9 Joint #9 VLV1 Valve #1 PMP1 Pump #1

Figure 2 Damaged Water Delivery System (P111 Damaged)

7 Four event scenarios are considered and modeled as follows. 1. pipe P111between Joints 9 and 11 (initially not damaged) suffers a break in the middle at time t=5 sec. We create Joint 10 at this mid point servings as an orifice through which the water leaks as shown in Figure 2. Pipes 11 and 12 are also created in place of P111 as a part of the network. No repair is made. 2. The same damage scenario as (1) above, but repair is made at t= 15 sec* (*unrealistic, analytical convenience only) 3. At t= 5 sec, the pump stops due to loss of power. The pump is not restarted. 4. The same power loss as (3) above, but at t=20 sec,* operator restarts it activating emergency power.

HAMMER code can provide a wide range of system performance information under these scenarios. However, only time histories of water head and flow rate are plotted here. The water head histories at J9, J11, J13 and J20 (the last two are consumption nodes) are plotted for these four scenarios as shown in Figures 3-6. The water head time histories show that significant hydraulic transients can occur at the nodes which are close to damage location, while the nodes far away from the damaged pipe experience less prominent transient behavior. This is obviously what we expect and yet these quantitative results are the pillars of the framework for the proposed advanced SCADA system.

Damage Detection A method of damage detection and localization, including the identification of malfunctioned equipment, is described here that is based on the comparison of the hydraulic parameters (the water head in this case and flow rate will also be considered in future study) before and after the event. For the primary purpose of a rapid detection and localization, it is most effective to catch the sign of change at the outset of the event. Fortunately for a sudden change such as a pipe break and pump stoppage, the response of the network is rapid particularly in the neighborhood of the source, as demonstrated in Figures 3-6. This suggests that some measurable signatures that indicate the rapidity of this change can be used for this purpose. One convenient quantity that serves this purpose is the water head gradient as defined below. − HH D −= 12 (2) − tt 12 Here H2 and H1 are the water head of a node at the time t2 , t1 respectively and t2-t1=0.2 second in this study. During the steady state normal operation, D is usually very small. In this paper, the water head gradient measured at the observation nodes are integrated into the GIS platform for real-time visualization and for other advantages. In fact, the time histories of water head is generated and shown in Figure 3-6 under the scenarios. Figure 7 and 8 are the water head gradient distribution under event scenario 1 and 3 respectively. They indicate that the point of highest value of D as the location where damage occurred.

(a) J9 (b) J11

Figure 3 Nodal Water Head Time Histories under Event Scenario 1

(a) J9 (b) J11

Figure 4 Nodal Water Head Time Histories under Event Scenario 2

(a) J9 (b) J11

(c)J13 (d) J20

Figure 5 Nodal Water Head Time Histories under Event Scenario 3

(b) J11 (a) J9

Figure 6 Nodal Water Head Time Histories under Event Scenario 4

Figure 7 Distribution of Water Head Gradient due to a Pipe Break

Figure 8 Distribution of Water Head Gradient due to Pump Losing Power

11 Development of Advanced SCADA

Damage localization in utility networks was studied in the context of SCADA [5,6,7]. At this time, we plan to instrument Irvine Ranch Water Distribution’s network using MEMS sensors at possible locations (e.g. surface of pipes at manhole and hydrant, [8] ) to observe sudden changes in the water pressure. This will not require invasive procedures and still identify the extent and location of the damage. The result of transient simulation analysis using a virtual network (Fig.1) demonstrated sharper changes in the pressure (Figs. 3-6) at the joints closer to the location of a pipe break. These changes observed at a large number of sensor sites through the SCADA consisting of the MEMS (Micro Electro Mechanical Systems) sensors can be used to identify the location and extent of the damage. The pressure and flow rate changes themselves are not necessarily the quantities to be monitored, but it is the sudden vibration activity that is induced by propagating waves originating from the point of pipe damage. In this context, acoustic emission techniques could prove to be useful depending of pipe material. Clearly, monitoring of other facilities in the water network systems such as central buildings and filtration plants must also be integrated in the SCADA system. We use again the Irvine Ranch Water District’s network as test-bed for this purpose as we make progress.

CONCLUSION AND FUTURE RESEARCH The purpose of this study was to develop a methodology to identify the location and determine the severity of damage in a water delivery system by monitoring water pressure on-line at densely installed sensor locations within the system. Water head gradient is introduced as index and used as a key parameter to locate damaged pipe when the water system exhibit acute transient behavior. Currently, only water head is considered for the stated purpose. Actually, the flow rate change also provides useful information and will be used to enhance the accuracy of the identification result for this purpose. The method, which is based on on-line water pressure variation before and after water system damage or malfunction, will be integrated with GIS-based SCADA system to provide a practical real-time damage identification method and a decision support tool for the effective disaster response. The future study will be focused on (1) optimizing the number of monitoring stations with careful selection of their locations, and (2) refining the methodology to achieve more accurate results, particularly with the effective usage of water flow information and (3) applying this method to regional water utilities (e.g. Irvine Ranch Water District).

ACKNOWLEDGEMENTS

This work was supported under National Science Foundation Grant CMS-0112665. REFERENCES [1] HAMMER User’s Guide (2003), HAESTAD Press, CT, USA, [2] National Academies (2003) “Making the Nation More Secure”, National Academy Press [3] Nishio, N. (1994) "Damage Ratio Prediction for Buried Pipelines Based on the Deformability of Pipelines and the Nonuniformity of Ground ," J. of Pressure Vessel Technology, ASME, 116, 459-466 [4] Tanaka, S., Shinozuka, M., and Hwang, H. H. M. (1993) "LIFELINE-W(2) User's Guide: a Program for Connectivity and Flow Analysis of a Water Delivery System Under Intact and Seismically Damaged Conditions." Technical Report, Princeton University

12 [5] Yamazaki, F., Katayama, T., and Yoshikawa, Y. (1994) "On-Line Damage Assessment of City Gas Networks Based on Dense Earthquake Monitoring." Proc. 5th U.S. Conf. on Earthq. Engrg., EERI, Vol.4, 829-837 [6] Eguchi, R. T., Chrostowski, J. D., and Tillman, C. W., (1994) "Early Post-Earthquake Damage Detection for Lifeline System." EQE Research Report prepared for National Science Foundation [7] Shinozuka M., Liang J.w. and Feng, Q., M., (2005). “Use of SCADA for Damage Detection of Water Delivery Systems”, Journal of Engineering Mechanics, ASCE, March,. [8] Takada, S., and Ogawa, Y. (1994) "Seismic Monitoring and Real Time Damage Estimation for Lifelines." Proceedings of the 4th U.S. Conference on Lifeline Earthquake Engineering, ASCE, 224-231 [9] Dong,X.J. and Shinozuka, M. (2004) “GIS-Based Seismic Damage Localization for Water Supply Systems” Proceedings of the 13rd World Conference on Earthquake Engineering, Vancouver, B.C., Canada, Paper No 520

13 Session 1: People’s Cooperation for Seismic Measures

S1-1 The Evaluation of the Public Awareness and the Measures against Earthquakes with the Use of a Survey Questionnaire” Presenter: Takashi Kashiwamura (Hachinohe Regional Water Supply Authority) Japan

S1-2 “Earthquake Countermeasures in Yokohama and Cooperative Activities with Residents” Presenter: Hironori Nukui (Yokohama Waterworks Bureau) Japan

S1-3 “Consideration for Resident Participation in Seismic Hazard Mitigation Measures to Secure Water Supply” Presenter: Masahiro Kimura (Osaka Prefectural Waterworks) Japan

S1-1

“The Evaluation of the Public Awareness and the Measures against Earthquakes with the Use of a Survey Questionnaire”

Presenter: Takashi Kashiwamura

(Hachinohe Regional Water Supply Authority, Japan)

The Evaluation of the Public Awareness and the Measures against Earthquakes with the Use of a Survey Questionnaire

Hachinohe Regional Water Supply Authority Takashi Kashiwamura

1. Introduction

Hachinohe is an earthquake-prone region, where an earthquake of more than a seismic intensity of four occurs once a year. The main measure against earthquakes taken by the Hachinohe Regional Water Supply Authority (HWSA) had been to employ earthquake-resistant pipelines since 1968, when the Tokachi-oki Earthquake (seismic intensity five) brought serious damage to the pipelines. In 1994, however, the Sanriku Haruka-oki Earthquake (seismic intensity six) brought serious damage to the pipeline network including the water distribution pipelines, but to the main facilities and the earthquake-resistant pipelines. As a consequence, nearly 30,000 households experienced a water shortage. Therefore, systems to minimize the earthquake-disaster, to restore lifeline functions quickly, to establish earthquake-resistant facilities and restoration systems, and to secure water supply systems in an extreme emergency have been performed as countermeasures.

Currently, only 22.1% of the pipeline networks are earthquake-resistant. Since a huge earthquake can occur without notice, there is an urgent need to establish the restoration systems and emergency water supply systems. Among them, the latter is especially important, because it is difficult for emergency water supply to perform effectively unless the people have clear information about it. In order to supply water quickly and effectively in the case of an emergency, there is a need to have good relations and to communicate effectively with its citizens, that is, the customers and/or clients. For this reason, one of the measures against earthquakes is to provide appropriate information and advice to its citizens.

In this workshop, public awareness of disaster prevention using regular survey-questionnaires, and the public relations activities of HWSA to citizens as to the measures undertaken against earthquakes, were reported.

-1- Table 1: The Rate of Earthquake-Resistant Pipelines of HWSA

Earthquake-resistant Total length of Rate of pipeline length the pipeline earthquake-resistant (m) (m) pipelines (%) Raw water transmission 4,268.7 16,037.9 26.6 main Transmission main 118,156.7 152,302.2 77.6 Water distribution pipe 303,776.9 1,760,580.6 17.3 Total 426,202.3 1,928,920.7 22.1

(According to surveys conducted until May, 2004)

2. Public reactions on the measures against earthquakes

On December 28th, 1994, the Sanriku Haruka-oki Earthquake caused serious damage to most of the facilities of Hachinohe City. Water supply services were suspended (about 30,000 doors in the main shock, 5,000 doors in the after shock.) because of the damage to the pipelines by the earthquake. After the earthquake, a questionnaire was distributed to the people who were unable to avail of water service – a service ever-present in their daily life. The questionnaire investigated the actual conditions such as the kind of concern and awareness they had about the damage the earthquake had on the water supply facilities and what they thought about the measures taken by HWSA to mitigate the effects of the damage caused by the earthquake.

-2-

The questionnaire regarding the Sanriku Haruka-oki Earthquake:

Term : 1995.9.14 ～ 1995.10.3 Area : Hachinohe Region (11 communities) Subject : Household Investigation Sampling Design : Random Sampling Method : Individual Survey-Interview No. of Households : 15,000 households Retrieval Rate : 95.6%

Items Investigated Concern over expected large earthquakes Whether suspended or reduced service occurred during the earthquake or not In case of water shortage, how to secure water Degree of satisfaction with the emergency water supply Kind of container used when provided with the emergency water supply Utilization of well-water Preparation of the emergency water Impact of suspended or reduced service to daily life Situation of flush toilets Recognition of the usefulness of the earthquake-resistant pipelines Recognition of the cost of the facility construction Concern over lifeline

-3- 1) Earthquake-resistant facilities

In this research, many Q11Q⑪災害に強い施設整備をすすめるにあたり、 What do you think about the duty of paying water charge 水道料金からの負担について on earthquake-resistant facilities? people showed concern about earthquakes. About 86% of the people thought it 32% 54% 4% 10% necessary to construct disaster-resistant facilities even if they have to pay for it. 0% 20% 40% 60% 80% 100% The people have understood, 負担は病やむをえないIt’s inevitable to bear the cost. 必要と思うが負担は少なくしてほしいIt’s necessary to bear the cost, but I want to reduce it. to some degree, the necessity 災害時の断水はやむをえないので負担したくないSuspension of service is inevitable; so I don’t pay the cost. of having disaster-resistant どちらとも言えないNo idea facilities.

In this way, HWSA adopted the earthquake-resistant pipelines with the joint structure (like S-type, SⅡtype, NS-type), which have had the best earthquake-resistant performance since 1996. This decision was based on their experience from the damage caused from earthquakes, the observation of earthquakes, and reports of the pipe accidents.

2) Emergency water supply activities

The respondents’ views 意見・要望 Opinion and Request and opinions were also

Request and complaint asked, in addition to the 地震・災害に対する要望・苦情about earthquake disaster 520 formal survey questions Construction,工事・維持管理 maintenance 73 (where the answers were Water-quality,水質・水源 source 97 made through the selection Sales, management営業・経営 320 of options). There were Suspended service, Reduced断水・減圧 pressure 24 no restrictions on the topic Thanks, encouragement感謝・激励 270 or subject matter of the その他Others 186 views and opinions, in 0 100 200 300 400 500 600 order for the respondents Numbers件数

-4- to be open and frank.

As a result, 1,490 or 10.4% of the respondents gave their views, where 520, or most of the answers were on earthquakes and disaster. The views centered on, their irritation and other complaints on the suspension of services and limited water supply. There were many complaints to the kind of water wagons and type of publicity made. In concrete terms, a sampling of views and opinions are as follows: “The service points should have been obvious”, “we weren’t informed of the time of emergency water supply”, “we had to wait for emergency water supply which was too long”, and “tell us when the water service will be reopened.” These complaints revealed that the people did not have appropriate and adequate information, which resulted in their irritation and complaints about the emergency water supply.

Through this investigation, it becomes important not only to undertake measures against damage caused by earthquakes, but also to inform and educate the people on the policies the water supply organization has made to mitigate the damages caused by earthquakes. By doing so, the people are made to understand and appreciate the efforts, consequently getting them to work together with the organization in the measures against earthquake damage, as well as improving earthquake-resistant facilities, emergency water supply systems, and restoration systems.

3. Public relations and public information schemes to disseminate earthquake-resistant measures

Public relations and information activities have an important role in making restoration from disaster quicker, in improving effectiveness of emergency water supply and in getting rid of anxiety on how to secure water supply. Therefore, HWSA prepared a PR manual in 1998, as one of the promotional activities on accident prevention in times of emergencies such as earthquakes geared to each resident.

(1) Public relations and information activities during normal situations

In order to make restoration activities smoother, HWSA informs the people of the

-5- disaster prevention system and the way of securing water supply so that people become more aware of disaster prevention.

1) Disaster awareness and prevention

HWSA disseminates information to each resident on accident awareness and prevention.

2) Public relations and information activities in times of emergency water supply during disasters

By providing and disseminating to the people the measures and the system on emergency water supply, it becomes easier to get cooperation of the people in times of disaster. The PR activities in times of emergency water supply are as follows:

Amount of drinking water to be secured from HWSA;

Manner of water supply distribution (water wagons, plastic canteens, cartons of water);

Identifying water supply station (showing water supply station signs clearly)

Amount of supplied water per capita;

Others such as water bottles, request and contact for the emergency water supply

3) Public relations and information activities on the measures to be undertaken by each individual on disaster preparedness.

By emphasizing that everyone may keep emergency drinking water computed at nine liters per three days per person for every member of the family

Keeping emergency drinking water;

Using the bathtub as a water container;

Information on water quality and the methods of maintaining water quality by boiling drinking water.

-6- 4) Public relations and information activities for water supply stations in local areas

HWSA prepares a disaster prevention map including the location of each local water supply station to be distributed to every home;

HWSA provides information on the water supply station through the use of proper signages.

5) Public relations and information activities to be undertaken on disaster prevention

Construction of earthquake-resistant water facilities and pipelines, promotion of the water supply block systems, and improving backup systems;

Construction of the ring-shaped earthquake-resistant distribution trunk pipelines;

Keeping water bags and plastic bottles.

6) Manner of public relations and information dissemination

Utilizing the print media through the newspapers and brochures to be distributed to every home on disaster prevention;

Utilizing the broadcast media, such as the TV and radio during “the disaster-prevention week”, “the water-supply awareness week”, etc;

Construction of emergency water storage tanks and storage of water bags and plastic bottles.

(2) Public relations and information activities in the event of an earthquake

HWSA offers information of the damage status, prospect of restoration, the emergency water supply to the media and the people in the event of an earthquake, thereby minimizing concerns over drinking water and water for domestic use. These activities will reduce the confusion especially to the marginalized sectors of the citizenry such as the elderly, the handicapped and foreign nationals, etc.

1) The content of the PR materials

Status of the damage on the water facilities and the pipelines;

-7- Coverage area where water service is suspended;

Places and measures where emergency water supply is available;

Schedule and prospects when water service can be restored;

Information on water quality.

2) Manner of public relations, public information and dissemination

Public information utilizing the broadcast media such as the TV and radio

Public information utilizing mobile loudspeaker vans;

Public information by community wireless

Public information using the print media such as the newspapers;

Public information by word of mouth through neighborhood associations;

Public information through the answering service.

3) Information for members of the media

HWSA actively offers information for the print and broadcast media and makes quick and accurate reports to remove any cause for concern from the people about drinking water and domestic water supply.

4. Public awareness regarding earthquake-resistant measures

Public awareness research is done every three years and there have been four investigations so far. The purpose of the research is to assess how much the people are familiar with and understand the water supply system through the organization’s public relations and information dissemination efforts. In the research design, the sampling method used is random sampling of 5% of the customers. The items in the questionnaire are on water quality, commodity charge, water service, publicity efforts, and so on. The results of the research conducted in 2001 are found in the succeeding charts.

-8- Q. What kind of information do you want to get about the waterworks system? Multiple answers are acceptable.

Others 5%

Water supply equipment 5%

Couseling window 10%

Waterworks 4%

Financial standing 5%

Trouble & reparing 10%

Preventing a freeze 7%

Saving water 8%

Against disaster 13%

Commodity charge 13%

Water quality 20%

0% 5% 10% 15% 20% 25% Percentage

Q. Do you prepare for the emergency of earthquakes and secure drinking water?

(In （全体集計）total) 36% 48% 16%

0% 20% 40% 60% 80% 100%

(Each（グループ別集計） group)

広報紙をいつもみる人Always read 42% 46% 12%

Sometimes広報紙を時々見る人 read 34% 50% 16%

広報紙を見たことがない人Not ever read 26% 49% 25%

0% 20% 40% 60% 80% 100% Not equipped 備えているEquipped 備えていないが確保は必要考えていないNot equipped but feel the need

-9- In the first question, the answer “Preparation against earthquakes” ranked second and equals to the result of the research conducted three years ago. The answer has the same level of interest compared with water quality and commodity charge. The result implies that people recognize earthquakes as an important concern. In the second question, the answer “Equipped water in case of emergency” counts only 36% of the answers.

On the contrary, 48% of the customers chose “Not equipped but feel the need of securing drinking water”. Therefore, a total of more than 80% customers feel the need for emergency water equipment. These results show that customers have become more aware of the importance of preparing themselves in the event of an earthquake, although they still have not taken action.

In addition, the group that answered “always read publicity” has higher awareness about storing drinking water than the group that answered “not ever read publicity”. This means public interest is promoted with adequate publicity of the measures regarding earthquakes

We have established “the waterworks cooperator* system”. The main activities of the cooperators are to conduct events such as visiting the water supply facilities and workshops, to participate in various events such as cleanup campaigns, and to claim their opinions in publicity. The result of the disaster-prevention-awareness research about the waterworks cooperator refers to the several charts below.

* Qualified waterworks cooperators are recommended by the people of the community. There are not less than 70 cooperators that serve for a fixed term of two years.

-10- Q. Do you prepare for emergencies such as earthquakes and secure drinking water?

2003H15-16年度～04 59% 41%

2001H13-14年度～02 57% 43%

1999H11-12年度～00 42% 58%

1997H 9-10年度～98 44% 56%

0% 20% 40% 60% 80% 100% 確保しているKeeping していないNot keeping

Q Since the Sanriku Haruka Oki earthquake, has your sense of crisis against earthquakes changed?

2003H15-16年度～04 10% 83% 7%

2001H13-14年度～02 16% 82% 2%

1999H11-12年度～00 34% 58% 8%

1997H 9-10年度～98 47% 53% 0%

0% 20% 40% 60% 80% 100% Having Little 危機感があるHaving sense 危機感が薄いsense 危機感がないHaving no sense

There is an overwhelming majority of the customer-respondents who actually keep drinking water in case of emergency compared with the research for other citizens. It seems that waterworks cooperators felt the need and the importance of taking preparations for earthquakes by visiting waterworks facilities and workshops. It would therefore be important to continue public relations and information activities.

-11- On the other hand, even the awareness of the cooperators also gradually began to wane as shown in the figures in the chart above. There is some difficulty in generating enough awareness in cases of emergencies.

5. The participation of public in training of earthquake-resistant measures

Every year general emergency drills are conducted in each municipality and three items are selectively operated, as shown below.

1) The outline on the explanation of the assistance and the material stockpile in restoration are as follows:

The storage situation of the emergency restorable materials:

DIP Straight pipe φ75～φ800 161 pipes

DIP Special fitting φ75～φ800 246 pipes

Joint for repairing ACP breakage 30 set

Joint and other repair material 1 set

The equipment for the emergency water supply:

Water Bags Carrying Bins Plastic Bottles

4 Liter 5 Liter 10 Liter 10 Liter 20 Liter 0.5 Liter 200 5,170 38,165 790 999 2,000

The details of the emergency water supply facilities:

Emergency water receiving tanks 5 places

Emergency water supply bases 31 places

The cooperative organizations that come into agreements:

-12- - The mutual water service disaster back-up agreement of Pref. Aomori:

Pref. Aomori (1969)

- JWWA Tohoku district branch Letter of agreement on mutual back-up in disaster:

JWWA Tohoku district branch (1997)

- Letter of agreement on mutual back-up during disaster:

Saitama-shi Pref. Saitama (1995)

Ninohe-shi Pref. Iwate (2003)

Taneichi-Cho Pref. Iwate (2003)

- Letter of agreement on restoration work in emergency and disaster:

Hachinohe Pipe Construction Society (1999)

Manufacturers which have an ability to repair main pipes: 46 companies

Manufacturers which have ability to repair pipe under 50: 25 companies

Earthquake-resistant joint official plumbers (JWWA authorized): 800 people

2) The emergency restoration training

Duplicating a water supply model (temporary pipe laying) on the ground and demonstrating water suspension and a restoration process in disaster.

-13- Emergency restoration training of waterworks facilities

Instructions on restoration Water service tank

Repairs inspection

Damaged point Pump operation

2 plumbers Sluice valve operation

3) Emergency water supply training

People actually experience using water supply tools to receive water from a tanker and emergency water supply equipments.

6. Final chapter

The chief factor that causes confusion, anxiety and irritation among the people in the case of suspension of water supply because of earthquakes is the lack of information on when and where to get water and on the prospect of restoration, rather than the amount of the water supply. In addition, although people do recognize the necessity of undertaking some mitigating measures against earthquakes and are always concerned about them, although the rate of actually taking concrete measures such as keeping drinking water isn’t sufficiently high. It is not easy to make the people aware of the measures against disasters

-14- without alarming them, while they’re losing the sense of crisis day by day. Therefore, a problem exists regarding getting the citizens to be adequately prepared for earthquakes.

To be able to improve the current situation, it is important to continuously disseminate relevant information to the people. It is not only the organization that can mitigate the circumstances; but also efforts undertaken by the individuals themselves. Hence, it is vital that the emergency activities are disseminated, such as the process of emergency water supply, the system of finding solutions in the event of suspension of water supply services. The suspension of water supply is critical to the supply of drinking water; but this also affects the daily domestic use of water such as bathing, doing the laundry, and flushing toilets. The longer suspension of water supply lasts, the worse things would get. Thus, there must always be a proactive exchange opinions with the people in order to get creative solutions that could enrich the disaster prevention measures.

-15- S1-2

“Earthquake Countermeasures in Yokohama and Cooperative Activities with Residents”

Presenter: Hironori Nukui

(Yokohama Waterworks Bureau, Japan)

Earthquake Countermeasures in Yokohama and Cooperative Activities with Residents

Hironori NUKUI Section Head Emergency Management Section General Affairs Division Yokohama Waterworks Bureau

Abstract

Water service plays a vital role in the lifeline that supports people’s living and social activities. Therefore, it is essential to supply a minimum amount of water necessary even immediately after the occurrence of an earthquake. Damage to waterworks facilities must be minimized. Furthermore, pipeline-based water supplies must be restored as soon as possible.

Based on the lessons learned from damage to the waterworks facilities during the Great Kobe Earthquake of 1995, the City of Yokohama has focused on three key measures of “strengthening the aseismatic properties of waterworks facilities,” “establishing earthquake disaster countermeasure sites” and “enhancing emergency activities.” At the same time, because association and/or collaboration with local residents is essential to facilitate emergency activities at the time of a disaster, Yokohama is also working to recruit people who can support and cooperate in emergency activities, and to establish a framework for facilitating emergency water supplies in cooperation with residents.

At this workshop, I would like to report on measures to deal with earthquakes that have been implemented by the City of Yokohama in collaboration with its residents.

1. Outline of Yokohama

Situated in the central part of the Japanese archipelago and some 20 km to 40 km southwest of metropolitan Tokyo, Yokohama has an area of about 434 km2 and a population of about 3.53 million as of March 2004. After Tokyo, it is the second largest city in Japan.

Since Yokohama established Japan’s first modern waterworks system in 1887, eight expansion projects have been undertaken to serve the growing population as well as the city’s increasing social and economic activities. At present, Yokohama waterworks operates water intake sources of 1,955,700 m3 and has a daily water supply capacity of 1,820,000 m3.

As of the end of March 2004, the total length of water pipes is approximately 9,080 km, which cover the entire city in a mesh configuration to provide a stable water supply throughout the area.

2. Basic Concept of Earthquake Disaster Countermeasures

Because Japan sits on the Pacific-rim earthquake belt, which is one of the world’s most active seismic zones, it is especially prone to earthquakes. Even after entering the modern ages, major earthquakes have shaken Japan many times. The Great Kanto Earthquake that struck the Kanto region in 1923 inflicted major damage on the entire Kanto area. In particular, the damage sustained by Yokohama was especially serious, and the waterworks facilities suffered catastrophic damage.

While Yokohama waterworks has promoted measures to deal with earthquakes based on these experiences, it has adopted three principal activities to implement such measures more effectively in view of the lessons learned from the Great Kobe Earthquake. These include:

(1) Strengthening the aseismatic properties of waterworks facilities in order to minimize damage. (2) Establishing earthquake disaster countermeasure sites to carry out emergency water supplies and emergency restoration in case water supplies are cut. (3) Enhancing emergency activities such as risk management at the time of disasters and implementing various training programs.

With a focus on these measures, the Yokohama Waterworks Bureau is operating its water service in consideration of the balance between hardware aspects such as the reinforcement of facilities and software aspects in terms of how to best utilize these facilities at the time of a disaster.

3. Earthquake Disaster Countermeasures

(1) Earthquake probability

Yokohama projects that the following major earthquakes may occur: a South Kanto earthquake with the epicenter located in Sagami Bay (magnitude 7 class, with a Japanese seismic intensity of 6 – 7 in Yokohama), a Tokai earthquake with the epicenter located in Suruga Bay (magnitude 8 class, with a Japanese seismic intensity of 5 in Yokohama) and a vertical-type earthquake with the epicenter located in the western part of Kanagawa prefecture (magnitude 6 class, with a Japanese seismic intensity of 4 – 6 in Yokohama).

(2) Probable damage

According to the projections of damage to waterworks facilities if a South Kanto earthquake occurs, which is expected to inflict the greatest damage among the projected earthquakes, damage is considered to be minor as backbone facilities that mostly consist of buildings have been sufficiently reinforced in terms of structure, materials, etc., as compared to those at the time of the Great Kanto Earthquake in 1923. However, water will be cut off in pump-fed areas if electric power is interrupted. Small- and medium-diameter water pipes are expected to sustain the greatest damage from an earthquake, and most damage will occur in these pipes.

(3) Strengthening the aseismatic properties of waterworks facilities

Measures to strengthen the aseismatic properties of facilities include replacing leaded joints with ductile cast iron and steel piping and reinforcing such piping, reinforcing aqueduct spans and open watercourses, providing duplexed electric power receiving systems and installing private power generators at purification facilities and pumping stations, reinforcing existing distribution reservoirs by constructing aseismatic embankments and reinforcing the interiors of major trunk pipelines and pipelines on slopes to increase their aseismatic properties. Through these measures, we are providing aseismatic reinforcement of facilities so that even a South Kanto earthquake whose seismic intensity is expected to be equivalent to that of the Great Kanto Earthquake could be endured. In addition, we are considering the prevention of secondary disasters.

(4) Establishing earthquake disaster countermeasure sites

To promptly and effectively conduct emergency water supply and emergency restoration activities, Yokohama has selected sites to provide for emergency water supplies at the time of a disaster in each area within the city, has stored materials and equipment necessary for emergency water supplies and emergency restoration, and has established disaster countermeasure centers to receive support staff members from water service operators in other cities. a. Establishment of emergency water supply sites

(a) Distribution reservoirs

Emergency shut-off valves have been installed in distribution reservoirs. If an earthquake of slightly less than the Japanese seismic scale of 5 or above occurs and if the water level declines to a pre-set level, earthquake management software that monitors purification plants is activated and an emergency shut-off valve of one of two tanks is closed to secure a supply of drinking water. At present, 24 distribution reservoirs secure a total effective water storage volume of approximately 980,000 m3. Of this volume, about 190,000 m3 (20% of the total volume) would be secured at the time of an earthquake. This amount corresponds to the volume of drinking water for all residents for one week (49 liters per person).

(b) Emergency underground storage tanks

Based on plans to enable residents to access drinking water within approximately 1 km, emergency underground storage tanks are being installed mainly at elementary and junior high schools that have been designated as medical aid centers among all local disaster preparedness centers (elementary and junior high schools). As of the end of March 2004, 133 tanks have been installed. Since the goal is to install 134 tanks, the last tank is now under construction.

To start pipeline-based emergency water supplies as soon as possible, earthquake-resistant pipes are installed in main distribution trunk lines that are relatively strong against an earthquake. Emergency water supplies are then connected to the earthquake-resistant pipes. Based on plans to enable residents to secure drinking water within approximately 500 m, most of these taps are being installed at elementary and junior high schools, etc. that have been designated as local disaster preparedness sited. As of the end of March 2004, these taps were installed at 324 sites. Plans call for installing taps at 358 sites by the end of March 2007. b. Establishing disaster countermeasure sites

(a) Storage sites for emergency water supply materials and equipment

Emergency water supply materials and equipment are stored at 14 sites in the city. Water supply tanks and polyethylene containers for emergency water supplies are stored at these sites.

(b) Storage sites for emergency restoration materials and equipment

Emergency restoration materials and equipment are stored at the premises of four distribution reservoirs within the city. With respect to the pipes buried underground in the city, those with diameters of 100 mm ~ 300 mm account for about 90 percent of the total length of ductile iron pipes. Since pipes with diameters of 400 mm or more are highly earthquake resistant, we stock about 6 km of pipes with diameters of 100 mm ~ 300 mm for emergency restoration. We also stock pipe fittings such as flange adaptors, rubber gaskets, bolt & nuts. In addition, to prevent the deterioration of stored materials, material yards are divided into two sections, with materials on one side used alternatively in each fiscal year for ordinary work so that the newest materials are always on hand.

Based on our experience in conducting support activities in Kobe, Yokohama started constructing facilities to accept support staff from other cities in 1995. By 1998, we completed eight buildings at seven sites in accordance with the development plans.

Each of these buildings has a total floor space of 170 m2 ~ 260 m2 and the capacity to accommodate approximately 30 persons. The facilities include warehouses, which can also be used for meeting rooms and storage space, kitchens, dining rooms, toilets, bathrooms and laundry rooms. The buildings are also equipped with the minimum daily necessities.

(5) Emergency water supply structure

With respect to emergency water supplies at the time of an earthquake in Yokohama, the Waterworks Bureau provides direct water supplies from emergency water supply sites and, at the same time, implements vehicle-delivered water supplies to local disaster preparedness sites with the cooperation of support personnel from other cities, etc. Each ward office is responsible for requesting support from local residents and volunteers in implementing vehicle-delivered water supplies to elderly people living alone, etc., in addition to direct water supplies from emergency wells, storage tanks, etc.

Additionally, to enhance emergency activities, the Waterworks Bureau has established a cooperative structure involving the cooperation of retired Waterworks Bureau personnel for emergency water supplies, etc., at the time of disasters. Now that the development of the facilities is nearly completed, the establishment of a framework of cooperative activities with residents for emergency water supplies must be promoted in order to further enhance emergency activities.

(6) Enhancement of emergency activities

Required earthquake countermeasures include hardware aspects such as the development of facilities and software aspects involving how to best use such facilities at the time of a disaster. What is required if an earthquake occurs and damage such as a water shortage occurs includes emergency water supplies and emergency restoration activities. As explained previously, the Waterworks Bureau is promoting the establishment of emergency water supply sites within areas that people can walk from their homes. Our plans call for the installation of underground storage tanks for emergency use at 134 sites and the installation of emergency water supplies at 358 sites. However, because a factor of uncertainty is involved regarding the number of bureau personnel that can be assembled, it appears very difficult to be able to conduct emergency water supply activities at all of these facilities by waterworks personnel alone, although this depends on the extent of the damage. Accordingly, it is vitally important to work for early restoration at the time of a disaster by conducting emergency activities under the combined efforts of government personnel and volunteer citizens.

To enhance emergency activities at the initial stage after a disaster occurs, the Waterworks Bureau is promoting the establishment of a framework under which emergency water supply activities are implemented in cooperation with residents. Measures to achieve this include not only the preparation of instruction manuals and the use of information systems, but also providing residents with training courses in conducting procedures for obtaining emergency water supplies from underground storage tanks. This will enable many waterworks personnel to engage in emergency restoration activities, aiding in quicker restoration of water service. a. Preparing manuals and establishing a structure for support from other cities, etc.

The basic action guidelines that indicate details about role sharing and what must be conducted are included in “Disaster Prevention Plans in the City of Yokohama (Earthquake Countermeasures)” and “Earthquake Countermeasure Handbook” prepared by the Waterworks Bureau. Based on these guidelines, we have prepared an assortment of manuals including the Manual for Requesting and Accepting Support from Other Cities, the Emergency Water Supply Manual and the Earthquake Countermeasure Manual (Purification).

These manuals clarify the duties assigned to each employee and facilitate a structure to enable each employee to act quickly and precisely.

In addition, because there are limitations in the capabilities of a disaster-stricken city to conduct restoration activities if a large-scale disaster occurs, a support structure has been established by concluding agreements on reciprocal support among cities. One such example is the Memorandum on Reciprocal Support at the Time of Disasters among Waterworks Bureaus in 13 Major Cities. Agreements have also been concluded with various organizations. b. Utilization of information systems

Rapid gathering and dissemination of accurate information is important in helping to prevent the spread of any disaster-induced damage and to conduct appropriate emergency activities, etc. For this purpose, Yokohama has developed the Yokohama Disaster Prevention Information System. In addition to meteorological data and earthquake data, this system supports emergency measures by collecting various kinds of information from organizations and bureaus handling lifeline services including water service on a real-time basis. At the Waterworks Bureau, the disaster countermeasure command headquarters (the Nishiya Command Office) consolidates such a variety of information, which is used for emergency activities in association with the Integrated Waterworks Facility Monitoring System.

Furthermore, in 2004, satellite telephones were introduced in the disaster countermeasure command headquarters (the Nishiya Command Office) and the Comprehensive Coordination Office (head office) for communications at the time of a disaster.

4. Cooperative Activities with Residents

The City of Yokohama has facilitated the establishment of local disaster preparedness sites in order to further strengthen the capability to cope with disasters within local communities based on lessons learned from the Great Kobe Earthquake.

Specifically, elementary and junior high schools (454 schools) that are located near citizens have been designated as places of refuge at the time of an earthquake. These schools have been equipped with emergency storage facilities to store emergency materials and equipment as well as foodstuffs, etc. As such, these schools are designed to function as local disaster preparedness sites at the time of a disaster to provide temporary shelter and to conduct rescue and relief activities by using stored emergency materials and equipment. Among these centers, local medical first-aid stations have been installed at 145 schools to provide emergency medical treatment and first aid for about three days after the occurrence of a disaster.

The local disaster preparedness centers are operated in cooperation with residents by establishing a Local Disaster Preparedness Steering Committee for each place of refuge. Each committee consists of members selected from the community, nearby schools and ward offices. The purposes behind the establishment of such committees include the promotion of rescue and relief activities through mutual cooperation among community residents by using the stored emergency materials and equipment, and the maintenance of safety and public order in the life of those forced to evacuate their homes.

From among the elementary and junior high schools that have been designated as local disaster preparedness centers, the Waterworks Bureau has installed underground storage tanks for emergency use at those designated as local medical first-aid stations. Most elementary and junior high schools that have been designated as local disaster preparedness centers have been equipped with emergency water supply taps. However, it is difficult to expect that only personnel of the Waterworks Bureau will be able to conduct emergency water supply activities at all of these sites.

To help prevent the spread of damage resulting from a disaster, it is important for each citizen to develop an ability to act calmly if an earthquake occurs, recognizing that “Each one of us is responsible for protecting our own body; all of us together will protect our town.” Furthermore, the combined efforts of government employees and the public at large aimed at rapid restoration of everyday life are indispensable in conducting emergency activities when a disaster occurs.

Accordingly, the Yokohama Waterworks Bureau is promoting the development of a framework for emergency water supply activities in cooperation with residents in response to a disaster. Specifically, training programs in handling emergency water supplies are being implemented at emergency water supply sites with the participation of local residents under association with each Local Disaster Preparedness Steering Committee. At the same time, the education of residents is being promoted to enable them to operate underground storage tanks for emergency use along with emergency water supply taps.

(a) Promoting the establishment of a framework to conduct emergency water supply activities using emergency-use underground storage tanks in cooperation with local residents

To support voluntary disaster preparedness activities by citizens, the Yokohama Waterworks Bureau has been implementing the Yokohama Disaster Prevention Licensing Project in cooperation with the City of Yokohama’s Emergency Management Operation Office of the General Affairs Bureau, the Fire Bureau and the Environmental Services Bureau. The purpose is to promote the establishment of a framework to conduct emergency water supply activities using emergency-use underground storage tanks in cooperation with local residents if a disaster occurs.

As part of this project, which started in 2004, a license of a “material and equipment handling instructor” is issued to each person who understands the handling of emergency-use underground storage tanks and the operation of power cutters, rescue jacks, portable rice cookers, temporary toilets, etc., which are stocked in local disaster preparedness sites. These centers also serve as shelters if a disaster occurs. This instructor then instructs residents who can serve as disaster preparedness leaders in the community.

The Yokohama Waterworks Bureau provides technical guidance to recruit instructors, conduct examinations and qualify citizens as instructors. A license is issued to a person who qualifies as an instructor. These instructors are expected to instruct disaster preparedness leaders in the community and to lead in disaster preparedness activities. Current plans call for recruiting about 100 instructors and about 6,000 disaster preparedness leaders.

(b) Developing a framework to conduct emergency water supply activities using emergency water supplies in cooperation with local residents

To enable the use of pipeline-based emergency water supplies as quickly as possible, earthquake-resistant pipes are installed in distribution trunk lines, which are relatively strong against an earthquake. Emergency water supply taps are then affixed to these earthquake-resistant pipes. Since these taps are essentially fire hydrants, great care is required in their operation. However, as is the case with emergency-use underground storage tanks, emergency water supply taps also constitute important emergency water supply sites in a disaster. Accordingly, we are now identifying problems in terms of handling these taps, etc., and studying ways to ensure that emergency water supply activities using emergency water supply taps can best be conducted in cooperation with local residents.

By registering volunteer retired personnel of the Waterworks Bureau who have experience and knowledge in water service operations, we are promoting the development of “support and cooperating personnel.” These volunteers would collect data regarding damage inflicted by a disaster on waterworks facilities and cooperate in the management of emergency water supplies from emergency-use underground storage tanks. This system was started in 1997 as the Support and Cooperating Personnel System at the Time of a Disaster. Currently, about 350 persons are registered under this system. Their term as a volunteer is three years.

(d) Strengthening disaster preparedness training

Under the recognition that regular training is important in preparation for disasters, the City of Yokohama implements training in emergency activities throughout the year to both improve our abilities to deal with disasters and to promote cooperation with local residents.

Training programs implemented on September 1 (known as Disaster Prevention Day) include a comprehensive disaster preparedness training event sponsored by the City of Yokohama and a disaster preparedness training program sponsored jointly by Tokyo and six other prefectures and cities. On January 17 (known as Disaster Prevention and Volunteers Day), citywide training in responding to a disaster is implemented in order to improve our ability in this area. Training programs independently conducted by the Waterworks Bureau include joint training events held every three years with the City of Nagoya by mutually dispatching support teams and training programs in handling emergency water supplies using distribution reservoirs and emergency-use underground storage tanks with the participation of residents. Through these training programs with resident participation, we are establishing a framework for emergency water supply activities that are conducted with the cooperation of local residents. We show the locations of emergency water supplies and explain how to operate emergency water supply taps.

5. Conclusion

In pursuing the goal of “Creating a city that can withstand disasters and achieving a comfortable life for its citizens,” which is the foundation of municipal administration in the City of Yokohama, the Yokohama Waterworks Bureau is promoting the creation of water services that can continue to serve its residents even at the time of an earthquake. Now that installation of most of the facilities has been completed, we intend to promote training on a regular basis and encourage “cooperation with the residents of Yokohama” as a key measure to promptly and effectively conduct emergency restoration and water supply activities after the occurrence of a disaster. S1-3

“Considerations for Resident Participation in Seismic Hazard Mitigation Measures to Secure Water Supply”

Presenter: Masahiro Kimura

(Osaka Prefectural Waterworks, Japan)

Considerations for Resident Participation in Seismic Hazard Mitigation Measures to Secure Water Supply

Masahiro Kimura Executive Director of Osaka Prefectural Waterworks (President of Osaka Prefectural Waterworks Service Corporation)

1. INTRODUCTION

Almost ten years have passed since the 1995 Hanshin-Awaji (Kobe) Earthquake (herein after represented as “the Kobe Earthquake”), and people’s awareness concerning seismic hazards appear to be fading. However, according to the latest government study, a Tohnankai area Earthquake, which is projected to be a massive ocean epicenter earthquake (Magnitude 8 in class), will occur within 30 years. This projection was upgraded from a 50% probability to a 60% probability. In addition, the Uemachi fault zone, which is expected to suffer the most serious damage in the Osaka area, has been classified as a zone with a high probability of earthquake activity. Therefore, in Osaka, much more strengthening of seismic hazard mitigation measures has become necessary.

Now, the Osaka Prefectural Waterworks (herein after represented as “OPW”) which supplies about 600 million m3 of water annually to the municipal waterworks in Osaka has so far implemented various seismic hazard mitigation measures, such as earthquake-proofing of water pipelines and main structures, installation of emergency water supply facilities, and signing of interagency mutual aid agreements, etc. Beginning in fiscal year 2005, OPW will start implementing permanent seismic hazard mitigation measures including the systematic upgrading of aging facilities.

In order to minimize the effects of earthquake damage, collaboration of local governments, not only with waterworks, but also with local residents and businesses is crucial. This paper, for the purpose of minimizing damages caused by water outage after an earthquake, introduces effective use of "water service specialist volunteers", in order to establish the collaboration between local residents and waterworks, and considers policy priorities concerning securing emergency water supplys in the state of confusion shortly after an earthquake occurrence.

2. THE PAST SEISMIC HAZARD MIIIGATION MEASURES TAKEN BY OSAKA PREFECTURAL WATERWORKS (OPW)

In the 1996 fiscal year, OPW established “the Basic Guidelines of Seismic Hazard Mitigation Measures in Osaka Prefecture” (titled the “Osaka Safe Water Supply Plan”) in alliance with municipal waterworks in Osaka Prefecture.

So far, several plan measures have been implemented, based on the seismic-proof evaluation (from 1996 to 1997) of water supply facilities. These measures included signing of mutual aid agreements, improving water supply bases and making emergency upgrades on 52 facilities (11 building constructions, 14 civil engineering structures, and 27 water pipe bridges). These measures were made so that the minimum functions of the water supply facilities could still be carried out at the time of an earthquake disaster. Moreover in the 2003 fiscal year, OPW packaged and distributed one million units of bottled water (500ml aluminum bottles with expiration date of 5 years; Fig. 1.) as disaster stockpiles for vulnerable groups, and, in the 2004 fiscal year, OPW launched the support and assistant system (Water Service Assistant System) which collects information and assists water supply at the time of an

1 earthquake disaster. So far OPW has been taking temporary measures from both “soft” and “hard” perspectives.

3. FUTURE PERMANENT SEISMIC HARZARD MITIGATION MEASURES OF WATER SUPPLY FACILITIES

In March 2003, OPW introduced "the Osaka Waterworks Future Design" (Water Way 21) which encompassed directions of water supply services. In the 2004 fiscal year, OPW began completing a long-term facility improvements master plan based on the "basic principles concerning facility improvements" specified in this design. In this master plan, facilities requiring improvements by 2030 are identified, assuming that the useful life of these facilities is generally 50 years. Moreover, in this plan, permanent seismic hazard mitigation measures from the perspective of ‘hard’ measures, such as reinforcement of the Murano Water Treatment Plant to increase its resistance to earthquake damage, the division of water distribution systems, the improvement of Fig. 1. Emergency Bottled waters of Osaka disaster prevention facilities and high volume Prefectural Waterworks (OPW) bypass water pipes with the addition of earthquake-proof water tanks and the mutual connection of water treatment plants, are incorporated as shown in Fig. 2. The estimated costs to implement these measures over the next 25 years are 540 billion yen or approx. 5 billion in US dollars, and the concrete implementation plans are currently being discussed in order to begin their implementation in the 2005 fiscal year.

4. AIMING FOR FURTHER IMPROVEMENTS OF SEISMIC HAZARD MITIGATION MEASURES

(1) Water Supply at the Time of an Earthquake Disaster Even when an earthquake, as a result of the Uemachi fault zone, occurs, the goal of "the Basic Guidelines of Seismic Measures in the Osaka Prefecture" is to restore all water supply services in the Osaka Prefecture within 28 days, and to greatly reduce the number of people who would, without these measures, experience an outage of water supply. If all of the policies in the guideline are met, a restoration period of water supply will be sharply shortened from the present projected 67 days, and the rate of water-supply outage in Osaka Prefecture(except for Osaka City) just after an earthquake disaster will decrease from 75% to 50%, as shown in Fig. 3. However, the initial number of people who suffer damages due to an outage of water supply will still remain at 3 million. These collateral damages caused from water outage will continue until the restoration is completed. Though the minimum needs for household water are to be served by water service support or temporary water hydrants to these people, to supply water at full-scale is difficult until the emergency water supply system of waterworks becomes functional. In order to reduce this gap, the immediate restoration of local government functions is necessary. However, until then local residents and local businesses need to organize plan and prepare to provide a secure source of emergency water for themselves. Furthermore, even if the emergency water supply system becomes functional, in order to supply water smoothly in the confusion inherent in the aftermath of a massive seismic disaster, resources in addition to local government and waterworks are necessary.

Japan

Aigawa Water Treatment Plant

Isojima Intake Plant ◇Emergency Power Generating Facilities Osaka Murano Water

Mishima Water Treatment Plant

◇ Treatment Plant Updating of planar System Connection Water ◇ Emergency Power Facilities etc. Distribution Pipe (new) Generatin g Facilities etc. φ1,200 ～1,000mm Niwakubo Water Wide Area Reservoirs of Treatment Plant Higashiosaka District ◇ Updating of Intake Facilities etc.

Disaster Prevention Facilities Hiraoka Pump Station

○ Target Year : 2030 (Heisei 42) ○ Estimated Project Cost : Fujiidera Pump Statio n About 540 billion yen or approx. 5 billion in US

dollars Booster Pump Station (new) Bypass Water Distribution Pi pe (new) φ2,400～2,000mm

Senboku Reservoir

Izumisano Pump Station Water Treatment Facilities Intake/ Booster Facilities Sennan Reservoir Wide Area Reservoirs Conduit/ Connection/ Distribution Pipes

Highest Priority Improvements Improvements to be made over the 25-year after Kinokawa Water Treatment Plant Expansion Projects

Fig. 2. Fundamental Long-Term Improvement Plan of Osaka Prefectural

Waterworks (OPW) (proposed)

) 80 70

60 50

30 20

10 0 start of day 7 day 1 4 day 2 1 day 2 8

Rate of Water-Supply Outage (% restoration

pres ent s tate

after emergency restoration begins after permanent measures taken

Fig. 3. Change in Outage Rate of Water Supply in Osaka Prefecture after an

Earthquake (caused by Uemachi fault zone; not including Osaka City)

In order to improve seismic hazard mitigation measures further and to supply emergency water to people who need water most in a timely manner, the collaboration of local residents, business enterprises and volunteers, as well as fast and appropriate responses by local governments and waterworks have to be part of the hazard mitigation plan.

(2) Establishment of the "Triage Principle" at the Water Supply Service after an Earthquake Disaster When Niigata Prefecture suffered extremely heavy rain in July, 2004, 9 people fell victims of flooding from the Shinano River tributary, and 6 of them were 70 years old or older. Although the city had a safety measures manual with plans to prioritize the rescue of the most vulnerable groups during disasters, it was not effective.

In case of the Kobe Earthquake, although seismic hazard mitigation manuals were included in the local disaster prevention plan, the shear enormity of the earthquake overwhelmed the first responders and the initial response by the local administration was not carried out smoothly1). Two primary factors were identified as reasons for the ineffectiveness. Firstly, the local government employees were not familiar with the emergency manuals. Secondly, personnel in charge panicked not knowing where to start since the disaster was much more than anticipated. In the initial response right after an earthquake disaster, organizing of the personnel takes time and can be stalled as an overwhelming amount of damage information reports and inquiries from panicked residents flow in.

To prepare first responders better for these types of circumstances, disaster safety manuals to be implemented right after earthquake disasters are currently being created to be a part of the disaster prevention plans. However, since a disaster could develop into unforeseeable situations, it is next to impossible to have all the perfect disaster preventions and responses, even if manuals are improved.

4 Moreover, because detailed manuals anticipating all cases are not practical to use at the time of disasters, manuals should be as simple as possible. Therefore, identifying what immediate action the waterworks department should take right after an earthquake disaster is crucial.

"Triage" of medical treatments in disasters is "the evaluation and prioritization of critical need, so as to identify and treat the most serious injuries first” The International Committee of the Red Cross sets priorities of medical treatment as (1) life, (2) preservation of limbs, (3) preservation of functions, and (4) cosmetics. Immediate measures which the waterworks personnel should implement right after earthquake disasters should be limited to the most efficient and effective ones based on the best evaluation of the system conditions at the time. For example abandoning attempts to salvage or repair systems that have failed and concentrating our first efforts to preserve systems that are damaged but still operable.

①Priorities of measures carried out by waterworks immediately after earthquake disasters Measures against disasters are classified into the following 4 phases from the viewpoint of the purpose and urgency2).

● Phase 0: Establishment of organization for initial response ● Phase 1: Urgent measures ( preservation of life) ● Phase 2: Emergency measures ( Triage - preservation of any functional portions of the system) ● Phase 3: Rehabilitation and reconstruction measures to restore full function of the system

Phase 0 was newly classified, based on the experience of the Kobe Earthquake, where (a) personnel was not able to arrive at disaster headquarters, (b) space and materials in order for the headquarters to function were not secured, (c) personnel was overwhelmed with telephone inquiries by the residents and media. In the phase 1, measures taken are to be concerning the preservation of life, such as disaster medical treatments, fire extinguishing, prevention of secondary disasters (i.e. epidemics). In the phase 2, measures taken include supplying food and water, establishing shelters, distributing relief goods, distributing and providing emergency management information, and restoration of a lifeline. These phases are applied to seismic hazard mitigation measures of the water supply service as follows (TABLE 1);

● Phase 0: Establishment of organization for initial response 、 gathering of personnel, information gathering and establishment of countermeasure headquarters (emergency water supply is to be provided by using stockpiles created and maintained though the planning and cooperation of local government, businesses, and residents) ● Phase 1: Urgent measures: Preservation of water for extinguishing fires, prioritizing water supply to medical facilities, restriction on drinking water, etc. ● Phase 2: Emergency measures: water supply to shelters, restoration of water supply facilities

After the Kobe Earthquake, full-scale emergency water service began 12 to 24 hours after the earthquake, and restoration began 5 days after in Kobe and Ashiya Cities, and 7 days after in Nishinomiya and Takarazuka Cities. In order to carry out the effective emergency, communication concerning the establishment of organization, information collection and assignment of staff and materials is essential.

Waterworks personnel needs to focus, for a definite period immediately after the occurrence of a powerful earthquake, on the preparation of full-scale evaluation, planning , and communication (about 1 day), and after that, to start actual water supply restoration services, where residents are informed of this enough in advance. In the preparation period, while collecting information and drawing up a water

5 supply/restoration plan, they should limit their activities to information distribution to residents and the press, preservation of water for extinguishing fires, water supply to medical facilities directly related to the protection of life.

In the Kobe Earthquake, shortage of lavatory water was also a serious issue. Taking into consideration that some degradation of the sewerage treatment system is inevitable, portable lavatories should be available (for one to three days for mass distribution thus lowering demand on water supplies while helping to reduce the health risks associated with exposure to raw sewage.

TABLE 1. Priorities of Water Service at Massive Earthquake Disasters

Phase ０１２３

Days elapsed Day １ Day ２～３ Day ４～７ Day ８～28 drinking, cooking, lavatory washing. bathing Purpose of use drinking water drinking water water at shelters, houses Amount of use ３L/person・day ３L/person・day ３～20 L/person・day 20～250 L/person・day Emergency water Supply Conveyance water supply, Conveyance water supply, Water for securing measures Individual stockpile equipments, Temporary water tap Water tap life Conveyance water supply Shelters, Water supply bases, Shelters, Welfare facilities, Shelters, Elderly person living alone, Supply place Individual Elderly person living alone, Sites ( 250ｍ～10m apart Disabled person, Disabled person from living places) Upper-layers story residents Individual Fire-extinguishing (3.3m 3/one affair) Fire companies Fire companies Fire companies actor Resident aid ( 300m 3/one affair ) 300m 3/one affair 300m 3/one affair (Water consumption) (20m 3/one affair) Water for Buckets, extinguishing fires Instruments used Fire extinguishers, Pumps, Hoses, etc. Pumps, Hoses, etc. Pumps, Hoses, etc. Hoses, etc. Fire prevention tanks, Bath water, Hydrants, Ponds, Rivers, Hydrants, Ponds, Rivers, Water used Hydrants, Fire prevention tanks, etc. Sea water Sea water Ponds, Rivers, Pools, etc.

Well for emergencies, Well for emergencies, Well for emergencies, Water used Stockpile water, Stockpile water, Tap water Tap water Conveyance water Conveyance water Water for Water supply Water supply Water supply method Conveyance water supply Conveyance water supply medical use by temporary pipes by temporary pipes Required amount of water, Amount of water 50L /one operation such as washing of medical Inpatients’ life water etc. Water service restoration supply 150L/one dialysis equipments Back up of water service by Water tap, Water for Stockpile water for related companies, Water supply Water used Conveyance water supply by business use emergencies Conveyance water supply by by temporary pipes employees etc. employees etc.

② Preservation of water for extinguishing fires Because the Kobe Earthquake occurred in predawn hours and wind was gentle, the spread of fires was limited, and the death caused fire was about 10% of total fatalities (559 people). According to a study by Professor Takada of Kobe University, in the Kobe Earthquake, the spread of fires would have been suppressed and the number of fatalities would have been halved if the fire hydrants were functioning properly. Therefore, the role of water supply is significant.

Also, according to the report from the Japan Water Research Center (January, 2001)3), fire was extinguished by fire-engine vehicles and residents immediately after the Kobe Earthquake occurred and half of the 285 fires(146 cases) were put out only by residents and about 40% of those were extinguished. Residents used an extinguishing agent in 81 cases and water in 40 cases. Water is generally used for initial fire extinguishing by residents at 70 % of the time. Therefore, water is considered an excellent extinguishing agent having both high fire-extinguishing capability and low

6 toxicity.

Since fire fighting power is reduced after earthquake disasters because of the simultaneous occurrence of fires and the street debris that delays fire department response. Initial fire extinguishing by local residents has been found to be a crucial component reducing the spread of fires that occur immediately after a large earthquake.

The Fire Defense Agency notes that the response time of fire extinguishing activities by individuals after fire outbreak is 3 minutes and the response time of fire extinguishing with support by neighbors is 5 minutes. Water for fire fighting in homes should be secured with top priority after earthquakes. Although the disaster prevention organization of local residents needs to be in charge of fire extinguishing before arrivals of fire engines, installation of usually available hydrants and preservation of water are essential. According to the research of the Kansai Waterworks Research Committee4), the upgrading of water pipes to earthquake-proof pipes and flexible mutual utilization of water supply systems improve the dependability of hydrants substantially. However, in the case of earthquake disasters, fractures of water pipes and leakage of water may be unavoidable. Therefore, the increasing number of waterworks are adopting the automatic water emergency shutoff valves to secure water in reservoirs. For this reason, in order to secure water for extinguishing fires right after earthquake disasters, it is important to promote installation of fire prevention water tanks combined with rainwater storage, and bathwater storage at each home preparing for water-supply outage. However, waterworks need to advance the practical use of earthquake-proof water tanks and re-installation of hydrants onto high volume, large caliber main water pipes with comparatively high earthquake-proof capability, as well as improvements of mutual connected pipes and private electric generators. Furthermore, through damage prediction of pipelines and the introduction of water line status information systems, even if there are water leakages from pipelines, it is necessary to consider continuation of water supply, to provide a minimum emergency water supply to reduce the risk of secondary disasters.

③ Water supply to medical institutions An exhausted surgeon working without water after the Kobe Earthquake said “the hospital shut off from the lifeline was unable to fulfill its fundamental function.” He said with great regret5).

Damages to the lifelines will enormously impact medical practices, and may cause “preventable death”. According to the questionnaire about the Kobe Earthquake conducted by Kanazawa University6), water could not be obtained at all in 60% or more of regional hospitals. The lack of water had a serious impact on sterilization of medical equipment and inability of doctors and nurses to wash their hands, as well as surgery and dialysis therapies Since an outage of water supply due to earthquake disasters will seriously affect the survival of not only patients in hospitals but earthquake victims, waterworks should consider water supply to medical institutions secondary only to fire fighting measures.

After the Kobe Earthquake, many hospitals themselves are striving for duplication of water resources, such as the installation of emergency water tanks and utilization of ground water. While we support the development of emergency water supply systems by medical institutions, these measures will remain back up systems. Therefore, it is necessary for waterworks to be prepared to manage water supply right after earthquake disasters through prioritization and by effective management of water reservoirs and switching of water distribution systems, while implementing earthquake-proof upgrading and duplication of pipelines so that the water supply to medical institutions can be more secure during and after a large-scale earthquake.

Although sales of private water supply systems using ground water to commercial users is intensifying recently, ground water is a finite resource and leaving pumping of ground water uncontrolled will lead

7 to the eventual exhaustion of water resources at the time of disasters. Therefore, sustainable utilization of ground water pumping systems for continuous use as a opposed to emergency use must be researched so that recommendations protecting this resource can be implemented.

(3) Improvement of Local Disaster Prevention Capability through Residents Participation According the report of Japan Water Research Center (January, 2001)3), it is estimated that 6,000 people lost their lives to the Kobe Earthquake. 90 percent died the day of the earthquake and 80 percent of these died from suffocation within collapsed structures while another 10 percent were lost to fire. Of the many thousands who survived the quake, 90 percent were rescued from collapsed structures by families and neighbors. First responders could only reach and rescue about 3 percent of those who were trapped and in need of help in order to survive. These findings reinforce the practical notion that the role which local residents can play, such as fire fighting and rescue efforts in the crucial time period immediately after an earthquake disaster.

In the traditional Japanese society, there have always been voluntary disaster prevention systems organized by neighborhood associations and resident associations, and those played big roles in local disaster preventions and or response until now. It has now been observed that in recent decades societal changes in Japan, such as the general aging of the population and lifestyle changes, that mean fewer long-term ties to neighborhoods have greatly reduced the participation in neighborhood and community organizations that traditionally would have been the source of trained volunteers responsible for emergency prevention and emergency response. The White Paper on the Fire and Disaster Management (2003) mentions that from 1956 to 2003, while the number of the fire-fighting personnel across the county increased by 160,000 from 32,000, the number of local volunteer fire fighters declined was reduced by half from 1,800,000 to 930,000.

After the Kobe Earthquake, the emergency response manual of seismic measures were created by each prefecture based on a local disaster prevention plan. Our regional plans must evolve to accommodate the natural phenomena reflecting regional characteristics, and ever changing conditions of human society as well. Because there is a limit in disaster relief by local governments, it is crucial for local residents to be prepared for natural disasters and help themselves and their communities when disasters occur.

①Efforts and direction for improvement of local disaster prevention capability The research report by the Central Disaster Prevention Council (July, 2002)８) about the future direction of seismic hazard and mitigation measures states that the changes in the stationary nature of our conventional communities, in urban as well as rural areas, is beginning to cause concern within the traditional local disaster prevention organizations for their disaster prevention capabilities. As a result, training of voluntary fire fighters and voluntary disaster prevention organizations and the formation of new communities adapted to the current circumstances are vital. Therefore, the formation of the “Disaster Prevention Collaboration Society” is needed, where various participants must include not only administration but residents, businesses and NPO’s to cooperate in disaster measures that their own communities have identified. The voluntary disaster prevention organization is based on identifying, enrolling, training, and engaging the cooperation of local residents as they are made up today. This means all able bodied persons whom are located in specified zones during the work day and alternately those whom are present in the off hours. The Kobe Earthquake proved the necessity of the Disaster Prevention Collaboration Society. The Disaster Prevention Collaboration Society will be the source of emergency training, education of disaster prevention information, disaster prevention inspections, group purchase of materials. The goal is that at the time of disaster, the volunteers of the Collaboration Society will be prepared to perform initial fire extinguishing, emergency evacuation of residents, rescue and relief of injured, collection and transfer of information, supply of food and water

8 and inspection of damaged structures. According to the White Paper on Fire and Disaster Management (2003), the rate of organization of such voluntary disaster prevention groups is 61.3% of 3,213 municipalities across the country in April, 2003, while the rate was 43.8% prior to before the Kobe Earthquake. The maximum rate is 98.4% in Shizuoka Prefecture, while the rate in Osaka is 58.3%.

The Fire and Disaster Management Agency (FDMA) is advancing the awareness of the public to the need to further develop community based Disaster Prevention Collaboration Societies through a television campaign. The FDMA is also promoting the creation and maintenance of disaster prevention activity bases to serve the developing voluntary disaster prevention organizations. However, at this time, there are many inactive and titular organizations. Activation beyond systematization is required. The followings are specified as methods to activate these organizations9).

● Training of people who can actively participate in earthquake disaster prevention measures ● Developing basic measures to combine crime prevention and welfare activities ● The creation of disaster prevention organizations as described above that are made up of a consortium of those able bodied persons present in neighborhoods today such as school employees, local company employees, merchants etc.

As examples of this important trend, in Hyogo Prefecture, trainings of leaders for new voluntary disaster prevention organizations, the holding of activity promotion conventions, support programs creating safe community files, etc. are already being performed. In addition, reviving some of the valuable original activities of disaster prevention organizations in the prefecture. These activities include the refuge experience meetings where local residents stay at the gymnasium of elementary schools, the resident registration system of skilled persons residing in apartment communities, the cooperative disaster prevention studies by residents, companies and junior high school students and activities aiming at the upsurge of disaster prevention technology and knowledge where NPO’s and ladies’ in the society practice water bucket brigades, wheel barrow competitions and a disaster prevention quiz all as new activates in neighborhood athletic festivals10). In Kyoto City, the disaster prevention chart program using personal computers to be set up as early warning systems during disasters, and as public notice boards listing the necessity of disaster measures and to motivate disaster prevention actions11). In Shizuoka Prefecture, the local disaster prevention coach system has been recently established and the introduction of “the Disaster Imagination Game (DIG)” has been implemented 12).

②Measures of waterworks towards earthquake disaster prevention After the Kobe Earthquake, each waterworks is developing various measures, such as the duplication and reinforcement of facilities resistant to earthquakes, the improvement of back-up facilities, etc., as “hard” measures, and the emergency restoration and emergency water supply planning, etc., as “soft” measures. Concerning the implementation of some of the mechanical aspects of the soft system improvements for the emergency water supply needed right after a large earthquake, some progress has been made. Examples of the progress include earthquake-proof water tanks, emergency wells and emergency hydrants installed in the range of about every 1km around elementary school area, and distributions of potable water for emergencies.

With regard to advancements in developing the all important human element of the soft measures aspect of our agenda, A few very exciting example are the emergency water supply training with citizens in Yokohama City, which features the establishment of the support assistant member system at the time of disaster by waterworks retired employees and implementation of the "Aisui Volunteer" (April, 2003) in Aichi Prefecture. In the 2004 fiscal year, OPW has launched a support assistant system (Waterworks Assistant System) at the time of disaster, and is soliciting applicants with waterworks

9 experience who perform information gathering and reporting when massive earthquakes or water leakage accidents occur and assist with water supply activities at the recently installed waterworks water supply bases.

③ Measures of waterworks towards promotion of local involvement in local disaster prevention organizations Since collaboration with local residents and volunteers is indispensable to mitigate damages of earthquake disasters, waterworks must identify and promote ideas that support the goal of re-creating local organizations of first responders to disasters as follows; (a) Establishment of volunteer’ organizations of retiring water service specialists One of the recent evolutions in our society is the growth of work-related association and a corresponding diminishment of traditional neighborhood and family association. Many waterworks staff are and will be retiring soon. It is anticipated that waterworks retirees will return to local communities. This is a favorable opportunity as these individuals can be recruited into the new volunteer organizations we are promoting within our local communities. Water service specialist volunteers in local communities can be used as leaders or advisers concerning water related issues. (b) Implementation of everyday activities in community organizations dedicated to emergency water supply In order that local disaster prevention organizations are truly effective at the time of disasters, regular involvement and frequent practice are required. At present, although activities of volunteers of waterworks are restricted to the field of earthquake disasters, it has become crucial to promote the development of everyday activities in the local community, as well as conducting trainings in the use hydrants and earthquake-proof tanks with fire fighting and disaster prevention organizations. The following activities can be considered as important services and activities that will strengthen to local volunteer organizations; ・ Reporting sites which receive complaints and requests daily about water supply service in communities ・ Reporting of water leakages and accidental water contaminations of water resource rivers to waterworks bureaus ・ Performance of public relations and educational activities to local residents, such as the proper location and storage of emergency water stockpiles, storage of bath water, promotion of rain water storage and water sprinkling, etc. ・ Performance of visiting services cooperated with local welfare organizations, such as regular distributions/ exchanges of the emergency water stockpiles, adequate services of water supply installations, etc. to vulnerable groups, such as elderly and disabled peoples who have troubles with securing water by themselves in disasters Since such activities will complement waterworks businesses, on the occasion of achievements, it is desirable to establish new organizations of NPO’s as water service specialist volunteers’ groups and to carry out a part of those businesses as outsourcing from waterworks bureaus. (c) Cooperation with other waterworks Because the use of many public volunteers is necessary to aid water service after earthquake disasters, roles as coordinators will most often fall to water service specialist volunteers. It is desirable for waterworks to become resource support sites for water service specialist groups and to assist in the establishment of special organizations for coping with local issues, as well as for educating and developing them. In the educational training program, it is necessary to incorporate the sharing of the accumulated knowledge of post systemic conditions and challenges, as well as information on how to best organize to cope with other disasters. (d) Cooperation with other local organizations Many local disaster prevention organizations, resident associations and water supply bases have selected every elementary school area as their activities and maintenance unit. Since education may

10 play a big role in disaster measures, it is considered very effective for elementary schools to make maps of preventing disasters and securing of their school area and to carry out “the Disaster Imagination Game” as a part of their lessons, as shown in Fig. 4.

When making such maps, investigations on the stock status of portable water in convenience stores in school areas, the amount of water in local pools and reservoirs, the location of hydrants and wells, etc. are included, culminating in a complete assessment of available water resources that will be known to the local disaster volunteer members. By incorporating this program into the curriculum of upper elementary grade classes, every year the updating renewal of local water resource data is attained with the continuous promotion of disaster prevention education. These school based organizations cooperating with local welfare organizations and companies, local disaster prevention organizations will add new data and information, such as distribution maps showing the location of the most vulnerable groups such as elderly residents, and then use to carry out map exercises and simulations. In local communities, water service specialist volunteers will participate in such map making activities and trainings, and learn actions required at the time of disasters for themselves, while promoting home disaster prevention preparedness, the location of hydrants, the quantity of stockpiles of emergency water, the dissemination of this information at the time of disasters, etc. Furthermore, interacting on an ongoing basis with waterworks staff helps to resolve general ongoing issues with water services as they work to improve their disaster response preparedness.

Administration and Disaster Board of Education Prevention Section in charge Waterworks

Damage Anticipation Local Disaster Prevention Plan

Elementary School Local Prevention・Safety Map Upper Classes Schools, Escape routes for Evacuees, Port of Distress, Gardens, Hospitals, Police Boxes, Ponds, Rivers, Map Exercise of Fire Hydrants, Convenience Stores, Wells, Disaster Imagination Game Earthquake-proof Tanks, Dangerous Sites, High Voltage Cable (DIG)

Social Welfare Councils, Local Water Service Disaster Prevention Specialist Organizations, Volunteers Residents’ Associations, Local Prevention・Safety Map 2 Neighborhood Associations Elderly person living alone, Disabled person living home, Gas Stations, Companies, Offices, Aging Buildings, Medical Facilities, Dangerous Map Exercise and Care needed Facilities and Real Training

District Disaster Prevention Manuals

Volunteers Groups District Disaster Prevention Plan

Fig. 4. Concept of Mapping Exercise for Improvement of Local Disaster Prevention Capability

11 ④Toward the Integration of water supply systems In Japan, while the slowdown of the growth in new water demand has been forecasted, the projected expense of updating many facilities which were constructed amid the high-growth period of the Japanese economy is increasing, The cost of these improvements will be a significant challenge for many water works in the near future. Furthermore, an increasing number of water works engineers are reaching retirement age. Thus, there is a concern about the loss of knowledge and experience concerning the techniques of distributing safe water .

The “Vision of Water Supply” of the Ministry of Health, Labor and Welfare points out the necessity of the promotion of the integration of the water various supply systems. Also in Osaka Prefecture, while 44 waterworks supply water by themselves, the integration through cooperation of the separate water supply systems is considered unavoidable, in order to achieve efficiency and advancement of waterworks and to advance seismic measures to any meaningful degree. On the other hand, the integration of the water supply system may cause a sense of disconnect between the residents and what were before locally run and controlled waterworks. This may adversely affect the communication and cooperation with local resident volunteer organizations we are proposing. This makes the creation of the water service specialist volunteers system taking advantage of waterworks retirees even more important. These water works retirees, while blending in local communities through everyday activities, are expected to become a bridge between local residents and waterworks through the participation in trainings and information by waterworks, hopefully offsetting any disconnect with the water works during the system wide integration process.

5. CONCLUSION

Japan is experiencing a significant aging of its population. We know that at some point, massive earthquakes, floods, and unforeseen disasters will strike. In the past, a collateral damage due to water supply outage caused by these disasters has also occurred magnifying the lethality of the disaster. Even if Japanese waterworks have continued challenges aiming at the first runner in the world in the future as mentioned in the “Vision of Water Supply”, waterworks planners are committed to identifying and answering the many challenges we face now and in the future such as transferring the knowledge that only experience can provide to a new generation of engineers, the seismic hardening of installations and the replacement of aging facilities to name a few. All of these, and similar measures are limited by revenue availability. In order then to reach past these revenue limitations and to improve our emergency response, we will expend a great effort to assist the re-development of a grass roots movement of locally based (waterworks trained) and organized neighborhood first responders. Hence, it is required to improve practical and operational emergency response manuals as well as to introduce an accurate damage assessment and information transmission system. Because it is difficult for a public organization including waterworks to perform direct and sufficient correspondence right after earthquakes, residents’ self-help and their mutual aids are still more indispensable.

From the above viewpoints, this paper has considered the thought of the initial correspondence of waterworks at the time of earthquake disasters aiming to level up seismic measures and the application of water service specialist volunteers for the improvement of the local disaster prevention capability. After carefully examining the many studies of the events during, and immediately after the Kobe Earthquake we have determined that our planning for the hardening, duplication and integration of our water supply infrastructure is validated. The point of this presentation is that it has also become indisputably clear that in terms of the initial survival of the population in the affected area nothing we could do will have a more positive outcome than to vigorously support the re-development of community based emergency response organizations. In short, we need treat both goals, i.e. improvement of “hard” and “soft” measures, as equal in importance for our future.

[REFERENCES]

1) Junichi Dodo: The Damage of Waterworks in Hyogo Prefecture and Lessons of Emergency Management in the Great Hanshin-Awaji Earthquake, Water Supply, Vol.18, No.3, pp.112-117 (2000). 2) Disaster Prevention Research Institute of Kyoto University: The Approach to Disaster Prevention Research, (January, 1996). 3) Japan Water Research Center: Fire Disaster and Water Supply (January, 2001). 4) Kansai Waterworks Research Committee: Seismic Measures against the 1995 Hanshin-Awaji (Kobe ) Great Earthquake, Suido Kouron (June, 2002). 5) Shinko Hospital: the Record of Correspondences of Shinko Hospital against the 1995 Hanshin-Awaji (Kobe) Great Earthquake (July, 1995). 6) Masakatsu Miyajima et al.: Earthquake–proof Evaluation of Water Supply System of Medical Facilities, Journal of Japanese Waterworks Association, Vol.72, No.11, pp-95-99 (2003). 7) Kazuyuki Nakagawa: Improvement of Disaster Prevention by Cooperation Workshop, Monthly Journal of the Fire Protection Equipment and Safety Center of Japan (March, 2004). 8) Central Disaster Prevention Council: The Research Report on the Future Scheme of Seismic Measure (July, 2002). 9) Quarterly Journal of the Institute for Fire Safety and Disaster Preparedness, Spring edition 2, 5, 6 (2003). 10) Quarterly Journal of the Institute for Fire Safety and Disaster Preparedness, Summer edition, 5 (2003). 11) Quarterly Journal of the Institute for Fire Safety and Disaster Preparedness, Spring edition, 5 (2004). 12) Manual of independent Disaster Prevention of Shizuoka Prefecture, (March, 2002).

13 Session 2: Risk Assessment

S2-1 “Verification of Effectiveness of Pipeline Renewal Based on Damage Analysis on Recent Earthquake” Presenter: Toshio Toshima (Japan Ductile Iron Pipe Association) Japan

S2-1

“Verification of Effectiveness of Pipeline Renewal Based on Damage Analysis on Recent Earthquake”

Presenter: Toshio Toshima

(Japan Ductile Iron Pipe Association, Japan) The 4th US-Japan Workshop on Seismic Measures for Water Supply, January 2005

Verification of Effectiveness of Pipeline Renewal Based on Damage Analysis on Recent Earthquake

Toshio Toshima, Shogo Kaneko

Introduction

The diffusion of drinking water supply service in Japan reached 96.7% in 2003. Today, the majority of construction works on water pipelines involve the renewal of long-used aged pipes, such as the asbestos-cement pipes and cast iron pipes, rather than the installation of new pipelines. Since a large number of pipelines that were installed during the period of rapid economic growth in 1960s and 70s to meet the sudden rise in the demand for water then are presently in need of renewal, the renewal of aged pipelines is one of the most important issues today.

The aged pipes are not only likely to suffer serious damages in the event of an earthquake, but are also susceptible to leaks and breakages even under normal circumstances and can seriously affect hydraulics and water quality. For this reason, the improvement of earthquake resistance is advanced not only as a measure against earthquakes but also as a part of overall functional improvement of water pipelines.

Because of the increasing need to reduce cost and give adequate explanation to the affected residents in recent years, it is particularly important today to clearly set forth the expected effectiveness of pipeline renewal from the planning stage. The purpose of this paper is to examine the effectiveness of pipeline renewal works that were carried out after some of the recent earthquakes. The effectiveness, particularly in terms of alleviating the damages by earthquakes, is evaluated by comparing and analyzing the damages on pipelines caused by the Tokachi-oki Earthquake of September 2003 and those by other previous earthquakes.

Toshio Toshima & Shogo Kaneko, Member of Technical Committee of Japan Ductile Iron Pipe Association, 8-9, 4-Chome, Kudan Minami Chiyoda-ku, Tokyo 102-0074, Japan

- 1 - Outline of Pipeline Damages Caused by the 2003 Tokachi-oki Earthquake

Fig. 1 shows the outlines of the Tokachi-oki Earthquake (Magnitude 8.0) that occurred on September 26, 2003 and the distribution of seismic intensities reported by the Japan Meteorological Agency. A seismic intensity of 6-lower was recorded in the towns along the coastline in eastern region of Hokkaido island, located at the northern part of Japan.

Hokkaido island

Asahikawa Kobe

● 4 Obihiro ● Sapporo Tokyo 5+ 5+ 4 - 6 Kushiro

6- Toyokoro Urakawa 5+ :Seismic Intensity 5-upper Hakodate 4 6- :Seismic Intensity 6-lower ×Epicenter

Date September 26,2003 4 Aomori Epicenter E144.1°N41.8° Magnitude 8.0 Fig.1 Seismic Intensity Distribution of the 2003 Tokachi-oki Earthquake

According to the statistics taken by the Hokkaido government, the water supply stopped in approximately 16,000 households in a total of 6 cities, 23 towns and 1 village due to serious damages on the facilities. While the supply was restored in most of the municipals by the end of September 26, or within the day of occurrence of the earthquake, it took several days to resume the supply in Toyokoro-town (refer to Fig.1, service population 3,700) and Urakawa-town (refer to Fig.1, service population 12,900). The stoppage was due to the damages on the pipelines in most cases.

Table 1 shows the total numbers of damages on water pipelines (transmission and distribution pipelines) in major cities and towns hit by a seismic intensity of 5 or higher. The pipelines were damaged in eastern part of Hokkaido, including the central city Kushiro-city (refer to Fig.1, service population 192,100), as well as in other cities and towns. The damages were particularly prevalent in Toyokoro-town and Urakawa-town. The “damage rate,” which represents the number of damages per 1 km of pipeline, is

- 2 - generally employed to assess the severity of damages on water pipelines. The damage rate was 0.24/km in Toyokoro-town and 0.13/km in Urakawa-town. Since the damage rates reported in the 1995 Kobe Earthquake (maximum seismic intensity 7) 2) were 0.31/km in Kobe and 0.72/km in Nishinomiya, the damage rates of Tokachi-oki Earthquake were high but not as high as those of the Kobe Earthquake. Finally, the damage rate in Kushiro-city was low at 0.01/km.

Table 1 Number of Damages on Water Pipelines Number of damages Pipeline Damage rate Ductile iron Polyvinyl Asbestos- length＊２ pipe Others＊１ Total Number of Damages chloride pipe cement pipe km (general joint) (km) Urakawa-town 5 9 2 6 22 176 0.13 Samani-town 1 0 001390.03 Mitsuishi-town 1 0 102850.02 Shizunai-town 0 3 0 0 3 172 0.02 Taiki-town 2 0 1 1 4 323 0.01 Honbetsu-town 1 0 001780.01 Ikeda-town 3 2 4 0 9 169 0.05 Urahoro-town 0 3 0 1 4 241 0.02 Toyokoro-town 21 23 5 9 58 244 0.24 Hiroo-town 5 0 005690.07 Shikaoi-town 0 1 0 0 1 125 0.01 Churui-village 1 0 001260.04 kushiro-city 3 0 0 5 8 844 0.01 kushiro-town 3 4 0 1 8 115 0.07 hamanaka-town 0 3 0 1 4 113 0.04 Slip-out and Slip-out of joint Slip-out of breakage of joint Damage mode Breakage of joint Breakage of pipe －－ pipe body body *1 Breakage of screw steel pipe joints and attachments *2 References:"2000 National Waterworks Statistics"by Japan Waterworks Association.

The pipelines suffered a number of damages in Urakawa-town in the 1982 Urakawa-oki Earthquake and in Kushiro-city in the 1993 Kushiro-oki Earthquake. The asbestos-cement and cast iron pipes, both of which fall short of having the required earthquake resistance, were renewed in Urakawa-town and Kushiro-city after the earthquakes. In the following sections, the damages on the pipelines caused by Tokachi-oki Earthquake and those by other previous earthquakes are compared and analyzed to examine the effectiveness of the pipeline renewal.

- 3 - Urakawa-town

Outline of Urakawa-town water supply system

Table 2 shows the outline of Urakawa-town water supply system, including the water supplied population and the length of water pipelines.

Table 2 Outline of Urakawa-town Water Supply System

Water supplied population 12,879 Diffusion rate 92.7 % Water pipeline length 147 km

Maximum water supply/day 5,620 m3 ● Urakawa-town

Pipeline Damages in the 2003 Tokachi-oki Earthquake (Seismic Intensity 6-lower)

A total of 22 damages on the pipelines were found in Urakawa-town after the earthquake, and the damaged places are plotted on the map in Fig. 2. The different symbols used for plotting indicate different types of pipes. “Others” include screw joint steel pipes and accessories such as air valves. The damages on ductile iron pipes were caused by slip-out of joints, while those on polyvinyl chloride pipes and asbestos-cement pipes were caused by rupture of pipe body itself as well as slip-out of joint. Photo 1 shows the damage on ductile iron pipe that occurred at the point indicated by the arrow in Fig. 2. The damages on ductile iron pipes were found in soft marl and fill-ups along the roads.

- 4 - ●

▲ ▲ ▲

★ ●● ▲ ★ Urakawa station ▲ ★ ▲

Urakawa-town office ★ ■■ ▲ ★ Urakawa seaport ▲

●…Ductile iron pipe ★ ▲…Polyvinyl chloride pipe ●● ▲ ■…Asbestos-cement pipe ★…Others Fig.2 Plotted Chart of Damaged Photo 1 Slip-out of Joint on φ 300 Water Pipelines in the 2003 Ductile Iron Pipe Tokachi-oki Earthquake

Pipeline Damages in the 1982 Urakawa-oki Earthquake3) (Seismic Intensity 6)

Fig. 3 shows the distribution of seismic intensities and profile of the 1982 Urakawa-oki Earthquake (Magnitude 7.3). Plotted on the map in Fig. 4 are the points where the pipelines were damaged.

In the 1982 Urakawa-oki Earthquake, there were 69 damages on water pipelines. In contrast, in the 2003 Tokachi-oki Earthquake, the water pipelines were damaged at 22 points, or approximately 1/3 the number of damages recorded in 1982. When the points of damages in Fig. 2 and Fig. 4 are compared, it is evident that the number of damages has decreased in the urban area.

- 5 -

Asahikawa

3 Sapporo Obihiro 4 4 3 Kushiro 6 Urakawa Hakodate 3 × Epicenter Date March 21, 1982 Epicenter E142.6°N42.1° Aomori 3 Magnitude 7.3 Fig.3 Seismic Intensity Distribution of the 1982 Urakawa-oki Earthquake

■ ▲ ■ ▲▲ ▼ ▼ ★★★ ■ Urakawa station ■ ■ Urakawa-town office ▲ ▼ ★ ■ ▲ ■ Urakawa seaport浦河港 ■■ ■■ ▼ ■ ▲ ■ ▼ ■ ■ ●…Ductile iron pipe ■■■■ ▲■ ▼…Cast iron pipe ▲■■■ ▲…Polyvinyl chloride pipe ■…Asbestos-cement pipe ★…Others

Fig.4 Plotted Chart of Damaged Water Pipelines in the 1982 Urakawa-oki Earthquake

Pipeline Renewal

In Urakawa-town, approximately 25 km of asbestos-cement and cast iron pipes (approximately 20% of total pipeline length), which are inferior in earthquake resistance, have been renewed mainly in the urban districts since the 1982 Urakawa-oki Earthquake. Fig. 5 shows the change in the lengths of different types of

- 6 - pipes between 1982 and today. The length of asbestos-cement pipes today is half that in 1982 and the renewal of cast iron pipes is completed.

160 (140km) (147km) 140 Ductile iron/cast 16.4% 21.8% iron pipes 120 Asbestos 100 cement pipes Polyvinyl chloride 80 pipes

60 73.5% 73.6% 40 Pipeline length(km)

20 0 1982 2002 Fig.5 Changes in Total Lengths of Different Types of Pipes (Urakawa-town)

Fig. 6 shows the layout of pipelines today. The thick lines indicate the pipelines that have been newly replaced by ductile iron pipes since 1982 and the plotted stars indicate the points where the pipelines were damaged in the 1982 Urakawa-oki Earthquake. From this map, one can see that the pipelines that were damaged in 1982 are now replaced with ductile iron pipes. While the difference in seismic motions needs to be considered to make a more precise evaluation, the renewal of the pipelines for improving their earthquake resistance is considered as one of the factors that contributed to the reduction of damages.

★ ★ ★ ★ ★ Urakawa station ★ ★ ★ ★ ★ ★ ★★ ★ ★ ★ ★ ★ ★ ★★★ ★ ★ ★ ★ ★★★★ ★★★ ★ ★★ Urakawa-town office ★ ★ ★ Urakawa seaport Ductile iron pipes (installed after 1982) ★…Points of pipeline damages Ductile iron pipes in 1982 Urakawa-oki Earthquake Asbestos-cement pipes Polyvinyl chloride pipes Fig. 6 Water pipelines Renewed since the 1982 Urakawa-oki Earthquake

- 7 - Kushiro -city

Outline of Kushiro-city water supply system

Table 3 is an outline of Kushiro-city water supply system including the water supplied population and the length of water pipelines.

Table 3 Outline of Kushiro-City Water Supply System

Water supplied population 192,100 Diffusion rate 100.0 % Water pipeline length 831 km ● Maximum water supply/day 73,375 m3 Kushiro-City

Pipeline Damages in the 2003 Tokachi-oki Earthquake (Seismic Intensity 5-upper)

A total of 7 damages were found on the water pipelines in Kushiro-city. The points of damages were as plotted on Fig. 7. The 3 points where the ductile iron pipes were damaged are indicated in the expanded map on the right. The damages were concentrated in the Midorigaoka district and all three of them were caused by slip-out of joint. Photo 2 shows the damage on ductile iron pipe that occurred at the point indicated by the arrow in Fig. 7, and Photo 3 shows the area where the damage as shown in photo 1 was found. The damages were made along a road with a 12-degree slope. A number of cracks occurred on the road surface and a level difference of 70 mm was found between a manhole and the ground surface. The trace of sand boil, which is an evidence of liquefaction, was also observed in the area.

- 8 -

Enlargement of A part R38

Nemuro main line R44 Ku ★ shiro sta Kushiro-nishi seaport tion ● ● ▼ ● ◎Kushiro city hall ●● ● Kushiro seaport ○A ◆ ●…Ductile iron pipe ▼…Cast iron pipe Harutori lake ◆…Steel pipe(SGP) ★…Others ★

Fig.7 Plotted Chart of Damaged Water Pipelines in the 2003 Tokachi-oki Earthquake

Photo 2 Slip-out of Joint on φ 150 Photo 3 Area where Damage shown Ductile Iron Pipe in Photo 2 was found

- 9 - Pipeline Damages in the 1993 Kushiro-oki Earthquake4) (Seismic Intensity 6 )

Fig. 8 shows the distribution of seismic intensities and general information about 1993 Kushiro-oki Earthquake (Magnitude 7.8). Table 4 shows a comparison between the ground surface acceleration recorded at Kushiro Seaport by Port and Airport Research Institute in the 1993 Kushiro-oki Earthquake and the 2003 Tokachi-oki Earthquake. 5) This report concluded that the ground acceleration in the east-west direction was greater in 2003, other items were larger in 1993.

Asahikawa 3 Sapporo Obihiro 5 6 3 Kushiro × 5 Epicenter Hakodate Urakawa 4 Date January 15, 1993 Epicenter E144.4°N42.9° Magnitude 7.8 Aomori 4

Fig.8 Seismic Intensity Distribution of the 1993 Kushiro-oki Earthquake

Table 4 Comparison of Maximum Acceleration at Kushiro Seaport3)

Maximum Acceleration(gal)

East-West North-South Up-Down Ground surface 576 347 149 2003 Tokachi-oki Earthquake Underground 202 154 66 Ground surface 343 450 362 1993 Kushiro-oki Earthquake Underground 268 203 121

Table 5 compares the damages on pipelines. While the pipelines were damaged at 31 points in 1993 Kushiro-oki Earthquake, the number of damages reduced to 7, or 1/4 that of 1993, in 2003 Tokachi-oki Earthquake.

The points of damages made in 1993 are plotted in Fig. 9. As in the case of 2003

- 10 - earthquake, the damages were concentrated in Midorigaoka district. The district is found on a reclaimed land in a valley and a number of houses and roads were damaged due to liquefaction in the 1993 Kushiro-oki Earthquake.

Table 5 Comparison of Water Pipeline Damages in Kushiro-city

1993 2003 Kushiro-oki Earthquake Tokachi-oki Earthquake Ductile iron pipes 12 3 Cast iron pipes 10 1 Asbestos-cement pipes 4 0 Steel pipes 3 1 Polyvinyl chloride pipes 0 0 Others(accessories) 2 2

Total 31 7

●

8 R3 ▼ ▼ Enlargement of A part Nem ▼ ◆ uro mai n line ◆ ▼ R44 ●● ▼ ★ ● ▼ Kushiro-nishi seaport Kush ★ ● iro ● ▼ sta ● ●▼ ●● tion ●● ●▼● ◎Kushiro city hall ● ● ● ■■■● ● Kushiro seaport ▼ ■ ■ ■ ■■■ ■ ▼ ○A ● ▼ Harutori lake ●…Ductile iron pipe ▼ ▼… ■ Cast iron pipes ■ ■ ■…Asbestos-cement pipes ● ◆…Steel pipes ★…Others

Fig.9 Points of Pipeline Damages in the 1993 Kushiro-oki Earthquake

Pipeline Renewal

After the 1993 Kushiro-oki Earthquake, Kushiro city worked on the renewal of pipelines to improve their earthquake resistance. 6) The asbestos-cement pipes, which totaled approximately 15 km at the time, are mostly renewed today. 7)

Furthermore, with regard to cast iron pipes, the Pipeline Renewal Plan for renewing the trunk line pipes with a nominal diameter of 300 or more was established in 1999. 7),8)

- 11 - In preparing this plan, the priority of pipelines to be renewed was determined through comprehensive evaluation of the damage rate, hydraulic importance and the degree of deterioration of each pipeline. The damage rate of individual pipeline was calculated against an assumed seismic motion by using the damage estimation formula9) proposed by the Japan Waterworks Association, and the result was converted into a score. The hydraulic importance score was obtained by calculating the population that would be directly affected by the stoppage of water supply due to some accident along the subject pipeline. The score for the degree of deterioration of the pipeline was calculated on the basis of its nominal diameter, the year of installation and the used water pressure, by employing the method of diagnosing cast iron pipes proposed by the Japan Water Research Center.

The priority for renewal was classified into Rank A, B and C. A total of 14.6 km of pipelines was classified as Rank A, or as the pipelines that are urgently in need of renewal, 7.4 km as Rank B, or the pipelines that require planned renewal, and 62.6 km as Rank C, or those whose timing of renewal should be determined as required by re-evaluating their conditions according to the environment of use and their state of deterioration.

In Kushiro-city, the renewal began with Rank A pipelines that were considered to be at the highest risk. All of these pipelines are completely renewed today.

Fig.10 compares the lengths of different types of pipelines at the time of 1993 Kushiro-oki Earthquake and today. From this, it is obvious that the previous asbestos-cement pipes and cast iron pipes, which were inferior in earthquake resistance, have been replaced with ductile iron pipes in the past ten years.

Fig.11 compares the layout of pipelines in Midorigaoka district in 1993 and that of today. The renewal of asbestos-cement pipes has progressed in this district, with only a few pipes remaining today. Furthermore, it is worth noting that the newly installed 8.4 km of ductile iron pipes equipped with a joint with anti-slip-out mechanism did not suffer any damages in 2003 earthquake.

From the above observations, the renewal of the pipelines in Kushiro-city is considered to be one of the factors that contributed to the reduction of damages on pipelines.

- 12 - 700 (593km) 600 Ductile iron pipes (500km) Cast iron pipes 500 400 Asbestos-cement pipes 82.5% 68.7% Others(steel pipes) 300 200 Pipeline length(km) Pipeline 3.1% 0.3% 26.0% 100 15.5% 2.4% 1.7% 0 1993 2002 Fig.10 Changes in Lengths of Different Types of Pipes (φ75 or more)

Ductile iron pipes Cast iron pipes Asbestos-cement pipes (1993) (2003) Fig.11 Pipeline Renewal Since 1993 Kushiro-oki Earthquake (Midorigaoka District)

Conclusion

By comparing the pipeline damages in Urakawa-town and Kushiro-city in the 2003 Tokachi-oki Earthquake and those of the previous earthquakes, replacing asbestos cement and cast iron pipes, which are inferior in earthquake resistance, with more resistant ductile iron pipes in a well-planned manner was proved to be effective in reducing the damages on the pipelines caused by earthquakes.

In Japan, the Ministry of Welfare and Labor announced the “Waterworks Vision” in June 2004, which is based on the basic principle of “Waterworks that continue to explore new possibilities to be the top in the world.” 11) The vision sets forth some

- 13 - concrete targets, such as achieving 100% earthquake resistance rate of trunk pipelines (conveyance, transmission and distribution mains) in the next ten years, for the purpose of additionally reinforcing the measures against disasters.

We intend to exert our effort on various endeavors, including R & D related to ductile iron pies, in order to contribute to the realization of this waterworks vision.

【References】 1) Hokkaido government http://www.pref.hokkaido.jp/kseikatu/ks-kkhzn/h15jishin/suidouhigai04.pdf 2) Japan Waterway Association, “Pipeline Damages in 1995 Hyogoken Nambu Earthquake and Their Analyses”, May 1996. 3) Masami Kurochi, Toshiyuki Iwamomoto, “Survey of Pipeline Damages of 1982 Urakawa-oki Earthquake”, Journal of Ductile Iron Pipes, Vol. 33, October 1982. 4) Kushiro City Waterworks, “Damages to Waterworks Facilities in Kushiro-oki Earthquake and Responses, January 1994. 5) Port and airport research institute http://www.pari.go.jp/bsh/jbn-kzo/shindo/japanese/japanese_news/japanese_news_2 003/japanese_news7.htm 6) Takeo Odashima, “Establishment of Aged Pipeline Renewal Plans Using the Earthquake Damage Estimation System”, the 52nd National Research Conference on Waterworks , May 2001. 7) Japan Waterway Journal (January 5, 2004) 8) Yukio Hatanaka, “Damages on Water Pipelines in Kushiro City in the 2003 Tokachi-oki Earthquake”, the 55th National Research Conference on Waterworks, May 2004 9) Japan Waterworks Association: Projection of Damages on Pipelines by Earthquakes, November 1998. 10) Japan Water Research Center, “Manual for Cast Iron Pipeline Renewal and Establishment of Renewal/Replacement Plans”, March 2001. 11) Ministry of Welfare and Labor, “Waterworks Vision”, June 2004. http://www.mhlw.go.jp/topics/bukyoku/kenkou/suido/vision2/index.html

- 14 - Session3: Risk Managemant

S3-1 “Bringing Pragmatic Engineering to Earthquake Preparedness at Contra Costa Water District” Presenter: Stephen J. Welch (Contra Costa Water District) USA

S3-2 “Seismic Upgrade of Water Facilities ─ An Asset Management Approach” Presenter: William F. Heubach (Seattle Public Utilities) USA

S3-3 “Assessing and Managing Risk ― Planning Future Upgrades to San Diego’ Water System” Presenter: Michael E. Conner (City of San Diego Water Department) USA

S3-4 “Case Study of a Backup System for Water in Kinki Region Presenter: Eizo Seki (Japan Water Research Center) Japan

S3-5 “Uniform Confidence Hazard Approach for the Seismic Design of Pipilines” Presenter: Craig Davis (Los Angeles Department of Water and Power)USA

S3-1

“Bringing Pragmatic Engineering to Earthquake Preparedness at Contra Costa Water District”

Presenter: Stephen J. Welch

(Contra Costa Water District, USA)

BRINGING PRAGMATIC ENGINEERING TO EARTHQUAKE PREPAREDNESS AT CONTRA COSTA WATER DISTRICT

Walter J. Bishop, Thomas J. Linville, Stephen J. Welch Contra Costa Water District

INTRODUCTION

The Contra Costa Water District (CCWD, District) provides raw and treated water service to over 500,000 customers in Contra Costa County, California. Its primary water conveyance system is an over 60-year old, 85-kilometer long, open, concrete lined canal which serves two CCWD water treatment plants, and various cities, municipalities, industries and agricultural customers.

In 1994, following the Loma Prieta and Northridge earthquakes, CCWD undertook a comprehensive seismic assessment of its water conveyance, treatment and distribution systems. The study focus was to identify strategic improvements throughout the CCWD system to minimize water service interruption after a maximum credible earthquake from relevant faults. The study identified over $170 million (2004 U.S. dollars) of improvements required in the system, including improvements to existing pumping and piping, as well as construction of additional pumping, piping and canal improvements.

CCWD’s resulting Seismic Reliability Improvement Program (SRIP) was created to implement the major improvements in a timely manner. Over that time frame CCWD has implemented various design approaches to complete the important task of readying its water conveyance, treatment and distribution system for earthquakes. Its investment of over $170 million of improvements to pump stations and storage, lifeline pipelines, emergency operation facilities, water treatment and power systems were made with a major focus on post-earthquake response.

Through these investments, CCWD has held a strong focus on preparing for earthquake response using strong project management to develop cost-effective engineered solutions. CCWD’s engineering provides design solutions that maximize available funding by using simple, reasonable improvements to ensure its ratepayers do not fund overly risk-average, excessive improvements. CCWD believes its vision of strong management with a focus on cost, timely product delivery, simplicity and prudent risk management has created pragmatic engineered designs for reliability following any emergency, but especially earthquakes. This paper will present some of those solutions, including design details and costs.

This report describes how CCWD implemented effective, efficient and pragmatic engineering solutions to meet the objective of being ready for the next major earthquake in Contra Costa County.

OUTLINE OF CONTRA COSTA WATER DISTRICT

CCWD was established in 1936 to provide water to the central and northeastern regions of Contra Costa County, California. CCWD headquarters is in Concord, California, approximately 56

kilometers east of San Francisco, located in Central Contra Costa County, one of the fastest growing counties in California. (See Figure 1).

Figure 1 – Vicinity Map

CCWD draws its water from the Sacramento-San Joaquin Delta under a contract with the U.S. Central Valley Project (CVP). As part of the CVP, the 85-kilometer Contra Costa Canal was built in 1937. By means of the canal, water is diverted from Rock Slough (13-kilometers east of Antioch, California) through a 7-kilometer unlined channel into a 78-kilometer concrete-lined canal. Four stations lift water 37.8-meters above sea level to the canal's Antioch summit, after which gravity flows the water to its terminus in Martinez. The canal runs through Oakley, Antioch, Pittsburg, Concord, Walnut Creek, Pleasant Hill, Pacheco, ending at a terminal reservoir in Martinez, California (see Figure 2).

Contra Costa Water District Service Area Map

Pittsburg

Martinez

Oakley Antioch Concord

Clayton Brentwood

Walnut Creek

Los Vaqueros Reservoir

Legend CCWD Service Area ´ 012340.5 Canal Miles Figure 2 – CCWD Service Area

CCWD has 4 raw water storage reservoirs: Martinez, Contra Loma, Mallard, and Los Vaqueros, totaling a storage capacity of 1,233,487-litres (103,070 acre-feet). It also has two treatment plants; the Ralph D. Bollman Water Treatment in Concord, with a capacity of over 300 million litres per day; and the Randall-Bold Water Treatment Plant in Oakley, with a capacity of 150 million litres per day. CCWD is also in the process of constructing a new water treatment plant in Brentwood, California with an initial treatment capacity of 60 million litres per day. Roughly half of CCWD customers receive treated water directly from CCWD, the remaining from 6 local agencies who treat and distribute CCWD water.

Today, Contra Costa Water District serves a population of over 500,000 and includes treated water distribution facilities with 33 pump stations, 44 storage reservoirs, and 1,252-kilometers of pipelines. CCWD is managed by a General Manager appointed and governed by a 5-member, publicly elected, Board of Directors. CCWD’s 325 employees serve customers that include the municipalities of Antioch, Pittsburg, and Martinez, commercial and industrial companies, agricultural entities, and the residential communities of Clayton, Concord, Pleasant Hill, and Walnut Creek

The greatest challenge CCWD faces today is continuing to provide high-quality water with an ever-increasing customer demand in an environment with an extremely limited water supply. CCWD is committed to ensuring adequate water resources, high water quality, and reliability in for the present and future, including in times of emergency.

GOOD DESIGN STARTS WITH A FOCUS ON THE PRACTICAL

1994 was a very challenging year for CCWD. CCWD was in the early stages of construction of a $450 million (1994 U.S. dollars) reservoir (Los Vaqueros Reservoir) to address ever increasing water quality challenges, was beginning implementation of earthquake preparedness as a result of the Loma Prieta and Northridge earthquakes, and was in the early stages of a rate stabilization strategy to keep rates below inflation in response to nearly a decade of high, variable rate increases under previous management. The importance of strong, focused management was never more important for CCWD, including engineering management.

Unlike some of the other water utilities in California, CCWD is not yet optimally scaled. CCWD faces the organizational requirements of large organizations such as formal governance controls, public relations, complex financial systems, responsibilities to industry stewardship, investment in existing and future capital, and elected official oversight, yet has only a 500,000 customer base over which to spread its costs (total population served, not number of connections). In addition, as CCWD’s sole water supply is the Northern California Delta, a water source of decreasing quality, the challenges of supply and treatment are increasingly expensive.

CCWD has taken a two-fold approach to meeting these challenges: expand the level of service provided, and place a significant focus on efficiency and productivity. CCWD’s traditional treated water service area is reaching build out, but as noted above, Contra Costa County is still rapidly expanding in the east county. CCWD future service efforts are focusing on serving this area of increasing demand with strong success in the communities of Antioch and Brentwood. However, despite this future opportunity, CCWD aggressively manages resource productivity through a culture of efficiency and effectiveness (for example all key performance measures are continually benchmarked for improvement). For the seismic improvement efforts, the result was that for CCWD to deliver the goals of improving water quality at lowered rates, while still implementing long-range improvements for post-earthquake reliability, CCWD had to focus on efficient, well-managed, practical improvements.

As most designers can testify, creating sensible designs can be a test of will. For structural engineering the details of transferring complex loadings to a simple, efficient load path can easily distract a design engineer from the objective. For many water facilities such as pipelines in varying soils, above and below ground water storage tanks, and above ground piping systems, the evaluation of loads itself can be easier said than done as the dynamics of soil, fluids, and structure are complex and oftentimes still theoretical. On top of this complexity add the test of understanding and designing simple load carrying systems, and even experienced engineers can miss the mark. Yet as most contractors and engineers will confirm, the best design is a simple, easy-to-build design. CCWD designers and engineering consultants went into each seismic improvement with this thought in mind. The objective not only was to create an efficient, simple load path (or other simple solution as was the case for the Treated Water Generator and Seismic Valve Project (TWGSV)), but also to create designs that were easy to construct, and thus, lower in cost. CCWD engineers faced varying proposals for the projects presented below, and oftentimes with workable but complex solutions, but the guiding decision criteria for CCWD was and is a focus on the practical. The end results presented below were designs that optimized the CCWD investment in its seismic improvements.

OPTIMIZING THE INVESTMENT

Following the Loma Prieta and Northridge earthquakes, in 1994 CCWD undertook a comprehensive seismic assessment of its water conveyance, treatment and distribution systems. The study focused on strategic improvements throughout the CCWD system to minimize water service interruption after a maximum credible earthquake. The study identified over $170 million (2004 U.S. dollars) of improvements to existing pumping and piping, storage and power systems. CCWD’s resulting SRIP was created to implement these improvements.1, 2

The purpose of the SRIP was to identify a combination of capital and operational improvements that efficiently allowed CCWD to provide reliable water service. Starting in March 1994, CCWD began assessment of its overall water storage, treatment and distribution system. The objectives of the assessment were to maintain public health and safety, meet water needs of existing and future customers, and ensure operation of critical facilities following an emergency or natural disaster such as an earthquake. The resulting SRIP planned a cost-effective package of system improvements to better CCWD’s system for future growth demands and post-earthquake service.

The SRIP initially defined reliability criteria and seismic design criteria. This first step included defining the study area (CCWD existing and planned raw and treated water service boundaries, see Figure 2); the planning period (out to the year 2020); the specific facilities to be studied, and the analysis stress events by which to model (Concord fault, M6.5; Great Valley fault M7.0). The criteria was intended to be comprehensive of both expected deficiencies of the normal operating system, and emergency based deficiencies resulting from the stress event. A key aim of this step was ensuring that the criteria would be focused on obtaining an efficient and pragmatic strategy.

Since 1994 the District has effectively implemented the complete SRIP. The implementation was completed through two key efforts. The first was construction of the Multi-Purpose Pipeline (MPP), a 45-kilometer, 122-centimeter backbone transmission pipeline designed to ensure a reliable post-earthquake source of either treated or raw water. The second was construction of seismic improvements through the CCWD’s existing Capital Improvement Program. From 1994 to the present CCWD had planned out capital improvements at various sites throughout its treatment and distribution system. These improvements were not necessarily entirely for earthquake preparedness, but in the interest of efficiency by getting construction economies, interrupting the site and community just once, and ensuring upgrades were comprehensive, CCWD opted to implement the seismic improvements (as identified in the SRIP studies) through already planned capital improvements7 rather than as part of a seismic-only program. To ensure the upgrades were implemented with a criteria based on overall need, improvements were prioritized based upon both operational need (for example additional storage or pumping capacity need, or end of useful life replacement cycle) and seismic importance. As a result, though many sites were already planned for improvement over the 10-year horizon, some were re-prioritized to be completed earlier, while other sites were added as a result of no previous planned improvements during the implementation horizon. (Table 1 in the following section outlines the sites and years of implementation.) Additionally, as the improvements were being implemented within the view of an overall improvement to the facility (both seismic and operational), designs oftentimes solved both a seismic deficiency as well as operational (for example unanchored interior-to-tank inlet/outlet pipelines were simply replaced with new, anchored pipelines that provide multi-directional, diffused water distribution to address both seismic and water quality issues simultaneously). Again, the focus was pragmatic solutions, whether through implementation within an existing planned project, through designs that solved both operational and seismic problems simultaneously, or through basic, easy-to-install design details.

THE IMPROVEMENTS

The improvements implemented by CCWD over the last 10 years have been predominantly applied to CCWD’s treated water system. About 90-percent of the funding for the seismic efforts was applied to the treated water system (assuming the majority of the MPP benefits are treated water.) Table 1 provides an overview of the improvements made, both on the treated and raw water facilities.

Site Description Year Project Seismic Cost Implemented Bid (Construction) TREATED WATER Coating, structural, Nob Hill Tank 2000 $250,000 $35,000 operational Rehab Coating, structural, Divide Tanks 2001 $307,000 $25,000 operational Rehab Coating, structural, Clayton Valley Tanks 2002 $709,000 $135,000 operational Rehab Port Costa Tank Tank Replacement 2002 $500,000 $150,000 Coating, structural, Elderwood Tank 2005 $1,505,000 $350,000 operational Rehab Pump, electrical, Bailey Pump Station 2000 $755,000 $45,000 structural, coating Rehab Pine Hollow Pump Pump, electrical, 2004 $504,000 $55,000 Station structural, coating Rehab Clayton Valley Pump Pump, electrical, 2002 $954,000 $65,000 Station structural, coating Rehab Elderwood Pump Pump, electrical, 2001 $775,000 $85,000 Station structural, coating Rehab Paso Nogal Pump Pump, electrical, 2004 $659,000 $72,000 Station structural, coating Rehab Lime Ridge Pump Pump, electrical, 2005 $1,942,000 $285,000 Station structural, coating Rehab TW Backup generators to key Generators/Seismic 2004-2005 $1,675,000 $145,000 pumping facilities (seismic valves) Valves RAW WATER Martinez Reservoir Coating, structural rehab 2004 $501,000 $55,000 Outlet of outlet pipeline EOC Building Structural upgrade 2004 $394,000 $305,000 Upgrades RW Reservoir Seismic improvements 2001 $1,847,000 $1,847,000 Seismic Electrical, structural RW Pumping Plants 2003 $678,000 $55,000 rehab Mallard Slough PS New pumping and piping 1999 $9,037,000 NA

Table 1 – Seismic Projects Summary (excludes MPP)

Treated water

Core improvements to the treated water system consisted of facility structural retrofits (for example strengthening existing building walls, or roof to wall connections, etc…), backup power connections, addition of seismically activated valves, strengthening of pipeline supports, improved anchorage of system components, flexible connections at structure to pipeline connections, planning for interties with adjacent water systems, and external power generators at key pumping plants.

Table 2 provides a summary breakdown from the Table 1 list of sites for the seismic improvements made at each treated water facility. Details 1 to 15 provide some actual design details for some of the project sites. An observation in reviewing both Tables 1 and 2, as well as by the details, is the fairly low cost and simple design of the seismic improvements in relation to the remaining project scope. Except for the Lime Ridge Pump Station and EOC Building Rehabilitation projects which carried a significant share of cost for building strengthening, the projects generally required less than 15-percent of the project funding. Through the use of sensible, basic, and easy to construct seismic engineering and design, CCWD has been able to prepare its system for the design earthquake for a reasonably low cost. In fact, the upgrade costs necessary to restore the subject sites to original life and current operational standards far exceeded the cost for the seismic improvements in almost every case.

Site Description of Seismic Scope Only TREATED WATER Nob Hill Tank Strengthened column, flex-tend to I/O, (advanced analysis verified tank) Divide Tanks Strengthened column, (advanced analysis verified tank adequate) New tank holdowns, flexible overflow connections, strengthened rafter Clayton Valley Tanks to roof to wall connections Port Costa Tank Tank Replacement - meet current code, flex-tend I/O Strengthened rafter to roof to wall connections, added rock anchors and Elderwood Tank foundation, flex-tend I/O Strengthened pipe supports, flex connections at suction/discharge Bailey Pump Station piping, equipment anchorage Pine Hollow Pump Strengthened pipe supports, flex connections at suction/discharge Station piping, equipment anchorage, anchored surge tank Strengthened pipe supports, flex connections at suction/discharge Clayton Valley Pump piping, equipment anchorage, advanced analysis of pump building, Station anchored surge tank Elderwood Pump Strengthened pipe supports, flex connections at suction/discharge Station piping, equipment anchorage, strengthened enclosure walls Paso Nogal Pump Strengthened pipe supports, flex connections at suction/discharge Station piping, equipment anchorage, strengthened enclosure walls Strengthened pipe supports, flex connections at suction/discharge Lime Ridge Pump piping, equipment anchorage, lowered roof to lower response (avoided Station new building cost), strengthened building walls TW Generators/ See Tables 3 and 4 Seismic Valves

Table 2 – Treated Water Projects Summary – Seismic Scope

One of the more innovative approaches taken in implementing the seismic improvements was used for the Treated Water Generator and Seismic Valves (TWGSV) project. This project initially started as a project to add additional treated water storage throughout the distribution system for post-earthquake (and other emergency) events. The project also provided for the addition of seismically activated shutoff valves at key, redundant treated water storage throughout the distribution system. The concept of the project was to add emergency response flow reliability by adding new tanks, and preserving existing tank storage volume through seismic valves. However, the estimated cost for this approach, primarily driven by the additional storage needs, exceeded $10,000,000.

The novel approach: instead of constructing more storage that mostly remained unused under normal operations, and could still simply drain away through significant pipeline ruptures following an earthquake, construct more reliability into pumping capacity. The thinking was that because the majority of the initial demand to the water system following an earthquake comes from pipeline ruptures, not just fire fighting, additional storage turns out to be a one-use investment. Once the storage capacity is gone, whether for fighting the fires or spilling to the ground, if the pumping capability is suspect, all can be lost. As the major reason for pumping failures following an earthquake is from lost power, CCWD placed its focus on power reliability. CCWD performed optimization studies to identify the key pump stations, and the needed pumping capacity for emergency response demand conditions to develop a project that instead installed emergency power at nine key pumping stations throughout the system.5 Tables 3 and 4 provide a summary of the power systems implemented.6 Table 4 highlights the generator sizing chosen to optimize the flows as a replacement for storage (reference 5 provides the optimization criteria.) CCWD did not need to size the generators to supply full pump station capacity (as can be observed in Table 4). Depending on existing storage within the pump station zone, as well as the zone specific demands (the percentages listed provide the share of customers served by the pump station), pragmatic generator sizing could be optimized at less than full capacity, and was. The result of the TWGSV project approach was a savings of over $8,000,000 to CCWD.

Table 3 lists the other minor effort of the TWGSV project, addition of reservoir outlet seismically activated valves. Such valves were required at four reservoir sites. The actuators were installed on existing valves where possible, and were only installed on reservoirs that provided redundant storage within the distribution system pressure zone.

TREATED WATER GENERATORS AND SEISMIC VALVES PROJECT

Project Summary Generator Total Pumps Generator Separate Main Sitework Sitework by Comments Sites (Pump Pumps at Served by Size Generator Construction Special Stations) Station Generators Procurement Package Package 1 BAILEY 6 4 450 kW √ √ √ CLAYTON 2 5 3 300 kW √ √ √ VALLEY GREGORY 3 3 1 150 kW √ √ √ GARDENS Package with future 4 LIME RIDGE 4 2 400 kW √ √ √ TWFIP design bid documents. 5 PASO NOGAL 3 2 100 kW √ √ √ Portable Generator PINE 6 5 4 300 kW √ √ √ HOLLOW Defer Generator Purchase 7 SAN MIGUEL 5 3 300 kW √ √ √ to Future TWFIP Project. Eagle Peak will be served EAGLE by existing portable 8 3 ------PEAK generators. No further work in this project. Seminary will be deferred 9 SEMINARY 3 ------to future TWFIP design.

Reservoir Sites No./Size New No./Size New No./Size Practice Main Sitework Site work by Comments (Seismic Valves Operators Seismic Shutdown Test Construction Special Valves) Valves Req'd Package Package New valve and motor operator. Includes 1 BAILEY 2-24" 1-24" 1-24" √ √ √ replacement of one manual valve. CLAYTON Reuse existing valves. Add 2 n/a 2-12" 2-12" √ √ √ VALLEY new motor operators. Package with future 3 LIME RIDGE 1-36", 1-24" 1-36", 1-24" 1-36", 1-24" √ √ √ TWFIP design bid documents. Reuse existing motor- 4 TAYLOR n/a n/a 1-24", 1-16" √ operated valves.

Generator Sites Total Pumps Pumps Served Generator Separate Main Sitework Sitework by Comments (Other) at Station by Generators Size Generator Construction Special Procurement Package Package 1 BISSO LANE N/A N/A 750 kW √ √ √ Defer to Future Project.

Table 3 – Generator Summary by Site

TREATED WATER GENERATORS AND SEISMIC VALVES PROJECT GENERATOR CAPACITY AND SIZING COMPARISON

TANK SIZE @ 24 PUMP STATION LOCATION [XX %] = [Percent PUMP SIZE CALCULATED FUEL CONSUMPTION RUN TIME for 650 HRS FULL LOAD of all TWSA customers served by pump station] NO. PUMPS (HP) GENERATOR SIZE (KW) (GAL/HR) GAL TANK (HR) (GAL) COMMENTS

BAILEY 1 125 150 12 54 290 [21%] 2 125 300 24 27 580 Caterpillar requires use of a 400 KW unit to meet air 3 125 350 28 23 670 regulations 4 125 450 36 18 860 Caterpillar requires use of a 650 KW unit to meet air 5 125 550 41 16 980 regulations 6 125 650 46 14 1110 CLAYTON 1 100 -- N/A N/A N/A 2 Pump Minimum VALLEY 2 100 300 24 27 580 [9%] 3 100 300 24 27 580 4 100 400 32 20 770 5 100 450 36 18 860 GREGORY 1 60 150 12 54 290 GARDENS 2 60 300 24 27 580 [3%] 3 60 300 24 27 580 LIME RIDGE 1 200 300 24 27 580 [31%] 2 200 400 32 20 770 Caterpillar requires use of a 650 KW unit to meet air 3 200 550 41 16 980 regulations 4 200 750 60 11 1440 PASO NOGAL 1 40 75 [5%] 2 40 100 8 86 180 3 40 125 10 65 240 PINE HOLLOW 1 75 100 8 81 190 [13%] 2 75 150 12 54 290 3 75 300 24 27 580 4 75 300 24 27 580 5 75 350 28 23 670 SAN MIGUEL [9%] Zone 2 pump 1 125 300 24 27 580 300 kW starts 1x125 HP Zone 2 pump 3 125 300 24 27 580 300 kW starts 2x125 HP 450 kW starts 1 x 125 HP Zone 2 pump 5 200 450 36 18 860 plus 1 x 200 HP 450 kW Starts 1x 125,1 x 200 Zone 3 pump 7 75 450 36 18 860 and 1 x 75hp Ultimate capacity (all Zone Zone 3 pump 8 50 500 40 16 960 2&3 pumps) Firm Capacity Zones 2 & 3 300 kW starts 2 x 125 HP (Pumps 1, 3, & 8) - - 300 24 27 580 and 1 x 50 HP NOTES: 1 ) Generator sizing calculation based on Caterpillar “Electrical Power Design Pro” Software. 2 ) Highlighted row indicates selected generator size for TWGSV Project.

Table 4 – Generator Sizing Optimization Page 10 of 14

Raw water

Core improvements to the raw water system consisted of canal facility improvements (gates and lining replacements), slope modifications (shallow the slopes), backup power to primary pumping, pipeline and fault crossing improvements, and an additional redundant raw water pumping plant and piping.

Table 5 provides a summary breakdown from the Table 1 list of sites for the seismic improvements made at each raw water facility. Details 10 to 15 provide some actual design details for some of the projects. Again, the seismic component of cost for the raw water improvements was a small fraction of any project implementation cost. The only exception was for the RW Reservoir Seismic Improvements Project which was a project implemented entirely to add emergency drain capabilities to one of CCWD’s raw water reservoirs (added a new outlet structure and pipeline, valving, instrumentation and controls). CCWD’s resulting cost effective solutions for earthquake preparedness ensured the objectives outlined above could be secured without rate impacts, and without limitations on CCWD service. A focus on the practical paid dividends.

Site Description of Seismic Scope Only RAW WATER Martinez Reservoir Installation of a 24-inch flex-tend to outlet Outlet EOC Building Add shearwalls, roof to wall connections, additional roof ties, wall to Upgrades foundation connections, added foundation RW Reservoir New outlet structure, new outlet piping, seismic actuated valves, new Seismic controls and valving RW Pumping Plants Strengthened walls, anchored equipment and piping Mallard Slough PS New pumping plant and piping – meet current code

Table 5 – Raw Water Projects Summary – Seismic Scope

(Note: Several raw water projects have been implemented since 1998 to improve CCWD’s canal system. These projects have been primarily canal slope and canal liner replacements for over $7,000,000 over that timeframe. These projects improve CCWD seismic response, but have not been listed as these projects have no specific seismic component included. The Mallard Slough Pump Station Project listed in Table 1 was a project that added a new, reliable raw water pumping plant to the CCWD system. This new pumping plant was constructed to improve raw water operational reliability under all conditions, but has the added benefit of improving CCWD post-earthquake response by adding another source of raw water to system demands. This project has been listed because of its significant contribution to post-earthquake operations, but does not include a seismic cost as the project was new construction with all design requirements consistent with current building codes. The seismic-only project cost was not therefore identified.)

DETAILS

The following details provide a general overview of the type of details used to improve seismic performance during rehabilitation of the sites. The designs present a simple approach to improving performance at the sites.

CONCLUSION

The Contra Costa Water District prides itself on being one of the more efficient and effective public utilities in California. To a degree its challenging environment of being solely reliant on the California Delta for water supply tests CCWD to be efficient. The CCWD objective of continuously increasing the value provided to its customers at a stable cost below inflation additionally drives efficiency. For the important task of preparing its system for future earthquakes, the same practical approach used to manage the CCWD system on a daily basis proved to be the best approach to ensure an efficient implementation of preparedness.

CCWD designers and consultants held a strong focus on simple, reasonable improvements over the last 10 years. The presented design approach and details provided some general examples of the type of pragmatic effort used to implement seismic improvements at CCWD. By keeping a view of the many performance criteria of CCWD facilities when designing improvements, including operational, maintenance, as well as structural needs, and looking for opportunities to provide simple solutions to the various seismic problems, CCWD was able to save significant cost. Developing solutions that solve multiple problems (such as with the TWGSV project that solved both seismic power outage issues, general water storage deficiencies, as well as any other type of power outage issue), as well as combining design bid packages into comprehensive bids (such as with the many tank and pump station rehabilitations), allowed for both economy in scope and scale. The results are earthquake response preparedness with cost-effective engineered solutions.

While CCWD hopes that a major earthquake that seriously jeopardizes its water system and the general public never occurs, CCWD has implemented the necessary improvements to ensure it is ready to respond with success for such an inevitable event. By using simple, practical designs and project implementation approaches, CCWD additionally has ensured that its rate-payers have obtained that reliability in a very cost-effective manner.

REFERENCES

1. Contra Costa Water District, “Water Facility Supply and Reliability Improvement Strategy – Development of Seismic and Reliability Criteria Final Report – Volume 1,” August 1994 2. Contra Costa Water District, “Seismic and Reliability Improvements Project, Volume 1, Appendices”, September 1996 3. Contra Costa Water District, “Seismic and Reliability Improvements Project, Volume 2”, September 1996 4. Contra Costa Water District, “Seismic and Reliability Improvements Project, Volume 3”, September 1996 5. Contra Costa Water District, “2002 Treated Water Master Plan Update, Final Report”, December 2002 6. Contra Costa Water District, “Treated Water Generators and Seismic Valves Project, Preliminary Design Report”, August 2003 7. Contra Costa Water District, “Ten Year Capital Improvement Program Plan, For Fiscal Years 2005 to 2014”, February 2004 8. Contra Costa Water District, Project Designs: “Clayton Valley Tanks, 2002; Bailey Pump Station, 2000; Paso Nogal Pump Station, 2003; Lime Ridge Pump Station, 2004; Martinez Reservoir Outlet, 2004; Raw Water Reservoir Seismic Improvements, 2001”

S3-2

“Seismic Upgrade of Water Facilities – An Asset Management Approach”

Presenter: William F. Heubach

(Seattle Public Utilities, USA)

Title: Seismic Upgrade of Water Facilities – An Asset Management Approach

Authors: William F. Heubach (Contact person) Senior Civil Engineer Seattle Public Utilities Engineering Division 700 5th Avenue, Suite 4900 P.O. Box 34018 Seattle, Washington 98124-4018 Phone: (206) 386-1389 Fax: (206) 233-7241 E-Mail: [email protected]

William C. Steenberg (co-author) Associate Civil Engineer Seattle Public Utilities Engineering Division P.O. Box 34018 Seattle, Washington 98124-4018 Phone: (206) 386-1389 Fax: (206) 684-7612 E-Mail: [email protected]

Seismic Upgrade of Water Facilities – An Asset Management Approach

William F. Heubach1 William C. Steenberg1

ABSTRACT

Seattle Public Utilities (SPU) has recently initiated a comprehensive Asset Management Program. Detailed project development plans must be prepared for all projects, including seismic upgrade projects, over $250,000. The plans must identify alternatives and present business case analyses that include benefits and costs for all alternatives.

Seismic upgrade projects present unique challenges for incorporation into the Asset Management Program. The benefits and costs of seismic upgrades are heavily dependent on the system-wide response to facility behavior. Future system changes can also greatly influence the benefits and costs of seismic upgrades. Business loss-of-opportunity costs are difficult to estimate.

As part of an American Water Works Association Research Foundation tailored collaboration project, a methodology for assessing system-wide response from seismic damage was used to estimate costs associated with earthquake damage. The methodology included using standard seismic vulnerability assessment procedures and incorporating the vulnerability results into a hydraulic model. Assumptions about facility earthquake performance can then be readily changed in the model to reflect proposed seismic upgrades and their effect on post earthquake performance. Future system changes can also be incorporated into the model.

The model results are then used to estimate economic losses. Although seismic upgrade and earthquake repair cost estimation is (relatively) straight-forward, business loss-of-opportunity costs resulting from water unavailability is much more difficult to quantify. Because many current business economic loss models appear to greatly overestimate the effects of water unavailability, SPU has developed a procedure to more realistically estimate business economic losses from water unavailability. These loss estimates are then incorporated into the business case analyses to determine the merits of facility seismic upgrades.

INTRODUCTION

In 1990, Cygna Energy Services completed a seismic vulnerability assessment of Seattle Public Utilities’ (SPU’s) water system facilities. Based on facility criticality and seismic vulnerability, vulnerable facilities were prioritized for upgrade and a $20 million seismic upgrade program was initiated.

In the late 1990’s, SPU initiated a program to identify and seismically assess and upgrade the system of major in-town pipelines that directly connect the Tolt and Cedar River transmission pipelines to the in-town

1 Seattle Public Utilities, 700 5th Avenue, Suite 4900, P.O. Box 34018, Seattle, Washington, 98124-4018

1 reservoirs and tanks, and to important economic areas such as the downtown corridor. The intent of this program was to develop a system of “backbone” pipelines that would remain functional and still be capable of supplying the in-town reservoirs and tanks after a major earthquake.

In 2003, a comprehensive asset management program was initiated by SPU. This program requires detailed business case analysis of all projects over $250,000. Life cycle costs, expressed in terms of net present value, are estimated for different alternatives that can be used to satisfy project requirements. The business cases are presented for approval to an asset management committee comprised of SPU executives.

In order to estimate seismic upgrade benefits, the system performance must be determined assuming the upgrade(s) is implemented and assuming the upgrade(s) is not implemented. The mitigation alternative evaluation methodology involves developing a hydraulic model of the in-town water distribution system and running the model for the different mitigation alternatives. Based on the hydraulic modeling results, economic losses are estimated. Construction, operations and maintenance costs and the expected earthquake losses for different alternatives can be used to calculate the net present value of each alternative. The methodology is shown in Figure 1.

DATA COLLECTION

SPU maintains a Geographic Information System (GIS) with comprehensive data on its water system facilities. Pipeline vulnerability significantly affects overall system vulnerability. Consequently, information on pipeline characteristics that affect pipeline vulnerability such as pipeline materials, joint types, diameter and year of installation data is particularly important.

SEISMIC HAZARDS

For each upgrade/mitigation alternative, seismic losses must be estimated across the continuum of seismic events that may affect water system performance. Like many seismic areas, there are a variety of seismic source zones that may affect the SPU water system area. In order to simply the analysis, the assessment was performed for a representative scenario and the results were interpolated to cover the continuum of seismic events.

Although there are interplate subduction and shallow surface fault zones that may cause large ground motions in the Puget Sound region, deep intraplate earthquakes, 40 to 60 kilometers below the earth’s surface, are the most likely to cause damaging ground motions in the Puget Sound. Deaggregation of the different source zones (Frankel) shows that intraplate events are the primary contributor to the 0.10 probability of exceedance in 50 years ground motions. The 1990 Cygna seismic vulnerability study also was based on the 0.10 probability of exceedance in 50 years ground motions ground motions. Consequently, these ground motions were used as a basis to define the representative earthquake scenario.

Although peak ground acceleration is often used to define building and structure vulnerability, permanent ground displacement and peak ground velocity are more indicative of pipeline damage. Permanent ground displacement is usually a much more significant contributor to pipeline damage than transient wave propagation effects. Permanent ground displacement can result from fault rupture, liquefaction/lateral spread, settlement, landslide or lurching. Peak ground velocity is often used to express transient wave propagation. A consultant, Zipper-Zeman, estimated the peak ground velocities and permanent ground displacements for 500-year ground motions (see Figures 2 and 3).

2 Pipeline Properties Earthquake Hazards - Material - Wave Propagation - Joint Type - Permanent Ground - Size Displacement - Condition

ALA Pipeline Leak Rate Vulnerability Model Models

Expected Seismic Average Leak Rates as a Pipeline Vulnerability/ Performance of Non- Function of Pipe Size Pipeline Repair Rate/ Pipeline Water System and Permanent Ground Number of Repairs Components Displacement

EPANET Hydraulic Model of SPU Water System

Water System Performance As a Function of Time After Earthquake

Loss/Cost Estimates - Business and Fire Economic Loss - Repair and Life Cycle Cost - Seismic Upgrade Analysis - Maintenance

Life Cycle Cost Estimate

Figure 1. Backbone Pipeline System Study Methodology

Legend

Permanent Ground Displacement 0 - A 0.1 - B 0.5 - C 3 - D 6 - E 12 - F 24 - G 36 - H >36 - I

10000 0 10000 20000 Feet

W E

No guarantee of any sort implied, including accuracy, completeness, or fitness for use.

Figure 2. 500-Year (0.10 Probability of Exceedance in 50 Years) Peak Ground Displacements (Inches)

Legend Peak Ground Velocity 9.4 - A 9.7 - B 10 - C 10.2 - D 10.8 - E 11.9 - F 12.9 - G 13.3 - H

10000 0 10000 20000 Feet

W E

No guarantee of any sort implied, including accuracy, completeness, or fitness for use.

Figure 3. 500-Year (0.10 Probability of Exceedance in 50 Years) Peak Ground Velocities (Inches/Second)

5 FACILITY VULNERABILITY

Cygna Energy Services used structural analysis techniques to estimate the seismic vulnerability of concentrated facilities such as pump stations, tanks, treatment facilities, etc.

Pipeline seismic vulnerability models were used to estimate the number of pipeline breaks/leaks. The most significant pipeline characteristics that influence susceptibility to permanent ground displacement and wave propagation hazards are the pipeline materials, pipeline joint type, pipeline state-of-repair, the backfill/trench conditions, the pipe segment length, the presence of cross connections, and pipeline diameter. In general, pipelines made with ductile materials such as welded steel, ductile iron and polyethylene tend to perform better than more brittle materials like cast iron and concrete. Restrained joints that allow some joint movement also tend to perform better than rigid unrestrained joints.

The American Lifelines Alliance (ALA) pipeline vulnerability models (American Society of Civil Engineers/Federal Emergency Management Agency) were used to estimate the number of pipeline breaks/leaks. Numerous other pipeline vulnerability models have been developed. Although it was impossible to compare accuracy of the different models, ALA models were available for more pipe material and joint types than most of the other models. The ALA models are of the form

RRpgv = 0.00187 * K1 * PGV

0.319 RRpgd = 1.06 * K2 * PGD

where

RRpgv = the number of repairs per 1000 feet due to wave propagation effects K1 = a peak ground velocity constant based on the pipe material and joint type PGV = the peak ground velocity in inches per second RRpgd = the number of repairs per 1000 feet due to permanent ground displacement K2 = a permanent ground displacement constant based on the pipe material and joint type PGD = the peak ground displacement in inches

Regression analyses were used in the ALA report to calculate K1 and K2 . Typical K1 and K2 values are shown in Table 1. Like most water utilities, SPU has many materials and joint types in the SPU pipe inventory that were not addressed in the ALA report. Engineering judgment was used to assign K1 and K2 values to those pipeline material and joint types not covered in the ALA models.

The ALA report suggested that the data was inconclusive on whether pipe vulnerability was a function of pipe diameter. Larger diameter pipelines usually have stronger cross sectional properties. Because larger diameter water pipelines tend to be more critical than smaller diameter pipelines, construction quality control also tend to be better for larger diameter pipelines. There is data that suggests that larger diameter pipelines were less vulnerable than smaller diameter pipelines. Consequently, data presented in the ALA report was used to develop a least squares regression equation to express pipeline vulnerability such that

Fd = 1.586 * exp (-0.0770 * D)

where

Fd = a pipe size factor that was multiplied by RRpgv and RRpgd to obtain the failure rate for pipelines with diameter equal to D in inches

6 Fd was limited to a minimum value of 0.25 (pipe diameters of 24 inches and larger used Fd = 0.25) and a maximum value of 1.0 (pipe diameters 6 inches and smaller used Fd = 1.0).

Table 1. Representative K1 and K2 Values

Pipe Material Joint Type K1 K2

Cast Iron Cement 1.0 1.0

Cast Iron Rubber Gasket 0.8 0.7

Welded Steel Rubber Gasket 0.7 0.7

Asbestos Cement Rubber Gasket 0.5 0.8

Asbestos Cement Cement 1.0 1.0

PVC Rubber Gasket 0.5 0.8

Ductile Iron Rubber Gasket 0.5 0.5

An Arc Macro Language (AML) program was used within the GIS to calculate the failure rate for each pipe segment. For the SPU system, over 63,000 segments were analyzed.

Based on the permanent ground displacement and peak ground velocities predictions, and the ALA vulnerability models, approximately 1253 pipeline failures were estimated for the SPU water system. These failures are summarized in Table 2.

Table 2. Pipeline Failure Summary

Wave Propagation Permanent Ground Pipeline System Failures Displacement Failures Total Failures

Primary Backbone 10 57 67

Secondary Backbone 12 131 149

Distribution 109 928 1037

Total 137 1116 1253

7 PIPELINE LEAK RATES

The severity of the breaks/leaks were estimated. Although the ALA models predicted the number of expected pipeline repairs, these models do not indicate the severity of the repair. A failure that completely breaks open a pipeline would have far more severe consequences on the hydraulic performance of the system than a minor joint failure. A combination of engineering mechanics and engineering judgment was used to estimate the leak severity.

Based on judgement and the 1990 study of the southern part of the SPU water system, a basic failure was assumed to cause a 200 gallon per minute (gpm) leakage of water through an 8-inch pipeline at 60 psi. Greeley's Formula (AWWA M36) was used to calculate the orifice area needed to allow 200 gpm to leak through an 8-inch pipeline pressurized to 60 psi. For larger and smaller diameter pipelines, the leak area was assumed to be proportional to the pipe diameter.

The basic failure was assumed to be caused by transient waves. In areas where there was not any permanent ground displacements, after the transient waves have passed, there would not be any permanent ground displacements that would result in increased pipe separation beyond the initial break. In areas of permanent ground displacement, it was assumed that the permanent ground displacements would enlarge the original break.

HYDRAULIC MODELING

An EPANet hydraulic model of the water system was used to determine earthquake effects on water system operation. Post-earthquake facility operational status for the concentrated facilities was based on the 1990 Cygna Energy Services seismic vulnerability study findings and approved or completed seismic upgrades.

The pipeline leaks were also applied to the model. The leaks were applied to the EPANet hydraulic model as emitters at the end of a control segment and a check valve. Emitters were junctions within EPANet that release water based on the pressure differential across the emitter. The check valve was used to keep EPANet from allowing water to flow back into the system whenever negative pressures occurred. The emitter settings allowed for a coefficient C that was based on the expected flow in gallons per minute (gpm) with a one pound per square inch (psi) pressure drop and an emitter exponent typically set equal to 0.5 such that

Q = C * Py

where

Q = the flow through the emitter in gpm C = the emitter coefficient P = the water pressure in psi y = the emitter exponent, usually 0.5

Applying a large number of leaks/emitters to the EPANet model would have been very time consuming. Additionally, it was difficult to get EPANet to converge if there were a large number of leaks/emitters. The 1253 projected leaks were reduced to 77 equivalent leaks. The demand on the system equal to 50% of the maximum system demand was assumed.

The hydraulic model was run and the results were compared for different upgrade assumptions. Six different potential options have been identified:

8 1. Do nothing 2. Upgrade and/or replacement of existing pipelines 3. Implement isolation and control strategies 4. Implement seismic design standards for new facilities and pipelines 5. Emergency preparedness and response 6. A combination of Approaches 2 through 5.

Figure 4 compares the water pressure throughout the system 8 hours after the scenario earthquake under the current conditions and with a mitigation strategy that includes selected facility upgrades and isolation of vulnerable pipeline areas. Note that installation of the isolation valves keeps the central part of Seattle pressurized. This central corridor includes the important downtown and hospital districts. In addition to maintaining water pressure in this critical corridor, Seattle in-town reservoirs are prevented from draining. Water in the reservoirs can be accessed by the fire department to use in areas without water pressure and will also allow much quicker service restoration to those areas without major pipeline damage.

Figure 4. Comparison of Water System Pressure 8-Hours After Earthquake Without Any Seismic Upgrades and Using Isolation Valves to Isolate Damage

9 ECONOMIC LOSSES

Economic loss estimates will be used as input with other life-cycle costs to develop net present value estimates for different upgrade options. Three types of economic losses are being estimated:

• Pipeline repair cost • Fire losses due to water unavailability caused by pipeline failure • Reduction in gross regional product caused by water unavailability

Because SPU is essentially owned by its customers, SPU economists believe the net effect of loss of revenue from being unable to sell water due to earthquake damage is zero.

Pipeline Repair Costs

Pipeline repair cost data from the Nisqually Earthquake is being used to estimate the pipeline repair costs for the scenario earthquake. The repair costs are estimated as a function of break severity. The estimated break severity is based on the leak severity analyses.

Additional Fire Losses

The number of fire ignitions caused by the earthquake is based on the FEMA Hazus model. The ignition model was determined from a regression analysis. The data shows a great deal of scatter so a significant uncertainty should be assumed when using the model. According to the model:

I = -0.025 + 0.592a – 0.289a2 where I = the number of ignitions per 1,000,000 square feet of built land a = the peak ground acceleration expressed as a function of gravity

The 500-year peak ground accelerations in Seattle are expected to be around 0.3g. It is estimated there are 2.5 billion square feet of built space in Seattle. Consequently, the FEMA Hazus model estimates 69 earthquake-caused fires for Seattle after an earthquake that produces 0.3g ground motion.

Approximately 70% of earthquake-caused fires occur immediately after the earthquake and the other 30% of the fires occur in the days following the earthquake as electric power and natural gas service are restored. The number of ignitions is multiplied by 0.7 to determine the number of expected fires immediately after the earthquake. In Seattle, 48 of the 69 fires would be expected to occur immediately after the earthquake.

The number of fires is calculated as a function of land use zone. For each land use zone, an average exposed value of the structure (the potential fire loss - does not include the land value) is estimated. Fire fighting effectiveness is modeled by Hazus as

0.7 Effectiveness = (Rtruck Rwater ) where Rtruck = the number of fire trucks available to fight the fire divided by the number of trucks needed to fight the fire Rwater = the water available to fight the fire divided by the water needed to fight the fire

Because there are only 32 engine trucks (plus 11 ladder trucks), it is assumed that only 32 of the 48 fires can be effectively fought even if water is available. In other words Rtruck = 0 for 16 of the fires. Rwater is determined for

10 the different seismic upgrade/option scenarios so that the fire fighting effectiveness can be compared for each upgrade scenario. It is assumed that regardless of fire engine or truck availability, an average of 25% of the exposed value will be lost due to the fire. Additionally, it is assumed that the fires occur during benign weather conditions when fire spread is not likely. The loss for each fire is estimated as

0.7 FL = EV {0.25 + [1 –(Rtruck Rwater ) *0.75]} where FL = the fire loss EV = the exposed value of the structure

Monetary Losses to Business and Industry

Water utilities have traditionally used macroeconomic approaches (eg., see Chang) to estimate earthquake losses due to water unavailability. This approach looks at the gross regional product for each industrial sector and how water unavailability would limit production and lower the gross regional product.

However, there are several problems with the macroeconomic approach:

1. The reduction due to water unavailability must reflect the reduction in only the value-added portion of the economic figures 2. The economic loss must be deaggregated so that the loss due to water unavailability can be separated out from the loss due to damage to the facility/business infrastructure, unavailability of electric power, damage to the transportation, telecommunications, wastewater, etc. systems. Currently, there not any reasonably accurate methods to do this on a macroeconomic scale. 3. The geographic boundary for the loss estimates must be defined. For example, even if Factory A is forced shut down, there is likely a Factory B located somewhere that will be able to increase capacity to meet the demand from loss of Factory A. Although there will be some inefficiencies created by transferring the work from Factory A to Factory B, the overall economic loss will be much less than the value-added that Factory A is not able to produce. However, if Factory B is located in a different jurisdiction than Factory A, officials in Factory A’s jurisdiction will be primarily concerned with the gross loss in their jurisdiction and less concerned with the overall net effect. 4. The loss estimates must account for recovery effects. That is, if Factory A is forced to shut down for a period of time, the value added by Factory A may be able to be recovered by Factory A by increasing production rates, overtime, etc. after Factory A resumes production. 5. Additional losses may incur in the region affected by the earthquake because customers chose to continue patronizing businesses outside the affected region even after the businesses inside the affected region regains production capability.

Because macroeconomic models do not adequately account for these factors, the models overestimate the business interruption costs due to water unavailability (eg., ECONorthwest and National Resource Council).

A microeconomic approach is being used by SPU. The EPANet hydraulic model is used to estimate the volume of water that cannot be delivered to SPU customers because of earthquake damage. The value of this water is estimated as the cost to replace this water by emergency methods such as tanker truck. Although the replacement cost of the water still does not precisely reflect the business opportunity losses due to water unavailability, this replacement cost estimate is a much better indicator of the business opportunity costs than the estimates used by current macroeconomic approaches.

These business opportunity losses are combined with the other costs to estimate life cycle costs of the different alternatives. It is important to note that life safety issues are also considered when the mitigation alternatives are evaluated.

11 ACKNOWLEDGMENT

Some of this work was performed as part of an American Water Works Association Research Foundation tailored collaboration project. Don Ballantyne of ABS Consulting was the principal investigator. Funding was provided by the American Water Works Association Research Foundation and the following participating utilities:

City of Everett, Washington City of St. Louis Greater Vancouver (British Columbia) Regional District Los Angeles Department of Water and Power San Francisco Public Utilities Seattle Public Utilities Tacoma Public Utilities Thames Water

David Lee (East Bay Municipal Utility District), Charles Pickel (Memphis Light, Gas and Water) and Jim Doane (Portland, Oregon Water Bureau) served on the Project Advisory Committee.

REFERENCES

American Lifelines Alliance, Seismic Fragility Formulations for Water Systems, Part 1 – Guideline, American Society of Civil Engineers/Federal Emergency Management Agency, April 2001.

American Lifelines Alliance, Seismic Fragility Formulations for Water Systems, Part 2 – Appendices, American Society of Civil Engineers/Federal Emergency Management Agency, April 2001.

American Water Works Association, Water Audits and Leak Detection, AWWA M36, 1999.

Baska, David A., PGV and PGD Estimation Project, Report prepared by Zipper Zeman Associates for Seattle Public Utilities, June 26, 2003.

Chang, Stephanie E., Evaluating Disaster Mitigation: A Methodology for Urban Infrastructure Systems.

Cygna Energy Services, Seismic Reliability of the Seattle Water Department’s Water Supply System, February 6, 1990.

ECONorthwest, Valuing the Benefits of Earthquake Protection: Literature Review and Recommendations, Report prepared for Seattle Public Utilities, October 21, 2004.

Federal Emergency Management Agency (FEMA), HAZUS Technical Manual.

Frankel, Arthur D., Powerpoint Presentation on Puget Sound Probabilistic Ground Motions at the October 21, 2004 United State Geological Survey Workshop, Seattle,Washington.

National Research Council, Commission on Geosciences, Environment and Resources, The Impacts of Natural Disasters: A Framework for Loss Estimation, National Academy Press, 1999. 3, 1996.

12 S3-3

“Assessing and Managing Risk – Planning Future Upgrades to San Diego’s Water System”

Presenter: Michael E. Conner

(City of San Diego Water Department, USA)

Assessing and Managing Risk – Planning Future Upgrades to San Diego’s Water System

Michael E. Conner P.E. City of San Diego Water Department

Abstract

The San Diego Water Department serves over 1.3 million people including the cities of Coronado and Del Mar. The distribution system includes over 3,300 miles of pipelines serving 99 different pressure zones. The Department’s nine raw water reservoirs supply three water treatment plants, 45 pump stations, and 31 treated water reservoirs. The City of San Diego does not have a separate capital improvements program for seismic upgrades to its water system. The funding for improvements requires an assessment of competing needs based on:

Risk Assessment – Vulnerability Assessment/Security Seismic Vulnerability Operational Need – Reliability Capacity – Raw, treated, and transmission capacity Condition Assessment Regulatory Requirements – Water Quality – Meeting future drink water standards Water Reuse – Local, State, and Federal mandates for recycled water

The San Diego Water Department incorporates the various needs via condition assessment, vulnerability studies, hydraulic modeling, and a formal prioritization process. This paper presents how the CIP Planning and Engineering Sections work cooperatively with Water Operations to devise the best management practice for upgrading the City of San Diego’s water system.

INTRODUCTION

The Capital Improvements Program (CIP) originated in 1996, when a Strategic Plan for Water Supply was initiated with the support of a City wide Public Advisory Group and the City Council. The two groups participated in water supply workshops, discussing the need for significant capital improvements to ensure a cost effective, safe, and reliable water supply. The strategic plan process identified the need to implement a Capital Improvements Program. The CIP Program Management Plan was developed in 1997 and established a list of capital projects to be implemented starting in 1998 and continuing through 2006. The projects are being financed through the sale of investment quality municipal bonds, capacity charges on new development, and water rate revenues. The repayment of the bonds and costs associated with the new CIP are being paid for by an increase in water rates approved by the City Council. Due to limited financial resources the CIP could not incorporate all of projects identified during the initial development of the program. Subsequent studies have added projects to the CIP including the projects identified as a result of the Seismic Assessment Project. A priority listing of CIP Projects is updated annually to best utilize the funding available for capital improvements. The CIP Planning and Engineering Sections work cooperatively with Water Operations to devise a best management practice for upgrading the City of San Diego’s water system.

THE WATER SYSTEM

The City of San Diego’s water system serves 1.3 million people and the Cities of Coronado, Imperial Beach, Solana Beach and Del Mar. The San Diego Water Department (SDWD) Distribution System is composed of about 3,300 miles of distribution pipe, 3 water treatment plants, 51 pump stations, 31 treated water reservoirs, 14 pressure tanks, all serving a total of 99 pressure zones. The 3 water treatment plants get raw water either from the City’s own raw water reservoirs, or directly from the San Diego County Water Authority aqueducts, which receive water from the Colorado River and the State Water project through the Metropolitan Water District. These aqueducts are also used to fill some of the City’s raw water storage reservoirs. The City imports approximately 90 percent of its annual water supply from the San Diego County Water Authority.

SEISMIC ENVIRONMENT

The City of San Diego is exposed to potential earthquakes from a number of local area faults. The major faults include Rose Canyon, Silver Strand, La Nacion, Coronado Banks, San Diego Trough, San Clemente, Newport Inglewood, Elsinore, San Jacinto, and San Andreas. Figure 1 shows the major San Diego Faults considered in this analysis.

Five possible earthquakes represent a bounding set of events that could generate the worst impacts to the water system. These possible earthquakes are shown in Table 1.

Table 1. San Diego Area Possible Earthquakes

Earthquake Magnitude Estimated Where Source MW Recurrence (yrs) Rose Canyon/Silver Strand 7.2 2000 La Jolla to Mexico Rose Canyon 6.5 400 La Jolla to Mission Bay Silver Strand 6.5 1000 Downtown to Mexico Elsinore 7.4 2500 East of San Diego La Nacion 6.6 10,000 Alvarado WTP to Mexico

Figure 1. Major Faults Near San Diego Service Area Figure 2 shows the mapped liquefaction and landslide zones. The liquefaction zones are characterized as being “high” or “low” susceptibilities. The landslide zones are characterized as being in one of nine susceptibilities. Each landslide or liquefaction zone can be digitized as a polygon with a typical boundary accuracy of approximately 20 feet. Figure 2. Liquefaction and Landslide Zones in San Diego

Analysis of City of San Diego Water Department Facilities

A Water Department’s Seismic Assessment Project analyzed the City’s entire water system including the raw water piping, water treatment plants, and the treated water distribution system. Additionally, an analysis reviewed the reliability of power after an earthquake. The analysis of the electrical grid was compared to the Water Department’s pumped pressure zones. It was concluded that most pump stations would have electrical service restored within 24 hours of a major seismic event. The Seismic Assessment Project identified improvements needed throughout the water distribution system. Most of the pump stations, reservoirs, and water treatment plants appear to be rugged enough to withstand a probable earthquake. However, many pipelines are subject to damage in the case of such an earthquake.

Buried Pipe Vulnerability Assessment

The inventory of water system infrastructure was exposed to simulated earthquakes using a Monte Carlo simulation model (SERA). The SERA model incorporates the location of faults, attenuation models (which account for spectral variations with distance from the fault and local soil type), landslide models (which account for the proportion of each slide zone that can actually slide), liquefaction models (which assess the chance of liquefaction given local accelerations and soil profile types), and the amount of surface fault offset at locations where pipes cross faults. For each site location, the peak ground acceleration, response spectral ordinate, peak ground velocity and permanent ground deformation is calculated. Pipe damage algorithms were adapted from published works by Eidinger (Eidinger 1998, Eidinger, in press). For fault crossings, the amount of offset and the pipe material are the critical parameters in determining whether the pipeline will break. Other parameters (soil backfill, angle of pipeline crossing, depth of burial) were also considered.

The performance of the system was measured using a mitigation and cost benefit analysis technique considering life safety, fire loss, hospitals, and customer service requirements on a pressure zone basis. Different seismic improvement packages were developed, analyzed and potential reductions in water related losses were estimated. Based on the results of the cost benefit analysis, recommendations have been made and are being implemented in the Capital Improvements Program. Seismic upgrades are included in all facilities being rehabilitated or replaced. Individual projects are being scored in the CIP prioritization process.

CONDITION ASSESSMENT

All water agencies are trying to stretch their budgets to maximize ratepayer value. Condition assessment is a key part of any resource allocation strategy. For buried pipelines corrosion is a major factor affecting service life and a pipeline’s ability to withstand the forces of an earthquake. For small water distribution mains (16” and less), the cost to excavate and perform a condition assessment is high relative to replacement costs. However, for large transmission pipelines (30” and larger), a significant amount of capital costs can be saved via a condition assessment. The City of San Diego utilized such an approach on the Otay 2nd Pipeline.

Background

The Otay 2nd Pipeline conveys water from the Otay Water Treatment Plant to service customers in the southern areas of San Diego. Built in the 1920s, the 36-54" pipeline extends from the Otay Reservoir 19 miles northward to the University Heights Reservoir. The pipeline is a key link between areas served by the Alvarado and Otay Water Treatment Plants, and plays a vital role in reservoir management strategies. Figure 3 shows the service areas fed by San Diego’s three water treatment plants.

In 1989, based on the assumptions that the 60 year old Otay 2nd Pipeline had outlived its useful life and lacked adequate capacity, a capital improvement project was initiated. In 1995, this project was revised and consisted of two phases: the first was to install approximately 30,000 lineal feet of 48 or 54-inch pipeline between Telegraph Canyon Road and State Highway 54 and the second was to install approximately 6,725 lineal feet of 42-inch pipe to replace portions of the 36-inch pipeline north of State Highway 94.

Figure 3. Water Treatment Plant Service Areas

In 1997, a Value Engineering (VE) review was conducted on the Otay 2nd Pipeline project to ensure that its function would be achieved at the lowest overall cost without sacrificing quality. The VE evaluated the alignment study for the first phase and the 30% design for the second phase. The VE recommended the size of the Otay 2nd Pipeline be studied to determine if there was sufficient hydraulic capacity to meet future demands. In addition, the VE recommended that the entire pipeline’s condition be thoroughly investigated to more accurately assess the need for replacement. If this investigation showed that the pipeline’s condition was not as deteriorated as anticipated, and the existing capacity was adequate, then a significant reduction to the $108 million cost to replace the pipeline could be realized by extending the pipeline’s life using a cathodic protection system.

A pipe sizing study was completed in May 1999. Hydraulic modeling determined that the capacity of the existing pipeline was adequate for the next 20 to 25 years. Accordingly, the Water Department decided to analyze the condition of the entire pipeline to determine the need for replacement.

Field Investigations

The investigations included excavation and extensive tests and analyses at determined test sites. Each excavation was approximately ten feet by fifteen feet in plan, centered on the pipe, and extended to a depth of approximately two feet below the bottom of the pipe. The pipeline’s coating was examined to determine its condition, thickness, adhesion to the pipe and dielectric insulation effectiveness. Laboratory tests were conducted to determine the coating’s water absorption and resistivity characteristics, which are indicators of its performance. The coating was then sandblasted off the pipe, and the pipe examined. Pipe tests included a corrosion pitting survey and ultrasonic testing to determine the remaining pipe wall thickness. The resistivity of the soil surrounding the pipe was tested to determine if it represented a corrosive environment to the pipe. Additional laboratory tests determined the soil’s chloride content and pH to further characterize the soil’s corrosive potential. After completing the inspection and tests at each excavation site, the pipe was recoated and backfilled.

Conclusions of Condition Assessment

The final report, prepared by Corrpro Companies, Inc. in November 2000, drew conclusions about the condition of the pipeline based on the tests and inspections done and made recommendations for further action. The report identified sections of the pipe in good condition, areas which require additional corrosion protection and sections which require replacement. Facilities to update or install a new cathodic protection system include rectifiers, test stations, electrical isolation of copper water services, and new coatings for 28 exposed sections of the pipeline. Installation of a cathodic protection system will extend the life of the existing pipeline and defer costly replacement of these sections. The Water Department used the conclusions and recommendations detailed in this report to develop an alignment and phasing program. The condition assessment resulted in a plan for $40-45 million in upgrades instead of spending over $108 million to replace the entire 19 mile pipeline.

VUNERABILITY ASSESSMENT

The events of September 11, 2001 heightened the public’s awareness of the security of the nation’s infrastructure to terrorist attacks. As a result, HR 3448, Title IV, Section 1433 was signed into law. HR 3448 requires that “each community water system serving a population of greater than 3,300 persons shall conduct an assessment of the vulnerability of its system to a terrorist attack or other intentional acts intended to substantially disrupt the ability of the system to provide a safe and reliable supply of drinking water.” HR 3448 mandates that water systems serving a population of 100,000 or more, complete vulnerability assessments (VA) by December 31, 2002.

The City of San Diego assembled a project team certified in the Sandia National Laboratories Risk Assessment Methodology for Water (RAM-W), which is a methodology acceptable to the U.S. EPA for this work. The first step of the VA was to divide the system into the three major supply systems that correspond to the treatment plant that is the primary supply for the area (Miramar (the northern system), Alvarado (the central system) and Otay (the southern system). The project team, along with select City staff, utilized this approach to identify essential facilities and assets within the City’s supply, transmission, treatment, and distribution system necessary to meet an overall mission objective. Site inspections of critical facilities were then completed to determine how an adversary might attack the asset and what current physical protection system elements are in place. Using the risk equation, relative risk values were then determined for each evaluated asset. The risks along with possible actions to help mitigate consequences of an attack or to improve the physical protection systems are contained in a confidential vulnerability report on file with the City.

As a result of the Vulnerability Assessment of its water system, the City of San Diego Water Department has been actively upgrading security at its facilities. A Security Advisory Group (SAG) comprised of CIP Engineering and Water Operations staff has worked together with a consultant certified RAM-W. Additional site inspections of critical facilities have been conducted to evaluate how an adversary might attack the asset and what protection system elements should be in place to deter, detect, and respond to an attack. SAG has implemented upgrades to essential facilities and assets within the City’s water supply, transmission, treatment, and distribution system. The ultimate goal is to provide a well-balanced risk for all critical assets and a prioritized implementation plan for security upgrades.

The Water Department has scored security upgrades as a top priority in its Capital Improvements Program. The objective is to deter, delay, and detect any malevolent acts towards the City’s water system including physical, biological, chemical, or cyber-related attacks. Deterrent measures include physical barriers such as k-rails and bollards for vehicles and fencing, razor wire, and signage for unauthorized personnel. Detection equipment includes cameras, intrusion alarms, sensor cables, laser and microwave technology. Delay measures can take many forms such as using shatterproof materials for windows and skylights. Card readers and electronic surveillance provide an added layer of security to protect employees in addition to the facilities where they work. The Security Advisory Group is actively pursuing State and Federal grants to provide matching funds in order to maximize ratepayer value for security upgrades.

UPDATE ON CIP PRIORITIZATION

Due to limited financial resources the CIP cannot immediately incorporate all projects identified during the initial development of the program and subsequent studies including the projects identified as a result of the seismic evaluation. A water rate case was developed in the Spring of 2002 to support a bond issuance for capital improvements from Fiscal Year 2003 through 2007. A prioritization process provides the documentation necessary for bond council, rating agencies, and future grant opportunities. Given that the needs of water utilities are constantly changing, a systematic, reproducible process helps to document priorities at any given time. This is extremely important when there is a difference between forecasted budgets and the actual funds available for capital improvements. The prioritization list is updated annually using the following approach.

The Water Department enlists a team of senior personnel from Operations, CIP Program Management, Finance and Planning, and Water Research & Development. The panel represents the diverse interests of the divisions within the department. However, all panel members have to possess a background in system operation, capital financing, and long term planning. The panel refines the prioritization criteria based on various water system needs. As the end-user of Water Department capital projects, Water Operations is given double voting power so that the planning and engineering functions do not outweigh the interests of long-term operations. After the top criteria are selected, the team reaches consensus on each of their relative weights.

Criteria & Weights

* Health & Safety 18 points * Regulatory Compliance 18 points * Operations Need/Reliability 18 points * Future Demand/Expansion 15 points * DHS Compliance 13 points * Project Status 9 points * O&M Cost Reduction 9 points

Each of the seven criteria is scored for each project on a scale of 1 to 10. The process uses the Simple Additive Weighting method. A total score for each project is obtained by multiplying the scale rating for each attribute value by the importance weight assigned to the attribute and then summing these products over all attributes. After the total scores are computed for each project, the projects with the highest score (highest weighted average) are the ones with the highest priority. The team is tasked with scoring over 150 different capital improvement projects. The merits of each project are discussed among the group before the project is scored. Many team members have specific projects that are vigorously defended. The relative importance of projects is sometimes debated. The end result is a fair process that balances the competitive needs of the different Divisions of the Water Department for a variety of water infrastructure upgrades.

The results of the prioritization are presented to the Water Executive Team (WET) for their review and final approval. The WET, made up of the Director and Deputy Directors, consider the political climate as well as prior commitments to City Council and other City Departments. Additional scenarios and alternatives are explored based on these considerations before the prioritization is finalized. Once the WET makes a final decision, the capital budget for the next fiscal year and the following 10 years is developed based on the project listing approved by the WET.

CONCLUSION

The City of San Diego Water Department is constantly faced with financial challenges to pay for upgrades to its system. The competing needs for new infrastructure, security upgrades, regulation driven treatment technologies and alternative sources of supply must all complete for limited funding. The City of San Diego does not have a separate capital improvements program for seismic upgrades to its water system. Assessing and managing risk is necessary in order to decide the most appropriate measures to improve the water system. The funding for improvements requires an assessment of competing needs based on condition assessment, vulnerability studies, hydraulic modeling, and a formal prioritization process. The CIP Planning and Engineering Sections work cooperatively with Water Operations to devise the best management practice for upgrading the City of San Diego’s water system. S3-4

“Case Study of a Backup System for Water in Kinki Region”

Presenter: Eizo Seki

(Japan Water Research Center, Japan)

Case Study of a Backup System for Water in Kinki Region

Akira SADO Water Supply Division Ministry of Health, Labor and Welfare Masahiro FUJIWARA Japan Water Research Center Tatsuyoshi FUJISHIRO Japan Water Research Center Eizo SEKI Japan Water Research Center

Abstract

Spurred by the Great Hanshin-Awaji Earthquake in January 1995, recognition was made of the necessity of linked cooperation within disaster prevention spaces, such as with wide-area disaster prevention bases beyond the prefectural borders. For a Tonankai/Nankai Earthquake which is highly likely to occur with the first half of the present century, too, the form of even safer and worry-free regions is called for through an orientation towards minimization of disaster damages and speedy emergency responses and recovery. Against the background, three governmental bodies - the Cabinet Office, the Ministry of Land, Infrastructure and Transport, and the Ministry of Health, Labor and Welfare - cooperatively performed a “Survey for the Establishment of Cooperative Policies and wide-area Disaster Prevention Bases within the Kinki Region.” Here, investigation was made of wide-area disaster prevention activities being performed, with consideration given for the unique municipal configuration, etc., of the Kinki region. This report is investigated within the Ministry of Health, Labor and Welfare Committee. The Japan Water Research Center (JWRC, a public foundation), as commissioned by the Ministry of Health, Labor and Welfare, performed an investigation regarding a desired backup system that would enable wide-area water sharing and circulation that goes beyond existing water utilities at the time of an emergency, with the goal of ensuring a stable supply of water at the time of a disaster.1)

1. Purpose to Establish of a Wide-Area Backup System and its Effects

The purpose of the establishment of a wide-area backup system is considered as follows: At the time of an emergency, the shared circulation of water among related utilities over a wide area, and the securement of large amounts of water and supply thereof as necessary to water utilities which have been subjected to damage.

The effects of the establishment of such a wide-area backup system are as follows. (1) Improvement of the backup system stability against earthquake

1 Although in the hypothesized earthquake, its damages are not expected to extend to the entire Kinki region, they are projected to extend over multiple metropolitan districts and prefectures. Wide-area cooperation would improve the backup system stability against earthquake. (2) Enable to use other water-systems in an emergency By the establishment of wide-area cooperation, share and circulation of water among water supply utilities enable damaged water supply utilities to use the water resources of other water-systems, that would be difficult to ensure on their own, in an emergency. (3) Enable to backup in other emergencies Wide-area backup system would not only be an effective means for an earthquake disaster, but also be used for other emergencies; a water shortage (drought), water-quality accidents or cryptosporidium, etc. (4) Decrease of demand fluctuations Furthermore backup-system would be applicable during the times other than emergencies; to decrease water demand fluctuations (just as interregional sales of electrical power take place during demand peaks in summer seasons), and to keep ability of supply water in case of renewal construction of main water treatment facilities(to avoid shutdown of the facilities). (Here, however, various issues would arise in tandem with such use, such as systematic problems regarding water-use rights, etc.)

2. Overview and Characteristics of Damages Occurring Due to the Great Hanshin-Awaji Earthquake

2.1 Characteristics of damages to water facilities

Looking at the disaster damage of the Hanshin-Awaji Earthquake as an example, much of the damage to concrete structures of water purification facilities and reservoir facilities, etc., occurred as water leakage resulting from the damage of expansion-joints, and these were especially characterized by their occurrence at old facilities constructed prior to the promulgation of the new anti- seismic design code, and in the case of poor ground conditions due to land-filling, etc.2) Nevertheless, since, in general, water facilities have as their sitting condition the existence of a good ground state, with the exception of only a portion of concrete-based structures, destructive-type damages did not occur. Conversely, since their functional requirements mean that poor ground conditions cannot be completed avoided for linear structures such as header and distribution pipes and conduits, etc., and especially pipelines, and due to the fact that joint structures for most pipelines were not made in consideration of the kind of earth-shaking that occurred in this earthquake, as well as other factors, there were numerous cases of damages occurring, such as pipe-body ruptures and

2 damages resulting from joint omissions and damage, etc. In the Hanshin Awaji Earthquake, the enormous number of distribution-pipe damage sites required the assistance of municipalities and other utilities other than those cities that directly experienced disaster-related damage; such assistance finally extended to a nationwide scale, and water-supply suspension lasted over a long period of time.

2.2 Impairments of the lives of citizens due to water-supply suspension

In those cases where there was long-term suspension of water supply due to disaster damages, in regards to the water required to maintain the lives of citizens and the water necessary for their daily lives, various methods were required to ensure obtainment, including: (1)Pre-purchased bottled water and reserve water within homes (leftover bathtub water); (2)The purchase of bottled water; (3)Reception into containers of water provided at emergency-relief water-supply sites, such as evacuation centers, etc. Further, even those citizens who would be sure to show considerable anger in regards to an unexpected water-supply suspension during ordinary times instead this time responded by surmising, “It can’t be helped, especially since all disaster victims are faced with the same circumstances,” and “We just have to put up with this.” Here, in addition to the fact that they were unable to enjoy the conveniences and pleasure of ordinary water-supply services during the suspension, there was also need for extra labor, such as the carrying of water and the waiting in lines, in addition to other free time-restraints. As the time to recovery became longer, there was increased sharpness in the complaints of citizens with suspended water supply; in the fifth week from the occurrence of the earthquake, distressful statements were heard, including, “It is extremely hard to participate in the transfer-to-containers water-distribution process,” and “I am tired of climbing the stairs of my apartment building,” etc. Other complaints were heard that were clear expressions of anger: “I’ve reached the limits of my patience,” and “Did you forget to ensure a water supply to my residence only?”3)

2.3 Emergency-response supply of water

Here, a summary is made of emergency-response water-supply activities performed by Kobe city directly after the earthquake3). (1)Since congested road traffic was the largest obstruction, water for supply- water tanker trucks, etc., was taken from fire hydrants located within respective water-service districts, thereby improving transport efficiency.

3 (2)As the number of supply-water tanker trucks and other equipment increased due to outside support, etc., more fine-tuned allocations were made; however, there was a limit of 10 liters to 20 liters per person per day. (3)In tandem with pipeline recovery, the emphasis was changed to point-site water supply, with the setting of temporary supply-water hydrants using firefighting water hydrants, the installation of temporary water-distribution pipes, and the securing of individual water-supply hydrants within residences. This reduced the need of citizens to carry water long distances.

2.4 Emergency-response recovery

There were great difficulties in the restoration of pipelines within disaster- impacted municipalities where destructive-type damage occurred to pipelines. Below, an overview of emergency-response recovery is presented, with the chief focus being the Kobe City Waterworks Bureau. Water supply from the Hanshin Water Supply Authority, which is the chief water source for water supply to Kobe City districts, was temporarily stopped on January 17, the day of the earthquake. From the 18th, when water supply was restored from the Hanshin Water Supply Authority, water was collected in distribution reservoirs, and an attempt was made to gradually restart water distribution. However, since “Major pipeline damage had occurred, and a large amount of water leakage was occurring,” water could not be delivered to terminals. Thus, the method was adopted such that water-leakage surveys were performed, and repairs were made wherever there was water leakage. Nevertheless, repair work did not progress efficiently, for the following reasons, among others: (1) The large numbers of leakage sites caused a decline in water pressure within pipelines, meaning that there was little leakage noise; (2) There were many fallen-down buildings and debris on roads, and the additional danger of earthquake aftershocks meant that it took a considerable amount of time to discover leakage sites; (3) Dirt and gravel that had fallen into pipeline breakage sites blocked existing cutoff valves, making it impossible to completely cut off water; (4) Water- supply pipe cutoff hydrants located under fallen-building debris could not be cut off, etc. Thereafter, nearly one month later, from February 19, a major increase in water received from the Hanshin Water Supply Authority became possible; restoration work was renewed, and there was an expansion of water-passage and supply districts. By the end of March, with the exception of a portion of the seaside area, emergency-response recovery work had been completed citywide4)5).

3. Stability Evaluation of the Current State of Water Facilities Installation

In terms of the current state of water supply facilities installation within the

4 Kinki region, using the current status of facilities and water resources, in order to determine to what extent water could be stably secured for which areas during a disaster, the area was divided up into 11 areas, and a evaluation was made regarding the stability of water supply thereto. Here will be an overview of the evaluation method, and results.

3.1 Overview of evaluation method

Fig. 3.1 and Table 3.1 show each area within the Kinki region investigated this time. Within these 11 areas, on the basis of 6 evaluation indicators shown in Table 3.2, point values were established for each area, and each indicator has weight for each area. Also, in regards to what type of functions would be performed for each area during a disaster, questionnaire- type surveillance was done to the main water utilities within the Kinki region, and comprehensive determinations of the relative importance for each area NN 50km were made. Fig. 3.1 Kinki Region Areas

Table 3.1 Determination of Areas Name Area A Outs City and beneficial local governments from Shiga Prefecture Water Supply Authority B Kyoto City C beneficial local governments from Kyoto Prefecture Water Supply Authority D Osaka City E beneficial local governments from Osaka Prefecture Water Supply Authority（ north side from Yamato River） F beneficial local governments from Osaka Prefecture Water Supply Authority（ south side from Yamato River） G Kobe City H beneficial local governments from Hanshin Water Supply Authority（ except Kobe City） I beneficial local governments from Hyogo Prefecture Water Supply Authority （ except G/H） J beneficial local governments from Nara Prefecture Water Supply Authority K Wakayama City

5 Table 3.2 Indicators for Water Supply Stability Evaluations Evaluation Item Evaluation Indicator Reserve quantity per person (m3/person) = Reserve quantity 1 Facility capacity totals (m3)/Water-supply population (persons) Current status of emergency Reserve quantity (m3/person) = Emergency reserve quantity 2 cutoff valve totals (m3)/Water-supply population (persons) installation Current pipeline Pipeline capacity density (m) = Pipeline capacity totals 3 installation (m)/Water-supply area (m2) status Level of quake- Quake-resistance rate (%) = (Pipeline with quake-resistant 4 resistance of joints (m) + Pipeline with ordinary joints (m) x 50%)/ Water- pipelines supply area (m2) Multiplication of water resources Amount of water resources within an area with water- 5 (Degree of risk withdrawal amounts of 0.5 m3/s and above spreading) Stability against As based on Ministry of Land, Infrastructure and Transport 6 liquefaction land-quantity value information

As a method for determining the relative importance of facilities, the analytic hierarchy process (AHP) was used. The AHP is a rational selection method that a decision-maker selects the optimal alternatives from multiple alternatives; this is an evaluation method, that evaluation categories are arranged in a hierarchy and mutual comparisons are made among these in a situation where it is difficult to make a direct comparison of the relative strengths and weaknesses of alternatives.

In Fig. 3.2, a hierarchy of evaluation categories, which is used to evaluate the appropriateness of facility settings, is shown.

Optimal Pipe-setting Sites

Current status Current Level of Multiplic Stability Facility of emergency pipeline quake- ation of against capacity cutoff valve installatio resistance of water installation n status pipelines resources liquefaction

A B C D E F G H I J K

Fig. 3.2 Hierarchy Diagram of Evaluation Items for Optimal Pipe-setting Sites

3.2 Evaluation results

6 Investigation results are shown in Table 3.3 and Fig. 3.3. From the investigation results, a stability evaluation of water supply of wide-area disaster prevention bases shows, in general, cities where large populations are concentrated—Osaka city (D), Kobe city (G), Nara Prefecture (J)—have high total scores, a tendency of high stability in the area was observed. This is thought to have been impacted by the fact that in urban areas, there is vigorous promotion of quake-resistance infrastructure creation for responding to disasters such as earthquakes, etc., and there is implementation of effective facility establishment. As for areas in urban peripheries—Outs city and Shiga Prefecture (A), Kyoto Prefecture (C), and Wakayama city (K)—, the results show low stability. Thus, in order to fulfill the functions of a broad-area network for water supply, it is essential that water-supply stability is improved via vigorous progress in quake-resistant infrastructure creation within water facilities of the water utilities which is located in a midway portion between urban areas and their peripheries, namely, Kyoto Prefecture(C), Osaka Prefecture (E, F) and the Hanshin Water Supply Authority with the exception of Kobe City (H).

Table 3.3 Evaluation Result Totals Current status Current Level of Multiplic Facility of emergency pipeline quake- ation of Liquefa Name Total capacity cutoff valve installatio resistance water ction installation n status of pipelines resources A 0.0137 0.0095 0.004 0.0033 0.0078 0.0013 0.040 B 0.0067 0.0019 0.0139 0.0468 0.0078 0.0013 0.078 C 0.0142 0.0097 0.0065 0.0038 0.0078 0.0013 0.043 D 0.0112 0.028 0.0592 0.0256 0.0078 0.0013 0.133 E 0.0242 0.0099 0.0273 0.025 0.0078 0.0048 0.099 F 0.0471 0.016 0.0098 0.0112 0.0078 0.0139 0.106 G 0.0256 0.0021 0.0098 0.0653 0.0234 0.0139 0.140 H 0.0045 0.0039 0.0366 0.0167 0.0078 0.0048 0.074 I 0.0256 0.0057 0.004 0.0071 0.0549 0.0139 0.111 J 0.0702 0.0311 0.0062 0.0071 0.018 0.0013 0.134 K 0.0065 0.0111 0.004 0.0109 0.0086 0.0013 0.042

0.160

0.140

0.120

0.100

0.080

0.060

0.040

0.020

0.000 ABCDEFGH I JK capacity emergency valve pipeline quake-resistance water resources liquefaction Fig. 3.3 Evaluation Result Totals

7 4. Investigation of the Effects of Wide-Area Backup System Installation

Here, in order to perform an investigation of the effects of wide-area backup system installation, in regards to the case where the wide-area connection pipes are installed as currently conceived, economical and stability evaluation of water supply in the case where an earthquake occurred was made.

4.1 Concept of wide area connection pipes

Wide area connection pipes, which are considered a key structural element of a wide-area backup system, we used to the facilities based on the basic concept of the Kansai Water Supply Research Committee5), and on the results of the questionnaire surveys of each water utility. It is assumed that hypothesized water-circulation (passage) capabilities is a mean velocity of 1.5 m/s. Candidate sites for connections are shown in Table 4.1 and Fig. 4.1.

Table 4.1 Candidate Sites for Connections Scale of Water- Connecting connection passage Connection sites water utilities pipe capabilities (1,000 m3/d) Kyoto City — Rakusai water-distribution site a Kyoto inflow pipe ø600—Otokuni water ø600×500m 37 Prefecture purification plant ø1000 Shinyamashina water Kyoto City— purification plant distribution ø 600 b Kyoto 37 pipe ø600—Uji water purification ×2,000m Prefecture plant conveying pipeø700 Shinyamashina water Kyoto City— purification plant distribution ø 500 c Kyoto 25 pipe ø600—Otokuni, Uji, Kitsu ×5,700m Prefecture ø700 Kyoto Kitsu water purification plant— ø 700 d Prefecture— Northern-portion adjustment-site 49 ×7,600m Nara Prefecture system conveying pipe ø700 Kyoto Otokuni water purification plant Prefecture— ø700—Shimamoto conveying pipe ø 500 e 25 Osaka ø500 ×4,800m Prefecture Kyoto Takigi intermediary pumping Prefecture— station ø700—Murano water ø 700 f 49 Osaka purification plant ø2200 ×10,600m Prefecture Osaka Shijonawate pumping station— ø 900 g Prefecture— Northern-portion adjustment-site 82 ×8,800m Nara Prefecture system conveying pipe ø900 Miryo pumping station— Osaka Western-portion adjustment-site ø 800 h Prefecture— 65 system, western-portion trunk ×8,500m Nara Prefecture line ø1200 i Osaka Takamiya connection pipe ø 1000×55m 200

8 Prefecture — Shimojima connection pipe ø 1000×38m 200 Osaka City Niwakubo water purification Osaka plant southward conveying pipes j Prefecture— ø1000, ø900—Toyono water ø 2000×20m 200 Osaka City purification plant east conveying pipes ø2000 x 2 lines Niwakubo water purification Osaka plant southward conveying pipe ø 800 k Prefecture — 65 ø800—Southern-portion trunk ×1,100m Osaka City line ø1500 Osaka Senri-to-Ikeda conveying pipe Prefecture — ø700—Nishitada branch line ø350 ø 500 l 25 Hyogo ×6,100m Prefecture Osaka City — Kunijima water purification Hanshin Water plant distribution trunk line ø 1200 m 150 Supply ø1200—Amagasaki water ×6,500m Authority purification plant Hanshin Water 5th phase Inagawa conveying pipe Supply ø1650—Tada system conveying n Authority— pipe ø800 ø 800×50m 65 Hyogo Prefecture Kobe City— Nishitarumi no. 2 elevated ø 800 o Hyogo distribution-water pond—Kande 65 ×10,100m Prefecture system conveying pipe ø700 (i), (n) are already installed pipe

a,b,c Hyogo Kyoto Kyoto Prefecture Prefecture City o l e,f n d i,i Hanshin m Kobe Water Osaka Osaka Nara City Supply City Prefecture Prefecture Authority j,k g,h

Fig 4.1 Schematic Diagram of Connecting-pipe Network (Broken line: Already installed pipe)

4.2 Evaluation of economical efficiency

In this section, as an economical evaluation of the previously described connection pipes, an investigation was performed regarding the impacts on supply water unit prices and cost effectiveness on the Great Awaji-Hanshin Earthquake as an example.

(1) Calculation of estimated project costs, and impacts on supply-water unit prices When the cost for installation of a practical infrastructure was calculated, as

9 shown in Table 4.2, it shows that expenses would be totally an amount of approximately ¥81.5 million, including expenses for connecting pipes, pumping stations, and land-site fees. On the basis of this project costs, a trial calculation of the impact on water supply cost under the following assumed conditions shows that it increases ¥1.9/ m3.

Trial-calculation conditions Condition 1: 1/4 of project costs were considered to be paid from national government grant and [national budget] general account capital, while the burden of water utility was considered to be 3/4. Condition 2: The financial source for the burden of water utilities was assumed to be entirely covered by bond issues; its annual interest rate was set at 3%. Condition 3: The residual ratio of tangible assets with the exception of land was set at 10%, and the period of deprecation useful life was set at 38 years. Condition 4: A maintenance and management expense of 0.5% of project costs was included. Condition 5: Annual revenue water was assumed as the actual annual revenue water of the eight concerned water utilities.

Also, one can imagine that, depending on the specific method used for allocating the expense-burden amounts among water utilities, major differences will occur in terms of the relative impacts on each utility. Here, the assumption was made that the expenses for each specific connecting pipeline would be shared only among those utilities which are in direct connection with each other. When the relative burden among such pairs is assumed to be 50%:50%, then the trial calculation results show a considerable diversity in relative burdens, as shown in Table 4.3.

Table 4.2 Wide-Area Connection Pipes, and Project Expenses for their Related Facilities

Project expenses (x ¥1 million) Connec- Pumping ground Connecting utilities Notes tion station rent total pipes Kyoto City̶Kyoto a 350 - - 350 - Prefecture Kyoto City̶Kyoto b 1,370 - - 1,370 - Prefecture Kyoto city̶Kyoto c 3,198 - - 3,198 - Prefecture Kyoto Prefecture ̶ d 6,018 2,810 526 9,354 - Nara Prefecture

10 Kyoto Prefecture ̶ e 2,694 410 134 3,238 - Osaka Prefecture Kyoto Prefecture ̶ f 8,388 2,860 348 11,596 - Osaka Prefecture Osaka Prefecture ̶ g 8,830 6,016 1,032 15,878 - Nara Prefecture Osaka Prefecture̶ h 7,585 1,794 354 9,733 - Nara Prefecture Osaka Prefecture̶ - - - - already i Osaka City - - - - installed pipe Osaka Prefecture̶ j 252 - - 252 - Osaka City Osaka Prefecture̶ k 999 - - 999 - Hyogo Prefecture Osaka City̶Hanshin l Water Supply 3,422 690 140 4,252 - Authority Hanshin Water Supply m Authority̶ 9,810 - - 9,810 - Hyogo Prefecture Kobe City̶Hyogo already n - - - Prefecture installed pipe Kyoto city̶Kyoto o 9,009 2,074 390 11,473 - Prefecture 61,925 16,654 81,503 -

Table 4.3 Impacts on Supply-water Unit Prices Amount of Increase Water utilities （\/m3) Kyoto City 0.6 Kyoto Prefecture 18.4 Nara Prefecture 9.1 Osaka Prefecture 1.7 Osaka City 0.6 Hanshin Water Supply Authority 0.8 Kobe City 1.3 Hyogo Prefecture 4.1 Total Area 1.9

From these results, it is apparent that an issue of major importance will be the research of allocation methods that will enable water utilities participating in backup system projects, and their users, to obtain a certain degree of benefits from their participation.

(2) Cost-effectiveness analysis For the cost-effectiveness analyses, evaluations were performed using the “fixed quantity-response method” according to the Japan Water Works Association (JWWA)6). As for the method of treating benefits, the amount of the reduction in water-suspension damage amounts resulting from the avoidance of water-source

11 limitations (reduction in the number of days to recovery) was added up. Trial calculation was made of benefits using as an example the delay to recovery of water supply system in Kobe City caused by water source insufficiencies due to the Hanshin-Awaji Earthquake. Fig 4.2 is a graph of hypothesized fluctuations in water- circulation rates in the case where there were no limitations on water sources. The total population with suspended water service in this portion rose to 2.7 million persons. In this case, at a suspended water damage amount sum-total using damage amounts of ¥33,000/person･day, total damages are equivalent to ¥89,100 million. Insufficient water amounts at that time are 111,000 m3 per day; however, if the connection pipe between Osaka City and the Hanshin Water Supply Authority existed (water circulation capability: 150,000 m3/day), then one can consider that this water- amount insufficiency could be well covered thereby. Here, when a disaster-outbreak frequency of one occurrence within an evaluation period of 50 years is hypothesized (for the sake of convenience), the cost-benefit ratio (B/C) becomes 3.1, leading to the inference that such would have sufficient positive effects.

800,000 Assumed water in a 200.0 水源水量の制約が無 200 800 case with い場合の推定配水量no water 180.0 700,000 source limitations最大不足水量 180 102,000m3／日 700 Insufficient water 160.0 600,000 160

/day) 140.0 3 600 1995年実績配水量Water distribution 140 500,000 Water passage amounts in 1995 120.0

/日）水源水量の制約が無 120

3 500 rate in a case 400,000 with noい場合の推定通水率 water 100.0 400 source limitations 100

配水量（m 80.0 300,000 Water passage 1995年実績通水率通水率［復旧率(%)］ rate in 1995 80 300 60.0 200,000 60 suspended断水人口の 40.0

200 water-supply:減少効果部分 2.7 rate (%) passage Water 100,000 40 million延べpersons270 万人 20.0 100 20 0 0.0 Water distribution (1,000 m (1,000 distribution Water 1/171月17日 1/24 1月24日 1月31日1/31 2月7日2/7 2月14日2/14 2月21日2/21 2月28日2/28 3月7日 3/7 3月14日3/14 3月21日 3/21 3月28日 3/28

Fig. 4.2 Example of Insufficient water amount (Kobe City)

4.3 Case studies of the situation of assumed earthquakes

For the multiple earthquakes under consideration at the Cabinet Office, from their estimated damages, required water quantities in time sequence were computed, and a stability evaluation was performed for assumed hypothesized wide-area connection pipe routes.

(1) Forecasts of required water quantities at the situation of earthquake As for required water quantities following earthquake outbreak, forecasts were made for each of the five categories below; totals were computed by added these all together. Table 4.3 is a hypothesized example of required water amounts in

12 the case of a Median Tectonic Line earthquake. Immediately after the earthquake, since water leakage amounts account for a relatively large portion, one recognizes that the control of leaking water occurring due to an earthquake is an important issue.

6,000 Domestic water Emergency water Water used for recovery 5,000 Fire demand Leakage volume Total 4,000

3,000

2,000 (1,000 m3 /day)

Required water amounts Required 1,000

- 0 7 14 21 28 Number of days after earthquake Fig. 4.3 Case Study Showing Results of Estimated Required Water Fluctuations

(a)Domestic water Here, computation was made by assuming that the water-supply amount per person (including for business use, etc.) just prior to the earthquake would be required. This water amount includes water-leakage amounts occurring from the ordinary [i.e., pre-earthquake] time. (b)Amount of emergency-response supply water Emergency-response water quantities for general purposes were calculated by multiplying required emergency-response supply water amounts per person per day (3 liters/person･day) for the population with suspended water supply. Further, as required emergency-response supply water amounts for medical institutions, this was estimated at 100 liters per day, and added to the amount. This amount, when converted to the general suspended-water population, is equivalent to 1.5 liters per day per person. 7) (c)Amount of water required for recovery work For water quantities required for recovery work, here it was hypothesized that water amounts three times the water-distribution pipe capacity would be required, for the tasks, in order, of water filling, discovery of damaged sites, water removal, repair work, and water for cleaning repaired pipes. (d)Amount of water required for firefighting purposes 8)9)10)11) Water quantities for firefighting usage were calculated by multiplying the number of hypothesized flame-outbreak fire incidents by the fire-fighting water

13 amount per single fire incident. (e)Amount of water leakage Here, “water leakage” was determined to be unavoidable occurrences whereby, in the process of expanding the sphere of water-circulation in tandem with the distribution pipe recovery work for waterways that have experienced major damage, due to various reasons such as insufficient water pressure, human- resources restraints, reduced cutoff capacity of cutoff valves, etc., complete discovery and repair of water-leakage sites is not possible. Here, the water-leakage amount per unit-extension of circulating-water distribution pipes as calculated for estimate water leakage amounts at Kobe city (218 m3/day/km) was used as the function for maximum surface speed at time of earthquake, and a water-leakage ratio was set for hypothesized earthquake intensities.

(2) Case study of sending of support water As a result of simulation of the water demand-supply balance for each hypothesized earthquake, within the period of post-disaster confusion, fault-zone quakes that would cause water insufficiencies exceeding 10% of facilities capacity were those along the Median Tectonic Line, in the Ikoma fault zone, the Uemachi fault zone, and the Hanaori fault zone. Among these, shown is one example regarding water-circulation-capability amounts and required support-water amounts (support-capability water amounts) within the recovery process for a Median Tectonic Line earthquake (Fig. 4.4). These results show that, for Osaka Prefecture, required support-water amounts for the first day after disaster-outbreak would be 537,000 m3/day. For this, support-capability water amounts from surrounding municipalities (cities, towns and villages) would be 25,000 m3/day from Hyogo Prefecture, 400,000 m3/day from Osaka City, and 25,000 m3/day from Kyoto Prefecture. Further, for the 324,000 m3/day of water quantities required for Nara Prefecture, support capability-water amounts would be 49,000 m3/day from Kyoto Prefecture and 82,000 m3/day from Osaka Prefecture. Here, in regards to the support capability-water amount of 82,000 m3/day from Osaka Prefecture to Nara Prefecture, one observes that a portion of this water quantity is passed via Hyogo Prefecture or Osaka City, which are located in the vicinity of Osaka Prefecture. In other words, for disaster-impacted Osaka Prefecture and Nara Prefecture, expected can be water shared from neighboring enterprises, and water shared from non-neighboring enterprises yet via neighboring enterprises; here, one learns that positive effects from a wide-area backup system could be realized via “billiard ball”-like effects.

14 Kyoto City Support capability water ＝ｍ 87,000 3/day

9,000 ｍ3/day

Osaka Kyoto Hyogo Prefecture 25,000 Prefecture 25,000 ｍ3/day ｍ3/day Prefecture Required Support Support capability Support capability water water 3 (passage (passage water ＝ｍ 79,000 /day ＝ｍ 537,000 3/day capability) capability) ＝ｍ 65,000 3/day （69%） 82,000 ｍ3 /day 49,000 400,000 (passage ｍ3/day ｍ3/day (passage capability) (passage capability) capability) Support capability Osaka City Nara Prefecture Required Support Hanshin Water Support capability Kobe City Supply Authority water water ＝ｍ 324,000 3/day water ＝ｍ 981,000 3/day ＝ 349,000 （40%）ｍ3/day

Planned connection Planned connection Already pipes installed pipes connection (Not used in this pipes (Used in this case) case)

Fig. 4.4 Results of Sending of Support Water

5. Conclusion

5.1 Methods for a stable, wide-area water-system water supply

(1) Realization of a wide-area backup system Towards the concrete establishment of a wide-area backup system, it is important for both the national government and water utilities to investigate such things as methods, procedures and costs for the establishment. Further, it is necessary to, and it would be effective to, improve the quake- resistance capabilities of water facilities within each water utility, to secure storage quantities for emergency-time uses, and to adopt other measures for securing facility capabilities that would ensure water supply at levels well above certain levels. (2) Stability improvements of water supplies From the results of the overview evaluation of the supply-stability characteristics of water-supply water, within those areas having low supply- stability characteristics, it will be necessary to maintain high levels of supply stability characteristics for trunk facilities.

5.2 Future issues

(1) Establishment of connection pipes between trunk facilities or lines As a prerequisite for the actual realization of the intended effects of a wide-area

15 backup system, it will be necessary to secure water-sharing capabilities and functions among water utilities. Especially water utilities that have not yet established connection pipes between trunk facilities or trunk lines within themselves should be undertaken and completed in a planned fashion on the basis of ensuring their effectiveness.

(2) Expenses allocation for system creation In regards to the allocation of expenses engendered in the establishment of a wide-area backup system, there exists no rational, reasonable, officially sanctioned method. Including an investigation of similar cases both within Japan and abroad, required is the launching of efforts to search for, and realize, a variety of approach methods.

(3) Water-distribution system with consideration of disaster-related aspects At the time of a large-scale earthquake, there is frequent occurrence of water- supply pipeline damage, and it will become difficult to rely solely on fire-fighting hydrants alone for the firefighting water required for the simultaneously occurring fire outbreaks. Therefore, for water supply that is based mainly on a water-distribution system for naturally flowing water, here it will be necessary to investigate response strategies, including the laying of large-capacity water- sending pipes having storage-capability functions, and/or the diffused establishment of water-storage tanks for earthquake responses, etc. There is a further need to investigate response strategies for pump-pressurized water- distribution systems, such as reduced-pressure supply water with the goal of controlling water-leaks, and the setting of water-supply limitations for general uses in order to secure water for firefighting purposes, etc.

(4) Establishment of an information system In order to achieve the goals of speedy (early-period) comprehension of disaster status, plus correct initial-mobilization responses and sharing of information among concerned utilities, etc., desirable is the creation of an earthquake disaster-prevention information system within the water supply, one that effectively utilizes geographical information systems and disaster-prevention information data bases and the like.

16 References 1) Japan Water Research Center: “Survey for the Establishment of Cooperative Policies and Wide-Area Disaster Prevention Bases within the Kinki Region” March 2003, March 2004 2) Japan Water Works Association (JWWA): “Regarding the Survey Concerning the State of Damage, etc., of Water Supply Facilities Due to Earthquake in the Southern Portion of Hyogo Prefecture (Part I)” Journal of JWWA, Vol. 65, No. 4, June 1996 3) Kobe City Water Supply Reconstruction Plan Investigation Committee: “Guidelines for Making the Kobe City Water Supply Earthquake Resistant” May 1995 4) Kobe City Waterworks Bureau: “The Great Hanshin-Awaji Earthquake: Records of Water Supply Recovery” Feb. 1996 5) Kansai Water Supply Research Committee: “One Concept concerning the Establishment of Connection Pipes for Use in an Earthquake Disaster, and Water Supply Earthquake Damage Prediction as Based on the Perspective of Local Citizens” March 1996 6) Japan Water Works Association (JWWA): “Manual for Cost-Effectiveness Analyses of Water Supply Projects (Draft)” March 2002 7) Dodo Junichi: “Damage to Water Facilities in Hyogo Prefecture and the Lessons Thereof” Japan Water Research Center, Technical Report No. 26 (Great Hanshin- Awaji Earthquake and Water Supply”, pp. 22-23 8) Kobe City Fire Bureau: “Record of Firefighting Activities (Kobe City Area) During the Great Hanshin-Awaji Earthquake Disaster” 9) K.Yasuno, K.Hayakawa: “Fires and Water Supply” Technical Report No. 34, p. 91, Japan Water Research Center 10) K.Yasuno, Y.Nanba: “Record of Firefighting Activities (Kobe City Area) During the Great Hanshin-Awaji Earthquake Disaster,” p. 159, Kobe City Waterworks Bureau 11) Central Disaster Management Council: “Expert Investigation Committee for Response Policies for a Tokai Earthquake” No. 6 (Jan. 9, 2003) Distribution Documentation 12) Y.Sorakuma:”A Study on the Development of a Backup System in a Big Urban Area”3rd US-Japan Work shop on Water System Seismic Practices, August 2003

17 S3-5

“UNIFORM CONFIDENCE HAZARD APPROACH FOR THE SEISMIC DESIGN OF PIPELINES”

Presenter: Craig A. Davis

(Los Angeles Department of Water and Power, USA)

UNIFORM CONFIDENCE HAZARD APPROACH FOR THE SEISMIC DESIGN OF PIPELINES

Craig A. Davis

ABSTRACT Earthquake hazards to water pipelines are not evaluated consistently. There is a need for a methodology to evaluate earthquake hazards such that all pipes in a water system are designed with a uniform confidence to withstand damage and consistent with their importance to post-earthquake response and recovery. Some methodologies presently exist that are useful for the seismic hazard evaluation and pipeline design. The present methodologies for pipeline evaluation and design can adequately account for the seismic hazards, but are not performed in a manner that provides consistency throughout a water system. There is also no uniform procedure for evaluating pipelines based on their importance to post-earthquake response and recovery. The purpose of this report is to present the initial development of a method for evaluating earthquake hazards, including transient and permanent ground movements, so that all pipes in a water system are designed consistent with their intended function and with a uniform confidence that the actual forces a pipe may experience during an earthquake are less than or equal to the design forces.

The American Lifelines Alliance is supporting a project to develop pipeline seismic design guidelines. This project is a direct outcome of information and communication that has transpired in the previous workshops on water system seismic practices sponsored by the Japan Water Works Association and the American Water Works Association Research Foundation. Some of the information presented in this report will be utilized in the new pipe design guidelines. Any comments or suggestions on the proposed approach are welcome and can be submitted to the author.

______C. Davis, Ph.D., PE, GE, Waterworks Engineer, Geotechnical Engineering Group, Los Angeles Department of Water and Power, 111 N. Hope Street, Room 1368, Los Angles, CA, 90051, [email protected].

1 INTRODUCTION Numerous pipelines serve different functions in every water system. In different areas of a water system, different pipes will be subjected to different seismic hazards. In addition, individual supply lines extending over long distances may be subjected to several different seismic hazards. Seismic hazards include ground shaking, ground failure, and tsunami. Ground failure hazards includes fault movement, landslide, permanent ground deformations in weak soils, and liquefaction related problems such as lateral spreading, bearing failure, and settlement. Floatation from pore pressure increases also poses a seismic hazard. The seismic design of pipelines must consider effects of the different hazards in a uniform manner. Current methodologies for assessing seismic hazards do not provide for a consistent design approach. There is no uniform approach to the design of network components subjected to different seismic sources having very different recurrence intervals. Most procedures currently in practice advocate scenario earthquakes as a design basis. Scenario earthquakes would typically require similar design loads for pipes crossing two different faults having different probabilities of generating the same level earthquakes, (e.g., one generating an earthquake on average of 1,000 years and the other on average of 10,000 years). Probabilistic methods can account for the different earthquake return periods, but typically are not used to derive the necessary parameters needed for the engineering evaluation of all seismic hazards; leading to non-uniform seismic design criteria. Thus, currently if a single pipe is designed for different hazards, it is likely that some of the hazards would be developed on a less conservative basis than others. There is no consistent approach for considering the pipes’ importance to the post- earthquake operations of a water system. Some pipes are inherently more important to supply and distribution in post-earthquake disaster conditions than others. It is apparent that pipes serving a more important function need to be designed with stricter seismic design criteria in order to ensure that the design value is equal to or greater than the actual seismic force. There is a need for uniform seismic design among all hazards along a pipe and all pipes in a system performing similar functions. To meet this need, a uniform confidence hazard evaluation methodology is presented for buried pipelines. PURPOSE FOR UNIFORM CONFIDENCE HAZARD EVALUATION The purpose of a uniform confidence hazard evaluation is to provide a methodology to assess earthquake hazards so that all pipes in a water system are designed, for life safety and economy, consistent with their intended function and with a uniform confidence (reliability) that the actual forces a pipe may experience during an earthquake are less than or equal to the design forces. The method also provides design uniformity between different water systems. Seismic hazard analyses are generally performed using deterministic (DSHA) or probabilistic (PSHA) methods. A DSHA considers the effects from a particular earthquake scenario on a pipe, but it cannot adequately account for uncertainties in the evaluation and does not account for the risks associated with the accumulation of all seismic sources potentially affecting the pipe. Water systems located in highly seismically active regions would necessitate multiple DSHA scenarios having different recurrence intervals and leading to inconsistent results in the pipeline design. A PSHA simultaneously considers the effects from multiple earthquake source hazards on a pipe and the probabilities of a likely range of magnitudes over the length of each seismogenic source. A PSHA accounts for uncertainties in the evaluation and the risks associated with the accumulation of all seismic sources potentially affecting the pipe. As a result, the PSHA is recommended for use to develop uniform confidence hazard analyses. There are no existing requirements or guidelines for the seismic design of pipelines. However, the American Lifelines Alliance (ALA) is currently supporting a project to provide pipeline seismic design guidelines [1], for which much of the uniform confidence hazard approach presented herein will be utilized. Existing building codes provide a useful basis for establishing a hazard evaluation, but unfortunately hazard evaluations for pipelines must consider

2 transient and permanent ground movements and are much more complicated than that for buildings, which only consider transient ground movements. Pipelines are linear facilities that makeup portion of a water supply and distribution system. Single pipelines may be exposed to several earthquake hazards and different pipelines in a system of different importance may be exposed to the same or different earthquake hazards. A water system provides service to a greater community and the importance of each pipe must consider the community as a whole. For example pipes providing fire suppression for large communities or a supply to an emergency facility (e.g., hospital) are inherently more important than that supplying irrigation water to recreational parks. As a result, the importance of each pipeline should be determined based on its needed performance following an earthquake. All pipelines of similar importance should be designed with a consistent set of parameters to establish a uniform confidence that each pipe will survive an earthquake. The acceptable pipeline risk exposure may be assessed using analogous community risk acceptance for similar important building facilities, which are documented in existing building codes. Each hazard for which a pipe may be exposed can then be evaluated, accounting for uncertainty in the earthquake parameters, and related to the acceptable risk. A primary benefit for establishing a uniform hazard approach is the improved capability of water pipelines to function and operate during and following design earthquakes. This is accomplished using a performance based design methodology that provides cost-effective solutions and alternatives to problems resulting from seismic hazards. Improved water pipeline performance will help create a more resilient community for post-earthquake recovery, which is the ultimate reason why water pipelines are considered for improvement; therefore portions of the uniform hazard approach inherently considers the community impacts if pipeline damage were to occur.

PIPE FUNCTION AND CONFIDENCE LEVEL

Pipe Function Different types of pipelines in water supply and distribution networks serve different functions. Aqueducts, trunk lines, mains, and service connections identify the pipe type. The pipe function is related to its importance in providing water supply to the community and individual facilities and may not have a direct correlation to the pipe type. Table I classifies pipes into four functions related to their importance in aiding a community in post-earthquake response and recovery. The ALA Commentary [1] provides additional guidance on how to classify pipes as Function I, II, III, or IV based on how critical they are and consequences of failure, with consideration of: the facilities they serve; importance to the community for fire fighting, health, and post-earthquake emergency response and recovery; potential for secondary disasters resulting from pipe damage or failure; difficulty in making repairs; effects on community socio-economics; and a pipes location and size considering disruption of emergency response or evacuation. Pipelines are essential for providing domestic water supplies to the community. As a result, pipelines are critical for helping communities recover from an earthquake and to help prevent secondary disasters, such as fire and disease, following an earthquake. It is also important to develop consistent seismic design criteria for the community as a whole; that is, on a conceptual level a pipeline need not be designed with greater seismic criteria than the facility(ies) in which is serves or for its intended post-earthquake use, nor should it be design to a lesser standard than what the community expects. The pipeline Function classification in Table I are consistent and analogous with building code definitions[2][3][4], which establish higher level seismic design criteria for facilities that are more important to the community. For example, pipelines servicing facilities defined as essential in the building code are similarly defined as essential in the methodology proposed herein.

Table I. Pipe Function Classifications. Function Seismic Importance Description I Very low to None Pipelines that represent very low hazard to human life in the event of failure. Not needed for post earthquake system performance, response, or recovery. II Ordinary, normal Normal and ordinary pipeline use, common pipelines in most water systems. All pipes not identified as Function I, III, or IV. III Critical Critical pipelines and appurtenances serving large numbers of customers and present a substantial hazard to human life and property in the event of failure. IV Essential Essential pipelines required for post-earthquake response and recovery and intended to remain functional and operational during and following a design earthquake.

Confidence Level For operational purposes, a pipeline has a minimum performance reliability following an earthquake. The need for operational reliability in any given pipe increases with increasing functional importance. For seismic design, the reliability of a pipe being operational following an earthquake is expressed in terms of a confidence level C defining the confidence that the pipe design parameters will not be exceeded. C is an indicator for the probability of not having pipe damage and is related to the probability P that a design parameter may be exceeded when the design earthquake occurs: 1−= PC . C increases and P decreases with increasing pipe functional importance. P is defined in relation to a time period t, where t identifies the time basis for facility design. A 50-year design basis is recommended and used for this evaluation procedure to be consistent with standard engineering practice, and P is related to the t = 50-year timeframe. Table II identifies the recommended design basis for each pipe function.

Table II. Recommended design basis for each Function. Function Probability of Confidence Return Period Exceedance P Level C T (years) in 50 years I Undefined Undefined Undefined II 10% 90% 475 III 5% 95% 975 IV 2% 98% 2475

The average return period T is shown in Table II and related to P and C through: =−−= − ln/)1ln(/ CtPtT . T identifies the average time between seismic hazard occurrences. For practical design purposes, P and C are more important than T because engineers are more concerned with a probability of a design parameter being exceeded during an earthquake than considering the time it takes for the hazard to recur. T is only presented in Table II for descriptive purposes because hazard parameters are often presented in terms of T and this parameter is useful for quantifying hazards in terms of a single number. For earthquake hazards, T is more directly related to geological and seismological factors than engineering factors and should be considered in relation to a geologic time scale rather than a facilities useful life. The design basis for Function II pipes is determined consistent with common codes [2][3] for building structures performing similar functions. The design basis for Function IV pipes is determined from a geologic earthquake recurrence interval of 2,475 years, which is considered a rare but possible event and one that must be considered for design of essential facilities [5]. The design basis defined for Function III pipes is considered a reasonable intermediate value between Function II and IV pipes. The seismic design basis for Function I pipes is undefined because

4 even pipes that are not specifically designed for earthquake effects hold an undefined level of seismic resistance. Other Function and Confidence Considerations It is recognized that the pipeline Function concept presented in Table I may be difficult to apply for some pipelines that are part of a system delivering water throughout large portions of a community. Especially where there are mixed facility types within the distribution area. This is mainly because water systems are developed to provide service within large blocks and not just to a single facility, and many times there will be different facility types of varying importance within the distribution zone. For this reason additional seismic design recommendations are developed considering pipeline redundancy, isolation, continuity, etc. In addition, water systems and facility uses are complicated and it is difficult to identify all variations of use with general guidelines and for these reasons these general Functions may not conceptually apply to all pipelines; for example if a critical or essential facility can provide a complete self supporting water supply following an earthquake without the need for any domestic supply though normal pipe distribution, then it is possible the pipe seismic design criteria can be altered from the recommendations presented herein. The pipe Function classification and corresponding seismic design confidence level are specific to individual water supply and distribution systems. The following subsections prove for customization of the proposed methodology for specific system conditions. They also allow owners to consider cost-effective options in water system seismic improvements through use of redundancies, isolation capabilities, emergency response, etc. as alternatives to hardening specific pipelines.

Multiple Use Pipelines and Continuity Pipelines providing water service for multiple uses are recommended to be classified under the highest functional use in Table I. Where pipe connections and branches come from a higher Function pipeline to serve a lower Function, the branch pipe is recommended to be designed as the higher Function or be equipped with isolation capabilities in the event of damage. In addition, pipelines and pipe systems should be designed for the higher Function for which service is provided from the supply and water treatment source to the point of service. This recommendation applies to all pipes providing supply to the facility, regardless of ownership.

Redundancy Redundant pipelines increase the reliability of post-earthquake operations, provided the redundancy meets the following criteria: 1. A leak or break in one pipe will not lead to damage on other redundant pipes, 2. All redundant pipes can provide a minimum needed flow to meet post-earthquake operational needs, 3. The redundant pipes are spatially separated by an adequate distance through potential ground deformation zones such that, should ground deformation occur, each redundant pipe may not be subjected to the same amount of ground movement due to the natural variation in movement across a deformation zone, regardless of the actual design parameters.

Table III. Function classification for redundant pipes. LR Function as defined in 01 2 Table I I I I I II II II II III III II II IV IV III II

5 Redundancy provides an increase in confidence, for example two redundant Function II pipes provide an overall 99% confidence level. It is therefore acceptable to reduce the seismic design criteria for redundant pipes. Table III presents the recommended reclassification of pipe Function based on the redundancy level LR.

Branch Lines and Isolation Supply and distribution pipelines often have other supply lines, distribution lines, and service connections branching from them. Post-earthquake reliability may be compromised in pipes having branching lines that are designed to a lower confidence level. To ensure post- earthquake operational reliability the following procedure is recommended for evaluating branch pipe design requirements and isolation capability. This procedure is only applicable to pipelines of a lower Function branching from pipes of a higher Function.

1. Determine the Function for the branch pipe using Table I. 2. Determine the Function of the pipe it is branching from. 3. Design the branch pipe for: a. The lower Function if isolation valves are installed or an engineering analysis is performed and shows the branch pipe(s) will not disrupt post-earthquake performance of the higher pipe Function. This evaluation must account for the cumulative effect of potential damage on all branch pipes. b. The higher function if (a) is not satisfied. 4. Perform a check on isolated portions of the system to ensure that there are no significant life safety issues if a temporary water system outage resulted due to pipe damage or isolation. If there is a concern, design the branch pipe for higher function. EARTHQUAKE HAZARDS The primary earthquake hazards of concern for water pipes are transient and permanent ground movements. Tsunami poses a hazard along coastal regions, especially for above ground pipes, but will not be addressed further in this report. Buoyancy may affect a pipeline where there is an increase in subsurface pore water pressure, especially in areas prone to liquefaction, but will also not be addressed in this report. Transient ground movement describes the shaking hazard by waves propagating from the energy source and the amplifications due to surface and near surface ground conditions and topography. Permanent ground movement describes the ground failures resulting from surface fault rupture, slope movements and landslides, shear deformations, liquefaction induced lateral spreading and flow failure, and differential settlement. Table IV summarizes the transient and permanent ground movement hazards that may damage water pipelines, the earthquake parameters needed, including peak ground acceleration (pga) and velocity (pgv), for an engineering evaluation for each hazard, recommended methods for obtaining the earthquake parameters, and geotechnical parameters needed for a proper engineering evaluation of the earthquake hazard. A good overview of earthquake hazards and effects on pipes is provided in [6]. Table IV shows that an understanding of geotechnical conditions along a pipe alignment is necessary for a hazard assessment. The level of geotechnical understanding necessary for a proper assessment is dependent upon the pipe Function; the more important Functions should have greater detailed geotechnical assessments. Buried pipe performance is largely governed by the induced ground strains. Transient ground strains are generally smaller than those from permanent ground deformation. Pipeline evaluations should consider all potential strain sources.

Table IV. Earthquake hazards and parameters needed for pipeline design. Hazard Earthquake Obtain from: Geotechnical Parameters Parameters Transient Ground Movement General Shaking pga, pgv, PSHA Soil/rock conditions, depth, spectral response Vs Near-source Fault distance PSHA, fault map Fault type, orientation, directivity rupture direction Ground pga, pgv , PSHA Site soil and rock amplification spectral response conditions, Vs Impedance pga, pgv PSHA Soil/rock interface boundaries conditions, depth, Vs Topographic pga PSHA Topography amplification Basin edge pga PSHA Basin subsurface geometry, soil & rock properties, source distance Ground Oscillation Acceleration time history PSHA, site specific Soil profile, strength, Vs, analysis groundwater Permanent Ground Movement Faulting Magnitude, length Deaggregate PSHA Fault type, orientation or geologist Tectonic Fault slip, magnitude, Deaggregate PSHA Fault type, orientation deformations distance and geologist Liquefaction pga, magnitude PSHA, deaggregate Soil type, relative density, thickness, groundwater Lateral spread and pga, magnitude, distance PSHA ,deaggregate Topography, soil type, Flow failure strength, thickness, groundwater Slope movement, pga, acceleration time PSHA Topography, ground landslide history strength, groundwater Settlement pga PSHA Soil type, strength, thickness, groundwater Shear deformations pga or pgv PSHA Soil type, strength, thickness, groundwater Ridge shattering Pga PSHA Topography, rock/soil properties, rock fractures & orientation

EARTHQUAKE PARAMETERS FOR HAZARD ANALYSES As shown in Table IV, nearly all of the Earthquake parameters needed for analyzing seismic hazards to a pipeline can be obtained from a probabilistic seismic hazard analysis (PSHA) and deaggreation of the PSHA [7][8]. The PSHA results provide a consistent set of seismic design parameters having a uniform confidence that each parameter will not be exceeded. PSHA can be performed with computer programs [9] or through the United States Geological Survey (USGS) interactive deaggregation web page http//eqint.cr.usgs.gov. A PSHA can be performed for 2, 5, and 10% probability of exceedance in 50 years for Function II, III, and IV pipes to evaluate pga and spectral accelerations at various frequencies. The pgv may not be standard output for common PSHA results, but can be estimated by factoring the 5% damped acceleration at 1 Hz (SA1) from:

7 2 = SAscmpgv 1 /1.1)/( π (1) derived from relations presented in [10][11]. As shown in Figure 1, PSHA deaggregation will provide results for mean and modal fault distance R, magnitude M, and ε representing the ground motion distribution in relation to M and R.

Figure 1. Probabilistic Seismic Hazard Deaggregation for site “Pipe Example” presented for pga with a 2% chance of exceedance in 50 years.

The PSHA results do not change over short distances and therefore only a limited number of locations need be evaluated for each pipe. The total number of PSHA evaluations is dependent upon the total pipe length and number of seismic hazards the pipe crosses. Consideration should also be given to performing a more detailed grid of sites near fault crossings, in landslide hazard zones, in liquefaction hazard zones, and areas suspect of large shear deformations. The PSHA does not account for all active or potentially active faults that pipelines may cross. When crossings are encountered that need evaluation, but the active fault is not included as part of the PSHA, an engineering geologist is recommended to evaluate the characteristic earthquake magnitude consistent with methods used for the PSHA. Alignment Specific Transient Ground Movement Design Parameters A ground motion evaluation along a pipeline should consider the transient ground movement effects presented in Table IV for all surface and subsurface conditions along the alignment. In some PSHA analyses attenuation relationships may be specified to match existing subsurface conditions [9]. In other cases the PSHA results will be given for a standard site condition (e.g., USGS results for NEHRP site class B). In the latter case site specific ground motion parameters, Fag for short period (e.g., pga) and Fvg for long period (e.g., pgv) motions, may be obtained by using the NEHRP site coefficients [2]. However, the NEHRP amplification factors do not have the same confidence levels as recommended in Table II [7] and additional research is necessary for adjusting the site coefficients. Topographic amplifications can be estimated by applying a factor of Fat=1.5 to 2.0 [12][13] for acceleration on steep ridges; Fat=1.0 for no topographic amplification. Velocity amplifications are considered negligible for topographic effects. Alternatively, site specific topographic evaluations are recommended. Amplifications for basin edge effects and ground oscillation require specific alignment evaluations. Pipelines located within 15 km of the seismic source may be subjected to near-source seismic shaking, resulting in significantly larger ground motion parameters and ground strain than

8 pipes located further away from the source. Near source factors Na and Nv from Table V are applied to Function III and IV pipes when R < 15 km and M > 6.5. Near source factors are applied because of the PSHA uncertainty in ground motions: 1. Near earthquake sources due to sparse data. 2. in the direction of fault rupture; they are know to be greater, 3. oriented normal to the fault plane; they are known to be larger than those oriented parallel to the fault plane 4. at sites on the hanging wall of non-vertical faults; they have greater motions.

Table V. Near-source factors Na and Nv (modified from [3]) Factor R ≤ 2 km 5 km 10 km R ≥ 15 km 1 Na 1.5 1.2 1.0 1.0 1 Nv 2.0 1.6 1.3 1.0 1Near source factors may be linearly interpolated .

The design peak transient movements, PGAD and PGVD, are determined from:

aatagD ∗∗∗= pgaNFFPGA (2)

vvgD ∗∗= pgvNFPGV (3) Alignment Specific Permanent Ground Movement Design Parameters Permanent ground movements (PGD) pose the greatest hazard to pipelines, even though they are more localized and usually present themselves with less exposure to pipelines than transient movements. The primary hazard results from the large ground strains associated with permanent movements, which is largest at the movement boundaries. For liquefaction, this occurs at the interface between liquefied and non-liquefied materials; for faulting it occurs at the primary trace of surface rupture; for landslides it occurs at slide boundaries; for settlement the greatest hazard results at locations of greatest differential settlement; and for ridge shattering it occurs when large differential horizontal movement manifests to the depth of pipeline. Table VI presents the recommended pipeline design parameters for fault, slope, and liquefaction induced lateral spreading permanent ground deformation hazards. All Function III and IV pipelines, including redundant pipes reclassified to Function II using Table III, crossing active faults, landslides, or liquefaction hazards are recommended to be designed for the movements in Table VI. Function II pipelines are recommended to be designed for the movements in Table VI or have the capability to be isolated from Function III and IV pipes in the event of a permanent ground movement.

Table VI. Recommended design movements for fault, slope, and lateral spread. Pipe Function Design Movement Fault Slope Lateral Spread1 (see Eq. 4) (see Eq. 7) (see Eq. 8)

II PGDF PGDS PGDL III 1.5* PGDF 1.6*PGDS 1.35*PGDL IV 2.3* PGDF 2.6*PGDS 1.5*PGDL 1Lateral spread design movements may be reduced, as presented in Table VII, for soils having lower susceptibility to liquefaction.

The greatest differential settlement hazards are usually also associated with slope or lateral spreading deformations, and may be identified as a proportion of the movement in Table VI. Differential settlements not associated with these hazards should be evaluated on an alignment specific basis, further research is necessary to develop a design basis having a uniform confidence as identified in Table II. Tectonic deformations associated with general ground

9 warping, compression folding, and extension usually occur over large distances resulting in low pipe strain, however alignment specific concerns should be evaluated. Permanent ground movements resulting from shear and ridge shattering require alignment specific evaluations. Shear deformations may occur in relatively flat ground when cyclic inertial loads exceeding the reduced effective soil strength in non-liquefied soils. Ridge shattering typically is not of concern to pipelines except for instances where there are large continuous vertical or near vertical fracture planes in a ridge allowing amplified ground motions to develop out-of-phase movements resulting in large transient and permanent ground strains [14]. The design movements in Table VI are conditional upon the triggering of fault rupture, slope movement, or liquefaction. The design coefficients account for the conditional probability and were determined by using mean deformations for Function II pipes. Movements for Function III and IV pipes were proportioned to Function II by assuming a maximum of 20% probability of the movement condition occurring, which allows Function III and IV coefficients to be approximated by a half and one standard deviation, respectively. As a result, the coefficients are related to the analysis method and will therefore change with different methods used for evaluating fault, slope, and lateral spread movements from those presented herein. The coefficients provide a uniform confidence between each hazard and the different pipe Functions. Regional mapping may be used for PGD assessment, and is utilized in the ALA Guidelines [1], but is not further described in this report. PGD maps should account for uniform confidence in a similar manner as used in Table VI and described above.

Fault Movement A fault is considered active if it has moved within the past 11,000 years and potentially active if it has moved within the past 190,000 years. All fault crossings are recommended to be considered along the pipeline, regardless of whether the fault was included in the ground shaking hazard evaluation, by designing pipes for active fault movements as shown in Table VI and considering for sympathetic movements on potentially active faults. An Engineering Geologist should be consulted for all fault activity determinations. The surface fault rupture hazard evaluation procedure needs to consider all fault crossings, determine movement orientation of each and if the faults are active of potentially active, estimate the expected surface displacement, and estimate the width in which movement is expected to occur. Most fault rupture occurs over a zone and not along a single line. Pipe designs need to account for the total fault rupture zone and distribution of fault rupture throughout the zone. The average surface fault displacements PGDF can be estimated from [15]:

()PGDLog F +−= 69.080.4 M [Log(σ F ) = 36.0 ] (4) where σ F is the standard deviation of average displacement regression. Equation 4 was developed from data regressions using a combination of strike-slip, normal, and reverse faulting. Displacements represent the vector sum of horizontal and vertical movements along fault strike.

Slope Movement The assessment of slope movement resulting from earthquake shaking first requires an assessment of the static slope stability factor of safety FS. The slope, soil or rock resisting shear strength, groundwater conditions, bedding, jointing, fracturing, and other pertinent factors depending on the slope conditions need to be considered. The critical acceleration at which slope movements initiate is determined from:

c ()FSa −= sin1 α (5) α is the thrust angle. The average coherent landslide induced permanent down-slope ground displacement PGDs in cm can be estimated from [16]:

10 ()PGDLog S 460.1546.1 ∗+= (ILog A )− 642.6 ∗ ac [Log(σ S ) = 409.0 ] (6)

σ S is the standard deviation of mean displacement regression and IA is the Arias intensity in m/sec, which is estimated from:

I A −+−= 21.4 ∗ ()RLogM (7)

Lateral Spreading Lateral spreading is the liquefaction induced down slope movement occurring when cyclic inertial loads exceed the reduced effective soil strength and is generally associated with shallow surface ground slopes (as low as a fraction of a percent slope). An assessment of the liquefaction potential must first be made followed by an evaluation for the possibility of a lateral spread along the pipe alignment. Preliminary regional assessment of soil susceptibility to liquefaction can be made based on geological characteristics, age, and deposition mode [17] and grouped as having a very high, high, moderate, low, or very low chance of liquefaction. Table VII presents the recommended design movements as a function of liquefaction susceptibility, and expands the design considerations beyond that presented in Table VI by considering a range of possible soil susceptibilities.

Table VII. Liquefaction induced permanent ground movement design recommendations. Design Movement for Liquefaction Susceptibility Pipe Very High High Moderate Low Function

II PGDL 0 0 0 1 III 1.35*PGDL PGDL PGDL near-source zone only 0 IV 1.5*PGDL 1.35*PGDL PGDL PGDL near-source zone only1 1 Design pipes for PGDL if Na>1.0 or Nv≥1.3. Otherwise recommend using PGDL = 0.

The average ground displacement PGDL in meters can be estimated from [18]: ()PGDLog + = +− ∗ M − 278.0017.1280.701.0 ∗ (RLog )− 026.0 ∗ R L (8) 497.0 ()WLog 454.0 ()SLog 558.0 ∗+∗+∗+ (TLog 15 )

W is the free-face ratio (%), S is the ground slope (%), and T15 it the total thickness of all liquefiable layers in meters having SPT blow counts of N < 15 blows per foot. Parameter descriptions and allowable range of data is provided in [18]. Geotechnical engineers and engineering geologists are recommended to be consulted for evaluation of potential liquefaction hazards. Use of Equation 8 indicates a certain amount of field investigations are necessary for proper liquefaction evaluations. Investigations may be difficult for some pipes, but proper mapping and initial reconnaissance can reduce the amount of field work needed. Proper investigations, with due consideration of the pipes importance, are necessary for an adequate hazard assessment. An improper assessment of groundwater, soil type, T15, etc. may greatly overestimate of underestimate the hazard. As an alternative to Equation 8 and Table VII, liquefaction and lateral ground movements may be evaluated with more advanced methods. CONCLUSIONS An approach for evaluating earthquake hazards to pipelines with uniform confidence has been presented. The procedure ensures consistency for assessing permanent and transient ground movements. The goal is to have a uniform confidence level for the design of a single pipe crossing many hazards and for all similar pipes subjected to similar hazards throughout a water system. Pipes serving similar functions will be designed similarly, with more stringent

11 design requirements applied to the more important pipes applied in uniform manner. Different earthquake hazards are evaluated for pipeline design with a uniform confidence that the design forces will not be exceeded in future earthquakes. The information presented herein represents an initial work in developing a uniform confidence hazard approach. Additional research is needed to refine the hazards considered and to add additional hazards to the approach. ACKNOWLEDGMENTS The support from the Los Angeles Department of Water and Power and the American Lifelines Alliance under NIBS contract G&E-1-2004 are gratefully acknowledged. The guidance and input from J. Eidinger, D. Honegger, B. Maison, M. Connor, M. Matson, and L. Cheng in developing the evaluation procedure is acknowledged. J. Hu is acknowledged for providing lateral spread ground deformation analysis and confidence and redundancy evaluations. REFERENCES [1] American Lifelines Alliance (ALA), 2005 “Design Guidelines for Seismic Resistant Water Pipeline Installations,” FEMA and national Institute of Building Sciences, draft. http://homepage.mac.com/eidinger/ [2] Building Seismic Safety Council (BSSC), 1998, “1997 NEHRP Recommended Provisions for Seismic Regulations for New Buildings and Other Structures,” Part 1 (Provisions) and Part II (Commentary), FEMA 302/303, Washington D.C. [3] International Code Council (ICC), 2003, “International Building Code” Building Officials and Code Administrators International, Inc., Internationals Conference of Building Officials, and Southern Building Code Congress Internationals Inc., Birmingham, Al. [4] International Conference of Building Officials, 1997, “Uniform Building Code,” Whittier, CA. [5] Leyendecker, E. V., R. J., Hunt, A. D., Frankel, and K. S. Rukstales, 2000, “Development of Maximum Considered Earthquake Ground Motion Maps,” Earthquake Spectra, Earthquake Engineering Research Institute, Vol. 16, No. 1, pp. 21-40. [6] Bird, J., T. O’Rourke, T. Bracegirdle, J. Bommer, and I. Tromans, 2004, “A Framework for Assessing Earthquake Hazards for Major Pipelines,” Proceedings Intl Conf. on Terain and Geohazard Challenges facing Offshore Oil and Gas Pipelines, Institution of Civil Engineers, London, UK. [7] McGuire, R., 2004, “Seismic Hazard Analysis,” MNO-10, Earthquake Engineering Research Institute, Oakland, CA. [8] Bazzurro, P., and C. Cornell, 1999, Disaggregation of Seismic Hazard,” Bull. Seism, Soc. Am., 89, pp. 501-520. [9] Blake, T. F., 1995, “FRISKSP A Computer Program for the Probabilistic Estimation of Peak Acceleration and Uniform hazard Spectra using 3-D Faults as Earthquake Sources,” Computer Services and Software, Newbury Park, CA. [10] Newmark, N. and W. Hall, 1982, “Earthquake Spectra and Design,” MNO-3, Earthquake Engineering Research Institute, Oakland, CA. [11] Naeim, F., and J. Anderson, 1993, Classification and Evaluation of Earthquake Records for Design,” 1993 NEHRP Professional Fellowship Report, Earthquake Engineering Research Institute, Oakland. [12] Bouchon, M, 1973, “Effect of Topography on Surface Motion,” Bull. Seism, Soc. Am., 63, 3, 615- 632. [13] Yuan, X, and F. Men, 1992, “Scattering of Plane SH Waves by a Semi-Cylindrical Hill,” Earthquake Engineering and Structural Dynamics, Vol. 21, pp. 1091-1098. [14] Davis, C. A. and S. R. Cole, 1999, “Seismic Performance of the Second Los Angeles Aqueduct at Terminal Hill,” Proc. 5th U.S. Conf. on Lifeline Eq. Engr., ASCE, Seattle, Aug., pp. 452-461. [15] Wells, D., K. Coppersmith, 1994, “New Emperical Relationships among Magnitude, Rupture Length, Rupture Width, Rupture Area, and Surface Displacement, Bull. Seism, Soc. Am., 84, No.4, 974-1002. [16] Jibson, R., 1994, “Predicting Earthquake-Induced Landslide displacement using Newmark’s Sliding Block Analysis,” TRR 1411, Transp. Res. Board, National Academy Press, Washington, D.C., 9-17. [17] Youd, T. L., and D. M. Perkins, 1978, “Mapping Liquefiable Ground Failure Potential,” Journal of the Geotechnical Engineering Division, ASCE, 104, 4, 433-445.

12 [18] Bardet, J. P., T. Tobita, J. Hu, and N. Mace, 2002, “Large Scale Modeling of Liquefaction-Induced Ground Deformation, Earthquake Spectra, EERI, Vol. 18.

Session 4: Seismic Performance

S4-1 “Measures Against Active Faults for Distribution Trunk Lines and Seismic Observations” Presenter: Takashi Furuya (Yokosuka City Waterworks and Sewerage Bureau)Japan

S4-2 “Water System Seismic Performance, 1994 Northridge－1995 Kobe Earthquakes” Presenter: Le Val Lund (Los Angeles Department of Water and Power) USA

S4-1

“Measures against Active Faults for Distribution Trunk Lines and Seismic Observations”

Presenter: Takashi Furuya

(Yokosuka City Waterworks and Sewerage Bureau, Japan)

Measures against Active Faults for Distribution Trunk Lines

and Seismic Observations

Takashi Furuya

ABSTRACT

Yokosuka City is located in the middle of the Miura Peninsula in the southeast of Kanagawa Prefecture, facing Tokyo Bay to the east and Sagami Bay to the west, and has an area of about 100.67 square kilometers and a population of about 430,000. The Southern-Kanto district, of which Yokosuka City forms a part, is a region subject to frequent earthquakes compared to other regions in Japan, as it is situated on a boundary point where three plates meet (the Pacific Plate, the Philippine Sea Plate and the Eurasia Plate). This paper describes measures against active faults taken for the Otawa distribution trunk line, which passes through one of the active faults within the city, the Takeyama Fault, and seismic observations at the point passing through the fault.

Takashi Furuya, Engineer, Planning Division, Yokosuka City Waterworks and Sewerage Bureau 11 Ogawachou , Yokosuka City, Kanagawa, Japan 1. SEISMIC MEASURES FOR WATERWORKS IN YOKOSUKA CITY

1.1. Anticipated earthquakes

The Great Kanto Earthquake of September 1, 1923 (M. 7.9) caused more extensive damage to Yokosuka than any other earthquake that Yokosuka has experienced since it was designated as a city in 1907. In that earthquake, as Yokosuka was in the area near the epicenter, about 70% of its waterworks facilities were damaged and emergency water supply was needed for a long period of time. Yokosuka has the possibility of suffering earthquakes that would cause damage (damaging earthquakes), such as (1) earthquakes with epicenters on the Sagami Trough (ones similar to the Great Kanto Earthquake), (2) earthquakes known as Tokai earthquakes with epicenters on the Suruga Trough, and (3) near-field earthquakes with epicenters within plates. As Yokosuka has, within its limits, the Kitatake Fault, the Takeyama Fault and the Kinugasa Fault, all of which are considered as active faults with high activity[1], Yokosuka has the possibility of suffering near-field earthquakes with epicenters in these faults in the future.

1.2. Necessity for Seismic Measures

As Yokosuka’s waterworks facilities are located near large cities such as Yokohama City, Kawasaki City and the Tokyo Metropolitan Area and situated on a peninsula, it would take a great deal of time to receive assistance and supplies of emergency materials from other cities should a damaging earthquake occur. Furthermore, as more than 99% of the total water sources come from rivers situated far from Yokosuka—the Sagami River and the Sakawa River—and less than 1% comes from within the city limits, the water supply would inevitably be cut off for a long period of time if conveyance facilities or transmission facilities were damaged. These facts have also led to the debate on the importance of seismic measures.

YOKOSUKA CITY

KINUGASA Fault KITATAKE Fault

TAKEYAMA Fault

MIURA CITY

Fig.1-1. Yokosuka City Fig.1-2. Active faults in Yokosuka The Yokosuka City Waterworks Bureau prepared the “Yokosuka City Waterworks Bureau Disaster Countermeasures Guidelines” in 1978 and started to take full-fledged measures against disasters. As a basic policy, the Waterworks Bureau established the “Five Key Strategies for Seismic Measures” and started to take concrete actions according to the strategies.[2] As part of these countermeasures, measures against active faults were taken for the Otawa distribution trunk line, which passes through the Takeyama Fault.

HEMI Filtration Plant VERNY PARK

SHITAMACHI KOYABE

OTAWA

Fig.1-3. Utilization of distribution trunk lines

Table 1-1. Five Key Strategies for Seismic Measures Contents Major measures Reinforcement of Major transmission/distribution mains - Construction to increase transmission/distribution reinforced for earthquake resistance. earthquake resistance of the mains Arima water transmission system - Construction to fortify distribution trunk lines against active faults Development of new Two wells for emergency use constructed in - Development of water supply emergency water sources the city. sources for emergency use at Hashirimizu and Hayashi Securing drinking water Water storage facilities installed for - Shitamachi Distribution trunk purpose of securing as much drinking water lines as possible within the city. - 100m3 Tanks (Water storage equipment for emergency use directly connected to water pipes) - Installation of emergency stop valves in distribution reservoirs Establishment of water Emergency restoration plan and emergency - Preparation of necessary supply system water supply plan developed. mechanical equipment and Necessary equipment and materials resources prepared. - Development of information materials on facilities Clarification of roles Emergency roles of individual employees - Creation of operation manual of the bureau clarified. and implementation of emergency drill - Motorbike investigation group 2. STUDY OF THE TAKEYAMA FAULT

2.1. Location and Form

The Takeyama Fault is a right strike-slip fault extending from Akiya in Yokosuka City to Tsukui in Yokosuka City in the direction of WNW-ESE. The confirmed length of the fault is about 9 km on land, and the fault may extend into the ocean. The Takeyama Fault consists of four fault lines (eastern part of the Takeyama Fault, western part of the Takeyama Fault, northern Takeyama Fault, and southern Takeyama Fault). In view of the continuity and the articulation of geography, two lines in the eastern and western parts of the Takeyama Fault are the main faults of “CertaintyⅠ”and of the degree of “Activity A”.[1]

2.2. Implementation of a Trenching Study

2.2.1. Purpose of and Preparation for the Trenching Study

Yokosuka City implemented a trenching study on the Takeyama Fault in 1997. The purpose of this study was to estimate the detailed location of the fault, the timing of the latest activity, the history of past activities, and the timing of earthquake reoccurrences to use these estimations as basic data for a disaster prevention plan.[3] We conducted the following to determine the locations of trenches. 1) We extracted the region where the Takeyama Fault passes through, based on documents and interpretations of aerial photographs. We then implemented a surface geology survey within the range of about 500 m north and south of the fault (the survey area was about 9 km2). 2) We collected existing materials on boring, and conducted a new survey through boring at one location to estimate the location of the fault and the underground structure.

Fig.2-1. Conditions of trenching in the Tsukui district 2.2.2. Details of the Trenching Study

To gain a highly accurate assessment of the timings of fault activities, it is necessary to select places where fairly recent sediments exist as locations for the trenching study. For this study, we therefore selected two locations (Tsukui district and Sugaruya district), and conducted the study. Table 2-1 shows the scale of trenching. In the Tsukui district, horizontal trenches were excavated additionally along the fault lines observed at Trenches No.1 and No.2, for the purpose of measuring the horizontal displacement of the fault. We estimated the timings of activities by applying the Carbon-14 dating method to the collected samples.

No.１ Trench

No.２ Trench

Horizontal Trench Horizontal Trench（2－E）（） center Table 2-1. Scale of the trenching study

Length（ｍ） Width（ｍ） Depth（ｍ） Fa -Fault TSUKUI District Trench No．1 25.5 5～12 2～4.2 Fb -Fault No．21053 Horizontal Trench (Center) 5.7 2.2 0.8 Horizontal Trench Horizontal Trench (１－W ) 4 3 1.5 Horizontal Trench (２－E ) 4.1 3 1.6 （1－W） Fc -Fault SUGARUYA District Trench 16 6 3～3.5 0 5 10m

Fig.2-2. Plan view of trenching in the Tsukui district

2.2.3. Results of the Trenching Study

As a result of the study, we confirmed that there had been three fault activities at the Takeyama Fault from 5,600 years ago up to the present, and could figure out the ages of these activities. Calling the three fault activities I, II, and III in the reverse chronological order, I occurred from 2,000 to 2,200 years ago, II from 2,200 to 4,600 years ago (II may have occurred from 2,200 to 3,000 years ago) and III from 4,700 to 5,300 years ago. Incidentally, the periodicity of activities is unclear because the age of II could not be narrowed down. However, the maximum interval of activities in the past was estimated to be around 3,000 years with uncertainty, and the average interval of reoccurrence was estimated to be around 2,000 years. It has become clear that there is the possibility that activity will occur in the near future since more than 2,000 years have already passed since fault activity I. In addition, displacement was studied on Fa-Fault at the Horizontal Trench (center), on Fb-Fault at the Horizontal Trench (1-W), and on Fa-Fault and Fc-Fault at the Horizontal Trench (2-E). The vertical displacement was 0.4 m, and the horizontal displacement was right strike-slip of around 2 m. These displacements are due to one or two earthquakes, so the displacement due to one earthquake is the half of these values if two earthquakes had occurred.

Fig.2-3. Sketch of the trenching situation

Table 2-2. Results of the trenching study

Layers Name 14C date （y．B．P) Age of fault activities and interval of activities A 1530±50 B1 2010±60、1960±70 B2 2180±60 Activity Ⅰ （2000～2200 y．B．P) B3 2140±60 B4 2310±50 Activity Ⅱ B5 2580±90 （＞2200～＜4600 y．B．P) B6 2860±60、2920±70 C1 - D1 4740±70、4650±60 D2 - Activity Ⅲ E1 5350±70、4750±70 （4700～5300 y．B．P) E2 5070±80 E3 (Top ) 4710±50 E3 (Lower ) 5610±70 E3 (Lowest ) 5530±50 E4 - [y.B.P] : Indicating how many years ago from the base year (1950)

Observed Faults Observed Displacement Activity Frequency

Horizontal Trench (Center) Fa - Fault 1.1m～1.3m right strike-slip Once or Twice

Horizontal Trench (１－W) Fb - Fault -

2m right strike-slip Horizontal Trench (２－E) Fa - Fault、Fc - Fault Once or Twice 0.4m vertical uplift on the north side

2.3. Evaluation of the Takeyama Fault

There are many study reports on the Takeyama Fault, including those prepared by Yokosuka City. In 2002, the Earthquake Research Committee (belongs to the Ministry of Education, Culture, Sports, Science and Technology) published the long-term evaluation of faults on the Miura Peninsula, including the Takeyama Fault.[4] According to the evaluation, the Takeyama Fault belongs to a group of active faults with the high possibility of occurrence of an earthquake in the next 30 years among major active faults in Japan. The committee estimated the probability of occurrence of an earthquake of magnitude 6.5 or greater.

Table 2-3 Evaluation of the Takeyama Fault Item Characteristic Length About 11 km or longer Timings of past Activity 1 (latest activity): Between about 2,300 years ago and about 1,900 years ago activities Activity 2 (previous activity): Between about 3,100 years ago and about 2,300 years ago Activity 3 (activity before the previous activity): Between about 5,600 years ago and about 5,400 years ago

It is estimated that the emergence of a fault extending around 1 km due to the Great Kanto Earthquake of 1923 was not an activity inherent in the Takeyama Fault. Displacement by Right strike-slip: around 1m or more one activity Average interval of Around 1,600 to 1,900 years activities Scale of envisioned Around magnitude 6.5 or greater earthquake Probability of In the next 30 years 6-11% occurrence of an In the next 50 years 9-20% earthquake In the next 100 years 20-30% In the next 300 years 50-70%

3. MEASURES AGAINST ACTIVE FAULTS FOR THE OTAWA DISTRIBUTION TRUNK LINE

3.1. Otawa Distribution Trunk Line

For waterworks in Yokosuka City, the 9th expansion project was carried out from FY 1992 to FY 2000 for the purpose of using water from Miyagase Dam, which was completed in FY 2000. As one of the major operations in the project, the Otawa distribution trunk line (with the bore diameter of 1,350-1,000 mm and with total length of about 6.3 km of which about 2.2 km is yet to be undertaken) was constructed in cooperation with the Miura City Waterworks Department. About 2.8 km of this distribution trunk line was made with steel pipe with the bore diameter of 1,200 mm and was installed in a duct shared with a sewage pipe (shield tunnel). Since the installation route is bound to pass through the Takeyama Fault, we installed flexible pipes that can follow fault displacement during an expected earthquake, as a measure against active faults.[5] 3.2. Setting of Flexibility

We calculated estimated displacement due to fault activities by using an estimated formula since this project was implemented before the trenching study mentioned in 2.2. was conducted.[1] Assuming the length of the fault as L (km), displacement due to one activity as d (m), and the magnitude of an earthquake as M, they are related as indicated by the following formulas:

log10 L = 0.6M - 2.9, log10 d = 0.6M - 4.0. Assuming that the length of the Takeyama Fault, including the undersea part, is 20 km (L), the magnitude of the earthquake is estimated to be magnitude 7 and the displacement is estimated to be around 1.5 m if a fault activity occurs at the Takeyama Fault. Thereby, we determined the design flexibility of the part of the Otawa distribution trunk line where the line passes through the fault as 1.6 m. It can be said that this displacement is at the same level as displacement measured by the trenching study implemented afterwards.

3.3. Implementation of construction

For this construction, a shield, of which the outer diameter is 3,480 mm and total length is 2,830 m, was propelled at a depth of 9 to 66 m by the mud water pressure shield method in 1994. After the tunnel was completed, a pipe for water supply was installed in the lower half of the shield segment with the internal diameter of 3,050 mm while a sewage pipe, cases of communication cables, and a passage for inspection were installed in the upper half thereof. Figure 3-2 is a geologic cross-section drawn based on shield-tunnel boring. To confirm the position where the shield tunnel passes through the fault, we analyzed the components of soil from boring. As a result, we confirmed a change from strong clay/crush zone (Mmd) to silt sand stone (Smd) and tuffaceous sand stone (Hsd). Flexible pipes were used for the pipe for water supply at the position where the pipe passes through the fault. A flexible pipe is a combination of a 10mm-thick stainless steel bellows pipe and

Geological type Mmd：Strong clay/crush zone Smd：Silt sand stone Msl：Sand soil Hsd：Tuffaceous sand stone

Tunnel

Earthquake observation site TAKEYAMA Fault NISHI Filtration Center TSUKUI Pump Station

Fig.3-1. Shield tunnel cross-section Fig.3-2. Geologic cross-section a steel pipe, and it assures capability to endure a deflection of 1,600 mm and an expansion or contraction of 80 mm. Flexible pipes were installed at seven points in total: three points on the faults which are believed to have been penetrated by shield construction (two points on the main fault and one point on the fault of the Great Kanto Earthquake of 1923), two points on a derived fault (a fault that may move at right angles to the main fault), and two points on the physical property boundary (where soil property changed during boring).

① Physical property boundary ② the Fault of the Great Kanto Earthquake ③ Main Fault ④ Physical property boundary ⑤ Derived Fault ⑥ Derived Fault ⑦ Main Fault ⑥ ⑤ ④ OTAWA ⑦ TAKEYAMA Fault Distribution ③ Trunk Line ② ①

Fig.3-4. Seismic Design of the Otawa distribution trunk line The Great Kanto Earthquake Fault at active faults Fig.3-3. Installed points of flexible pipes

4. IMPLEMENTATION OF SEISMIC OBSERVATIONS

4.1. Purpose of Observations

Since there are few seismic observations and cases to be considered in respect of the behavior of underground structures cutting across active faults during an earthquake, we installed seismic observation equipment in March 1996 and have been conducting observations ever since.[6] The purposes of the observations are (1) to understand the behavior of the shield tunnel due to an earthquake that occurs at the periphery of the Kanto area, (2) to figure out the conditions of activities of the Takeyama Fault, and (3) to give feedback for future anti-earthquake measures.

4.2. Layout of Observation Equipment

Out of three points where the shield tunnel crosses the Takeyama Fault, we selected one point where differences in geological conditions are clear, as an observation point. The width of the crush zone at this point is estimated to be about 10 m on the basis of the results of shield-tunnel boring. To clarify the behavior of both sides of the fault during an earthquake due to differences in geological conditions, seismic observation equipment was installed on five cross-sections in total, specifically, at the center of the fault, and at 10-m and 20-m upstream/downstream from the center. Three accelerometers in total were installed at the center of the fault and on two cross-sections at 20-m upstream/downstream from the center, for the purpose of observing responses during an earthquake in the direction of the axis of the tunnel, in the direction perpendicular to the axis, and in the upward and downward directions (at the center of the fault: two horizontal directions; at other two points: two horizontal directions and one vertical direction). Four displacement gauges in total were installed at two points on each of two cross-sections (at the center of the fault and at 10-m upstream from the center (segment connection)), for the purpose of measuring displacement in the direction of the axis of the tunnel between the fault and the crush zone. Four strain gauges in total were installed at two points on each of two cross-sections (at 10-m upstream from the fault and at 10-m downstream from the fault), for the purpose of measuring circumferential strain that acts on the segment on the rock and in the crush zone. Data observed by observation equipment are recorded with equipment installed at a neighboring sewage pump station through optical fiber cables laid within the shield tunnel. In addition, seismic records are automatically recorded onto IC cards, and the maximum value is transmitted from the pump station through optical fiber cables and indicated on the central monitoring board at the Nishi Filtration Center.

A ： Accelerometer S ： Stain gauge ※Observation Point ：⑦ D： Displacement gauge

S S S S D D D A D A A

NISHI Filtration Center Takeyama Fault TSUKUI Pump Station

Fig.4-1. Layout of seismic observation equipment

4.3. Seismic Observations and Response Behavior in the Direction of the Tunnel Cross-Section

The records of 235 earthquakes were obtained in the period from April 1996, when seismic observations started, up to the end of 2003. Through analysis of seismic waves obtained by seismic observations, it was found that vibration characteristic differs between points with different soil structures across the fault. Figure 4-2 shows a comparison of the maximum circumferential strain (position at 30°) on the rock and in the crush zone due to earthquakes shown in Table 4-1. Regarding earthquakes Nos. 10 and 13, the maximum strain is the same irrespective of observation point, but with these exceptions, the maximum strain is larger in the crush zone than on the rock. Out of observation records for earthquake No. 16, Figure 4-3 shows a comparison between the circumferential strain waves of the tunnel at the point 10 m away from the fault and spectra thereof. This also shows that the maximum strain is larger in the crush zone. Figure 4-4 shows acceleration waves on the rock and in the crush zone at the point 20 m away from the fault boundary and spectral ratio (transfer function) thereof. The peaks are at 18Hz and 6Hz, followed by 8.5Hz within the frequency region of 10Hz or lower. This indicates that vibratory behavior differs between the areas on the rock and in the crush zone.

Table 4-1. List of major earthquakes Epicen Acceler Displac Magn tral Strai 12 岩盤（右）

Year:date:tim a e ） No. Earthquake i distanc n Rock e tion ment tude e (μ) Crush破砕帯（右） zone (Gal) (cm) μ

(km) ( 10

1 Sagami-wann 96:07:10:23:06 4.1 12 2.5 2.0 0.001 2 Yamanashi-toubu 96:08:09:03:16 4.6 71 6.5 4.7 0.001 3 Chiba-toho-oki 96:09:11:11:37 6.2 148 6.6 13.8 0.006 8 4 Yamanashi-toubu 96:10:25:12:25 4.5 65 4.6 4.4 0.001 5 Chiba-nannbu 96:11:15:10:30 4.0 28 3.2 2.5 0.001 6 Boso-nanntou-oki 96:11:20:11:27 6.0 176 2.4 5.2 0.001 6 7 Boso-nanntou-oki 96:11:28:16:40 5.2 89 6.4 9.5 0.002 8 Ibaraki-nanbu 96:12:21:10:28 5.4 102 3.1 6.6 0.002 9 Kanagawa-toubu 97:01:14:15:53 3.6 47 9.6 8.5 0.002 4 10 Chiba-hokuseibu 97:07:09:18:36 5.0 60 4.2 1.8 0.005 Tokyo-wann 最大ひずみ（μ） 11 97:09:08:08:40 5.1 50 7.4 3.3 0.007 2

12 Chiba-hokuseibu 98:01:14:02:17 4.9 68 3.9 9.7 0.005 strain Maximum 13 Chiba-nannbu 98:01:16:10:57 4.6 59 11.2 3.2 0.005 14 Izu-hannto-toho-oki 98:04:26:07:37 4.7 50 3.0 5.4 0.005 0 15 Izu-hannto-toho-oki 98:05:03:11:09 5.7 51 4.7 8.9 0.005 0246810121416 16 Chiba-nannbu 98:05:16:03:45 4.8 38 14.1 10.8 0.006 17 Chiba-toho-oki 98:06:14:22:17 5.6 105 2.5 5.9 0.005 地震番号 18 Torishima-kinkai 98:08:20:15:40 7.1 701 2.3 - - 19 Tokyo-wann 98:08:29:08:46 5.1 57 14.8 - - Earthquakes No. 20 Chiba-hokuseibu 98:11:08:21:40 4.6 59 4.7 - - Fig.4-2. Maximum strain on the rock and in the crush zone

4.4. Analysis of Characteristics of Responses in the Direction of Tunnel Cross-Section

To find out the characteristics of earthquake responses in the zone subject to seismic observations, we prepared the 2-Dimensional Finite Element Model and clarified differences in responses.[7] Since we have not conducted the seismic observations of the basement at this observation point, we found an input acceleration wave by using seismic records that are simultaneously observed at the Hemi Filtration Plant about 9 km away from the point. Figure 4-5 shows circumferential strain response waves (position at 30°) on the rock/crush zone sides on the cross-section 10 m away from the fault. The maximum strain observed and the maximum analytical strain was 7.5μ and 1.5μ on the rock side and 7.9μ and 1.8μ on the crush zone side. The trend of slightly larger strain in the crush zone is also discovered from analysis. However, the results of analysis are around one-fourth of the observed values, and the method of setting a ground motion to be input in the basement and other matters require further detailed consideration. Figure 4-6 shows a comparison between observed acceleration/strain waves and the Fourier spectra of analytical waves. The spectrum amplitude is always larger for observations, but there are common spectral characteristics. This result proves that the analytical model was almost adequate, but that the magnitude of an input ground motion used for analysis was not appropriate. The collection of acceleration records at the basement during an earthquake is the future task.

8 6 5 μ）

( 4 岩盤 (R) 岩盤(R) 2 Rock 0 破砕帯(R) -2 】 4 Crush zone -4 -6 Rock Strain ひずみ（μ） -8 3 0 1020304050 Strain 【 Time時間 (sec) (SEC) 8 2 6 ひずみ振幅 4 破砕帯(R) μ）

( 2 0 1

-2 Amplitude -4 Crush zone -6 0 ひずみ（μ） -8

Strain 0 5 10 15 20 0 1020304050 時間 (SEC) Frequency振動数(Hz) (Hz) Time (sec) Fig.4-3. Strain waves and Fourier spectra observed on the rock and in the crush zone (No.16)

15 3.5 10 岩盤（Y) 破砕帯／岩盤(Y) 5 3 Crush zone / Rock 0 -5 】 2.5 -10 Rock 加速度(Gal) Acc. 【

Acceleration (gal) Acceleration -15 2 0 1020304050 TIME(sec) 1.5 15 10 破砕帯(Y) 加速度振幅比 1 5 Amplitude Ratio 0 0.5 -5 Crush zone -10 加速度(Gal) 0 -15 0 5 10 15 20 Acceleration (gal) Acceleration 0 1020304050 振動数(Hz) Frequency (Hz) TIME(sec) Fig.4-4. Acceleration waves observed on the rock and in the crush zone and transfer function (No.16) 2 1.5

μ） 1 岩盤(R30) ( 0.5 0 -0.5 -1 Rock

Strain -1.5 ひずみ（μ） -2 0 1020304050 TIME(sec) 2 1.5

μ） 1 破砕帯(L30) ( 0.5 0 -0.5 -1 Fig.4-5. Strain waves (Analysis) -1.5 Crush zone Strain ひずみ（μ） -2 0 1020304050 TIME(sec) Rock Crush zone Rock Crush zone 岩盤部破砕帯部岩盤部破砕帯部 4 5 5 5

】観測観測 3.5 観測】観測解析 4 解析 4 解析 4 解析 3 Acc. 】【 Acc. 2.5 】【 Acc. Strain 【 Strain 3 3 3 Strain 【 Strain 2 2 2 1.5 2 加速度振幅加速度振幅ひずみ振幅ひずみ振幅 1 Amplitude 1 Amplitude 1 1 Amplitude

0.5 Amplitude

0 0 0 0 0246810 0246810 0246810 0246810 振動数(Hz) 振動数(Hz) 振動数(Hz) 振動数(Hz) Frequency (Hz) Frequency (Hz) Frequency (Hz) Frequency (Hz) Observation Fig.4-6. Comparison of frequency characteristics (strain and acceleration) Analysis 4.5. Observation Results

Results obtained so far are summarized as follows. (1) Through observations of strain of the tunnel, it was found that in all observation records, the amount of strain is larger at the part of the tunnel where the tunnel passes through the crush zone. (2) Differences in responses between on the rock and in the crush zone were discovered through analysis of the 2-Dimensional Finite Element Model. (3) It was found that it is possible to elucidate observation records analytically by collecting acceleration records on the basement in the observation area. (4) In the past observations, there have been no records of earthquakes of which the epicenter is in the vicinity of the fault.

4.6. Reconstruction of the observation system

Due to recent progress in the real-time processing of seismic data, the delivery of information, and other IT technologies, it has become possible to reconstruct the observation system relatively easily. For these seismic observations, we constructed a system that delivers behavior during an earthquake to managers and other persons concerned by e-mail in real time and enables them to confirm behavior.[8] For example, regarding deformation of the tunnel during an earthquake, we adopted a system in which displacement waves are found by integrating acceleration records twice and damage to the tunnel can be evaluated by using the amount of strain obtained based on relative displacement between two points. Moreover, we improved the system to one which indicates the three-dimensional behavior of the tunnel by using observed seismic waves and thereby makes it possible to confirm the movements of structures during an earthquake and the evaluation of damage at a glance. Figure 4-7 is a screen displayed when a manager who received information about an earthquake views the data after a seismic observation system records the earthquake.

Fig.4-7. Display screen for seismic information 5. CONCLUSION

This paper described measures against active faults for the Otawa distribution trunk line passing through the Takeyama Fault and seismic observations at the point passing through the fault. There was not necessarily sufficient information about active faults when the construction of the Otawa distribution trunk line was planned, and the route through the Takeyama Fault, part of which moved due to the Great Kanto Earthquake of 1923, was selected on the assumption that it is safer than the route through the Kitatake Fault. Near-field earthquakes attributable to active faults have attracted attention since the Great Hanshin-Awaji Earthquake, and many active fault researches have been conducted nationwide. The Takeyama Fault was also re-evaluated through the trenching study and other studies, and it was found to fall into the category of faults requiring caution on a national basis. Although it is impossible to hold down seismic energy, we believe that we have taken proactive measures to be taken as anti-earthquake measures for a distribution trunk line. The future issue is how to use data obtained through observations for anti-earthquake measures in Yokosuka City in order to lessen earthquake damage.

REFERENCE [1] The Research Group for Active Faults of Japan, “Active Faults In Japan” (1991) [2] Furuya, “Yokosuka-Shi Suidōkyoku no saigai taisaku keikaku ni tsuite” (Yokosuka City Waterworks Bureau Disaster Prevention Plan), The Fourth International Symposium on Water Pipe Systems (1997) [3] Yokosuka City, “Heisei 9 nendo Takeyama dansō chōsa itaku hōkokusho” (Report on FY 1997 Takeyama Fault Consignment Research) (1998). [4] Earthquake Research Committee, “Miura hantō dansōgun no chōki hyōka ni tsuite” (Long-Term Evaluation of Faults on the Miura Peninsula) (2002). [5] Igari, “Katsudansō wo tsūka suru Ōtawa haisui kansen chikuzō kōji no hōkoku” (Report on the Laying of Distribution Trunk Lines in Ōtawa Passing Through Active Faults), 47th National Meeting for Reading of Waterworks Research Papers (1996) [6] Ohbo,Furuya,Takamatsu, “Dansō wo yokogiru Shield Tunnel deno jisin-ji kyōdō kannsoku” (Earthquake Observation at Shield Tunnel Crossing a Fault), Proc. of 24th Conference on Earthquake Engineering (1997) [7] Ohbo,Furuya,Takamatsu, “Dansō wo yokogiru Shield Tunnel no dannmen hōkō no jisinn ōtō tokusei” (Earthquake Response Characteristics of Shield Tunnel Crossing a Fault), Proc. of 25th Japan Symposium on Earthquake Engineering (1999) [8] Ohbo,Yamanobe,Kutsuna, “Doboku kōzōbutu no real-time sonsyō hyōka system no kaihatu” (Development of Real-time Damage Evaluation Systems for Civil Structure), Proceeding of the Symposium on application of real-time information in disasters management (2004) S4-2

“Water System Seismic Performance 1994 Northridge-1995 Kobe Earthquakes”

Presenter: Le Val Lund

(Los Angeles Department of Water and Power, USA)

Cover Page

Title: Water System Seismic Performance 1994 Northridge-1995 Kobe Earthquakes

Authors: Le Val Lund (Contact Person and Presenter) Civil Engineer 3245 Lowry Road Los Angeles, California, USA 90027 Phone: (323) 664 4432 Fax: (323) 664 4432 E-mail: [email protected]

Craig A. Davis (co-author) Waterworks Engineer, Geotechnical Engineering Group Los Angeles Department Water and Power Box 111, Room 1368 Los Angeles, California, USA 90051 Phone: (213) 367 0855 Fax: (213) 367 3792 E-mail: [email protected]

Martin L. Adams (co-author) Managing Water Utility Engineer, Water Quality & Operations Bus. Unit Los Angeles Department of Water and Power Box 111, Room 1449 Los Angeles, California, USA 90051 Phone: (213) 367 1014 Fax: (213) 367 1019 E-mail: [email protected]

Paper submitted to by November 11, 2004: Elizabeth Kawczyynski Senior Account Manager American Waterworks Association Research Foundation 6666 West Quincy Avenue Denver, CO80235-3098 USA Phone: (303) 347 6106 Fax: (303) 734 0196 E-mail: [email protected]

LL 11-15-04B

Water System Seismic Performance 1994 Northridge-1995 Kobe Earthquakes

Le Val Lund, P. E.1, Craig Davis, P. E., Ph.D2, Martin L. Adams, P. E.3

ABSTRACT

The 1994 Northridge earthquake caused considerable damage to the Water System in the urban San Fernando Valley of the City of Los Angeles, which is mostly served by gravity from the Los Angeles Aqueduct system. In the 1995 Kobe earthquake the City of Kobe also had damage to its Water System. The Hanshin Water Authority, a gravity aqueduct system, after pumping from the Yodo River, serves Kobe along with other cities. A summary is made of the earthquake performance of these two water systems.

1 Le Val Lund, P. E., Civil Engineer, 3245 Lowry Road, Los Angeles, CA 90027. 2Craig A. Davis P.E., Ph.D., Waterworks Engineer, Los Angeles Department of Water and Power, Box 111, Room 1368, Los Angeles, CA 90051. 3 Martin L. Adams, P.E., Managing Water Utility Engineer, Los Angeles Department of Water and Power, Box 111, Los Angeles, CA 90051

INTRODUCTION

The 1994 Northridge earthquake caused considerable damage to the water system in the urban San Fernando Valley of the City of Los Angeles, which is mostly served by gravity from the Los Angeles Aqueduct system. The aqueduct receives its supply from the snowmelt from the eastern Sierra Nevada. The higher level fringe area of the valley is served by pump-tank systems. The 1995 Kobe earthquake caused damage to their water system, located along the coastal area of Osaka Bay and is mainly served by a high level aqueduct system pumped from the Yodo Gawa from which the distribution system is fed by gravity to the city and the Port and Rokko Islands. A portion of the city, islands, and adjacent cities are located on land reclaimed from the Osaka Bay. Kobe also has pump-tank systems serving the Rokko Mountains, inland from the Bay. Significant damage occurred to older water mains in the Valley as well as water mains in Kobe especially in the areas subjected to liquefaction. The type of joint system had an impact on the performance of the pipelines. The pump tank systems had varying performances for different reasons in the two events. The emergency response and recovery activities varied because of the difference in concepts for emergency response planning and recovery. The difference was related to different infrastructure configuration, geotechnical siting, construction, social needs, political and financial philosophies. A summary will be made of the performance of these water systems in the two earthquakes.

EARTHQUAKES

The Northridge earthquake occurred on January 17, 1994 at 4:31 AM Pacific Standard Time. The hypocenter was about 32 km west-northwest of the Los Angeles Civic Center in the San Fernando Valley. The moment magnitude for the earthquake was Mw 6.7. The earthquake occurred on a blind south-southwest dipping thrust fault. The Kobe, sometimes referred to as the Hanshin-Awaji or Hyogo-ken Nambu, earthquake occurred on January 17, 1995 at 5:46 AM Japan time. The magnitude 7.2 (JMA), Mw 6.7, epicenter was located on Awaji Shima, 20 km southwest from downtown Kobe on the Nojima fault. Both earthquakes had numerous aftershocks.

TOPOGRAPHY

The San Fernando Valley of the City of Los Angeles is relatively flat surrounded by mountains on the north, west and south. The major watercourses are the Los Angeles River and the Tujunga Wash. Areas subject to liquefaction are generally along these and other smaller waterways. The Los Angeles Aqueduct system enters the Valley from the north. Kobe is located along the northerly shore of Osaka Bay. Kobe and other cities are located on a narrow strip of land at the foot of the Rokko Mountains adjacent to the Bay. Land for urban development has been created by landfill along the shore and for the offshore Port and Rokko Islands. These lands were subjected to liquefaction.

WATER SUPPLY

San Fernando Valley

The main water supply to the Valley is from the First and Second Los Angeles Aqueducts (LAA) from the slopes of the Eastern Sierra Nevada. Groundwater; Metropolitan Water District (MWD), a wholesaler; and water recycling also serve the Valley and other portions of the city. The water system is essentially a gravity system with higher elevation being served by pump-tank facilities. Most of the water in the valley comes from the LAA System with a small amount coming from groundwater. Los Angeles has one water treatment plant (WTP) for treating LAA water and numerous chlorinating stations for system disinfection. Los Angeles water system is designed for domestic supply and fire suppression. Los Angeles City has a resident population of 3.9 million and the San Fernando Valley has a population of over one million people.

Kobe

The main water supply for Kobe and other cities along Osaka Bay comes from the Yodo Gawa (River) and Lake Biwa. A pumping plant at the Yodo Gawa pumps water to three WTP of the Hanshin Water Authority, located in Amagasaki and Nishinomiya City. The Koutou Pumping Plant boosts the water from the Inagawa and Amagasaki WTP into two parallel open channel flow aqueducts (tunnels) operated by the Hanshin Water Authority, a wholesaler, which supplies purified water to Kobe and other cities. Other smaller streams and reservoirs also supply Kobe. Kobe receives 75% of its water from the Hanshin Water Authority. Kobe water system is essentially a gravity system with some higher elevations in the Rokko Mountains, approximately 15%, being served by pump-tank facilities. Kobe water system is designed for domestic supply and fire suppression. Kobe city has a population of 1.5 million.

AQUEDUCTS

Los Angeles Aqueduct System

The LAA system consists of two gravity aqueducts. The 1913 First Los Angeles Aqueduct (FLAA) consists of 3,050 millimeters (120 inch) diameter riveted steel and concrete sag pipes (inverted siphons) and concrete lined tunnels. Damage consisted of buckled and split bell and spigot joint on steel pipe and cracked concrete on sag pipes (See Figure 1). Only minor cracking occurred in the tunnels. Rewelding and welding external butt straps repaired the damaged joints of the steel pipes. Installing internal butt straps and grouting the cracks behind the butt straps repaired the concrete pipes. Mortar plastering repaired the tunnels. Temporary repairs to place the FLAA back in service were made in 58 days. Permanent repairs were made in the summer of 1994. Since this earthquake occurred in the winter, water demand was low and supply to the city was provided by redundant alternative sources: the Second Los Angeles Aqueduct (SLAA) following earthquake damage repair, MWD, surface and groundwater storage reservoirs. FLAA repairs weren’t high priority.

Figure 1. FLAA geyser and close up of split bell& spigot joint (DWP & Lund)

The 1970 SLAA suffered damage at Terminal Hill where the aqueduct enters the San Fernando Valley. The 1,950-mm (77-in) welded steel pipeline was mounted on ring girder supported piers. At several locations the pipe buckled at the bell and spigot joint and the mechanical coupling at the top of Terminal Hill pulled apart. The buckled joint did not leak and was left in place until the summer 1994 for permanent repairs. The mechanical coupling was replaced by a more elaborate long sleeve flexible coupling and the SLAA was placed back in service in 12 days. Water supply to the city during the wintertime demand was by redundant alternative sources: MWD and surface and groundwater storage reservoirs.

Hanshin Aqueduct System and Sengari Aqueduct

Hanshin Aqueduct System operated by the Hanshin Water Authority consists of two long concrete lined tunnels (approximately 2500 and 1800 mm (100 and 70 in) in diameter) in the Rokko Mountains supplied by a pumping station at Koutuo. These tunnels are approximately 90 m (300 ft) above sea level and supply water to the lower part of the city by gravity. The damage to this aqueduct system was not recognized, because Kobe and Hanshin did not stop the operation and they were not inspected. The Sengari Aqueduct, which transmits water to Uegahara WTP from Sengari Reservoir, was damaged by flaking and cracking of the concrete lining (See Figure 2).

Figure 2. Sengari Aqueduct damage (M. Matsushita)

WATER TREATMENT FACILITIES

San Fernando Valley

LAA Water Treatment Plant has a capacity of 2.27 million cubic meters per day (600 million gallons per day) received minor damage, such as ground settlement around the plant, leaks at construction and expansion joints, and breaks in plastic chlorine solution lines. The plant is a rapid anthracite coal filter with ozone predisinfection. This damage was repaired, but not as a priority because of the need to restore service in the LAA and distribution system. Supply was provided by groundwater and MWD where the distribution system was operational. There was no damage to chlorinating stations.

Kobe

Hanshin Water Authority’s three WTPs and Kobe’s five WTPs are slow and rapid sand filters for potable water and one plant for industrial water. Hanshin’s plants at the time of earthquake were Inagawa (675,000 cubic meters per day) (178.3 million gallons per day), Amagasaki (213,000 m3/d; 56.2 mgd) and Kabutoyama (160,000 m3/d; 42 mgd). The Amagasaki WTP had severe damage to treatment facilities. A landslide damaged the Kabutoyama WTP by undercutting the solids handling building. Kobe’s three largest WTPs are Okuhirano (60,000 m3/day; 15.8 mgd), Uegahara (70,000 m3/d; 18.5 mgd) and Sengari (108,000 m3/d; 28.5 mgd). Damage occurred at two plants; Uegahara had subsidence at the sludge thickener facilities and mechanical and electrical equipment damage, and Motoyama (8,000 m3/d) (2.1 mgd) WTP damage occurred to the plant piping and concrete cracking to walls. Interim and permanent repairs were made to these facilities; however, repairs made were not of priority since it was necessary to put the distribution system in service and it was wintertime demand. After the earthquake the Hanshin Water Authority decided to combine the Amagasaki and Kabutoyama WTPs and construct the Amagasaki WTP as a new WTP.

DISTRIBUTION SYSTEM

San Fernando Valley

The Valley distribution system consists of a number of transmission lines and distribution mains ranging from 3,000 mm (120 in) to 100 mm (4 in) of steel, cast iron and ductile iron. The damage to large diameter pipe was due to compression or tension at the bell and spigot joints. Damage of the smaller distribution mains was usually to rigid joint cast iron mains or due to corrosion of older steel mains. Generally ductile iron pipe with the flexible rubber gasket joint performed well. Pumping and chlorinating stations performed well after the seismic retrofit in the 1970 and 1980’s. Two old riveted steel tanks were damaged beyond repair along with tank roofs and inlet outlet lines. A summary of damage to Los Angeles Water System is shown on Figure 3.

Facility Damage (damage/total) Repair cost (million dollars) Dam 0 /14 dams 0.32 Treatment plant 1 /1 plant 2.1 Raw water conduit (aqueduct) 14 repairs / * 4.5 Transmission main (Trunk Line) 60 repairs /700 km ** Distribution Reservoir 7 /86 12.0 Distribution pipe 1,013 repairs /11,740 km 5.2 Service connection (w/o cust. pipe) 208+ repairs /700,000 lines ** Other (including building) District Yards, etc. 17.0 Total 41.0 * FLAA and SLAA totals 660 km (410 miles), a total of 65 km (40.5 miles) of aqueducts are considered within the earthquake affected region. * * Distribution pipe damage costs include transmission mains, service connections and fire hydrants. Data on distribution pipe and service connection damage by Selcuk Toprak, Cornell University.

Figure 3. Damage to Los Angeles Water System Facilities

Kobe

Kobe distribution mains consisted of approximately 86% ductile iron pipe, 8% cast iron, 3% steel pipe and 3% polyvinyl chloride pipe. Kobe distribution pipe damage included cracked pipe body (20%), failure of the joints (65%) and damage to fittings (15%). The number of damaged service connections includes the service pipe up to the meter and the property pipe beyond the meter to the customer. Many service connections, especially on the customer side, were damaged due to building collapse. Roughly the customer pipe damage is about 4/5 of the total. Damage generally occurred to older pipe with rigid joints and in areas subject to liquefaction. There was good performance for pipe with expansion and flexible joints. In Japan, seismic joints for ductile iron pipe distribution mains provide the best performance; the S-type is for larger mains and the S- II for smaller mains. Figure 4 shows a summary of the damage to the Kobe water facilities. There were no reported leaks to the approximately 300 flexible joints that allow for expansion and contraction as well as rotation. This joint adjusts for the differential movement between structures, but is more expensive.

Facility Damage (damage/total) Repair cost (billion Yen) Dam 1 /3 dams Treatment plant 2 /7 plants Raw water conduit (aqueduct) 2 lines /43 km Transmission main (Trunk Line) 6 lines /260 km 7.0 Distribution Reservoirs and tanks 1 /119 tanks 1.9 Distribution pipe 1,757 failures /4,002km 13.5 Service connection/customer pipe 89,584 repairs /650,000 lines 2.5 Other (including building) Head office, Tobu Branch, etc. 4.1 Total 29.0

Figure 4. Damage overview of Kobe Water System (Matsushita, M.)

EMERGENCY WATER SUPPLY

San Fernando Valley

Emergency water supply to the San Fernando Valley is provided by a redundancy of sources. The major supply is from the LAA system, but can be supplemented from MWD or the groundwater system, as was done after the Northridge earthquake since both LAA were out of service for a period of time. Reservoirs and tanks surrounding the Valley also supply the distribution system. Most of the pump-tank systems have emergency generators for operating the pumps in case of a power outage. In the Northridge event because of the number of main breaks and general power outage, emergency water was delivered by beverage tank trucks and sanitized construction water trucks to the residents. Tankers were fitted with spigots from which individuals could fill their containers (See Figure 5). Some beverage companies provided free bottled water.

Figure 5. Beverage water tanker truck with spigots (Lund)

In areas where water was still available with the potential for contamination from the broken mains, residents were ask to boil their water before using for drinking and cooking. In cooperation with the Los Angeles City Fire Department, Fire Engine pumpers were used to pump water from a lower pressure service zone to a higher- pressure zone. This was called an “Interdepartmental (Inter-system) Pumping Location”.

Kobe

Emergency water supply to Kobe also was provided from a redundancy of sources. The major supply is from the Hanshin Aqueduct system and several local small reservoirs and tanks. Pump-tank systems served the area in the Rokko Mountains above the hydraulic grade of the Hanshin Aqueducts. The tanks in most cases are dual tanks in which one leg of the inlet-outlet can be seismically activated from the Kobe Water Control Center. All the tanks performed well except one old concrete reservoir and all 21 seismically activated tank outlet valves operated, except three. Because of the broken

mains, the open gated tanks drained rapidly. Located throughout the city are underground cisterns {(10,000 liters) (2,600 gallons)} for emergency fire protection. Spigots attached to the fire hydrants provided emergency water supply to the citizens (See Figure 6). There was no boil water notice given in Kobe. Various water tank trucks and various types of portable storage containers where located at centralized locations where citizens filled their containers.

Figure 6. Kobe emergency fire hydrant water supply and portable tanks (Lund)

WATER SYSTEM RESTORATION

San Fernando Valley

In the San Fernando Valley, Water System personnel did restoration from the resources of seven operating districts. The districts have trained personnel, material and equipment to perform repairs. The Oakland based East Bay Municipal Water District, MWD, and others provided mutual aid. Damage was mostly to cast iron pipe with rigid (cement or lead caulked) joints, welded steel pipe at bell and spigot joints, or to old steel pipe due to corrosion. The greatest pipe damage occurred to those subjected to permanent ground deformations (pgd); fortunately pgd in Los Angeles was not wide spread and pgd-related pipe damage was limited to a few localized areas.. Private contractors were hired for specialized services. Repairs were made by replacement of pipe, repair clamps, or welding. In some locations where numerous breaks occurred the entire main was replaced. Repairs were made block by block from a restored portion of the system, which had been pressurized and cleared for drinking water quality. Figure 7 shows the days by area required for water service restoration in the Valley.

Figure 7. Days by area required for water service restoration.(Adams-Lund)

Kobe

Kobe Municipal Waterworks Bureau personnel performed the Kobe Water System restoration from five operating districts with mutual aid from other water agencies. Restoration was hampered because of the blocked streets with ruble from damaged and collapsed buildings. Since the city is located on a narrow strip between the mountains and the bay, the damage to the railroads and highway hampered access with heavy traffic and limited access for the reconstruction effort. Also there was not sufficient water pressure in the distribution system to locate leaks.

100

40 Recovery 20

0 012345678910

Weeks Figure 8. Percent water service recovery by weeks. (M. Matsushita)

Kobe pipe materials were ductile iron, cast iron, steel and other materials. Most of the city is located in a marine environment, where landfills were made to provide areas for development, which is subject to liquefaction and large pgd. This occurred along the shoreline and Port and Rokko Islands. Pipes were severely damaged with large ruptures and failures in pgd areas requiring lengthily repairs. Ductile iron pipe with the S and S-II seismic joints performed well in the liquefaction areas. Repairs were made using repair clamps and replacement of mains. The percentage water service recovery by weeks is shown in Figure 8.

FUTURE SEISMIC MITIGATION

San Fernando Valley

1. Partially buried seismically designed prestressed concrete tanks replaced several older riveted steel water tanks, which were damaged beyond repair. Roof structures have been replaced on several other tanks. 2. Ductile iron pipe with rubber gasket joint and welded steel pipe cement coated and lined are still being installed in the distribution system. Unfortunately the Japanese seismic joints S and S-II for ductile iron pipe are not available in the United States. Trenchless technology is used to upgrade seismic performance of existing distribution pipe in addition to improving capacity and water quality. The method requires the replacement of all valves, fittings and service connections and an inspection of the pipe condition. Trenchless technology methods include cement mortar lining, steel pipe insertion, High Density Polyethylene (HDPE) slip lining, pipe splitting and bursting. 3. In response to the damage to several tank inlet-outlet facilities in the earthquake, approximately 60 small reservoir and tank inlet outlet lines have been replaced over a long-term program by flexible mechanical joints. These joints allow contraction and expansion as well as rotation to adjust for the differential movement between the tank and buried inlet outlet pipeline. These joints were developed in Japan and were very successful in the 1995 Kobe earthquake. They are now manufactured in the USA under license from the Japanese manufacturer. 4. In the San Fernando Valley a number of the large diameter trunk lines are from 70 to 90 years old and are of riveted steel construction. The pipelines were originally unlined; however, since have been cement mortar lined, but have a minimal exterior coal tar enamel coating. These lines may be subject to corrosion and are minimum thickness. Welded steel pipe lines are being installed to replace these lines, which maximizes the opportunity to shift groundwater to the western portion of the Valley. This adds to the redundancy of the system. Some of the older lines will be slip lined to restore their reliability and provide additional seismic redundancy to the system. A transmission line testing program has been initiated to evaluate pipe reliability and to determine areas for modification. A portion of the Granada Trunk Line that was severely damaged in the 1971 and 1994 earthquakes was relocated. 5. A number of water quality improvements are being made to comply with the Federal Safe Drinking Water Act (SDWA). The SDWA requires full treatment from any

open reservoir before distribution to the customer. In order to do this it requires enlargement of trunk lines from the LAA Treatment Plant and modifications to the open reservoirs. This project incorporates the design and construction to comply with current seismic requirements. 6. Damage in 1994 to the SLAA at Terminal Hill above ground pipeline was almost identical to the damage that occurred in the 1971 San Fernando event. To avoid a repeat of this situation an alternative location for the pipeline is proposed. Terminal Hill tends to be an inverted pendulum and slightly unstable. It is being proposed to replace this portion with a tunnel. The tunnel portion of the FLAA suffered essentially no damage in the 1994 event as well as in the 1971 event. 7. But the Following the 1971 event dams were analyzed using the latest state of the art dynamic stability analysis procedures based on the lessons learned from the San Fernando event. In 1994 there was no damage to dams; however, there have been improvements to the dam stability analysis procedures. A program has been initiated to review all dams using the latest method of seismic analysis and make necessary modifications to protect against future events.

Kobe

1. Kobe proposes to make its water distribution system durable and reliable. With the good performance of ductile iron pipe with the S and S-II joints, many kilometers of this pipe is being installed over a period of time. Also where appropriate the Japanese manufactured flexible joint is being used, where differential settlement is expected. 2. The telemeter control system and emergency shut off valve system worked well for the distribution reservoirs (tanks). In Kobe each of the reservoir sites consist of two reservoirs and an emergency seismic shut off valve is installed on only one outlet pipe. When the intensity of the earthquake or the outflow of water reaches a certain level, the valve shuts-off saving one-half the water storage, while the other continues to provide firefighting water. In the earthquake 18 out of 21 valves on the tanks worked correctly. In order to provide water for the citizens for three days, approximately 30 to 50 systems (service zones) have been modified to include the seismic shut-off emergency valves, similar to those on the reservoirs. Also emergency storage cisterns are being placed through out these service zones. 3. Prior to the Kobe earthquake the city had planned to build a third tunnel through the Rokko Mountains to meet the increase demand from the Yogo Gawa. This alignment would have been parallel to the existing tunnels. After the earthquake the plan was changed to install a large capacity transmission main [(2,400-mm)(95-in)] in a tunnel beneath the downtown area, which has been partially completed. Several of the construction shafts have been converted to emergency water supply stations and will serve citizens, water trucks, and fire engines. It will also serve pressure to the distribution system to better determine the locations of pipe leaks. It is anticipated that the installation of this new main would shorten the restoration time by providing a supply to the lower end of the distribution system.

CONCLUSIONS

Los Angeles and Kobe experienced earthquakes of the same magnitude (Mw 6.7) in a modern built urban environment on January 17, 1994 and 1995, respectively. Prior to these earthquakes, the LADWP and Kobe Waterworks Bureau had implemented seismic improvement programs to resist earthquake effects. All improvement measures implemented in both water systems performed well. Certain facility types performed better in each of the systems (e.g., less distribution reservoir damage in Kobe and less pipe damage in Los Angeles). In comparing the system performances two aspects stand out: the differences in number of pipe breaks and the total restoration times in each system. These aspects are related to the different geologic and built environments in which the two water systems reside. Permanent ground deformations affected pipe performance in both water systems; however ground movements were larger and more widespread in Kobe, causing more and larger pipe failures, as a result of greater liquefaction potential in soils within the damage stricken region. Kobe also sustained numerous service line breaks from building to collapse, whereas this was virtually nonexistent in Los Angeles. The building collapse also hampered Kobe restoration efforts. Water service was restored to all Los Angeles customers in 1 week while it took Kobe nearly 10 weeks for full service restoration. Restoration efforts in Los Angeles were more easily achieved due to a greater source supply redundancy and access to pipe damage. Overall, the difference in water service restoration is related to different infrastructure configuration, geotechnical siting, construction, social needs, political, and financial philosophies.

ACKNOWLEDGEMENTS

Shirley Wineman, LADWP Water Utility Manager-Costs; Jerome Chen, LADWP Engineering Designer-Graphics; Fang Lui, Graduate Student, University of Southern California –slide conversion; Makoto Matsushita, Kobe Municipal Water, Reviewer.

REFERENCES

[1] Personal field inspection and knowledge of the authors. [2] Takada, S. and J. Ueno, 1995, Performance of Lifeline Systems during the Great Hanshin Earthquake, Proceedings of Sixth U. S.–Japan Workshop on Earthquake Disaster Prevention for Lifeline Systems, Osaka, Japan, July 18 and 19, 1995. [3] Schiff, A., 1995, Northridge Earthquake: Lifeline Performance and Post Earthquake Response, ASCE, TCLEE Monograph No. 8, New York, NY. [4] Schiff, A. , 1995, Hyogken-Nambu (Kobe) Earthquake of January 17, 2005 Lifeline Performance, ASCE, TCLEE Monograph No. 14, New York. [5] Matsushita, M., S. Morita, and S. Ogura, 1998, Post-Earthquake Reconstruction of the Kobe Water System Based on the Lessons Learned from the 1995 Hanshin-Awaji (Kobe) Earthquake, Proceedings of the IWSA Workshop on the Anti-Seismic Measures on Water Supply, Tokyo, Japan, Nov. 1998. [6] Matsushita, M., 1998, Restoration Process of Kobe Water System from the 1995 Hanshin-Awaji Earthquake, ASCE, TCLEE, Proceedings of Fifth U.S. Conference on Lifeline Earthquake Engineering, TCLEE Monograph No. 16, New York, NY, September 1999. [7] Kitaura, M., M. Miyajima, , and H. Nakagawa, 1999, Lessons Learned from Damage to Water Supply Pipelines Due to the 1995 Hyogo-ken Nambu Earthquake in Japan, 8th Canadian Conference on Earthquake Engineering, Vancouver, BC, 1999. LL 11-28-04

Session 5: Mitigation and Prevention of Damage

S5-1 “The Outline of Seismic Measures of Tokyo Waterworks Bureau” ―From the Point of View of Cooperation with Regions― Presenter: Hiroshi Yamada (Bureau of Waterworks, Tokyo Metropolitan Government)Japan

S5-2 “Emergency Operation and Countermeasures for the Water Supply System in Taiwan Learned from 1999 Chi-Chi Earthquake” Presenter: Wei-Sen Li (National Science and Technology Center for Disaster Reduction) Taiwan

S5-3 “Evaluation of Seismic Upgrade Construction of Hanshin Water Supply Authority “―In the case of pipe line― Presenter: Keiichi Murakami (Hanshin Water Supply Authority) Japan

S5-4 “Current Status and Subject of the Seismic Upgrade of Kobe Water System after Ten Years from the 1995 Hanshin-Awaji Great Earthquake” Presenter: Tetsuro Kijima (Kobe Municipal Waterworks Bureau Planning Division)Japan

S5-1

“The outline of seismic measures of Tokyo Waterworks Bureau” - from the point of view of cooperation with regions –

Presenter: Hiroshi Yamada

(Bureau of Waterworks, Tokyo Metropolitan Government, Japan)

The outline of seismic measures of Tokyo Waterworks Bureau - from the point of view of cooperation with regions -

Hiroshi Yamada

ABSTRACT

It is about ten years since the occurrence of The Great Hanshin earthquake. However many big earthquakes hit the Japanese Islands one after another. If the waterworks facilities in Tokyo were directly hit by a big earthquake, heavy influence would be brought to the city lives and urban functions. Therefore seismic measures have been positioned as one of the important subjects of Tokyo Waterworks Bureau. Tokyo Waterworks Bureau has to two plans about seismic measures. One is The Plan of Projects for Seismic Measures of Tokyo Waterworks Bureau and the other is The Plan of Emergency Seismic Measures of Tokyo Waterworks Bureau. The Plan of Projects for Seismic Measures was drawn up for the purpose of reinforcing seismic resistance of the facilities and establishing the facilities for securing drinking water in case of an earthquake. The measures in this plan are divided roughly into "reinforcement for seismic resistance of the facilities", and "securance of drinking water". Tokyo Waterworks Bureau has been making efforts of public relations about this plan through various media. However, the new effective P.R. methods should be needed in future in consideration of the lowness of the degree of recognition. The Plan of Emergency Seismic Measures aims to restore the facilities and carry out emergency water supply quickly. The existing plan was revised based on the instruction from The Great Hanshin earthquake. For example, several methods of emergency water supply were added according to the requests from citizens. In order to carry out the emergency measures smoothly, the Bureau has established the cooperative system with the related companies and other big cities. Moreover, emergency water supply activities are supposed to be carried out after allotting the role to special wards and cities.

Hiroshi Yamada, Director for design and coordination, Construction Division, Bureau of Waterworks, Tokyo Metropolitan Government, 2-8-1 Nishi-Shinjuku, Shinjuku, Tokyo, Japan 163-8001

INTRODUCTION

About ten years have passed since the occurrence of The Great Hanshin earthquake which took thousands of precious human lives, houses and properties. For the meantime, many big earthquakes, The Western Tottori Earthquake, The Hiroshima Geiyo Earthquake, The Northern Miyagi Consecutive Earthquake, The Niigata Chuetsu Earthquake, hit the Japanese Islands one after another. Tokyo may be hit by a big earthquake at any moment. If Tokyo, which is the capital of Japan, were directly hit by a big earthquake, the influence would not remain in domestic but reach the whole world. Especially water service is bearing the very important indispensable roles of supporting daily lives and social economic activities. If the water supply facilities suffered destructive damages, serious troubles for the city life would arise. Moreover, it is assumed that heavy influence is brought to the urban functions. Therefore, in Tokyo Waterworks Bureau, seismic measures have been positioned as one of the important subjects of the Bureau on a par with "making sure of water resources", "supply of safe and palatable water", etc. And we are making efforts of carrying out seismic measures from the point of views of both hard and soft. This time, I will introduce about the whole seismic measures of Tokyo Waterworks Bureau, with the PR to the citizens etc.

OUTLINES OF WATER UTILITY AND WATER SUPPLY FACILITIES IN TOKYO

Tokyo Waterworks Bureau started water supply in 1898. The water resource volume is 6,230,000 m³ per day and the capacity of purification facilities is 6,860,000 m³ per day by the end of today. And water is supplied to the area of 1,222km², and for about 12 million people in 23 special wards and 25 cities in the Tama area. Moreover, provisional division of water is performed to three cities of Musashino, Akishima and Hamura, which are not included in the service area of Tokyo Waterworks Bureau. In fiscal 2003, the distribution amount is 1,613,000,000 m³, the maximum daily amount is 4,960,000 m³, and total length of distribution pipes is 24,800km.

Water resource facilities (Fig. 1)

The water resource in Tokyo can be divided roughly into two, the Tama River system and the Tone River and the Ara River system. As to mention about the Tone and the Ara River system which occupies 80 percent of the water resources in Tokyo, there are eight dams etc. in the upper stream of the Tone River, which the Ministry of Land, Infrastructure and Transport and the Japan Water Agency manage. Moreover, the Urayama Dam is in the upper stream of the Ara River. On the other hand, the Ogouchi Reservoir and the Murayama-Yamaguchi Reservoir are in the upper stream of the Tama River, which Tokyo Waterworks Bureau manages. In addition, there are the groundwater source facilities and the facilities of the Sagami River system, that water are subdistributed from Kawasaki City Waterworks.

Yagisawa Dam Naramata Dam dam 115 mill.ｍ³ 85 mill.ｍ³ Fujiwara Dam water conveyance Aimata Dam 31 mill.ｍ³ Kusaki Dam 20 mill.ｍ³ Sonohara Dam channel 13 mill.ｍ³ 50 mill.ｍ³ Yamba Da m Watarase Reservoir purification plant under construction Tone River 26 mill.ｍ³ Shimokubo Dam 120 mill.ｍ³ Takizawa Dam Musashi Channel under construction Urayama Dam Ara River Naka River Ogouchi Dam 56 mill.ｍ³ 185 mill.ｍ³ Yamaguchi Res. 34 mill.ｍ³ Murayama Res. Tama River Edo River

Sagami Dam Shiroyama Dam Sagami River

Figure 1. Outline of water resource facilities in Tokyo

Water purification facilities (Fig. 2)

Tokyo Waterworks Bureau operates 11 main purification plants. Total capacity of these plants is 6,860,000 m³ per day. Among these water purification plants, the Higashi-Murayama Purification Plant and the Asaka Purification Plant are connected by raw water conveyance pipe so that cross-supply of raw water can be performed. Moreover, purified water transmission from the Higashi-Murayama, the Nagasawa and the Sakai Purification Plants to Tokyo metropolitan area can be performed using gravitational force.

Water transmission and distribution facilities (Fig. 2)

The water supplying stations are organically connected with the water purification plants, and have performed the big role on water distribution control. Tokyo Waterworks Bureau has 38 main supplying stations which have capacity of distribution reservoirs of more than 10,000 m³. The total capacity of water distribution reservoirs is about 3,180,000 m³. Now, we aim at building water transmission and distribution system which is efficient and has highly seismic resistance. To realize reinforcement of backup function and block distribution system, we are establishing and rebuilding several water supplying stations, and establishing cross-supply facilities between water supplying stations.

Kanamachi Ara River

Asaka Misato Ogouchi Res. Yamaguchi Res. Misono Ozaku Murayama Res. Tama River Higashi-Murayama

purification plant Sakai Suginami water supplying station Edo water supplying station River (expansion, establishment)

transmission and Kinuta distribution pipe Nagasawa Tamagawa transmission and distribution pipe (suspending operation) (establishment) Sagami River

Figure 2. Outline of water transmission and distribution facilities in Tokyo

SEISMIC MEASURES OF TOKYO WATERWORKS BUREAU

Tokyo Metropolitan Government is positively pushing on their comprehensive seismic measures according to the ordinance of prevention of earthquake disaster of Tokyo Metropolitan Government, which was put into operation in October, 1971 and revised as the ordinance of seismic measures of Tokyo Metropolitan Government in December, 2000. Moreover, on the basis of the basic principle of this ordinance, Tokyo Metropolitan Government drew up the plan of projects for seismic measures of Tokyo Metropolitan Government in March, 2002. Based on these facts, Tokyo Waterworks Bureau is pushing on seismic measures according to two plans. One is The Plan of Projects for Seismic Measures of Tokyo Waterworks Bureau, which was drawn up to reinforce seismic resistance of the facilities and to secure drinking water. The other is The Plan of Emergency Seismic Measures of Tokyo Waterworks Bureau, which was drawn up to carry out restoration of damaged facilities and emergency water supply quickly in case of an earthquake.

THE PLAN OF PROJECTS FOR SEISMIC MEASURES OF TOKYO WATERWORKS BUREAU

In order to mitigate the damage of water supply facilities and secure water supply to citizens as much as possible, The Plan of Projects for Seismic Measures of Tokyo Waterworks Bureau was drawn up on the basis of the basic principle of the ordinance of seismic measures of Tokyo Metropolitan Government. In addition, this plan had been called The Plan for Prevention of Earthquake Disaster from the first plan (drew up in 1973) to the 7th plan (in 2000). In those days when the first plan was drawn up, it was thought that water transmission and distribution pipes are considerably damaged and extensive water

suspension is not avoided on the occasion of a big earthquake. And the main purpose of seismic measures was to secure drinking water for citizens, or to establish the system for emergency water supply. Therefore, securance of drinking water is placed as a mainstay of measures together with reinforcement of facilities also in the present plan. In recent years, from ten to twenty percent of bureau budget is spent for this plan. And the total amount of investment applied to these plans within the period from fiscal 1973 to fiscal 2001 is over 1 trillion yen.

Outline of the plan and systems of the measures

The term of this plan is three years from fiscal 2002 to fiscal 2004. And to adapt to the changes of the situation, revision can be made within the term of the plan.

The system of measures

The measures in this plan are divided roughly into "reinforcement for seismic resistance of the facilities", and "securance of drinking water". Moreover, measures coming into the system of reinforcement for seismic resistance of the facilities are divided according to the kinds of facilities. The system of measures is shown in Fig. 3.

reinforcement for seismic resistance of facilities reinforcement construction reinforcement of facilities of water storage,intake, and conveyance of facilities embankments of reservoirs construction of purification facilities purification plant private power plants reconstruction of chlorine feeding facilities water transmission and water transmission and replacement of transmission distribution facilities distribution pipe and distribution pipes expansion of water supplying station water supplying station assesment of seismic resistance and repair seismic diagnosis of facilities reinforcement for seismic resistance of reinforcement for seismic water supply equipment service pipe resistance of service pipes replacement of service pipes by stainless steel pipes reinforcement of water transmission and establishment of network of water transmission pipe water supply systems distribution facilities transmission pipes

securance of dirinking water preparation of materials for construction of base for emergency water supply emergency water supply establishment of warehouse for materials improvement of water supply equipment construction of emergency water tank construction of small tank

Figure 3. The system of measure in The Plan of Projects for Seismic Measures

Reinforcement for seismic resistance of the facilities

Reinforcement for seismic resistance of the facilities is carried out for the purpose of securing water supply as much as possible in case of an earthquake. For this reason, measures to reinforce seismic resistance of the facilities are divided into two types. One is the reinforcement construction of the facilities to reduce the damage in case of an earthquake disaster, and the other is the reinforcement of the water supply systems to reduce the shutdown area and shutdown time. As for the reinforcement construction of the facilities, the water supply facilities are classified according to their functions. And basic ideas about performing the measures and details of the projects carried out within the term of three years plan are arranged in the plan. And as the measures to reinforce the water supply system, establishing the water transmission and distribution network and block distribution system are intended to be carried out in this plan. The concrete contents of the projects carried out within the term of three years plan are shown in Table 1.

TABLEⅠ. BASIC IDEAS ABOUT MEASURES AND CONCRETE CONTENTS OF THE PROJECTS measures basic ideas contents of projects *carry out reinfoecement work of the *reinforcement work of the embankment facilities of water embankment supposed to be damaged of the Yamaguchi Reservoir (fiscal 2002) storage, intake, in case of a big earthquake according *reinforcement work of the embankment and conveyance to seismic resistant analysis of the Murayama-Shimo Reservoir *construct private power plants *construction of private power plants purification for power failure (Asaka and Misono P.P.) facilities *prevent secondary disaster *reconstruction of chlorine feeding facilities due to leakage of chlorine (Asaka and Higashi-Murayama P.P.) *construction of water supplying stations reinforcement water transmission *rebuild the old water supplying station (Kinuta,Koemon,Kamikitadai,Kamiishihara) construction of and distribution *replace old pipes reinforcement *replace transmission and distribution pipes acilities facilities to seisimic resistant pipes for seismic (259km,fiscal 2002 to 2004) resistance of *carry out seismic resistant analysis *seismic resistant anarysis facilities assesment of by new design standard (21spots,fiscal 2002) seismic resistance as for structure by old standard *reinforce for seismic resistance *repair low seismic-resistant structures (19spots,fiscal 2002) *replace servise pipes (over 75mm in diameter) *replace old service pipes to pipes water supply for seismic resistance (180spots,fiscal 2002) with seismic joint equipment *replace service pipes to stainless steel pipes or stainless steel pipes (88,183spots, fiscal 2002) *reinforce for seismic resistance *establishment of network of reinforcement of of the water supply system by water transmission pipes water supply facilities establishing water transmission (8.5km, fiscal 2002 to 2003) network and block distribution system *prepare materials and warehouse *replacement of old materials construction of base for for them at the bases for for emergency water supply emergency water supply *construction of warehouse (9spots) securance of emergency water supply drinking *improve the equipment *improvement of equipments for emergency water supply for emergency supply (7spots) water construction of *construct emergency water tanks *construction of emergency water tank in the blank area small emergency water tanks (3spots)

Securance of drinking water

It is assumed that pipeline will suffer a certain measure of damage on the occasion of a big earthquake even if it is reinforced for seismic resistance as much as possible. And it is also assumed that temporary or local shutdown can not be avoided. For this reason, the Bureau prepares many bases for emergency water supply and emergency water tanks for the occurrence of an earthquake so that drinking water may be secured and emergency water supply can be carried out in case of an earthquake. By the end of today, 196 bases for emergency water supply (including 73 emergency water tanks) have already established. The total capacity of reserved water is about 1,000,000 m³, which is equivalent to the amount for four weeks of emergency drinking water use in Tokyo. The concrete contents of the projects for securance of drinking water carried out within the term of three years plan are also shown in Table 1.

Public relations activity to the citizens in Tokyo about seismic measures

P.R. method to the citizens

Public relations activities about seismic measures are important in order to raise the effectiveness of measures and to ask the citizens to understand investing respectable expenditure of the Bureau for seismic measures. Currently, the Bureau is making efforts of public relations about seismic measures of Tokyo Waterworks Bureau through various media as follows.

1 exclusive pamphlets 2 The Tokyo Water News, which is a quarterly newsletter 3 educational materials for primary school pupils and junior high school students 4 the homepage of Tokyo Waterworks Bureau

Moreover, the location about bases for emergency water supply and emergency water tanks is often printed in the pamphlets, such as "the convenient book of life" and "guidance of disaster prevention", which are published by each special ward and city in Tokyo. Furthermore, at the time of an emergency drill, the park in which the emergency water tank is set up is often used as a site of drill, and drill for emergency water supply is also performed. By this drill, many participants can understand not only the location about emergency water tanks but also how to get emergency drinking water.

The degrees of knowing about the base for emergency water supply

Tokyo Waterworks Bureau has introduced the monitor system so that the needs of citizens can be reflected for management of waterworks. And questionnaire surveys to the monitors have been carried out several times in a year. Among these questionnaire surveys, the replies to the question about securance of drinking water are added up and the total result is shown in Fig. 4

％ 70 keeping drinking water knowing the base for emergency water supply 60

0 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003

Figure 4. The situation of keeping water and the degree of knowing about the base for emergency water supply

According to the results, the rate of monitors who keep drinking water in their houses was about 40% until 1994. After The Great Hanshin Earthquake, the rate rose to about 50% after 1995 and it is roughly maintaining the level in recent years. On the other hand, the rate of monitors who know the base for emergency water supply was about 20% before the Great Hanshin Earthquake, and rose to about 40% after that. Unfortunately in recent years, the rate is relapsed into the level of 20% again and it has stopped at the low level. As mentioned before, the Bureau has been making efforts of public relations about the base for emergency water supply through various media. However, the new effective P.R. methods should be needed in future in consideration of the lowness of the degree of recognition.

THE PLAN OF EMERGENCY SEISMIC MEASURES OF TOKYO WATERWORKS BUREAU

In case of an earthquake, it is necessary to re-supply water to citizens as soon as possible in spite of the social confusion just after the occurrence of an earthquake. Therefore, Tokyo Waterworks Bureau drew up the plan of emergency seismic measures so that the restoration of water supply facilities and emergency water supply activities can be performed smoothly, and so that the Bureau can begin the emergency seismic measures quickly and exactly by clearing up the role of the water supply headquarters which the Bureau organizes. This plan was a part of The Plan for Prevention of Earthquake Disaster in fiscal 1973 when the first plan was drown up. The existing plan was revised in January, 2000 after six times of revision based on the changes of the social background and the instruction from The Great Hanshin earthquake.

The contents of the plan

The system of measures in this plan is shown in Fig. 5.

to carry out emergency seismic establishment of basic organization measures quickly and exactly water supply headquarter initial actions at night or on holiday activities in case of official announcement of precaution communication system to perform allotment of roles emergency water supply plan gradational water supply between the Bureau and cities restoraton plan to restore as soon as possible dicision of priority of restoration keeping cooperative relations with related organs drill and in-service training to prepare for disaster comprehensive drill separate drill in-service training for staffs

Figure 5. The system of measure in The Plan of Emergency Seismic Measures

Basic organization

About the basic organization, some rules as follows are decided in the plan. When the anti-disaster headquarters of Tokyo Metropolitan Government is established with the occurrence of a big earthquake or the official announcement of precaution, Tokyo Waterworks Bureau immediately establishes the water supply headquarters in the Bureau, and makes efforts to clarify the line of command and to make preparations for the staff to be active. When a big earthquake occurs at night or on a day off, the staffs who live in the anti-disaster staff house and the named administrative staffs gather to the T.M.G. building first of all. And they attend initial activities with the staffs on standby alert or working in rotation at the branch offices. About the gathering of general staffs at night or on a day off, five grades emergency disposition systems are decided on to make preparations according to the situation of damage. If a big earthquake occurs during the business hours, general staffs attend emergency measures assigned in advance at their belonging posts.

Emergency water supply plan

Emergency water supply is basically carried out by gradational methods based on the needs of citizens. They are as follows: (1) water supply from the emergency water tanks

(2) water supply in the purification plants or the water supplying stations, etc. (3) water supply by water trucks (4) water supply from temporary water taps (hydrants) (5) water supply by emergency temporary piping

Restoration plan

About restoration, the following is decided in the plan. Several restoration groups for each kind of facilities are organized in the central office. Each group grasps the situation of damage of related facilities and draws up the concrete restoration plan with coordinating with other groups and the related organs. And the branch offices which take control of each facility carry out restoration activities according to that plan. As to the restoration of water transmission and distribution pipes, the priority of restoration is decided as "first important routes" and "second important routes" in advance, and the restoration activities are carried out efficiently based on this priority. At the same time, the Bureau makes efforts for reduction of shutdown area by water distribution control. Furthermore, restoration activities are carried out with cooperation of the construction companies connected with the Bureau.

Drill and in-service training

Drills about the emergency activities are performed so that various kinds of them can be carried out smoothly. Two types of drills are preformed separately. One is comprehensive drill which includes the operation of the water supply headquarters and the emergency attendance, and the other is separate drill which includes communication, restoration and emergency water supply. Moreover, in-service trainings are performed to have all staffs fully understood about their roles in case of an earthquake. In addition, as for emergency water supply among various drills, participation of not only the staffs of the Bureau but also the citizens who receive water is effective. In recent years, emergency drills which are performed on the disaster drill day are often carried out with participations of the government, T.M.G., cities and citizens.

Cooperation with the related organs

In order to carry out the above-mentioned emergency measures smoothly, the Bureau has established the cooperative system by asking the related companies to the Bureau for their cooperation and by concluding an agreement about emergency support with other big cities. About the related companies, the Bureau requests their cooperation in supply of pipe material for restoration and supply of vehicles in addition to emergency restoration works. Moreover, the waterworks bureaus in 12 large cities conclude an agreement mutually about emergency support. Furthermore, Tokyo Waterworks Bureau concludes an agreement about mutual emergency support with Chiba Prefectural Waterworks Bureau.

Cooperation with the cities and communities in emergency water supply

In order to carry out emergency water supply efficiently and effectively, it is necessary to get required information about not only the situation of damage of water supply facilities but also the situation of the disaster-stricken area and refuge of citizens and draw up the emergency water supply plan based on these information. Moreover, unlike the restoration of damaged facilities or water distribution, emergency water supply is a work to supply emergency water to citizens and sufferers directly. So, it can be said that emergency water supply is an activity which sticks to the community extremely and which needs cooperation with regional residents. The special wards and cities participate in the establishment of the regional disaster prevention organizations and their activities at ordinary times. And they are good at getting regional information necessary for emergency water supply. Therefore, emergency water supply activities are supposed to be carried out after allotting in detail the role to special wards and cities beforehand. The concrete allotment of roles is shown in Table Ⅱ.

Table Ⅱ. The allotment of roles in emergency water supply allotment of roles methods of emergency water supply setting up equipments water supply activities emergency water tank special wards and cities special wards and cities purification plant or water supplying station the Bureau special wards and cities water truck the Bureau (transport) special wards and cities temporary water taps the Bureau mutual cooperation emergency temporary piping the Bureau mutual cooperation notes; The Bureau carry out emergency water supply by water trucks in the following cases *at the refuges more than 2km away from bases for emergency water supply *for medical facilities (by the request of administrative organs) *for temporary dwellings (by the request of administrative organs)

Revision based on the opinion of citizens

Immediately after the occurrence of the Great Hanshin Earthquake in January, 1995, Tokyo Waterworks Bureau received the request from the disaster-stricken area and dispatched 1,249 staffs and they carried out emergency water supply activities and restoration activities. In these circumstances, the requests from citizens in the disaster-stricken area about emergency water supply proved to change as days passed. The mimetic diagram about that change is shown in Fig. 6. In January, 2000, at the time of the last revision of The Plan of Emergency Seismic Measures, several methods of emergency water supply and required organizations for restoration and emergency water supply were added so that the gradational water supply can be performed based on the above-mentioned changes of requests. The improved methods of emergency water supply are described in the above chapter.

just after restoration stage early of restoration middle of restoration latter of restoration earthquake 4 days 10 days 20 days 31 days passed days later later later later target volume 3 ３～２０２０～１００１００～２５０ (lit./person*day) for drinking for washing up, laundry, and toilet supposed usage for bath and shower ordinary carrying distance ２km ２km～２５０ｍ２５０ｍ～１００ｍ１００ｍ～temporary water (within) taps within the housing lot bases for

main methods emergency water supply

water trucks

temporary water taps

Figure 6. Change of the requests from citizens about emergency water supply

CONCLUSION

Tokyo Waterworks Bureau has prepared for the occurrence of a big earthquake previously by investing a large amount of funds in the improvement of their facilities and water supply systems and deciding on the plan about measures in case of an earthquake. In the future, we have to execute The Plan of Projects for Seismic Measures steadily and make efforts to build water supply system with highly seismic resistance. Also we have to study about closer cooperation with citizens as for The Plan of Emergency Seismic Measures. Furthermore, we want to tackle toward "seismic resistant Tokyo Waterworks", with reviewing the plans according to the future situation.

REFERENCES

[1] Tokyo Waterworks Bureau. 2002. “The Plan of Projects for Seismic Measures of Tokyo Waterworks Bureau” [2] Tokyo Waterworks Bureau. 2000. “The Plan of Emergency Seismic Measures of Tokyo Waterworks Bureau” [3] Tokyo Waterworks Bureau. 2004. “The Outline of Waterworks in 2003”

S5-2

“Emergency Operation and Countermeasures for the Water Supply System in Taiwan Learned from 1999 Chi-Chi Earthquake”

Presenter: Wei-Sen Li

(National Science and Technology Center for Disaster Reduction, Taipei, Taiwan)

Emergency Operation and Countermeasures for the Water Supply System in Taiwan Learned from 1999 Chi-Chi Earthquake

Ban-Jwu Shih Division Head of Management System and Policy, National Science and Technology Center for Disaster Reduction, Taipei, Taiwan, and Associate Professor, Dept. Civil Eng., National Taipei University of Technology, Taipei, Taiwan, [email protected]

Wei-Sen Li Postdoctoral Research Fellow and Research Assistant, National Science and Technology Center for Disaster Reduction, Taiwan [email protected]

Pei-Chung Hsu Deputy Director General, Taipei Water Department, Taiwan [email protected]

In 1999, the Chi-Chi Earthquake (ML=7.3), the most devastating one during the 20th century occurred in Taiwan. Beside the death toll over 2,500 and US $10.7 billion direct property loss, the infrastructure’s destruction such as lifeline systems had brought a major impact on livelihood and economic activity. According to the surveyed data announced by Taiwan government, the Chi-Chi Earthquake caused lots of damages in the water supply facilities and pipelines including as followings: Shih Kang Dam’s partial failure by the tremendous fault’s uneven lifting damaged the water intake for Taichung; the destruction of First Fengyuan Water Purification Plant caused the disruption of water supply for about 700,000 households in the Taichung municipal area; the distribution pipeline network in severely disastrous area suffered tremendous damages due to the excessive ground movement and dislocation. Therefore, the methodologies to minimize the damage of water supply systems after a major seismic event are crucial for the general public’s daily lives and the industrial productions. As a consequence, the higher seismic design standard for the water plants and pipelines, the more effective techniques for earthquake loss estimation at early stage and the better countermeasures are the goals that we have to develop for the future hazard mitigation. This paper first focuses on the earthquake damage and the emergency operation for water supply systems experienced in the Chi-Chi Earthquake and then presents a GIS-based scenario simulation. The suggested strategies and countermeasures for the water supply systems based on scenario analysis and the experience of the past events will also be discussed as a lesson learned from the Chi-Chi Earthquake.

1. INTRODUCTION

Taiwan is located on the circum-pacific earthquake belt, and the period of one destructive earthquake might be expected every ten years on average. The Chi-Chi Earthquake (ML=7.3) that took place on September 21, 1999 was the particularly ravaging one. As a result, the central part of Taiwan was the most disastrous area with casualties and property loss. According to the statistics, the death toll climbed to 2,490 victims, and 11,300 residents were severely injured. The damages of buildings included 51,753 totally collapsed and 54,406 partially collapsed (NFA, 2003). Based on the Directorate-General of Budget (2003), Accounting and Statistics, Executive Yuan, ROC, up to February, 2000, the direct economic loss amounted to $10.7 billion dollars. From the viewpoint of livelihood and development, the earthquake had great influence on the society of Taiwan.

Among the destructions induced by the earthquake, the water supply systems endured a lot of breakages and failures. Both the water treatment facilities and underground pipelines suffered widespread damages. From the investigations conducted by the National Science Council of ROC (Shih et al., 1999), many water treatment facilities in the affected areas were out of order after the earthquake. As for underground pipelines, Wang (2000) attempted to conclude the causes of damage, and his finding is shown in Figure 1. The major causes for break and ruptures of water pipelines were vibration/ground shaking (48%), vertical ground movement (16%), and ground collapses (11%) with other minor factors including ground cracking and opening, horizontal ground movement, and liquefaction (Shih et al., 2000a; data from Wang, 2000). However, this result was summarized from the repair work records which were filled by mechanics. Its correctness was doubtable because the mechanics lacked the background of seismic knowledge. Figure 2 to Figure 9 show the destruction of Chi-Chi Earthquake.

Due to the well-prepared plan for future seismic hazards, the earthquake loss estimation system is an important tool for Taiwan government especially. Taiwan Earthquake Loss Estimation System (TELES), similar to HazUS, was then developed for this purpose. However, the original fragility functions employed by HazUS should be verified and modified to accommodate the Taiwan’s local circumstances. The Chi-Chi earthquake provided such a chance (Shih et al., 2000b) to calibrate the relevant parameters. One goal of this paper is intended to examine the HazUS empirical formulas for predicting the damage of underground pipelines in a major earthquake. With the aid of a GIS-based system, two different approached are applied to derive the fragility relations of repair rates of PVC water pipes having nominal diameter (φ) (approximately inner diameter) larger than or equal to 65mm with respect to the peak ground acceleration and peak ground velocity. On the other hand, the emergency response systems were found to fall behind the realistic demand. The suggested improvements and strategies are proposed for increasing capacity to satisfy the future demand.

Figure 1: The reasons of failure for water pipelines (Shih et al., 2000; data from Wang, 2000)

Figure 2: The damage in First Figure 3: The damage of 2000mm SP distributing Fengyuan Water Purification Plant. pipe in First Fengyuan Water Purification Plant.

(Wang, 2000) (Wang, 2000)

Figure 4: The excessive dislocation of 1000mmSP Figure 5: The breakage at the joint of 100mmDIP overpass pipe in Yijiang Bridge, Taichung. (Wang, pipe in Jhuolan Town. (Wang, 2000) 2000)

Figure 6: The buckle and breakage of pipe. (Photo Figure 7: The breakage of ductile cast iron pipe. by Hong, Ruei-Huang, 2000) (Photo by Hong, Ruei-Huang, 2000)

Figure 8: The buckle failure of stainless iron pipe Figure 9: The compression failure of deformable (Photo by Hong, Ruei-Huang, 2000). steel pipe. (Photo by Hong, Ruei-Huang, 2000)

2. THE COUNTERMEASURES, STRATEGIES AND RESPONSE FOR SEISMIC HAZARD

Public water supply systems is one of the fundamental public utilities for residents’ daily reliance, as the consequence, the damages causing by earthquake would bring the broad inconvenience in restoring victims back to normal lives. In the worst case, the long-lasting recovery period could paralyze the routine functions of cities, because the distributed networks in metropolitan areas are inter-related and vulnerable to destructions of other lifeline systems.

The basic emergency response strategies of public water supply system after earthquake should be composed of loss estimation, urgent repair and water-supplying procedures, furthermore, the volume for extinguishing fire accidents should be anther issuer to be mentioned. With all pre-prepared plans and measures, the less social loss would impact on the economy.

2.1 PRIMARY DIRECTIONS FOR DISASTER MITIGATION

At the moment of disaster taking place, the local governments should adopt the effective countermeasures to cope with the situation and cooperate with central government and water supply company. Besides the launching of emergency center, the local government should react efficiently with transition in bureaucracy organization and management pattern to neutralize the disaster. For public water supply company, the tight cooperation with governmental sectors should be established by the fluent channel of communication. The followings are the general steps for acting responses:

2.2 STRATEGIC DIRECTIONS

Each running unit should execute the necessary procedures according to the emergent response plans and organize the Emergency Operation Squad, through the coordination with the Emergency Operation Center; the degree of damage could be identified by the information of database. From the records with the analysis from database, the allocations of rescue resources including personnel, materials and machineries could be determined with the maximum benefit to disastrous area. To ensure the implementation of the rescue work and prevent the further damage of the secondary disaster, the executions of the restraint of traffic, the quick survey of destruction, the temporary restoration and the examination of safety are top-listing targets to achieve at the first stage of disaster happening.

From the experiences of other countries and Taiwan, the 6-stage emergent response procedures are described briefly as Satge1: The identification of damage intensity; Stage 2: The transmission channel of emergent report; Stage 3: The preliminary moves for safety check; Stage 4: The urgent countermeasures; Stage 5: The initial repair works and Stage 6: The restorations of full functions for water supply systems. The lasting period of each stage would depend on the realistic circumstances, and the inter-dependence and overlap are the related factors to be considered. The details of each stage are depicted below.

1. Stage 1: The identification of damage intensity

As the water supply system could be endangered by natural disaster, the administrative level should recognize the intensity of damage by disaster information transmission systems or on-site reports and launch reactions according to the disaster-proof capacity, the surrounding qualifications and the conditions of geology. There are three sub-subjects that should be encircled as the classification standards.

(1) Minor damaged and usable: It means the system maintains its normal functions without alarming or rescue act, but the observation of safety should be placed. After the stability of disaster control could be reached, the personnel for testing and mending should be dispatched.

(2) Damaged and repairable: It represents part or whole function of system could be inactive temporarily and would be shut down for a short period. For decreasing the impact of the secondary damage, the necessary restriction of traffic should be

enforced with the announcement of warning alert and the implementation of emergent response plan.

(3) Seriously damaged and irreparable: It tells the system could suffer from critical destructions with malfunction and should be closed immediately. The prohibition of access by individuals and vehicles should be strictly carried out. The utilities for substitution should be built for provisional use. As the stable control of disaster could be reached, the original system might restore to the minimum running standard or should be reconstructed.

2. Stage 2: The transmission channel of emergent report

The administrative level or emergency response centers would determine the degree of damaged areas then launch the urgent announcement to three categories of personnel.

(1) Rescuing personnel: The members listed in emergency response plan contain the maintaining staffs, the police and fire departments, rescue organizations and medical institutes.

(2) Supporting personnel: The members listed in emergency response plan include all staffs of water supply system, the aids from other engineering units and the labors contracted in advance.

(3) The general public: On the processing of hazard, the warning alert should be released to the general public for drawing precaution and demanding cooperation.

3. Stage 3: The preliminary moves for safety check

At the initial start of this stage, staffs from local water supply agencies should rush into the damaged site for investigating the extent of destruction and making an evaluation. Then the follow-up steps, like support request and traffic restriction, could be executed in sequence.

4. Stage 4: The urgent countermeasures

This stage would initiate with Stage 3 simultaneously for the consistence of procedures. In dealing with emergency to recovery the function of system temporarily or partially, the trained technicians would examine the degree of impair by the basic equipments. For providing information of repair and restoration, the rapid survey of damage should be completed in 12 to 24 hours to meet the demand of judgment. The allocations of manpower, resource and provisions could be modified in accordance with the survey.

5. Stage 5: The initial repair works

Following the operations of previous stages, the condition of hazard could move into stability gradually and the initial repair works would be proceeded to sustain the function and prevent the

further loss. The quick diagnosis manuals and rapid retrofitting preparation should have been arranged in advance. The related notices could be put into considerations which are characterized as follow:

(1) Confirmation of Stage 4, if there some locations could be neglected or seriously damaged.

(2) The propitiate adoption of repair measures could uphold the fundamental function of system.

(3) The proposed projects of preliminary repair could withstand the attack of aftershocks.

(4) The feasibility of proposed projects could be implemented or not and the outcome of the destruction of repaired utilities would worsen the disastrous areas.

6. Stage 6: The restorations of full functions for water supply systems

To solve the pending obstacles of operation and reboot the damaged water supply system back to the minimum capability of function standard, these solutions would contribute to regain the disciplines of social order and resume the activities of economy.

2.3 ADMINISTRATIVE DIRECTIONS

A well organized administration would be composed of four parts: the levels of emergency response, the countermeasures of emergency, the techniques of rapid diagnosis and the information transmission framework of disaster. For better reaction to future challenge, the improvements should be plotted. These will be illustrated below:

1. The levels of emergency response:

According to the classification of the intensity of earthquake by Central Weather Bureau (CWB, Taiwan), the suggested levels of emergency response are described in succeeding Table1.

Table 1 the levels of emergency response

Level of emergency response Intensity of earthquake Ground Acceleration

7：over 400gal Level I Intensity 6 or 7 6：250～400gal Level II Intensity 5 5：80～250gal Level III Intensity 4 4：25～80gal Level IV Intensity 3 3：8～25gal

2. The countermeasures of emergency:

The countermeasures could be tabulated for the each suggested level of emergency response, in Table 1, see Table 2.

Table 2 the countermeasures of emergency

Level of emergency response Countermeasures of emergency The Emergency Operation Center (EOC) would initiate responses in central government with the Level I allocation of manpower and materials in the relief of damage. The local government chiefs would be in charge of the emergency response command post to start Level II all necessary procedures and make a report to central government about the process. The directors of local maintenance units would organize the emergency response squads and Level III recruit all maintenance units, in the same period, make a report to local government about the process. The directors of local maintenance units would Level IV dispatch staffs in reacting to emergency.

3. The techniques of rapid diagnosis:

Due to the breakages and the damages of channels, distributing systems, purifying systems and utilities caused by earthquake and buildings’ collapse, the contaminations and interruption of water supply could be the major problems to be solved. For speeding up the recovery and ensuring the sanitation quality of water, the techniques of rapid diagnosis should be required after earthquake. Acting the primary survey, the visual inspection with check lists, measuring tapes and cameras would bring out the gross view and extent of damages. The checked items should enclose the manholes, pipelines along road side, leakage, abnormal water pressure and suspending of water supply. The categorization of damage could be identified in different levels.

4. Information transmission framework of disaster:

Basically the fundamental information transmission framework would consist of the city-level and township-level emergency response units with the assistance from police and fire departments, community, for the efficiency of operation, most of the framework components should undertake the process by the same web-based information transmission system to provide the latest development and situation for EOC. If the internet access was blocked, the fax machines and telecommunication should be the supplemental channels.

3. REGRESSION MODEL AND PROCEDURES FOR FRAGILITY ANALYSIS

Since the digitized maps are the elementary component of water supply pipelines in vulnerability analysis, but before the Chi-Chi Earthquake, the digitized maps are the research topic in academic filed only without practical applications. In order to study the representative fragility functions to reflect the characteristics of Taiwan , 11 cities/townships were chosen as the primary

study areas in accordance with the Chi-Chi Earthquake, which were Juolan Township, Dungshr Township, Shrgang Village, Fengyuan City, Wufeng Township, Puli Town, Lukang Township, Fusing Township, Huatan Township Mingjian Township and Touliu City (Figure 10). 6 areas of them are passed directly by the Chelungpu Fault while the rest are more than 12 km from the fault. A geographic information system (GIS) database established in MapInfo format was generated for analysis. The basic layer of pipelines network were digitized from the construction blueprints provided by the Taiwan Water Supply Corporation (TWSC), along with length of pipeline, materials and nominal diameters are the necessary input information. The repair data collected from the field investigation and TWSC were also digitized into the GIS database.

Figure 10: The study areas for seismic vulnerability functions

3.1 DISTRIBUTING LENGTHS AND MATERIALS OF PIPELINES

The water pipelines in the study areas were made of several kinds of materials; polyvinyl chloride (PVC), ductile iron (DI), cast iron (CI), steel (S), polyethylene (PE) and others. The pipes with nominal diameters (φ ) less than 65mm were mostly used in connection between the water distribution pipelines and the households, so TWSC did not sketch the exact length of these connecting pipes. As the result, the major focus would be pipes larger than 65mm. The possession ratio of different materials was shown in Figure 11.

Figure 11: Distributing materials of water delivery pipes (φ >=65mm)

For the lack of digitized info-map of pipeline network, an alternative method was introduced to estimate the length of pipeline. In Taiwan’s urban area, the usage percentage of the tap water is almost 100%, and TWSC buried new pipes along the streets to meet the growth of population rather than replaced the old pipes. On the contrast, the population of small townships and cities didn’t increase or even declined in recent 20 years. Therefore, for small cities in Taiwan the street layout can be regarded as the layout of water pipes (φ >=65mm). The previous research revealed that the damage number of smaller pipes (φ <65mm) were about 4 times of larger ones (φ >=65mm) (Shih,

2002). By the ratio of pipe length (φ >=65mm) to street length, according to the population of

100 thousands, the result was about 1 for 100 thousands below, but it would be around 1.6 or abound for larger cities like Fengyuan and Chiayi (in Figure 12).

Figure 12: Pipe length/street length vs. population

3.2 FRAGILITY ANALYSIS OF PVC WATER PIPELINES

The nominal diameters of PVC water pipes were from 25mm (1”) to 500mm (20”). The PVC pipe joints were made by stuffing a glue-coated portion of one pipe into a heated end of another pipe; thus, the assumption was made that the joint was a continuous portion of the pipe body and the joint effect was ignored. Since the peak ground acceleration (PGA) and peak ground velocity (PGV) were the most frequently used parameters for fragility analysis, during the Chi-Chi Earthquake, the strongest vibration was along East-West direction. Therefore the E-W component of PGA and PGV would be the ground shaking parameters suggested by Chen et al. (2002). These PGA and PGV values were obtained from National Center for Research on Earthquake Engineering (NCREE).

3.3 UNIT OF FRAGILITY ANALYSIS

For better understanding of seismic activity around the island, the CWB had accomplished a rapid report system with the intense allocation of sensors. In our study areas, 9 among eleven cities/towns have one strong ground motion station individually. Only Dungshr Township has no station but it is very close to Shrgang Township; therefore, these two towns were considered as a unit. Fengyuan, which is a big city, had two stations; it was then separated into two town units denoted as the east and west sides. Therefore, regression analysis could be conducted by using ground shaking parameters recorded directly from stations other than from interpolation process. The repair rate was calculated by dividing the repair numbers by the pipe length for each town/city unit. Regression analysis was carried out and the results were as follows,

log (RR) = 1.10*log(PGA) - 3.05 (R2=0.77) ...... (1)

log (RR) = 1.57*log(PGV) - 3.52 (R2=0.69) ...... (2)

2 2 in which, RR=number of repairs/km, PGA=cm/s and PGV=cm/s, respectively, and R =square of the multiple correlation coefficient.

3.4 ANALYSIS BY CONSIDERING CIRCULAR ZONE FOR EACH STRONG GROUND MOTION STATION

Since there were 11 strong ground motion stations in our study area and all of them were located in populated districts, we could narrow down the area for calculating repair rate in regression analysis and adopt ground shaking parameters recorded directly from stations. This might be more reliable than methods mentioned above. A circular zone with a station as its center and 2 km in diameter was considered as a basic regression unit. This was selected by attempting different diameters, 1 km, 1.5 km. 2 km, 2.5 km, 3 km and 4 km, and checking the size parameters suggested by Toprak (1998) as well as the resulted R-squared values. The repair rate was calculated by dividing the repair numbers by the pipe length for each circular zone. The regression results were as follows,

log (RR) = 1.66*log(PGA) – 4.7 (R2=0.91) ...... (3)

log (RR) = 2.55*log(PGV) – 5.3 (R2=0.71) ...... (4)

where, RR, PGA and PGV have the same units as equations (1) and (2)

3.5 PREVIOUS ANALYSIS BY USING GRID CELLS (Shih, 2002)

The water delivery blueprints and the repair records of Dungshr Township, Shrgang Township, Fengyuan City, Wufeng Township and Puli Township were completed in 2000 and 2001. Therefore previous analysis was carried out only for these five towns by using grid cells. All water pipes with

φ >=65mm were considered in the analysis. Since PVC possessed about 80%, the results of all water pipes should be close to PVC only. Each city/township was divided into 1km x 1km grid cells after comparing different grid sizes with respect to the threshold area coverage value suggested by Toprak (1998). The PGA and PGV values at the center of each grid were calculated by linear interpolation from the records of strong ground motion stations. Figure 13 showed the grid cells, water pipelines, and locations of repairs and strong ground motion stations at Fengyuan City. In order to perform regression analysis, the grid cells were categorized into groups according to their 2 PGA/PGV values. Each group was made up of grid cells with PGA falling in an interval of 50 cm/s starting from 100 cm/s2or 20 cm/s for PGV starting from 40 cm/s, respectively. For each group, the repair rate (RR, number of repairs/km) was calculated by dividing the total repair numbers by the total pipe length of grid cells forming that group and PGA/PVG was taken as the average value from that group. The regression analysis for RR with respect to PGA/PGV based on groups was then conducted. The results were as below,

log (RR) = 2.13*log(PGA) – 6.1 (R2=0.4) ...... (5)

log (RR) = 3.79*log(PGV) – 7.7 (R2=0.62) ...... (6)

where, RR, PGA and PGV have the same units as equations (1) and (2).

Figure 13: The grid cells, water pipelines, and locations of repairs and strong ground motion stations at Fengyuan City 3.6 THE RESULTS OF REGRESSION ANALYSIS In this study, the seismic vulnerability functions for PVC water pipelines were established by the 1999 Chi-Chi Earthquake. A good correlation between repair rate and PGA/PGV was observed. The fragility curves of equations (1), (3), (5) and Toprak (1998) for RR versus PGA was drawn in Figure 14. Toprak equation was derived from cast iron pipes while this study focused on PVC pipes. Cast iron pipes usually have larger diameter than PVC ones. Therefore, the lower fragility of Toprak equation than our PVC results was reasonable. Figure 15 showed the fragility curves of equations (2), (4), (6) and HazUS (1999) for RR versus PGV. HazUS equations, taken from O’Rourke and Ayala (1993), were for ductile and brittle materials. Because PVC pipes were regarded as ductile in HAZUS manual, a simple comparison could be made as the following. In the Chi-Chi earthquake, the PVC water pipelines did not suffer so much as HazUS predicted; however, the difference was not quite significant. Statistically, the PGA had better correlation with the damage of PVC water pipelines than the PGV. For a 126 scenario earthquake event, we recommended the fragility equation of log(RR)=1.66*log(PGA)–4.70 and log(RR)=2.55*log(PGV)– 5.30 as the damage estimation for PVC pipelines with nominal diameter larger than or equal to 65mm.

Figure 14: Seismic fragility curves of repair rate Figure 15: Seismic fragility curves of repair rate vs. PGA vs. PGV

4. THE DEMAND ANALYSIS ACCORDING TO REGRESSIVE RESULTS

4.1 The emergent countermeasures of repair

The following items would be the necessary components for emergency response:

1. The estimation of required time

The duration of repair depending on the diameters of pipelines, in general, should include excavation, pumping water, disassembling pipelines, cleaning pipelines, recovering service and other factors that might be the critical path. According to the past experience, the durations required for mending are shown as Table 3 below.

Table 3 the time requirement for repairing pipelines in various diameters for Taiwan The diameter of pipeline Required time 90 mm and smaller 1.5 hours 100mm ~200mm 3.0 hours 250mm~400mm 6.0 hours 500mm~600mm 8.0 hours 700mm~800mm 10.0 hours 900mm~1000mm 12.0 hours 1100mm~1200mm 14.0 hours

Based on the table of required time, the sum of working hours from the estimated loss map of water supplying system could be calculated as the rules in Table 4.

Table 4 the relationship of the sum of working hours and repair rate Item Rule The total number of repair in each grid about different pipelines = the grid’s repair rate * the total length of different pipelines in each grid The required time for

repairing The required time for each grid = the total number of repair in each grid about different pipelines * the required time of different pipelines

2. The estimation of the demanded manpower

From the result of the required time in repairing, the demand of manpower could be determined for the effective labor allocation. The demanded manpower calculation would be summarized from the 282 records of water supply units’ dispatch orders after Chi-Chi Earthquake, then; the outcome could demonstrate the one repair hour request 1.2 labors on average. This result would be used to speculate the demanded manpower.

4. The estimation of the allocation and distribution of the demanded manpower

During the large-scale disaster, the prevention of insufficiency in assigning manpower could be avoided by the close coordination with the related engineering organizations. The distribution of the demanded manpower should be considered with the function of rescuing tasks, as the result, a safety factor, 1.25 assumed, should be applied to the demanded manpower.

4.2 The measures of urgent supply after earthquake

To satisfy the basic need of daily lives after earthquake, the fundamental ingredients should be taken into account.

1. The strategic plan of water supply after earthquake

After the attack of Chi-Chi Earthquake, the failure of Shih Kang Dam and the destructions of water supply system brought a hardship of distribution in the Taichung and Natou areas. For the future preparation, the required demand of water after major earthquake should be evaluated in advance according to the experience from Chi-Chi Earthquake.

(1) The estimation of the post-quake deficiency: Grounding on the research (Jiang, 2001), the failure ratio of water supply could be expressed as the function of damage ratio shown in Figure 16. With the estimation equation, the deficiency could be calculated in each grid.

1 Y = ()0473.01 ×+ X − 61.1

Figure 16 the failure ratio of water supply in function of damage ratio (Jiang, 2001)

(2) The estimation of out-of-service population after earthquake: The approximate population in deficiency could be figured out the multiplication of the population inhabiting in research areas with failure ratio.

(3) The recovery ratio and the objective of supply at urgent stage: From the practice of past earthquakes, the restoring degree of damaged water supply system would depend on the amount of involved personnel in repair, the skill of working staff and the performance of equipment in use. The statistic analysis of Kobe Earthquake depicts the recovery ratio and the days after quake in Figure 17. The required supply for life sustenance could be calculated in sequences as Table 5 to set the goals for each post-quake stage.

Y = 80.17 × X 435.0

Figure 17 the relation of recovery ratio and days after quake(高田至郎, 1996)

Table 5 the equations of the required supply for life sustenance after earthquake Item Equation for estimation of demand Repair rate = PGA substituted in the regressive Repair rate equation of repair rate Deficiency ratio = repair rate × equation of Deficiency ratio deficiency ratio Population in shortage at Population in shortage = deficiency ratio × present time Population Population in restored Population in restored service = population in service shortage × the percentage of recovery Demanded supply for life Demanded supply for life sustenance = population in sustenance in a single day shortage × the pressing target of supply per person

4.3 The contingency for the post-quake fire accidents

Beside the supply for life sustenance, the volume for fire extinguishment would be the essential issue to be discussed. The types and number of inflammation would affect the necessary amount for suppressing burning, and the fundamental requirement of fire engines and manpower could be estimated. From the past records of Taipei City, the average volume for quenching each fire is about 150 cubic meters. The post-quake requirement could be evaluated, suggested in Table 6 (Chang, 2004), the multiplication of the simulated result from GIS-based scenario program.

Table 6 the required volume for extinguishing after earthquake (Chang, 2004) Item Equation for estimation of demand Required water for fire Required water for fire fighting = the number of accidents fighting from simulation × the average volume per accident

5. CONCLUSIONS AND SUGGESTIONS

1. The upgrade of the seismic capacity of pipeline joint by improving design code and construction.

From the observation of Chi-Chi Earthquake, the most sever damage of pipeline concentrating in PVCP type due to the fragility of material. As the consequence, the seismic design provisions, including guideline and instruction, of water supply structures and utilities was first released in Taiwan, 2002. It defines the basic regulations and considerations from selection site to construction details. The design guide is based on JWWA’s research mainly

2. The adjustment and innovation of the management policy in selecting material.

With the facts of the aged pipeline systems and the fault line scattering in Taiwan, the possibility of crossing or near fault location is unavoidable. On the ground of risk management, the screening of appropriate material would be important. The following two points are required.

(1) For reducing the risk of quake, the broad collection of geologic analysis data would be the key element for the decision mechanism of pipeline system control.

(2) The utilization of geologic data would provide the potential risk analysis of earthquake.

3. The evaluation and retrofit of the existing pipeline systems.

The ductile joints should be installed to increase the flexibility of pipeline systems in the frequent earthquake area, especially, in district leaning to the high chance of liquefaction. The exposed joints should be adopted with the wrapped anti-leakage device, and the hangers should provide the sufficient supporting. If a bended type, the fixtures are necessary to prevent the excessive dislocation during shaking.

4. The modification of partition in distributing system with the concept of small division

For the complexity of maintaining the pipeline network systems, the capacity of emergency response is fragile to cope with the major earthquake. Acting the concept of small division would minimize the impact and the authorities could offer the effective and efficient countermeasures. In addition to contingency, the concept would benefit the preparatory works before hand. The aim of rapid recovery in service could be easily achieved.

REFERENCE

Chen, W.W, Shih, B.J., Chen, Y.C., Hung, J.H., Hwang, Howard H. (2002). “Seismic Response of Natural Gas and Water Pipelines in the Ji-Ji Earthquake,” Soil Dynamics and Earthquake Engineering, p1209-1214.

Directorate-General of Budget (2003), Accounting and Statistics, Executive Yuan, ROC, http://www.dgbasey.gov.tw/ HAZUS (1999). “Technical Manual, Earthquake Loss Estimation Methodology,” Federal Emergency

Management Agency, Chapter 8. National Fire Administration (2003), Ministry of Interior, Executive Yuan, ROC, http://www.nfa.gov.tw/ O＇

Rourke, M.J., Ayala, G. (1993). “Pipeline Damage due to Wave Propagation,＂ J Geotech Engng., 119（9）, pp.1490-1498.

Shih, B.J. (2002). “Renewed Damage Data and GIS Analysis of Lifelines in the 921 Chi-Chi Earthquake (I),” Research Report Sponsored by National Science Council, ROC, NSC 90-2625-Z-027-002. (Written in Chinese)

Shih, B.J., Chen, W.W., Wang, P.H., Chen, Y.C., and Liu, S.Y. (2000a). “Water System and Natural Gas Pipeline Damages in the Ji-Ji Earthquake—Calculating Repair Rates,” Proceedings of the Taiwan-Japan Workshop on Lifeline Performance and Disaster Mitigation During Recent Big Earthquakes in Taiwan and Japan, June 29-30, Tainan, Taiwan, pp.63-72.

Shih, B.J., Chen, W.W., Chang, T.C., Liu, S.Y. (2000b). “Water System Damages in the Ji-Ji Earthquake - A GIS Application,” Proceedings of the Six International conference on Seismic Zonation, Nov. 12-15, Palm Springs, CA, USA.

Shih, B.J. et al. (1999). “The 921 Ji-Ji Earthquake Investigation Report, Lifeline Damage,” National Center for Research on Earthquake Engineering, NCREE-99-056. (Written in Chinese)

Toprak, S. (1998). “Earthquake Effects on Buried Lifeline Systems,” PHD Dissertation, Cornell University.

Wang, B. (2000). “The Damage Report of Public Water System after the Ji-Ji Earthquake,” Journal of Water Supply, Vol. 19, No. 1, pp. 64-81. (Written in Chinese)

S5-3

“Evaluation of Seismic Upgrade Construction of Hanshin Water Supply Authority” －In the case of pipeline－

Presenter: Keiichi Murakami

(Hanshin Water Supply Authority, Japan)

Evaluation of Seismic Upgrade Construction of Hanshin Water Supply Authority－In the case of pipeline－

Ｋeiichi MURAKAMI, Keisaku DOHSEI, Kazuo OGURA , and Ken’ichi KOBAYASHI

Abstract In the Great Hanshin-Awaji Earthquake, the raw water pipes, transmission and distribution pipelines of the Hanshin Water, a total length of 181 km, were damaged at 129 locations. The water supply amount to the four cities (Kobe, Ashiya, Nishinomiya, Amagasaki City) dropped to 51% immediately after the quake and fell short of satisfying the demand. On the basis of this experience, an improvement plan for the earthquake resistance of the Hanshin Water ’s facilities and construct a “quake resistant, stable water supply system” was established, according to which various projects have been carried out to improve the resistance of pipelines/facilities and construct connecting pipe. Over the past ten years since the earthquake, the resistance of the pipelines has improved from 20% to 50%, as a result of the partial rehabilitation of the pipelines, which it is possible to construct by stopping the supply for a short period. The rehabilitation of raw water pipes and transmission pipelines, some part of which are crucial for maintaining stable water supply, however, has not been completed yet, with low-resistance iron cast/concrete pipes and plain concrete tunnels still accounting for approximately 20% (i.e. 20 km of transmission tunnels and 7 km of raw water pipes) of the total. The transmission pipelines that were newly constructed after the quake are to be connected to the large-capacity transmission pipe currently being constructed by Kobe City. This pipe, when completed, should be able to supply water to disaster-prevention facility and satisfy the needs for emergency water supply. Furthermore, also the rehabilitation of the transmission tunnel is scheduled in earnest, Because the fore-mentioned transmission pipeline will be used as the substitute for this tunnel during the construction. On the over hand, since it is not possible to stop the raw water pipeline of Inagawa Water Treatment Plant Supply System, which performs 80% of supply amount of the Hanshin Water, it is necessary to prepare a substitute facility for securing the required capacity, by constructing additional facilities at a different supply system. Under the recent economic conditions, however, this plan had to be postponed for financial reasons, and therefore, the plans to improve earthquake resistance are not progressing as they were originally scheduled. For this reason, in addition to making further effort to secure the funds and reduce the cost of construction, it is necessary to consolidate the water supply systems from a wide-area perspective and carry out the projects through close cooperation among the water supply and waterworks enterprises so as to achieve a supply system that effectively connects the important supply bases in the event of a disaster.

Ｋeiichi Murakami, Chief of the Maintenance Section, Distribution Division, Operation and Maintenance Department, Hanshin Water Supply Authority, 20-1Nishiokamoto3-chome, Higashinada-ku Kobe 658-0073 Japan Keisaku Dohsei, Chief of the Project Section, Planning Division, Construction Department, Hanshin Water Supply Authority, 20-1Nishiokamoto3-chome, Higashinada-ku Kobe 658-0073 Japan Kazuo Ogura Planning Division, Construction Department Hanshin Water Supply Authority 20-1Nishiokamoto 3-chome, Higashinada-ku Kobe 658-0073 Japan Ken’ichi Kobayashi, Manager of the Planning Division, Construction Department, Hanshin Water Supply Authority 20-1Nishiokamoto 3-chome,Higashinada-ku Kobe 658-0073 Japan

1 １．Introduction The Hanshin Water is a local public enterprise established in 1936 to supply drinking water to four cities (Kobe, Ashiya, Nishinomiya, Amagasaki City) in the Hanshin Region. The facilities of its Yodogawa Supply System (capacity: 373,000m3/day) were completed in 1956 and those of its Daido System (capacity: 595,000m3 /day) were completed in 1972. Furthermore, the Fifth Phase Expansion Project (hereinafter “the Fifth Expansion”) to increase the maximum total water supply capacity to 1,289,900m3/day has been underway since 1978. Presently, The Hanshin Water owns a total of 186 km of pipelines having a total capacity of 1,128,000m3/day and supplies approximately 80% of the total amount of water supplied in the four cities.

In the Great Hanshin-Awaji Earthquake (hereinafter “the previous earthquake”) in January 1995, the heavy damages on the facilities, particularly on the aged pipelines and structures, prevented The Hanshin Water from supplying adequate water. From this experience, the facilities are presently being rehabilitated according to the established a plan to improve earthquake resistance of Facilities. This paper discusses the achievements and the future issues of these rehabilitation works of the past ten years since the previous earthquake, with a focus on the improvement of pipelines.

Fig1. General Map of the Facilities of Hanshin Water Supply Authority

2. Establishment of the Plan to Improve Earthquake Resistance of Facilities 2-1) Damages of the Previous Earthquake Table 1 shows the condition of damages and the period required for restoring the major facilities. The damages were found at 6 locations of major structures other than those constructed by the Fifth Expansion. They involved cracking of the concrete structures at the places where stress was concentrated and breaking of expansion joints of the basin-type structures. １） Water leakage occurred at 129 locations along the weak and aged pipelines (mostly concrete or cast iron pipelines) unequipped with joints with flexibility or expansion functions.

2 Table 1. Major Damages by the Great Hanshin-Awaji Earthquake and Number of Days Required for Restoration

Number of days Name of facility Type of damage required for emergency repairs Yodogawa Raw Leakage from ψ1,200mm concret pipes 33 days Water pipes (23 locations) Damage on expansion joints of Nishinomiya Management Center (2 locations) 10 days Ponping Station Leakage from transmission/distribution pipes within the station (12 locations) Nishinomiya Slippage ofψ1,200mm ductile cast iron 10 days Transmission Main pipes mechanical joint (1 location) Flocculation basins: Protrusion of inlet conduit duct, damage to expansion joints, damage to foundation piles, etc. Sedimentation basins: Damage to Operation stoppage expansion joints, submergence of mud Inagawa Water impossible, large collecting facility, damage to foundation Treatment Plant number of damages, piles, etc. approximately 70 days Sand Filtration basins: Damage to foundation piles, submergence of flow meter, cracking on sidewalls of outlet conduit ducts, etc. Koto Transmission Slippage of ψ1,500mm ductile cast iron 15 days Main pipes mechanical joints (5 location)

The slippages of joints were caused by excessive ground motion beyond the allowable value of the ductile cast iron pipelines. Leakages also occurred not on the steel pipes themselves but on the expansion and flexible pipe of the water pipe bridge, due to a large relative dislocation of the substructure. Finally, in the concrete tunnel section the inserted FRP pipes were deformed and the covering concrete collapsed.

The amount of water supply between 24 hours and 48 hours after the outbreak of the previous earthquake (i.e. from 6 am of January 18 to 6 am of January 19), that is, during the 24 hours immediately after maximizing the capacity of the facilities and switching the supply systems, was 531,640m3/day, or approximately 51% of the maximum supply prior to the quake, which was 1,048,000m3/day. Because of the time required for completing the emergency repairs, the supply was not recovered to the original level until February 19.

2-2) Outline of the Plan On the basis of the lesson learnt from the previous earthquake, the Committee for Improving the Earthquake Resistance of Hanshin Water’s Facilities, comprised of experts, was established in May 1995 with an aim to construct a reliable drinking water supply system by making proposals for the implementation of the resistance

3 improvement plan2). The proposals were reviewed within The Hanshin Water before finalizing the details of the rehabilitation and maintenance works.

2-2-1) Targets In devising the plan, it was assumed that the subject earthquake, including marine earthquake, would have seismic motions equivalent to those of the previous earthquake. In the previous earthquake, it took approximately one month to recover the amount of supply back to the original level. In response to the requests from the citizens, the the Kobe Waterworks Bureau and the Nishinomiya Waterworks Bureau have respectively set their targets as “completion of emergency repairs in less than four weeks” ３） and “water supply to every household in three weeks.” ４） Likewise, the targets of the plan to improve earthquake resistance of facilities set by The Hanshin Water are as follows: ① construction of a system that is capable of supplying the needed amount of water even immediately after an earthquake, ② restoration within one week, should any part of the facilities be damaged, and ③ strengthening of the mutual information and communication system. In order to promote these targets on a practical basis, the projects were classified into short-term, mid-term and long-term, with the target water supply rate and days required for restoration set for each as presented in Table 2.

Table 2. Targets of the Plan to Improve Earthquake Resistance

Supply rate Number of days immediately after required for earthquake emergency repairs Hanshin-Awaji Earthquake (1995) 50% 1 month Short-term (-2000) 80% 14 days Mid-term (-2010) 90% 10 days Long-term (-2020) 100% 7 days

2-2-2) Improvement of Earthquake Resistance of Pipelines The plan to improve the earthquake resistance of pipelines (see Table 3) was prepared by considering the progress of the ongoing expansion project and the present state of renewal of aged facilities. Project priority was determined by considering the pipe type and the year of installation of respective pipelines, as well as the availability of alternative pipelines and the need for securing a substitute source of water supply during the execution of the project. Although K-type ductile cast iron pipes are not equipped with any quakeproof joints, their renewal was assigned low priority since they were little damaged in the previous earthquake. Furthermore, since some damages will inevitably occur even when the aged pipe are totally replaced or renewed for a greater resistance against earthquakes, also specified in the plan are the projects to effectively connect the pipelines and give redundancy to the routes in order to improve the reliability of the entire supply system as a total network.

4 Table 3. Plan to Improve Earthquake Resistance of Pipelines

Pipeline Diameter (mm) Type of pipe Length Improvement of earthquake resistance of damaged pipelines, etc. Yodogawa Raw Water Pipes φ1200・φ1350 Concrete/Cast iron pipes 12.2 km Amagasaki Transmission φ1350 Concrete/Cast iron pipes 6.4 km Main Short-term Kobe Transmission Main φ1,650～φ2,400 Newly installed 7.3 km Raw water connecting pipes between water treatment － Newly installed － plants Improvement of earthquake resistance of pipelines expected to be damaged Daido Raw Water Pipes φ1,350・φ1,800 Concrete pipes 6.5 km Inagawa Transmission Main φ1,650 Concrete pipes 4.3 km Mid-term Transmission connecting pipelines between pumping － Newly installed － stations Long-term Construction of wide-area mutual backup network among major facilities

Lenght （ｍ） Pipe-in-pipe ※5684 ※10244 4000 Pipe Replacement Pipe reverse/Hose-lining

※Including disaster restoration works after 3000 the Great Hanshin-Awaji Earthquake ※

2000

※※ 1000

0 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004

Year

Fig. 2 Adopted Pipe Rehabilitation Methods

3. Pipeline Rehabilitation Works Because of the urgent need to prevent the leakages, the partial reinforcement of pipe joints had been carried out since 1970. The overall rehabilitation of the pipelines began in earnest in 1980, and the construction of earthquake resistant pipelines started after the previous earthquake of 1995.

5 3-1) Pipeline Rehabilitation Prior to the Previous Earthquake As shown in Fig. 2, the rehabilitation, mainly of small diameter pipes, using the hose-lining and pipe reverse methods began in 1980 to prevent red water and leakages. The pipe replacement by open cut method using ductile cast iron pipes began in 1992 and earthquake resistant pipes (S, SII and KF types) have been used since 1993. Since large-diameter pipes are laid buried under major trunk roads, they were replaced with steel pipes by the pipe-in-pipe (hereinafter ”PIP”) method which has a relatively little influence on the surrounding areas.

3-2) Restoration Works The emergency repairing of the pipelines to stop the leakages was carried out immediately after the previous earthquake. It was not until after the completion of these repair works that the additional works were implemented to improve the earthquake resistance of the pipelines, particularly the resistance of those that were heavily damaged in the previous earthquake due to aging or to the complex ground characteristics of the surrounding area.

10 km of large-diameter pipes was inserted with steel pipes by using the PIP method and 7 km of small-diameter pipes was replaced with S-type ductile cast iron pipes and others. The works for these rehabilitations started in 1995 and lasted three years.

3-3) Short-term Projects Yodogawa Raw Water Pipes (hereinafter “RWPs”) was completed the rehabilitation by PIP method in 2002 later than previously planned. Since a new leakage had been found on the Yodogawa intake pipe which priority is higher than Raw Water Pipes. As for Amagasaki Transmission Main, in addition to the conventional steel pipes, the ductile cast iron pipes equipped with a newly developed earthquake resistant joint were partly used for the rehabilitation and the rehabilitation of the entire pipe, by employing the PIP method, was completed in 2003. The new pipeline connecting Inagawa and Amagasaki Water Treatment Plants, however, has not been started yet.

The project to redirect a part of the route of the newly constructed Kobe Transmission Main and connect it to the Large Capacity Transmission Pipe ５）which is being constructed by Kobe Waterworks Bureau, was completed in 2001. This Large Capacity Transmission Pipe, is intended to serve as a water source for supplying water to the disaster prevention facilities, as well as emergency water supply directly to peaple, in the event of a disaster. The Kobe Transmission Main, on the other hand, is connected also to the distribution mains in other cities of the consortium (i.e. connected to Nishinomiya at one point and Ashiya at two points) to ensure a greater emergency water supply capacity in the event of a disaster, and, at the same time, has the capacity for serving as a backup route during the rehabilitation works of the existing transmission tunnels of The Hanshin Water.

3-4) Mid-term Project Daido RWPs is a trunk pipeline leading to Inagawa Water Treatment Plant, where approximately 80% of the total water supply by The Hanshin Water is being carried out. While there had been several instances of leakage along this RWPs in the past, it is

6 difficult to renew its pipes on a full-scale because the operation cannot be stopped for a long period of time due to the lack of R RWPs backup capacity. For this reason, the leakages so far were corrected with an emergency repair of merely reinforcing the interior of the joint sections. At Inagawa Transmission Main, the rehabilitation of the pipes by PIP method is performed only during winter, when the demand for water decreases. The rehabilitation is scheduled to complete in 2005. The new connecting pipeline between Nishinomiya and Koto Pumping Stations has not been started yet.

3-5) Earthquake Resistance After 10 Years Since the Previous Earthquake In the ten years since the earthquake, the quake resistance rate has risen to 49% as shown in Table 4. It must be noted, however, that while the rate increased by around 18% in the first three years immediately after the earthquake, that is, up to 1998 (completion of disaster restoration project), the increase in the subsequent eight years remained at only about 16% because the plans had to be modified for financial reasons.

Table 4. Changes in Pipeline Length and Earthquake Resistance Rate by Pipe Types

(unit: m) Pipe type 1995 1998 2005 Asbestos-cement 9,632 8,636 1,758 Cast iron 33,384 18,662 11,148 Concrete 29,999 17,812 10,493 Concrete tunnel 20,558 20,558 20,558 Ductile cast iron 61,223 56,792 51,494 (without quakeproof joint) Ductile cast iron 1,420 12,632 22,244 (with quakeproof joint) Steel 24,880 46,847 68,063 Total 181,096 181,939 185,758 Earthquake resistance rate 14.5% 32.7% 48.6%

Note: Earthquake resistance = (Ductile cast iron pipe with quakeproof joint + Steel pipe) ÷ Total length

Fig. 3 shows a map of major pipelines that are still in need of rehabilitation works and Table 5 presents the breakdown of pipe types employed in those pipelines. A total of 44km of pipelines, or approximately 20% of the entire pipelines, still require rehabilitation and around 70% of those pipelines are constructed with large-diameter pipes. Aside from the concrete tunnel, which is planned to be backed up by Kobe Transmission Line and others during the works, the pipelines in question are not supported by adequate backup capacities to substitute their supply during the construction works, and therefore, it is difficult to rehabilitate in earnest.

7 Fig. 3 A Map of Major Pipelines Required Rehabilitation

Table 5. Lengths of Major Pipelines Required Rehabilitation

(unit: m) Pipe type RWPs TL DL Total Asbestos-cement 1,758 1,758 Cast iron 350 305 10,493 11,148 Concrete 6,579 3,707 207 10,493 Concrete tunnel 20,558 20,558 Total 6,929 24,570 12,458 43,957 RWPs:Raw Water Pipes TL:Transmission line DL:Distribution line

4. Evaluation of Earthquake Resistance Improvement 4-1) Evaluation Method As the method of evaluating the improvement in earthquake resistance, the supply rate immediately after an earthquake and the number of days required for restoration were calculated on the basis of the records of damages and restoration period required after the previous earthquake, and the obtained values were compared with the target values set in the Plan to Improve Earthquake Resistance presented in Table 2. ６） The supply rate immediately after an earthquake was expressed by the ratio of the possible amount of supply to the present total maximum capacity of the facilities (1,128,000m3/day). The possible supply amount was set as being equivalent to the amount of water supplied on the second day after an earthquake, because comparatively stable amount supplied. Furthermore, in calculating the possible supply amount of supply for the above ratio, it was assumed that the capacity of the damaged facilities would be halved after an earthquake. An allowance was set for 25% extra capacity for the facilities constructed by the Fifth Expansion Project and 10% for all the other facilities.

8 Table 6. Set Values for the Number of Days Required for Emergency Restoration of Pipelines (Assumption: One damage/pipeline) Pipe type Asbestos- Item cement Cast iron Concrete Cast iron φ400 φ1500 φ1500 Time required according to type of work Preparation 666 Drilling 624－ Unwatering 2 5 168 Pipe connection 89－ Interior repair －－30 Flooding 61616 Total (hours) 28 60 220 Set value for the number of days of 1.2 2.5 9.2 restoration

The respective numbers of days required for emergency restoration by pipe types were set as shown in Table 6. Since concrete pipes is different with of the castle iron pipe that of ductile cast iron pipes, it is difficult to perform the repair merely by cutting the leaking sections. Therefore, it is necessary to repair the leakage from inside the pipe, by closing the valves on both ends of the damaged section and discharging the entire water inside it. Furthermore, with regard to the cases of more than two damages on a single pipeline, it was decided that the total number of days required for restoration would be calculated by adding the time required for repairing the additional damages.

4-2) Examination of Different Scenarios of Damages by Earthquake 4-2-1) Preconditions First of all, it was assumed that the structures without adequate earthquake resistance and the aged asbestos-cement, concrete and cast iron pipes would most easily suffer damages by an earthquake.

The three cases (scenarios) presented in Fig. 4 are all based on an assumption that two structures (Yodogawa Pumping Station and Inagawa Water Treatment Plant) and one pipeline are damaged simultaneously. The one pipeline that is assumed to suffer damage is an Daido RWPs in Case 1, an Amagaski Transmission Main, in Case 2 and Chubu Distribution Main in Case 3.

Fig. 4 Diagram of Case Studies

4-2-2) Calculation of Restoration Period In the previous earthquake, it took 10 days to restore Nishinomiya Pumping Station. Since Yodogawa Pumping Station was built around the same time as Nishinomiya Pumping Station, the expected number of days required for its restoration was set also as 10 days. The rapid filtration basin７) is likely to be damaged since its earthquake resistance has not been improved in the course of restoration. For this reason, it was assumed that the restoration of the entire treatment plant would take around 2 weeks.

The respective period for restoring pipelines was determined by referring to their damage rates recorded after the previous earthquake. Table 7 shows the assumed extent of damages and the assumed number of days required for restoring respective pipelines1). Consequently, it was assumed that it would take 11 days to complete the emergency restoration of Daido RWPs and 9 days for Amagasaki Transmission Main (both comprised of large diameter pipes) and 5 days for Chubu Distribution Line (comprised of small-diameter pipes).

10 Table 7. Assumed Damage of Pipeline

Number of Damage Diameter Subject days for Damaged pipeline Pipe type Locations (mm) length(km) restoration (locations) (days) Daido Law Water Concrete φ1,800 3.6 2 11 Pipes Amagasaki Concrete φ1,200 0.6 1 9 Transmission Main Asbestos- Chubu Distribution cement/ φ200～φ800 4.1 6 5 Line Cast iron

4-2-3) Calculation of Supply Rates With the above assumptions, the supply rate after a disaster for each of the three cases was calculated. The results of calculation are presented in Table 8. If the required amount of water supply is 1,128,000m3/day, or equivalent to the maximum capacity of present facilities, the greatest shortage of 262,000m3/day and the lowest supply rate will be experienced in Case 1 where the damages are made along Daido RWPs.

Table 8. Supply Rate Immediately After Earthquake and

Assumed Amount of supply Supply rate(%) damage shortage(m3/day) CASE １ 262,000 77 CASE ２ 245,000 78 CASE ３ 169,000 85

4-3) Evaluation of Earthquake Resistance Improvement Project While the assumed number of days required for restoring Inagawa Water Treatment Plant has to be taken into account in estimating the number of total days required for emergency restoration of all of the three model cases, the maximum number of days required for restoring the pipelines alone is 11 days. Moreover, the supply rate immediately after an earthquake averages 80% as shown in Table 8. Therefore, it can be said that the targets set as the goals of short-term projects planned up to year 2000 have been generally achieved at present. However, because of the longer time required for achieving these targets, the rehabilitation of the pipelines is expected to take longer than it was originally planned. To prevent the delay in the earthquake resistance improvement works, additional efforts, including the effort to explain the necessity of such projects to the users, are highly called for.

Furthermore, the citizens will be affected if the water supply is completely stopped even for a short period of time, and in the event of a disaster, it is important for them to have

11 access to reliable water supply, even in smaller amounts. Since there are multiple pipelines installed along the aqueduct and the transmission line, the water supply may reduce but is highly unlikely to completely stop. On the other hand, when the distribution pipeline is damaged, since The Hanshin Water has no backup pipelines, the stoppage of water supply will be inevitable during the restoration works, and therefore, the citizens in the subject area will be greatly affected. For this reason, it is considered important to evaluate and examine the earthquake resistance improvement not only from the supplier’s side but also from the citizens’ perspective.

5. Future Issues 5-1) Redundancy of Water Supply System For The Hanshin Water serving as the water source of the four cities to ensure adequate water supply even after a disaster, high priority must be given to the reinforcement of trunk key pipelines. However, because of the lack of redundancy among the facilities, there are no means of backing up the supply, and therefore, it is not possible to carry out the rehabilitation works by stopping the water supply for a long period of time. In the recent years, when the emergency work was performed to repair a newly found leakage, the supply from the facilities respectively owned by the 4 cities had to be increased to compensate for the loss in supply. In other words, The Hanshin Water and its facilities as a system are not properly fulfilling their function of supplying enough water.

Furthermore, as it is mentioned above, since the short-, mid- and long-term projects have been either modified or postponed for financial reasons, the construction of connecting pipes has not been started yet. In order to improve the supply rate and reduce the period required for restoration according to the Plan to Improve Earthquake Resistance, it is regarded important to secure alternative sources of water supply and add redundancy to the overall system by promptly constructing connecting pipes among different supply systems and water treatment plants.

5-2) Disaster Prevention Function As shown in Table 5, 30% of the pipelines that will have to be rehabilitation in the future are distribution pipes. While the amount supplied by these pipes are less than that by the trunk pipelines, since the distribution pipelines are directly connected to the distribution pipelines of the four cities, their rehabilitation must be carefully planned by thoroughly discussing their routes and the timing of construction with the subject cities. In making the plan, considerations must be given also to the improvement of services to the citizens. By giving high priority to the security of water for firefighting and medical uses, it is necessary to improve the level of service provided to citizens as protecting the their lives and properties in the event of a disaster. The cost of improving the earthquake resistance of distribution pipelines is less and the period of construction is shorter compared to those of key pipelines. The renewal of distribution pipelines, therefore, is an effective way of contributing to the improvement of the earthquake resistance rate.

With regard to the mid-term and long-term projects, it is considered important to adjust them to the renewal projects of other waterworks enterprises and cooperate with the authorities in charge of disaster prevention in order to prepare an effective plan for

12 improving the earthquake resistance particularly of the pipelines that have a high disaster prevention effect and serve to provide the needed supply of water８） in the event of a disaster.

6. Conclusion After the Great Hanshin-Awaji Earthquake, supported by the Ministry of Labor and Welfare and other national and administrative bodies, the earthquake resistance of the pipelines was improved through the restoration from disaster projects and the projects to rehabilitate the aged pipelines, and the initial short-term targets have been generally achieved so far. Given the severe economic conditions of the recent years, in order to continue with these costly and time-consuming these projects, it is important to exert additional effort to raise the funds and reduce the costs. At the same time, in order to support safety and well being of the citizens, it is necessary to reexamine the supply system from a wide-area perspective and give considerations to the construction of a life-supporting line that connects the important water supply bases in the event of a disaster. It is believed that it will become increasingly important for the water supply and waterworks enterprises to closely work together in order to effectively carry out the projects and maximize the investment efficiencies.

References １)Kazuo Mishima, "Restoration and Anti-seismic Measures of Water Supply Facilities of Hanshin Water Supply Authority" U.S-Japan Anti-Seismic Measures Workshop 2000 ２ )The Hanshin Water Supply Authority Facility Seismic Capacity Improvement Plan Committee, “Proposal for the Seismic Capacity Improvement of Facilities” June 1995 ３) Kobe Waterworks Bureau Restoration Improvement Plan Committee, “Seismic Design and Construction for Water Supply Facilities” June 1995 ４) Nishinomiya Waterworks Bureau Restoration Improvement Plan Committee, “Seismic Design and Construction for Water Supply Facilities” June 1995 ５)Shozo Morita, ”Re-engineer of Waterworks System for the New Century – Kobe’s New Waterworks system Centering on Large-Capacity Water Mains” The 5th International Symposium on Water Supply Technology, November 2000 ６ )Takashi Hanamoto, “ A Preliminary Evaluation of the Water Supply System Restoructurization” The 5th International Symposium on Water Supply Technology, November 2000 ７)Keiichi Murakami, "Anti-Seismic Measures of Existing Water Supply Facilities －A case study of an- anti-seismic plan of Inagawa water treatment plant－" U.S-Japan Anti-Seismic Measures Workshop 2003 ８ )Shirou Takada, “ Waters system as life-supporting lifeline ” The SUIDO KORON 、 September 2002

13 S5-4

“Current Status and Subject of the Seismic Upgrade of Kobe Water System after Ten Years from the 1995 Hanshin-Awaji Great Earthquake”

Presenter: Tetsuro Kijima

(Kobe Municipal Waterworks Bureau, Japan)

Current Status and Subject of the Seismic Upgrade of Kobe Water System after Ten Years from the 1995 Hanshin-Awaji Great Earthquake

Tetsuro KIJIMA Kobe Municipal Waterworks Bureau Planning division

1.Introduction Kobe Water System was started in 1900, and is the seventh modern water system in Japan. At that time, it could supply water to 250,000 customers, 25,000 m3 per day. It has been supporting the citizen of Kobe and industrial activity by supplying safe and good water for 100 years with proper maintenance. Kobe Water suffered serious damages by the Hanshin-Awaji Great Earthquake in 1995 and we received nationwide aids. But restoration took ten weeks and it had big influence on the citizen. Kobe Water adopted “Basic Plan for Earthquake-resistant Kobe Water System”. This Plan is reflected on the lessons learned from the earthquake and requests from citizens by the process of the restoration. These are two of five aims in the Plan.. (1) To complete temporary restoration within 4 weeks. (2) To increase emergency water supply with the passage of time. Aiming to transform present system into “disaster-resistant and easy-to-repair” water system, we settled the seismic plan during 15 years. Emergency Water Supply System, Large Capacity Transmission Main (LCTM), and Seismic Pipe Network System are three Major reconstruction projects. In recent years, the amount of water supply and the water supply profit are decreasing by the spread of improvement in the water-saving consciousness of a company, and domestic water-saving apparatus etc. On the other hand, the reconstruction of the water supply system that passed 100 years such as the management on the citizen’s viewpoint, countermeasures to meet the aged facilities are required. Therefore investment decision is very important and will be made correctly. When we decide investment, explanation to the customer about the reason why we invest and what the benefit is, are very important. In this report we would like to evaluate and mention about the present situation of these projects after ten years from the earthquake disaster and

1 also describe about the system extension beyond city border, the mutual aid training, emergency water supply training with citizen. Finally, we mention about the subject such as water resources and measure to the aged facilities.

2. Overview of the Kobe Water Kobe is a city that is full of the charm and it is as warm as the average temperature of 17 degrees Celsius, and there is little rain, is blessed with rich tourist attractions. Kobe prospered as a trading port for many years. Kobe city was born by population of 130,000 in 1889. Now, it has developed into the city where 1,510,000 people live in an area of 550km2. 2.1 History of Kobe water Kobe water constructed three dams (Nunobiki, Karasuhara, Sengari) from 1897 to 1931. They are dams only for water service. But it could not catch up with the water demand because the population increases rapidly with city development. The water from Lake Biwa and Yodogawa ensured stable water supply. It was about 40 years after since Kobe water was started. Henceforth, construction of water transmission tunnel became important as well as maintenance of distribution system with expansion of a city region. It has 900,000m3 of water supply capability, and life and industrial activities of Kobe are supported now. Table-1 Kobe Water Supplies as of April ,2004 Facilities Number Capacity Reservoirs 3 Effective volume 13,000,000m3 Purification Plants 6 Filtration capacity 250,000m3/d ay Pump Stations 48 Distribution Reservoirs 123 Effective volume 550,000m3 Distribution pipes (km) 4,841 Include the transmission tunnel48km

2 Table-2 Daily Water Supply Capacity (m3/day) Hanshin Water Authority 672,000m3 Hyogo Prefecture Water Supply Project 28,000m3 City-owned water sources 200,000m3 Total 900,000m3 2.2 The feature of Kobe water Kobe city is divided by Mt. Rokko which runs to the east and west of a city region into 3 areas, Shigaichi,Hokushin, and Seishin. Although Kobe water conquered the geographical feature which is rich in diversity. The feature of the system is as follows. 2.2.1 Water Source and Transmission Tunnel 75% of Kobe water comes from the Hanshin Water Authority with primary sources including Lake Biwa and the Yodogawa River. 3% of water comes from Hyogo Prefecture Water Supply Project in the Hokushin area and the Seishin area. The remainder comes from Sengari Dam, Nunobiki Dam, Karasuhara Dam and rivers and springs in the city. The transmission tunnels which are two open channel which penetrate the Mt. Rokko from east and west has played the important role as trunk which supplies water from Hanshin Water Authority to Seishin area. 2.2.2 Layer and Block Distribution System Because of the geographical feature, a large number of facilities, such as 123 distribution reservoirs and 48 pump stations, are required. Water runs by gravity from the distribution reservoir. The water distribution system is divided into layers at 30m each and also divided by into blocks at 2 to 4km each in Shigaichi area.

From Aono dam 県営水道 From Dondo Dam 県営水道 (Hyogo Prefecture) (Hyogo Prefecture)

Urban district-Hokushin conveyance facilities

阪神水

Distribution Rsservoir Large Scale Transmission Main From the Yodo river Transmission Tunnel (via HWSA) transmission Main

Large Scale Transmission Figure-1 Schematic of Kobe Water System

3. Damages and Restoration Plan 3.1 Damages of Kobe water The Hanshin-Awaji Great Earthquake occurred at 5:46 a.m., January 17, 1995. (Ms 7.2) The earthquake caused a lot of terrible damage to lifeline and water supply was cut off simultaneously by the power failure and damage of facilities. As shown in Table-3, it was comparatively few in purification plants and distribution reservoirs and concentrated on distribution pipes and service pipes.

Table-3 Damage overview of Kobe Water System Facilities Damages Expense for repair (damaged/total) Dam 1 /3 dams (Billion yen) Purification Plant 2 /7 plants Raw water conduit 2 lines /43km 7.0 Transmission main 6 lines /260km Distribution Reservoir 1 /119tanks 1.9 Distribution Pipes 1,757failures /4,002km 13.5 Service Connection 89,584 repairs/65,000lines 2.5

4 Others(including Head office, Tobu branch, 4.1 Building) etc. Total 29.0 Since approximately 90% of the pipe in the distribution system was ductile iron or steel pipe, the pipe break was only 17%. But the largest cause of failure was joint separation and it was about 55%. The breakage of heavy fittings (for example air release valve) of the transmission main, was conspicuous. Most of service connection failures were caused by pipe breaks and joint separation following fall down of houses and road deformation. This tremendous number of failures was the main reason of rapid water loss from distribution pipeline. Moreover, although we tried hard to identify leakage points, as water pressure was low, it was extremely difficult. It was one of the main reasons to take 10 weeks for recovery. We received support from many workers. It mounted up 410,000 workers. The emergency shut-off valves closed on 18 of 21 distribution reservoirs following the earthquake. We intended to secure water as one-person one-day 3L and secured 42,000m3 volume of water in total. This amount is equivalent to the amount need in Kobe City for 9.3 days. Aid from many cities and the Self-Defense Forces, etc strengthened emergency water supply and it was continued till the end of March. 30,000 people and 14,000 tank trucks in total provided emergency water supply. Traffic congestion and collapsed houses covered the road were to be reasons why emergency water supply from the distribution reservoir to evacuation centers by tank trucks had been disordered. Little public relation to the citizen about water-supply truck caused number of complaints. The amounts of water which citizens need -- lives, such as not only for drinking but also for toilets, for baths, etc., -- were increased. And much amount of water for hospitals including for dialysis was demanded. 3.2 Restoration Plan Kobe Water established the "Kobe City Waterworks Restoration Planning Committee" early in March. Based on the proposal of the "Guideline for Earthquake-resistible Waterworks in Kobe City", Kobe Water decided upon the "Basic Plan for Earthquake-resistant Kobe Water System” in July

5 The Guideline is aiming to transform present system into “Disaster-resistant and Easy-to-repair” water system and it defines the directivity of Kobe Water. It is based on the customers’ voice collected during the disaster and reflects the experiences during the disaster. Targets are (1) To complete temporary restoration within 4 weeks. (2) To increase emergency water supply with the passage of time (3) To secure water at the disaster prevention center etc. We made “Basic Plan for Earthquake-resistant Kobe Water System” as a master plan of Kobe Water to achieve the targets shown in the Guideline. The contents are maintenance of (1) Emergency Storage System. (2) Large-Capacity Transmission Main (3) Seismic Pipe Network System (4) Seismic Upgrade of Principal Facilities (5) Seismic Upgrade of Monitoring and Controlling Facilities, etc. In Kobe Water, the enterprise, which should be improved preferentially before 2010, is decided upon and undertaken.

Table-4 Seismic-upgrade Program for 1996-2010 Seismic Upgrade Main Purpose & Contents Investment Items (Billion yen) Emergency Storage Securing Emergency Water 0.8 System Maintenance of Storages Large Capacity Backup for the Existing Transmission 45.0 Transmission Main Tunnels 13.8km Seismic Pipe Network Reducing recovery period 50.0 System 400km Seismic upgrade of Upgrade of Aged Facilities 18.5 Principal Facilities Maintenance of transmission pipe Seismic upgrade of Improve reliability 15.0 Monitoring and Upgrade of Telemeter Facilities Controlling Facilities Others Monitor of Water Quality 1.7 Maintenance of Monitoring Facilities of Water Quality Total 131.0

4.Enforcement and Evaluation Emergency Storage System, Seismic Pipe Network System, Large Capacity Transmission Main are three major reconstruction program. Progress of the seismic upgrade projects of Kobe Water is shown in Table-5. We will introduce progress of these projects and evaluate their results. Table-5 Progress of Seismic Upgrade Project （Billion yen） Project Total March 2004 % Investment Emergency 0.8 0.18 23 Storage System Large Capacity 45.0 13.73 31 Transmission Main Seismic Pipe 50.0 21.16 42 Network System Others 35.2 16.46 47 Total 131.0 51.53 39

4.1 Emergency Storage System This system is aiming to secure emergency water in the distribution reservoirs and Large Capacity Transmission Main in early stage of a disaster. And it is used for as a water distribution point or a tank truck operation base. It is a main project for emergency water supply. This system is intended to secure amount of water of 3L/person/a-day × 7days including the water for the evacuation center and the hospitals, etc. Although it was planed to improve 33 systems covered by circles of 2km radius before the earthquake, we made a change to 47 systems. 34 systems are installed at the end of the 2004 fiscal year. Four systems are scheduled to install in Seishin area in four years. In addition to this, five systems are installed in collaboration with the public construction project bureau in Kobe City.

7 Covered Area Installed Planned

→

March 1995 Installed 21/Planned 33

March2003 Installed 34/planned 47 Figure-2 Emerrgency Storage System

Furthermore, Kobe City adopts Seismic upgraded cistern for fire protection or Rain water storage system, Use of the advanced treated water of sewer, and Registration system of a well, etc. Based on lessons learned from the Hanshin-Awaji Great Earthquake, Kobe City organize " Community for Disaster Prevention and Welfare " in order to make citizen can support each other and prevent disaster in everyday life. It is formed for every elementary school division. There are 183 communities. Members of the community have communication through activity of disaster prevention training, making hazard map, crime prevention patrol, etc. Kobe Water supports emergency water supply training of the “Community for Disaster Prevention and Welfare” positively so that emergency water supply at disaster can be done smoothly. 4.2 Large Capacity Transmission Main In addition to the existing two transmission tunnels inside Mt. Rokko, Kobe Water plans Large Capacity Transmission Main (LCTM) underground deeply across urban area. LCTM is the facilities with various functions such as diffusion of risk of existing two tunnels, large storage-of-water capability, an alternative facility at the time of renewal of existing tunnels. And LCTM enables the emergency water supply and early restoration. It is the Kobe project after the great earthquake.

Table-6 Overview of Large Capacity Transmission Main Total Plan Ashiya Boundary to Myoudani Pump Station 30.4km Stage 1 Ashiya Boundary to Okuhirano Purification Plant Length: 13.7km Diameter: 2.4m Conveyance Capacity: 400,000m3/day Investment: 45Billion yen Period: 1996 to 2010 Storage Capacity: 59,000m3 The section between Ashiya boundaries and Sumiyoshi River Shaft (3.8km) was completed, and has started water transmission. Two shafts have emergency water supply facilities, and the “Community for Disaster Prevention and Welfare” practices emergency water supply training. “To complete temporary restoration within 4 weeks” is mentioned as the one of the targets in the Guideline. In the section in use, the restoration period is estimated about 6 weeks based on the situation of distribution pipe network of the 2003 fiscal year Furthermore, in the investment effect analysis of this section, by the manual of Japan Water Works Association, Benefit by Cost (B/C) is calculated as 1.87. In this calculation method all values are converted to Net Present Value. For instance, “Emergency water” is replaced into bottled water for 50 years. Fire suppression, Reduction of recovery period, Disruption of existing tunnels and so on. Although it is not satisfied enough because other important benefits such as loss of prolonged disruption are not considered, it can be utilized as one of the indices for a citizen. Now, the section between Sumiyoshi river to Okuhirano purification plants (9.9km) is under construction , and it will finish in 2010. In some sections, we have a plan to apply law for “the usage of large depth underground”. 4.3 Seismic Pipe Network System

9 Kobe Water has an aggressive pipeline replacement program to ductile iron pipe that began in 1962. Consequently, pipe failures and red water have been decreased and pipe network has been the earthquake-resistible. Change of Pipe Materials and Types are shown in Figure-3.

1970 1276 296 Level 3 (CIP,ACP,VP） Level 2 (DIP) 1980 684.8 1618.6 65.7 Level 1 (S-DIP,Steel）

1995 445 3198.2 358.8

2000 395 3112 805.6

2002 357.9 3030.3 1075.4

0 500 1000 1500 2000 2500 3000 3500 4000 4500 5000 Pipe length(km)

Figure-3Change of Pipe Materials and Types

March 1995

March 2003

Figure-4 Replacement of aged pipe(1995→2003) Seismic Pipe Network System (SPNS) is carried out so that the emergency water supply and restoration at the disaster can be performed promptly. The goals of this project are (1) Seismic upgrade of distribution main pipe by adopting 500m grid in order to perform pipeline which constitutes 500m grid in order to shorten

10 the period of restoration, to increase water for living use , and to shorten of conveyance distance. (2) Seismic upgrade of pipeline to disaster prevention bases, such as the evacuation center, e.g. schools, and a core hospital. (3) Installation of the emergency water supply point using fire hydrant of the distribution main. Seismic upgrade of distribution pipe is achieved by replacement of aged pipe a pipeline. Table-7 shows the plan and progress. In addition, in selection of replacing route, pipeline information, such as diameter, water quality, pipe break history, the degree of corrosion, and age, and soil condition is considered. Table-7 Progress of Seismic Upgrade of Distribution Pipe Plan Length: 400km Investment: 50 Billion yen Period: 1996～2010 2003 fiscal 189km(21.2Billon yen) year

4.4 Mutual Backup pipe Kobe Water will have the mutual backup pipe between contiguous cities and towns (6 city 2 town) to realize these aims. (1) Reliable supply to the boundary area of city. (2) The emergency water supply in the case of drought or suspension and quick start to repair work (3) Strengthen cooperation with contiguous cities and towns by joint disaster prevention training. Kobe Water received emergency water supply by laying pipe to Miki city and Sanda city, which was installed after the great earthquake. In 2004, mutual back-up pipe with Nishinomiya city was installed, and connection with Akashi-city, Ashiya city, and Miki city will be scheduled in 2005. After back-up pipe is installed, emergency water supply training of 2 cities is planned.

5. Subject Kobe Water has realized the seismic upgrade of facilities based on “Basic

11 Plan for Earthquake-resistant Kobe Water” for ten years after the Hanshin-Awaji Great Earthquake. Water saving and aging society causes decrease in water consumption and water supply profit and it makes situation surrounding water business complicated recently. Kobe Water also is required to invest in more beneficial project. Following on city development of Kobe city on the 1965s, Kobe Water laid a lot of pipelines and built many facilities. Replacement or renewal will be needed near future with many facilities. On the other hand, great earthquakes, such as Tokai earthquake and To-Nankai earthquake are forecasted. Realization of aggressive earthquake-resist projects is needed. Moreover, because of the shortage of rainwater in these days, the measure based on the characteristic of area against the drought is needed. In 1975 the frequency of drought was generated 1 time in 10 years, but in 2000 it was generated 1 time in 4years. In this section we describe the subject in these severe situation surround Kobe Water. 5.1 Water Sources While Kobe Water has no large rivers and lakes, Kobe Water draws water from Lake Biwa and the Yodo River. In addition Kobe Water ensures supply from new water sources. Three fourths of the Kobe Water resources depend on Yodogawa River. Prospect of water demand and supply, or the stability for water supply are discussed in reexamination of “Water-Resources Development Master Plan (Full Plan). Also in Kobe Water, there is deviation with forecast and result in water demand. In the reexamination of “Full Plan”, the water shortage frequency, upgrade of facilities, quality of resource water and disaster should be reconsidered. Kobe Water and the Hanshin Water Authority and its member cities held conference and began examination about the wide mutual support for water quality examination organization, water resource, and facilities. Developing the measure for stability of water supply and improvement of economical efficiency is needed with concerned cities. 5.2 Replacement of distribution facilities

12 Kobe Water started its water supply in 1900 and achieved 100% of service coverage by locating many facilities. However, from now on, aged facilities are increased and the replacement will be the subject

50 配水池Distribution Reservoir 1,500 Facilities aged 送水トンネルTransmission Tunnel 40 over 50 years 導・送・配水管Transmission 1,000 30 Main (km)

20 500

10 導・送・配水管(km) Transmission main (km) (km) main Transmission 配水池（箇所）、送水トンネル Transmission Tunnel (km) Tunnel Transmission Distribution Reservoir (place) 0 0 0～10 11～20 21～30 31～40 41～50 51～60 61～70 71～80 81～90 91～100 101～110 Figure-5図－５本市水道施設の経過年数 Increase of aged Facilities

In replacement of facilities, we take it as the opportunity for review the existing water supply system. (1) Check the facilities and give the priority. Then replace them intentionally. (2) Give higher investment effect by seismic upgrade and addition of new function. (3) The prolongation of life of the existing facilities by suitable maintenance and cost reduction by introduction of new technology and method of construction, etc. These three and fiscal plan must be considered in replacement project. Furthermore, strengtheningFigure-5 of financia Increasel ofaffaires aged Facilities is also an important for replacement. Moreover, after evaluate by index easy to understand for citizen, we have to publish the necessity of project, investment and effect of project, and promoting the mutual understanding with citizens is required. Although the existing transmission tunnels have played the big role in stable water supply, Kobe Water cannot stop and inspect the tunnels due to the increase of water demand. Aim for the inspection, reinforcement and disruption of existing tunnels,

13 Large Scale Transmission Tunnel is under construction as the alternative tunnel. Efficient and effective water supply system is aimed at the project. In the seismic upgrading of distribution pipe, selection of the route is performed by evaluating individually the function, degradation, and earthquake resistance. It cannot be said optimum method from the site of the investment effect. After improving the old selection / evaluation method, we make the “Reconstruction Plan of distribution network” which decide the priority of the route by assessment of cost-benefit of replacement and by evaluating from various fields, such as water quality , water quantity capacity, earthquake resistance, age and importance. 5.3 Mutual aid The response operation to a big earthquake exceeds capacity of the disaster city, so the mutual aid with other cities or towns is important. In this section the mutual aid agreement and activity are described. Kobe City has revised "Agreement on the mutual aid at disaster between 13 big cities” in 1997 based on lesson learned from the Hnshin-Awaji earthquake. Also between contiguous 7 cities and 2 towns, Kobe City concluded the agreement in 1996. In addition, Kobe City concluded mutual aid agreement about fire fighting etc. For the mutual aid of water supply, Kobe Water concluded "Note on the mutual aid between 12 big Waterworks Bureau" in 1996, and concluded “Note on the mutual aid between 13 big Waterworks Bureau” in June, 2004. In the Note the chief city is appointed so that the mutual aid can response quickly at the disaster. Kobe Water and Osaka Water are appointed as the chief city each other, and we have mutual aid training in 1997 and afterwards. In November 2004, in Osaka city, mutual aid training is scheduled between Kobe Water, Osaka Water and Fukuoka Water that is another chief city. We had mutual aid training between Hiroshima Water that is secondary chief city in 1999. Furthermore, mutual aid training between contiguous cities is scheduled

14 after the mutual back-up pipe is completed. Kobe Water always strengthens communication by various opportunities and we aim at quick response for disaster.

6. Conclusion Kobe Water started its service in 1900 and Hanshin –Awaji Great Earthquake hit Kobe Water after 95 years from its start. Kobe Water suffered serious damages by the earthquake. Based on the lessons learned from the earthquake Kobe Water adopted “Basic Plan for Earthquake-resistant Water System” In this report we mentioned about the current status and subjects about the seismic upgrade of facilities. After only a few years from the earthquake, the environment surrounds Kobe Water, such as reduction of water consumption or water supply profit, increase aged facilities, and risk, is changed. To realize the reconstruction of the water supply system, which described as the “Subject” a lot of investment and long years are needed. Therefore, customers’ viewpoint, environment, and mutual back up should be considered. Moreover, it is important to explain using index that is easy to understand for citizens about investment and effect of the project. We have to promote communication with contiguous cities, mutual use of facilities, cooperation with citizens and we must construct the efficient and effective water supply system by dispersion of risks, avoidance of repeat of investment. In order to raise the effect and the efficiency of the project, and to make customer satisfaction higher, Kobe Water publishes its target as the ”New Management Targets aim at 2004-2007”. And make effort to achieve the targets. Next year is the 10th year from Hanshin-Awaji Great Earthquake disaster, and various conferences and meetings including the “U.N.’s International Conference on Disaster Reduction” will be scheduled in Kobe city. Kobe Water will hold “the Kobe Public Forum on Experience during Earthquake Disaster” on January 29, 2005. We consider the ten years after the Earthquake as the re-start of the earthquake-resistance of Kobe Water, and make effort to realize the sustainable water supply.

15 Session 6: Seismic Proof Design of Waterworks and

Other Facilities

S6-1 “Design Guideline for Seismic Resistant Water Pipeline Installations” Presenter: John M. Eidinger (G&E Engineering System Inc.) USA

S6-2 “ALA Guidelines for Pipeline Analysis Methods and Appurtenance Design Methodology” Presenter: Bruce Maison (EBMUD) USA

S6-3 “Seismic Design Issues on Water Transmission Pipelines” Presenter:Luke Cheng (San Francisco PUC) USA

S6-4 “Seismic Diagnosis of Extensive Water Distribution Network” Presenter: Nobuhisa Suzuki (JFE R&D Coroporation)Japan

S6-5 “Seismic Upgrade of Prestressed Concrete Water Tanks” Presenter: David Lee (EBMUD) USA

S6-6 “Evaluation of Scenario Earthquakes and Examination of the Seismic Resistant Design Method of Waterworks in Hiroshima City” Presenter: Kenji Totoki (Waterworks Bureau, The City of Hiroshima) Japan

S6-7 “Seismic Upgrades of Pump Station Located on Liquefiable Soils in Portland, Oregon” Presenter: Tim Collins (City of Portland Water Bureau) USA S6-1

“DESIGN GUIDELINE FOR SEISMIC RESISTANT WATER PIPELINE INSTALLATIONS”

Presenter: John Eidinger

(G&E Engineering Systems Inc., USA)

DESIGN GUIDELINE FOR SEISMIC RESISTANT WATER PIPELINE INSTALLATIONS

John Eidinger1

ABSTRACT

Seismic design for water pipelines is not explicitly included in current AWWA standards. Compounding this problem, standard water pipeline materials and installation techniques are prone to high damage rates whenever there is significant permanent ground deformations or excessively high levels of ground shaking.

To help improve this situation, a new Design Guideline for Seismic Resistant Water Pipeline Installations (the Guidelines) has been developed. It is intended that the Guidelines be issued in March 2005. For the period from November 2004 through January 2005, the Guidelines are available in draft form for public comment. Comments from U.S., Japanese, Canadian and all other water utilities, pipeline manufacturers, AWWA, JWWA and other interested parties are welcomed.

The Guidelines provide direction for three situations:

• When the pipeline engineer has just rough estimates of the earthquake hazard, does not have the resources to do design by analysis, and wishes to rely on standardized pipeline components. The Guidelines provide the Chart Method. This is the preferred approach for common pipeline installations like 6-inch to 8-inch diameter pipes, fire hydrants and service laterals.

• When the pipeline engineer wishes to perform a limited design by analysis. The Guidelines provide the Equivalent Static Method. This is the preferred approach for medium important pipelines like 12-inch to 24-inch installations, or as a preliminary approach for major transmission pipelines.

• When the pipeline engineer has the resources to perform detailed subsurface investigations, geotechnical engineering and pipe stress analyses. The Guidelines provide the Finite Element Method. This is the preferred approach for essential non-redundant installations, like 36-inch to 120-inch pipelines.

INTRODUCTION

In most every severe earthquake, the largest negative impact to water utilities has been the damage to buried water pipelines. At the past three JWWA-AWWARF workshops (Oakland

1 President, G&E Engineering Systems Inc., 6315 Swainland Rd, Oakland CA 94611 USA. [email protected] 1 2000, Tokyo 2001, Los Angeles 2003), a great emphasis was placed by many participants on the rate of pipe damage, the causes of pipe damage, and the improved earthquake performance of new types of pipe.

After the Los Angeles workshop, many US participants got together and decided something ought to be done about this. Accordingly, in concert with FEMA, NIBS and the ALA, a team of engineers was assembled to put together the first ever US seismic design guideline for buried water pipelines. The American Lifelines Alliance (ALA) was formed by the Federal Emergency Management Agency (FEMA) in 1998 as a public-private partnership whose goal is to reduce risk to utility and transportation systems from natural hazards and manmade threats. In 2002, FEMA contracted with the National Institute of Building Sciences (NIBS) through its Multihazard Mitigation Council (MMC) to, among other things, assist FEMA in developing these Guidelines. The ALA sponsors this work through funding from NIBS and FEMA.

AmericanLifelinesAlliance

AUTHORS

The following people and their affiliations contributed to the Guidelines.

Person Affiliation Mr. John Eidinger (Chairman) G&E Engineering Systems Inc. Mr. Bruce Maison East Bay Municipal Utility District Mr. Luke Cheng San Francisco Public Utilities Commission Mr. Frank Collins Parsons Mr. Mike Conner San Diego Water Department Dr. Craig Davis Los Angeles Department of Water & Power Mr. Mike Matson Raines, Melton and Carella, Inc. Prof. Mike O'Rourke Rennselaer Polytechnic Institute Prof. Tom O'Rourke Cornell University Mr. Alex Tang Nortel Networks, Retired Mr. Doug Honegger Consultant (Technical Oversight) Mr. Joseph Steller NIBS (Project Management)

The Guidelines would not have been possible without the contributions from numerous staff of the San Francisco Public Utilities Commission, East Bay Municipal Utilities District, City of San Diego Water Department, the Los Angeles Department of Water and Power, and many other participating agencies.

2 OUTLINE OF THE GUIDELINES

The Guidelines describe the various steps in seismic water pipeline design, with commentary. The main topics included are: Goals; Performance Objectives; Earthquake Hazards; Subsurface Investigations; General Pipeline Design; Analytical Models; Transmission Pipelines; Bypass Pipelines; Distribution Pipelines; Service and Hydrant Laterals; Distribution Pipelines; and Other Components. The Guidelines are meant to be a self-standing document that can be used by pipeline designers in water utilities; as such, it is geared to provide simple procedures to achieve the overall goal. The Guidelines always allow for more detailed procedures to be used by geologists, geotechnical engineers and pipeline engineers when suitable. A link to obtain the entire draft Guidelines is listed in the Conclusions.

For the 4th AWWARF-JWWA workshop, four papers cover the major topic areas of the Guidelines. This paper describes performance goals and the design-by-chart method. The paper by Dr. Craig Davis covers reliability goals and definition of geotechnical hazards. The paper by Mr. Luke Cheng covers design issues for transmission pipelines. The paper by Mr. Bruce Maison covers the two design-by-analysis models and design issues for service laterals.

GOAL OF SEISMIC DESIGN FOR WATER PIPELINES

The goal of the Guidelines is to improve the capability of water pipelines to function and operate during and following design earthquakes for life safety and economic reasons. This is accomplished using a performance based design methodology that provides cost-effective solutions and alternatives to problems resulting from seismic hazards. Improved water pipeline performance will help create a more resilient community for post-earthquake recovery; therefore portions of the Guidelines inherently consider the community impacts if pipeline damage were to occur. The Guidelines do not intend to prevent all pipelines from being damaged.

To achieve this goal, the fundamental intent of the Guidelines is to assure a reasonably low rate of water pipeline damage throughout a water utility system, such that about 90% of customers in a system can be restored with piped water service within about three days after a design basis earthquake.

To achieve this level of performance, an acceptable damage rate will be about 0.03 to 0.06 breaks per 1,000 feet (0.1 to 0.2 breaks per kilometer) of equivalent 6-inch diameter pipe. The commentary of the Guidelines provides a calculation to convert a network of pipes of different diameters that may suffer both breaks and leaks, in conjunction with network redundancy, into a single equivalent break rate per equivalent 6-inch diameter pipe. By minimizing pipeline damage after earthquakes to this level of damage, a typical water utility serving a population of 150,000 people could expect to:

• Deliver water at serviceable pressure to 65% to 90% of all hydrants within the first hours after the earthquake, as long as there are adequate supply sources; and

• Deliver water via the pipe network to at least 90% of all customers within 3 days following an earthquake;

3 as long as the utility can isolate most of the leaking and broken pipes within one day or so, and repair equivalent 6-inch diameter pipes at a rate of about 20 within the first three days after the earthquake, and 20 per day thereafter.

For water utilities with limited post-earthquake repair capability, or serving pipe networks with limited or no redundancy, it is important to limit the damage rate to the lower range. For water utilities with much greater post-earthquake repair capability, it might be acceptable to sustain damage to the higher range.

NEW INSTALLATIONS AND REPLACEMENT / RETROFIT

It is the intent of the Guidelines that they be used for all new pipeline installations. Over a period of many years, a sufficiently high percentage of pipelines in a network will eventually have been designed per these Guidelines. Thus, it may take decades for some utilities to ultimately achieve the goals, unless a pipeline replacement / retrofit program is also adopted.

The decision to replace older pipes is a complex one. In many networks, many existing pipelines (such as cast iron pipe with caulked joints) will not meet the seismic design capability recommended by the Guidelines. Still, the Guidelines do not recommend replacing older 4-inch to 10-inch diameter cast iron pipes solely on the basis of earthquake improvement, since this is not thought to be cost effective. However, as old pipeline are thought to need replacement because they no longer provide adequate fire flows, or have been observed to require repair at a rate of more than once every 5 years, then the added benefit of improved seismic performance may justify pipe replacement. When replaced, the new pipes should be designed per the Guidelines.

Replacement of larger diameter pipelines (12-inch diameter and upwards) may be cost effective just from a seismic point of view, in areas prone to PGDs.

PIPELINE FUNCTION CLASSES

A pipeline's function within the system identifies its importance in achieving the system performance goal. Table 1 provides the 4 function classes. A pipe function identifies a performance objective of an individual pipe, but not that of an entire system.

Function Seismic Importance Description I Very Low to None Pipelines that represent very low hazard to human life in the event of failure. Not needed for post earthquake system performance, response, or recovery. Widespread damage resulting in long restoration times (weeks or longer) will not materially harm the economic well being of the community. II Ordinary, Normal Normal and ordinary pipeline use, common pipelines in most water systems. All pipes not identified as Function I, III, or IV. III Critical Critical pipelines and appurtenances serving large numbers of customers and present significant economic impact to the community or a substantial hazard to human life and property in the event of failure. IV Essential Essential pipelines required for post-earthquake response and recovery and intended to remain functional and operational during and following a design earthquake. Table 1. Pipe Function Classifications

4 THREE DESIGN APPROACHES

The Guidelines provide three approaches can be used in the design of buried pipelines.

• Chart method. The simplest approach. Avoids all mathematical models, and allows the designed to pick a style of pipe installation based on parameters such as regional maps for PGV and PGD hazards, and the pipeline function class.

• Equivalent static method. Uses simple quantifiable models to predict the amount of stress, strain and displacement on a pipe for a particular level of earthquake loading. The pipeline can then be designed to meet these quantified values, or pipe styles can be selected that presumably meet these quantified values without a formal capacity to demand check. Pipe selection is usually made by specification from available manufacturer's catalogs.

• Finite element method. This method uses finite element models to examine the seismic loads (whether PGA, PGV or PGD) over the length of the pipeline, and then uses beam on inelastic foundation finite element models (or sometimes use two- or three- dimensional mesh models) to examine the state of stress and strain and displacement within the pipeline and pipeline joints. Pipe design is often shown on contract drawings, covering material selection, joint preparation, trench design and other factors.

Figure 1. Direction of Permanent Ground Deformation (PGD)

5 CHART METHOD

Transmission Pipelines

Transmission pipelines may carry raw or treated water. Due to their importance to a great number of people, Function Class I is generally to be avoided except for those pipes whose failure would not impact any customer for 30 days or more.

Tables 2 to 5 set the pipeline design category (A, B, C, D or E). Figure 1 shows the meaning of perpendicular (transverse) and parallel (along the axis) orientations. If a portion of a pipeline has two or more categories for the various hazards (ground shaking, transverse PGDs, parallel PGDs, fault offset PGDs), then the highest category controls.

Inch/sec Function I Function II Function III Function IV 0 < PGV ≤ 10 A A A A 10 < PGV ≤ 20 A A A B 20 < PGV ≤ 30 A A B C 30 < PGV A B C D Table 2. Transmission Pipelines – Ground Shaking

Inches Function I Function II Function III Function IV 0 < PGD ≤ 2 A A A A – welded steel B - segmented 2 < PGD ≤ 6 A A A B 6 < PGD ≤ 12 A A B C 12 < PGD A B C D Table 3. Transmission Pipelines – Liquefaction and Landslide Transverse to Pipeline Alignment

Inches Function I Function II Function III Function IV 0 < PGD ≤ 2 A A B B 2 < PGD ≤ 6 A B B C 6 < PGD ≤ 12 C C C D 12 < PGD D D D E Table 4. Transmission Pipelines – Liquefaction (Lateral Spread) and Landslide Along Axis of Pipeline

Inches Function I Function II Function III Function IV 0 < PGD ≤ 2 A A B B 2 < PGD ≤ 6 A B B C 6 < PGD ≤ 12 A C C D 12 < PGD ≤ 24 A D D E 24 < PGD A D E E Table 5. Transmission Pipelines – Fault Offset

6 Distribution Pipelines, Service Laterals and Fire Hydrant Laterals

In most cases, distribution pipelines are in networks. Failure of a single distribution pipeline will not fail the entire network (once the broken pipe is valved out), but the customers on the broken distribution pipeline will have no piped water service until the pipe is repaired. The engineer can assume that distribution pipelines are Function Class II, except in the following cases:

• The pipeline is the only pipe between lower elevation pump station and upper elevation pump station / reservoir in a pressure zone, and the failure of that pipeline will lead to complete loss of supply to the pump station serving a higher zone, or loss of the water in the reservoir for fire fighting purposes. For example, a 12-inch diameter pipe from lower elevation pump station that delivers water to a higher elevation tank within a pressure zone, and that also serves water to higher elevation pump stations.

• The pipeline is the only pipe delivering water to particularly important customers, such as critical care hospitals. For example, an 8-inch diameter pipe that has a service connection to a 200 bed hospital.

Past earthquakes have shown that there can be great quantity of damage to distribution pipelines, especially in areas prone to PGDs or high velocity pulses. While no single distribution pipeline is as important as a transmission pipeline, the large quantity of distribution pipe damage can lead to rapid system-wide depressurization, loss of fire fighting capability, and long outage times due to the great amount of repair work needed. Accordingly, we recommend that most distribution pipes be classified as Function Class II and very few as Function Class I (under ~5% of total pipeline inventory). A few distribution pipes serving essential facilities could be classified as Function III or IV; or they could be designated in suitable emergency response plans as prioritized for prioritized and rapid repair (generally under one day or two days at most). Once the Function Class is set, Tables 6 to 11 define the Design Category.

Inch/sec Function I Function II Function III, IV 0 < PGV ≤ 10 A A A 10 < PGV ≤ 20 A A A 20 < PGV ≤ 30 A A A (with additional valves) 30 < PGV A A (with additional B valves) Table 6. Distribution Pipelines – Ground Shaking

Inches Function I Function II Function III, IV 0 < PGD ≤ 2 A A A (with additional valves) 2 < PGD ≤ 6 A A (with additional B valves) 6 < PGD ≤ 12 A B C 12 < PGD A C C Table 7. Distribution Pipelines – Liquefaction and Landslide Transverse to Pipeline Alignment

Inches Function I Function II Function III, IV 0 < PGD ≤ 2 A A B (with additional valves) 2 < PGD ≤ 6 A B C 6 < PGD ≤ 12 A C D 12 < PGD A D D Table 8. Distribution Pipelines – Lateral Spread and Landslide Along Axis of Pipeline

Inches Function I Function II Function III, IV 0 < PGD ≤ 2 A B B 2 < PGD ≤ 6 A B C 6 < PGD ≤ 12 A C D 12 < PGD ≤ 24 A D E 24 < PGD A E E Table 9. Distribution Pipelines – Fault Offset

Service Laterals and Hydrant Laterals

Inch/sec Any Lateral 0 < PGV ≤ 10 A 10 < PGV ≤ 30 A 30 < PGV B Table 10. Laterals – Ground Shaking

Inches Any Lateral

0 < PGD ≤ 2 A 2 < PGD ≤ 12 B 12 < PGD C Table 11. Laterals – Liquefaction, Landslide and Surface Faulting

Design Categories

There are five design categories. Category A denotes standard (non-seismic) design. The following summarizes the general design approach for Categories B, C, D and E:

• B = restrained with extra valves • C = B + better pipe materials • D = C + quantified seismic design; or provide bypass system. • E = D + peer review (it is strongly recommended that FEM method be used for any pipe with Classification E)

8 Tables 12 to 19 provide guidance for seismic pipe design using the chart method based on the categories A through E. Note. This guidance is based on commonly available pipe and joinery as of 2004. As new pipe products become available, they can be used in the chart method as long as suitable justification (FEM, test, etc.) is provided to show that the pipe meets the intended reliability of the pipe and performance of the pipe network as a whole.

Design Category Cost Effective Design Notes Approach A Standard B Extended Joints C Restrained Joints D Extended and Restrained Joints Standard with bypass E Special Joints Standard with bypass Table 12. Ductile Iron Pipe

Design Category Cost Effective Design Notes Approach A Standard B Standard with extra insertion C Restrained Joints D Extended and Restrained Joints Standard with bypass E Not recommended Standard with bypass Table 13. PVC Pipe

Design Category Cost Effective Design Notes Approach A Single Lap Weld B Single Lap Weld Weld t = pipe t C Double Lap Weld Weld t = pipe t D Double Lap Weld / Butt Weld D/t max 110 in PGD zones E Butt Weld D/t max 95 in PGD zones Table 14. Welded Steel Pipe

Design Category Cost Effective Design Notes Approach A Standard B Extended Joints C Restrained Joints D Extended and Restrained Joints Standard with bypass E Not recommended Standard with bypass Table 15. Gasketed Steel Pipe

Design Category Cost Effective Design Notes Approach A Gasketed or Single Lap weld B Single Lap Weld Weld t = pipe t C Double Lap Weld Weld t = pipe t D Not recommended Standard with bypass E Not recommended Standard with bypass Table 16. CCP & RCCP Pipe

Design Category Cost Effective Design Notes Approach A Standard B Butt Fusion Joints C Butt Fusion Joints D Butt Fusion Joints E Butt Fusion Joints Table 17. HDPE Pipe

Design Category Cost Effective Design Notes Approach A Standard B Soldered joints C Soldered joints Expansion loop / Christie box / Other box Table 18. Copper Pipe

Design Category Cost Effective Design Notes Approach A Standard B Dresser-type coupling C Multiple dresser couplings D EBAA flextend type couplings E Not recommended Relocate hydrant Table 19. Segmented Pipelines Used as Hydrant Laterals

Design Category Cost Effective Design Notes Approach A Bolted, Single Lap Weld, Fusion Weld B Bolted, Single Lap Weld, Weld t = pipe t Fusion Weld C Bolted, Double Lap Weld, Weld t = pipe t Single Lap Weld with fiber wrap, Fusion Weld D Bolted, Double Lap Weld, Bolted, Double Lap Weld, Single Single Lap Weld with fiber Lap Weld with fiber wrap, Fusion wrap, Butt Weld, Fusion Weld Weld E Bolted, Double Lap Weld, Bolted, Double Lap Weld, Single Single Lap Weld with fiber Lap Weld with fiber wrap, Fusion wrap, Butt Weld, Fusion Weld Weld Table 20. Continuous Pipelines Used as Hydrant Laterals

In addition to the design categories in Tables 12 to 20, the following additional requirements are made. These recommendations are cumulative (For C, include B and C recommendations).

• B. Add isolation valves on all pipes within 50 feet of every intersection, for example, four valves on a four-way cross.

• C. Maximum pipe length between connections for segmented pipe is 16 feet, or as otherwise justified by ESM or FEM.

• D. Maximum pipe length between connections for segmented pipe is 12 feet, or as otherwise justified by ESM or FEM.

Bypass Pipelines

During design of a pipeline, it is typical to perform some preliminary seismic and hazard investigation. A geotechnical engineer can perform literature search of available publications and assess the seismic setting of the pipeline and identify potential hazards such as fault crossings, landslides, and zones of potential liquefaction.

With this information, the pipeline design engineer can often times route the pipeline to avoid well-defined hazards. This is the most cost-effective approach for minimizing seismic-related damage to a pipeline. However, sometimes there is no feasible way to avoid a hazard and the pipeline must be routed through the hazard.

Instead of using a higher Category Design (such as D or E), the owner can elect to provide a bypass capability, as long as the owner has the ability to install the bypass within about 1 day after the earthquake, and in consideration of the entire post-earthquake response. Bypass capability might be the most cost effective approach to mitigate many fault and landslide

11 crossings for Function Class III pipelines. Bypasses can be used in retrofitting existing pipelines or for new construction where loss of service cannot be tolerated for more than one day.

A typical bypass is illustrated in Figure 2, consisting of a line isolation valve, if none previously existed, and a 12-inch diameter connection and manifold assembly on either side of the defined hazard. In order for the bypass to be used effectively, the hazard must be relatively well defined. Each of the manifolds is configured to accept one or multiple large diameter hose connections. In the event of a seismic event that results in a pipeline failure within the bounds of the hazard, the hazard isolation valves are closed, thereby stopping leakage at the point of failure. The hose is then deployed across the ground between the two manifold assemblies and serves as a temporary pipe bypass, allowing restoration of flows through the system. Figure 3 shows a deployed bypass system at a fault crossing where deployment of three flex hoses was possible.

Figure 2. Bypass Manifold Assembly

Figure 3. Flex Hose Attached to Manifold Outlets

The criteria for the bypass system components are included in Table 21. So called "large diameter flex hose" (diameter ~5-inches) will generally not provide sufficient flow rate at a reasonable pressure drop, for distances on the order of 1,000 feet between manifolds. So called "ultra large diameter flex hose" (diameter ~12-inches) can provide high flow rates at separation distances of 1,000 feet (or more). There are pros and cons with using either 5-inch or 12-inch hose, including: flow rate and pressure drop; cost; storage life; deployment effort and time; hose breakage and resultant pipe whip; etc.

Description Criteria Pipe Materials Mortar-lined and mortar- or tape/epoxy-coated steel pipe Field joints shall be flanged, welded, or mechanically coupled with suitable restraint Design for anticipated internal, external, and transient loading conditions Provide cathodic protection as needed Manifold Pit Precast reinforced concrete with seismic design factors suitable for site Traffic rated steel plate cover Sized for easy hose deployment 12-inch Valves and Butterfly or Gate Smaller Flexible Hose 12 -inch flex hose, burst pressure ~ 400 psi, operating pressure ~150 psi. Distances up to 1,000 feet or more at flow rates of up to 5,000 gpm 5-inch fire hose from local Fire Department. Distances up to 1,000 feet at flow rates of up to 500 gpm Connections to be coordinated with manifold configuration Table 21. Bypass System Components Criteria

13 CONCLUSIONS

It is the intent of these Guidelines to provide a unified, comprehensive and simple approach that can be readily adopted by water utilities for the design of new pipeline installations. The draft Guidelines are available for public comment through January 2005. They may be obtained via the Internet at: http://homepage.mac.com/eidinger/ (follow the link to downloads, and then download Seismic Guidelines.doc.) Comments should be sent to any of the authors.

The Guidelines may result in changes in pipeline installations in moderate and high seismic areas throughout the United States. Given the large economic consequences of widespread pipeline damage, the authors believe that the extra reliability afforded by these changes is worthwhile and cost effective. We hope that the Guidelines will spur water utilities to procure better pipelines in high hazard locations; in turn, the pipeline manufacturers will manufacture and supply better products. This is, in part, a "chicken and egg" process, since prior to the current moment (late 2004 – early 2005) we have not had the Guidelines for water utilities; nor have we always had suitable cost effective pipelines provided by manufacturers to meet the Guidelines.

ABBREVIATIONS AND UNITS

Customary US units (inches, pounds, gallons) are used in this paper. Conversions to SI units are provided below. All pipe sizes are in customary US units; conversion of a customary pipe size (such as 12-inch diameter) to SI units has no precision, as a 12-inch pipe may often have outside diameter anywhere from ~12-inches to ~13-inches.

ALA American Lifelines Alliance AWWA American Water Works Association AWWARF American Water Works Association Research Foundation ESM Equivalent Static Method FEM Finite Element Method FEMA Federal Emergency Management Agency JWWA Japan Water Works Association MMC Multihazard Mitigation Council NIBS National Institute of Building Sciences PGA Peak Ground Acceleration (g) PGD Permanent Ground Deformation (1 inch = 2.54 cm) PGV Peak Ground Velocity (1 inch/sec = 2.54 cm/sec) inch inch (1 inch = 2.54 cm) feet feet (1 foot = 12 inches = 30.48 cm) g gravity constant (1g = 386.4 inch/sec2 = 981 cm/sec2) gpm gallons per minute (1 gpm = 3.785 liters per minute) psi pounds per square inch (1 psi = 6.895 kilopascals) sec second

14 S6-2

“ALA Guideline for Pipeline Analysis Methods and Appurtenance Design Methodology”

Presenter: Bruce Maison

(East Bay Municipal Utility District, USA)

ALA Guideline for Pipeline Analysis Methods and Appurtenance Design Methodology

Bruce Maison, Michael O’Rourke, and John Eidinger

ABSTRACT

The American Lifelines Alliance (ALA) Guideline offers recommendations for design of seismic resistant water pipelines. This paper summarizes two parts of the Guideline: Analytical Models (Section 7) and Service and Hydrant Laterals (Section 11). The Guideline contains three pipeline analysis models of progressively increasing complexity. The simplest is the Chart Method. A type of pipe installation (material and joinery) is selected from tables based on the intensity of the particular seismic hazard (peak ground velocity, or permanent ground displacement), and the relative importance of the pipeline. The Equivalent Static Method is based on simple analytical models of pipe response to seismic hazards. The models are used to estimate the pipe force and displacement. These quantities serve as the basis for selecting the type of pipe installation. The third and most complex method is the Finite Element Method in which the pipe and the pipe-to-soil interaction are accounted for in a detailed mathematical model (beam-on-inelastic foundation analogy). The results from the model (typically via computer analysis) serve as the basis for selecting the type of pipe installation. The Guideline also addresses pipe appurtenances that serve a variety of functions with the most common being customer service and fire hydrant lateral connections. Because significant numbers of appurtenances have been damaged in past earthquakes and the fact that these tend to be non- engineered for seismic conditions, the Guideline advocates specific design for these components.

INTRODUCTION

The American Lifelines Alliance (ALA) is a public-private partnership formed in 1998, funded by the U.S. Federal Emergency Management Agency (FEMA) and managed by the National Institute of Building Sciences (NIBS), with the purpose of reducing risks to lifelines. ALA is developing guidelines to improve the performance of lifelines in the event of natural hazards such as earthquakes or floods, and man-made threats including biological or radiological attacks. Seismic design for water pipelines is not explicitly included in current American Water Works Association (AWWA) standards. In 2004, ALA initiated a project to develop a Design Guideline specifically for seismic resistant water pipeline installations. The goal is to improve the capability of water pipelines to function and operate during and following earthquakes. The Guideline is currently in a working draft form, and is available for industry comment. This paper summarizes two parts of the Guideline: Analytical Models, and Service and Hydrant Laterals.

______Bruce Maison, American Lifelines Alliance c/o EBMUD, 375 Eleventh Street, Oakland, CA, 94607. Michael O’Rourke, Rensselaer Polytechnic Institute, 110 8th Street, Troy, New York, 12180. John Eidinger, G&E Engineering Systems, Inc., 6315 Swainland Road, Oakland, CA 94611.

1 ANALYTICAL MODELS

Section 7 of the Guideline provides three different approaches that can be used in the design of buried pipelines. The simplest is the Chart Method that avoids all mathematical computation, and allows the designer to select a style of pipe installation based on parameters such as regional maps for ground shaking intensity [e.g., peak ground velocity (PGV)] and ground failure [e.g., landslide peak ground displacement (PGD)] hazards, and the relative importance of the pipeline in context of the entire water system. The Equivalent Static Method (ESM) represents the next level of complexity. It uses simple analytical models to predict the amount of force and displacement in a pipe for a particular level of earthquake loading. The pipeline can then be designed to meet these calculated values, or pipe styles can be selected that presumably meet these calculated values without formal demand-to-capacity checks. Pipe selection may be made by specification from available manufacturer's catalogs. The most complex method is the Finite Element Method (FEM). This method uses finite element models to examine the distribution of loading over the length of the pipeline, and then uses beam-on-inelastic foundation finite element models (or sometimes two- or three- dimensional mesh models) to examine the state of stress and strain and displacement within the pipeline and pipeline joints. It is envisioned that the FEM approach would be used only for the PGD hazard. Pipe design typically will be indicated on contract drawings, covering material selection, joint preparation, trench design and other factors. The design may be accompanied by an engineer's certified stress report.

Chart Method

The Chart Method (Section 7.2) provides recommendations for pipe design via information contained primarily in a tabular format. The Chart Method is applied via the following steps. Step 1: Categorize the pipeline as one of three types (per Section 3): Transmission, Distribution, or Lateral. Step 2: Classify the pipelines as one of four function groups (per Section 3.2). The function groups vary from I (very low importance) to IV (essential). Step 3: Determine the seismic hazard (per Section 4). The Chart Method recognizes four types: ground shaking (defined in terms of PGV); liquefaction or landslide ground failure transverse to pipeline run (defined in terms of PGD); liquefaction or landslide ground failure along the axis of pipeline run; and fault offset (defined in terms of PGD). Step 4: Assign pipeline design classification (per Section 7.2). Five design classes (denoted by letters) with progressively increasing seismic ruggedness are defined with Class A having the standard (non-seismic) design to Class E representing the highest level of seismic design. The design class is dependent on pipeline category (from step 1), function (from step 2), and seismic hazard (from step 3). Below is the design classification table for distribution pipelines subjected to fault offset to illustrate the assignment process. TABLE 1. DESIGN CLASS FOR DISTRIBUTION PIPELINES SUBJECT TO FAULT OFFSET Fault Offset Function I Function II Function III, IV 0 < PGD ≤ 2 inches A B B 2 < PGD ≤ 6 A B C 6 < PGD ≤ 12 A C D 12 < PGD ≤ 24 A D E 24 > PGD A E E

2 Step 5: Select pipe material and associated design approach (per Section 7.2). The designer has the choice of numerous materials including: ductile iron, welded steel, concrete and PVC. Specific design approaches are indicated for each material and design classification. Below is the design approach table for welded steel pipe to illustrate the process.

TABLE 2. DESIGN APPROACHES FOR WELDED STEEL PIPELINES Design Pipe Segment Connection Notes Classification Type A Single lap weld type B Single lap weld type Weld thickness, t = pipe t C Double lap weld type Weld t = pipe t D Double lap or butt weld type Max D/t of 110 in PGD zones E Butt weld type Max D/t of 95 in PGD zones

Equivalent Static Method

The Equivalent Static Method (ESM) computes pipe seismic response quantities (forces, displacements) using idealized models describing the interaction of the hazard, soil and pipe (Section 7.3). The models are expressed in terms of simplified equations that the designer can rapidly apply by hand computation (use of computer-based software not necessary). The purpose is to account for the physical aspects governing the pipe response in a simplified manner so the designer can apply the method to specific situations with the understanding of the key mechanisms influencing behavior. The hazards may be estimated using regional maps. The axial force or moments in the pipe body, and displacement or rotation at the pipe joint are computed using the equations. The pipeline design requires the forces and displacements to be less than allowable values. The ESM has equations for calculation of pipe response resulting from ground shaking hazard that produces transient ground strains from seismic wave passage; and, ground failure hazards such as landslides, liquefaction, or surface faulting that result in permanent ground deformations. A distinction is made between continuous and segmented pipelines due to their very different response behaviors. A continuous pipeline has joints possessing significant strength and stiffness relative to the pipe barrel (often referred to as restrained joints). An example is a steel pipe having welded (single lap, double lap or butt welded) joints. A segmented pipeline has joints having low strength and stiffness relative to the pipe barrel (often referred to as unrestrained joints). Examples of segmented pipe are ductile iron or PVC pipes having push-on bell-and-spigot gasketed joints.

Intial Soil Final Soil Positon Positon

Pipeline (a) PGD

PGD

Soil Block PGD Distributon

Positon Along Pipe (b) Figure 1. ESM model for landside PGD oriented along the axis of pipeline run. 3 To illustrate the ESM process, the equations associated with buried pipe response to landslide PGD oriented along the axis of pipeline run (longitudinal) direction are explained here. The equations are representative of the degree of complexity inherent in the ESM. These are based on the model shown in Figure 1a. The soil movement is assumed to be a block of finite length that undergoes a uniform down-slope displacement (Figure 1b). The maximum pipe forces and displacements generally occur at the margins of the soil mass undergoing movement, causing either pipe tension (pull-out at the head of the moving soil mass), or pipe compression (push-in at the toe of the moving soil mass).

Typical ESM Equations for Continuous Pipelines

The force for designing the pipe barrel and joints may be taken as the smaller of F1 or F2 representing upper bound estimates of the axial force in the pipe. F1 is the force assuming the pipe is elastic and fully compliant with the soil, and F2 is the ultimate force the soil can transfer to the pipe.

F1 = AEtuδ

Lt F = su 2 2 where, A = cross sectional area of pipe, E = pipe modulus of elasticity, tu = ultimate axial frictional force per unit length at the soil-pipe interface, δ = estimated PGD displacement, Ls = estimated down-slope length of the soil mass undergoing movement. Force F2 assumes that half the total applied soil load is resisted in tension (at the head) and half in compression (at the toe). The calculated force, smaller of F1 and F2, need to be less than the ultimate capacity of the pipe as determined by the pipe manufacturer, tests or other methods.

Typical ESM Equations for Segmented Pipelines

The ground displacement is assumed to be accommodated by pipe joint expansion and contraction. The axial displacement that the joint must be able to accommodate may be taken as follows. For push-on type pipe joints (not having mechanical stops preventing pipe segments from pulling apart), the design displacement may be taken as:

∆ joint = δ

For pipe joints having mechanical stops (e.g., restrainer rings) preventing pipe segments from pulling apart, the PGD may be assumed to be distributed over several joints and the design displacement may be taken as:

δ =∆ joint n where, n = the number of restrained joints near the head or near the toe of the moving soil mass. The mechanical stops must be designed to accommodate the following force.

4 (ntL +1) F = up stop 2 where, Lp = length of the pipe segment.

Finite Element Method

The Finite Element Method (FEM) is potentially the most powerful analysis method that can estimate pipe response for site-specific situations. Theoretically, it could be used for the analysis and design of any pipeline, however typically it would be use only for the most important pipelines (transmission pipelines) subject to some type of PGD. Figure 2 depicts the typical form of FEM application. The model consists of beam-column and spring line elements. The pipe is discretized by a series of beam-column elements having the ability to resist axial, shear and bending moments via elastic and inelastic actions. The soil is modeled by transverse- horizontal, transverse-vertical and axial springs connected to the ends of each beam-column element. The soil spring behaviors are generally nonlinear reflecting actual soil characteristics [1, 2]. The seismic hazard is typically represented by PGDs that enter the analysis as applied displacements at the ends of the soil springs. Soil-pipe interaction is captured as the applied displacements distort the soil and pipe according to their relative stiffness and strengths. For brevity, detailed descriptions of the FEM are not contained here, but can be found in the Guidelines (Section 7.4).

Figure 2. Finite element method of pipeline analysis.

5 APPURTENANCE DESIGN CONSIDERATIONS

Appurtenances are those ubiquitous components connected to pipelines that serve a variety of functions with the most common being customer service and fire hydrant lateral installations (Figures 3 and 4). Post-earthquake water system surveys reveal a significant fraction of the total numbers of buried pipe repairs deal with appurtenances (as much as 20%). Appurtenances are typically non-engineered for seismic conditions having hardware governed by availability, ease of installation and maintenance economics. Section 11 of the Guideline advocates specific design consideration for appurtenances to enhance their seismic ruggedness.

Figure 3. Typical customer service lateral installation.

Figure 4. Typical fire hydrant lateral installation.

Figure 5. Example air valve installation and seismic hazards.

Figure 5 depicts an appurtenance consisting of an air valve located in a vault and associated piping connecting to a buried main to illustrate how the different seismic hazards affect the installation. The air valve suspended inside the vault can vibrate due to ground shaking like any free standing structure connected to the soil. The piping embedded in soil connecting the air valve to the water main is subject to transient strains generated by the soil from seismic wave passage. Should the soil mass at the vault moves relative to the main (such as from liquefaction, landslides, or faulting), the piping will be subject to applied deformations that could cause failure depending on the magnitude of the movement, soil strength, and pipe flexibility, strength and ductility. Past earthquakes have demonstrated that customer meters and fire hydrants generally are not vulnerable to ground shaking. However, other appurtenances have been shown to be susceptible to damage, especially components that are mounted in a relatively flexible manner (like inverted pendulums within or outside of a vault) having non-ductile connections. For example, the air valve mentioned above may have had no seismic consideration in the original design, yet it has the potential for dynamic amplification due to its support configuration acting as a flexible inverted pendulum, and the pipe connections are typically threaded having low ductility (brittle) behavior, may not have been totally engaged during installation, and may suffer from aging/corrosion. Should appurtenances be designed according to any modern design standard such as the NEHRP provisions [3], then the performance will in general, be excellent. However, the Guidelines do not require seismic design requirements at sites with peak ground acceleration, PGA < 0.15g based on experience. Past earthquakes also have demonstrated virtually all buried appurtenances, including customer meters and fire hydrants, when properly installed and not weakened by age or corrosion, generally are not vulnerable to transient ground strains. Accordingly, good installation procedures, quality inspection, and corrosion protection programs will mitigate damage to appurtenances. Permanent ground displacement represents the most serious hazard for buried appurtenances. Figure 6 illustrates one mechanism. The appurtenance is located in an unstable soil mass undergoing lateral PGD, and attached to a water pipe anchored in a stable soil mass. The pipe is anchored while the appurtenance is dragged along within the moving soil mass. The

7 relative motions cause stresses to develop in the appurtenance with the key location being at the attachment to the pipe that is the interface point between the relative motions (point A in Figure 6). Whether the appurtenance pressure boundary fails resulting in a leak, depends on the strength and flexibility of the attachment. A relatively strong attachment allows the appurtenance to shear through the soil. A flexible attachment can accommodate the relative displacements. Flexibility can be provided by mechanical hardware and/or material ductility. Recommended design for customer services and fire hydrant laterals follow. Customer Services. Main cocks, typically made of brass castings, are relatively weak and possess low ductility due to the threaded connection into the pipe. The strategy for PGD-tolerant design is to uncouple the main cock from the (moving) soil. This can be achieved by providing a soft void space around the main cock so that a modest amount of relative motions can be distributed over the relatively flexible and ductile service tubing. One such device is the service boot (Figures 7) that one west coast utility uses in areas of known ground movements having a history of main cock failures. Fire Hydrant Laterals. Fire hydrant laterals are typically connected to the pipe with tee connections that possess significant strength and ductility (especially if the lateral branch pipe is welded steel). Therefore, the standard installation, having no special mechanical couplings to provide additional flexibility, is able to resist modest levels of relative PGD. However, it is clear that under large PGD, it is likely that failure of the lateral will occur at the pipe-to-branch attachment point. Accordingly, the Guideline recommends use of flexible coupling devises at this connection when large PGDs are possible. However, the magnitude of PGD beyond which special flexible coupling devises are cost-effective is difficult to quantify. Life-cycle cost must be considered on a case-by-case basis. Dresser-type couplings have the potential for increased maintenance costs due to leakage over time (versus a continuous pipe). Special ball joint and slider couplings (EBAA flextend or equivalent) are relatively expensive leading to high installation costs versus the low likelihood that seismic PGD will affect a particular hydrant installation. Hydrant installations having histories of actual failures due to PGDs are candidates for coupling devices as these will likely experience additional PGDs in future earthquakes. The Guideline recommends one dresser-type coupling for PGDs up to 3 inches; two dresser-type couplings for PGDs up to 12 inches; and flextend-type couplings for larger PGDs.

Stable Soil Mass Water Main w/ restrained joints

PGD

Highly Stressed Point at Attachment Unstable Soil Mass A

Appurtenance

Plan View of Buried Water Main and Appurtenance

Figure 6. Example of PGD mechanism affecting appurtenance.

Figure 7. Image at left: Side view of service boot. Image at right: Photo of service boot components: HDPE drain pipe and end cap (upper left), two foam inserts (upper right), and visqueen plastic sheeting (foreground).

CONCLUSION

The American Lifelines Alliance (ALA) Guideline offers recommendations for design of seismic resistant water pipelines, and two sections of the Guideline are summarized in this paper: Analytical Models, and Service and Hydrant Laterals. The objective is to establish national consensus guidelines thereby improving the capability of water pipelines to function and operate during and following earthquakes. The Guideline provides a first step in achieving this goal.

ACKNOWLEDGEMENTS

The Guideline is written under contract to the American Lifelines Alliance, a public-private partnership between the Federal Emergency Management Agency (FEMA) and the National Institute of Building Sciences (NIBS). This report was prepared by a team representing practicing engineers in the United States water utility industry and academics.

The ALA “Design Guideline for Seismic Resistant Water Pipeline Installations” is currently available as a draft, for industry comment. The complete draft guidelines are available for download at: http:// http://homepage.mac.com/eidinger/Menu7.html (click on link: “downloads”, click on: Seismic Guidelines.doc). Comments should be sent by January 29, 2005 to John Eidinger at [email protected].

REFERENCES

[1] Response of Buried Pipelines Subject to Earthquake Effects, Monograph, Multidisciplinary Center for Earthquake Engineering Research, State University of New York, Buffalo, New York, 1999. [2] Guidelines for the Seismic Design of Oil and Gas Pipeline Systems, American Society of Civil Engineers, 1984. [3] NEHRP Recommended Provisions for Seismic Regulations for New Buildings and Other Structures, Federal Emergency Management Agency, FEMA 368, March 2001.

9 S6-3

“SEISMIC DESIGN ISSUES ON WATER TRANSMISSION PIPELINES”

Presenter: Luke Cheng

(San Francisco Public Utilities Commission, USA)

SEISMIC DESIGN ISSUES ON WATER TRANSMISSION PIPELINES

Luke Cheng San Francisco Public Utilities Commission/Engineering Design Bureau San Francisco, California U.S.A.

ABSTRACT

This paper is one of the presentations at the 4th JWWA-AWWARF workshop addressing the American Lifelines Alliance (ALA) Design Guideline for Seismic Resistant Water Pipeline Installations. This paper covers design approaches for 1 to 3-meter diameter transmission pipes. Specific design issues considered are seismic hazards, system redundancy, pipeline alignment, pipe material, pipe structural design and analysis, steel and ductile iron pipe joints, appurtenances, system modeling, corrosion control, internal water pressure, constructability, economic considerations, environmental issues, public relation or outreach, emergency response planning, repair priority and emergency response and preparedness plans. Design considerations for pipes on fault crossing are also briefly discussed.

INTRODUCTION

Several large US water utilities in conjunction with the American Lifelines Alliance (ALA) have decided to develop a Design Guidelines for Seismic Resistant Water Pipeline Installations. The authors include participants from US water utilities, universities and practicing engineers. This ALA Design Guidelines address various topics from earthquake hazards, analysis methods to different sizes and types of pipeline including appurtenances in both general and detailed discussions.

This paper summarizes and highlights the issues to be considered in the design of 1 to 3- meter diameter water transmission pipelines. A more detailed discussion is available in the Guidelines (see the end of the paper to see how to get a complete copy of the Guidelines). The chapter is divided into two sections: one on general seismic design issues and one specific to pipes at fault crossing. A detailed discussion on design of welded joints in steel transmission pipes is also included in this chapter.

SEISMIC DESIGN ISSUES RELATED TO TRANSMISSION PIPELINES

The general approach to design of transmission pipelines covered in the ALA Guideline includes (1) seismic hazard and geotechnical assessment, (2) pipe materials, (3) design earthquakes, (4) pipeline alignment, (5) soil mitigation, (6) pipe joints, (7) pipe structural design and analysis, (8) pipe supports, (9) pipe depth and trench backfill, (10) pipe bend

1 and thrust block design, (11) appurtenances, (12) system redundancy, (13) system modeling, (14) corrosion control, (15) internal water pressure, (16) constructability, (17) economic considerations, (18) environmental issues (19) public relation or outreach, (20) emergency response planning, (21) security, and (22) other special design issues.

Seismic Hazards and Geotechnical Assessment Past earthquakes indicated that site conditions such as topography, geography, terrain and soil, have great influence on seismic damage sustained by pipes.

For every transmission pipeline project, a geotechnical evaluation of the seismic hazards such as liquefaction, landslide, lateral spreading, seismic settlement, seismic wave propagation and fault crossing for each geologic area along the pipeline alignment should be performed. The evaluation should also include the impact from man-made features, such as existing retaining walls, transmission towers, cuts and fills, etc.

Pipe Materials Transmission pipelines in the US are most commonly built from steel, prestressed or reinforced concrete cylinder pipe. Smaller transmission pipelines could be built using ductile iron or high density polyethylene materials. In each case the design can use gasketed or various types of restrained joints.

One of the most important factors in designing an earthquake resistant structure is ductility of the material. Ductility refers to the ability of the material to sustain large plastic deformation without failure. Materials of high ductility include ductile iron, welded steel and some plastic; but in earthquakes, these materials will often only perform in a ductile manner if the pipe joinery also accommodates the forces needed to induce generally yielding in the pipe barrel.

Design Earthquakes Design earthquakes should be identified and the associated ground motion developed for each geologic area along the pipeline alignment. For high seismic hazard areas, the owner may wish to consider two levels of earthquakes that should be evaluated, if the owner wishes to have two levels of performance goals, such as:

o Extremely reliable under Maximum Earthquake representing an upper level that is unlikely to be exceeded during the remaining life of the pipeline.

o Reasonably reliable under Probable Earthquake representing an event most likely to occur during the pipeline’s life.

Pipeline Alignment Liquefaction and lateral spread susceptibility, landslide potential, seismic settlement, fault crossings, levels of expected ground motion and seismic wave propagation effects should be considered in pipeline alignment decisions. Alternate alignments to avoid high

2 seismic hazard potential areas, if possible, should always be investigated. The extra cost to align a pipeline to avoid a seismic hazard may be worthwhile when considering the extra reliability afforded post-earthquake.

Soil Mitigation When a pipeline alignment must go through soils with high liquefaction and lateral spread susceptibility or high landslide potential, soil stabilization should be considered. Alternatives for soil mitigation in this case might be soil nailing, vibroflotation, drainage wells, pressure grouting and underpinning the pipeline.

Pipe Joints It has been observed in past earthquakes, the pipes with flexible and restrained joints performed better than ones with rigid or non-restrained joints.

Welded Steel Pipe

Three types of weld are used for welded steel pipes: single fillet weld lap joint, double fillet weld lap joint and full penetration butt weld joint. In area with high seismic hazards (liquefaction, lateral spread, landslide and fault crossing), the double lap weld (up to a point) or full penetration weld (preferred) joint is recommended. A section on welded joints design of steel transmission pipes is presented at end of the discussion of general design issues. Mechanical joints can also be used in highly localized area like a fault crossing or for underwater installations with soils highly susceptible to settlement or other movements.

Riveted Steel Pipe

Riveted steel pipe is no longer being produced in the US. However, when retrofitting an existing riveted steel pipe, finite element analysis should be performed to quantify the load on the non-replaced riveted pipe if replacing the entire segment of pipeline through the high seismic hazard region is not feasible.

Ductile Iron Pipe

Ductile iron pipes can be used for smaller diameter transmission pipelines; the largest size available is 64 inches. Some of the joints or fittings are shown in Figure 1. Additional joints can be found in AWWA M41 or manufacture’s catalogs such as American Ductile Iron Pipe, US Pipes and others.

In high seismic hazard areas (such as locations with high liquefaction potential, high landslide susceptibility, fault crossing (with very limited offset) or very high ground motion coupled with poor soil condition), joints similar to KUBOTA S and SII Type joints can be used; these have been shown to perform very well in past Japanese earthquakes for pipes with diameter up to about 24" and sustaining PGDs of about 24".

Figure 1 – Ductile Iron Pipe Joints

4 Reinforce Concrete Cylinder Pipe (RCCP) and Prestressed Concrete Cylinder Pipe (PCCP)

In moderate and high seismic areas, the joints should be tied together to prevent the pull out of joints during earthquakes. This can be accomplished by using the “tied joints”. Generally, there are two types of tied joints – welded and harnessed. The welded joints are shown in Figure 2 and harness in Figure 3. For the welded joints, it is important to provide the weld completely around the joint.

Figure 2 - RCCP Welded Joints (from AWWA M9)

Figure 3 - RCCP Harnessed Joints (from AWWA M9)

Pipe Structural Design and Analysis Three types of analytical models for design or retrofit pipelines are presented: chart method, equivalent static method and finite element method.

In general, for designing transmission pipelines in moderate and high seismic zones, equivalent static and/or finite element method should be used. For the preliminary design purpose, the chart method is preferred due its great simplicity.

If the chart method is chosen in a high seismic area without further validation by ESM of FEM, then at a minimum, the designer is highly advised to adopt only materials and pipe joinery with high ductility. Ductility is a very important factor in designing an earthquake-resistant structure. Pipe tension, compression, bending, and shear must be taken into account in seismic design of continuous pipelines.

For pipe bends and joints experiencing large deformation, non-linear thin shell finite element models can be used to assess if actual stresses and strains are less than the allowable.

6 Pipe Supports Pipes have different types of support structures, depending on whether they are above ground or below ground. Figure 4 illustrates some possible support configurations.

Figure 4 – Pipe Support Configurations

Pipe Depth and Trench Backfill Weight of backfill is governed by pipe depth and backfill material. This determines resistance to pipe loading when subjected to differential ground movement. If engineering analysis indicates less resistance is desirable, shallow burial or above ground installations should be considered. If one side of the pipe is on the bottom of a sloping ground, a retaining wall may be required for this side of trench to prevent possible loading from slope movement.

Pipe Bend and Thrust Block Design Ideally, a thrust block should be placed at any horizontal and vertical pipe bend. Once the thrust forces (seismic and hydrostatic) are determined, design of the block can be followed by the procedures outlined in Chapter 9 of AWWA M9, or Chapter 8 of ASCE Manuals and Reports on Engineering Practice No. 79, Steel Penstock. The pipe joints on either side of the thrust block should be designed to take the thrust load transmitted through the joints. Welded joints and/or mechanical restrained joints will be required. The Guidelines recommend that the welded / restrained joints be continued for a distance from the bend such at to provide a factor of safety of about 3 against hydrostatic thrusts; of a suitable FEM analysis done to confirm that seismic (ideally plus seismic transient pressure) forces do not lead to joint pullout in earthquakes.

If placing a thrust block is not an option, a detailed analysis including soil-pipe interaction at the bend location should be performed. Thicker pipe, tension joints, stiffener rings and soil hardening are few of design options to be considered.

7 Design Features and Appurtenances

Emergency Cross Connections The system should be designed with the assumptions that some earthquake damage will occur. If there are two or more parallel pipelines, emergency cross connections to the adjacent pipeline(s) should be constructed at selected locations. If possible, inter-tie facilities with adjoining water utilities should be considered.

Consideration should be made that damage to one parallel pipe will not induce failure to the adjacent parallel pipe. This type of failure mode has not been observed in past earthquakes when damage to one pipe has been limited to serious leakage. However, a blowout break at high pressure can result in rapid erosion of nearby soils, possibly undermining adjacent pipes.

Overflows At sites where pipe damage is likely, there should be design provisions for overflow protection to minimize the inundation potential to structures and streets, or erosion that would cause serious impacts. Overflows might include dewatering plans and drainage systems.

Isolation Valves (Shutoff Valves) Water system isolation valves should be installed to segregate pipelines with a high vulnerability from those with a lower vulnerability to earthquake damage. In the event of a pipe break, this will allow operators to close valves, segregating damaged portions of the system and more quickly restoring operation of the undamaged system. Valves should be maintained with periodically inspected, tested and exercised. The isolation valves should be closed quickly (possibly ~20 minute closure times on large pipes) but not to cause significant water hammer to prevent further damage from undermining and flooding.

It is recommended that both air vacuum valves and blow off valves be installed with isolation valves. All such assemblies should be designed for inertial loading and in consideration of long term corrosion impacts.

Seismic or Excess Flow Activated Actuators The isolation valves should be installed with seismic or excess flow activated actuators to prevent further damage from earthquake induced pipeline leakage or rupture. "Seismic Only" actuation (such as upon high PGA) should not be used; instead, actuation should be based on high PGA coupled with high flow / excessive pressure drop; or in many cases, only upon human operator action.

These actuators should be carefully designed so that they will not shut off in an unwarranted event.

8 Blow off (Surge) and Air Release/Vacuum Valves (Air Inlet) Surge and/or air release valves should be considered to accommodate flows resulting from breaks that could damage the system such as a large downstream break that could result in negative pressure upstream imploding the pipe.

On large diameter pipes, blow off and air release / vacuum assemblies are often housed in circular concrete vaults (made of circular concrete pipe) overlying the transmission pipe. If areas prone to PGDs, these concrete vaults should be suitable anchored to the concrete encasement / foundations around and beneath the pipe, to avoid the potential for them displacing relative to the pipe and causing damage toe the equipment within.

Seismic Design of Laterals All laterals attached to transmission pipes should be designed for seismic loads. Air vacuum valve assemblies should be designed with special attention to avoid failures between the valve assembly and the main pipe during severe ground motion or deformation.

System Redundancy Redundancy should be built into water transmission pipeline system. Additional pipelines, multiple smaller pipelines in lieu of a single large pipeline should be considered to minimize delivery reduction due to pipe rupture. Cross connections and isolation valves should be incorporated into the system.

System Modeling A system or network model for the pipeline segment being designed should be developed. The interrelationship of the segment being designed to the entire system needs to be included with flow and operation perimeters determined.

In order to perform such analysis, the following information will be required:

(1) Seismic hazard mapping or assessment (liquefaction, landslide, ground motion and fault rupture) for the design segment of pipeline.

(2) Scenario earthquake(s) to be considered.

(3) System hydraulic network distribution models.

(4) Flow and operation requirements.

(5) Pipeline inventory (pipe material, size, joints, age and corrosion).

The objective of the system model analysis is able to provide the following results:

9 (1) Identify seismically vulnerable segments of the pipeline.

(2) Locate potential water outage areas.

(3) Provide damage level and loss.

(4) Estimate possible repair efforts and repair times after an earthquake.

(5) Help establish suitable design criteria for the pipe to meet overall reliability targets.

With the above information, emergency response plans and mitigation procedures can then be developed.

Two examples of system models are Eidinger (2002) and Bllantyne (1990).

Corrosion Control Corrosion weakens the pipe’s strength. It is one of the major causes for pipe failure during an earthquake. The corrosive environments to which a pipeline exposed could be water, atmosphere, soil, adjacent pipeline and/or structures.

Corrosion control measures include providing linings and coatings to minimize corrosion, and controlling with cathodic protection.

Internal Pressure and External Loads Internal water pressure should include hydrodynamic and water hammer pressure. The pipe also needs to be checked for external loads such as dead weight of soil, live loads, thermal loads. In some areas, the pipe needs to be checked for frost heave, nearby blasting, or other special conditions.

Constructability Construction methods should always be considered during planning and design phases. The physical site conditions and environmental issues might dictate the type of construction. The construction methods for transmission pipelines include trenching and open cut, aerial crossings, horizontal directional drilling, boring and jacking, and tunneling.

Economic Considerations A cost-benefit analysis should always be performed for each transmission pipeline project. For a seismic project, the performance goals should also be established.

The following items might have the influence on the total cost of a transmission pipeline project: (1) pipe and casing materials availability, (2) design cost, (3) construction

10 methods, (4) construction inspection efforts, (5) site/work area access requirements, (6) dewatering requirements, (7) right-of-way required, (8) traffic disruptions, (9) permits needed, (10) special equipment needed, (11) availability of experienced contractors, (12) contaminated soils, (13) backfill material requirements, (14) environmental impacts, (15) dust control, (16) noise reduction, (17) restoration, (18) maintenance and (19) risk.

Environmental Issues Environmental issues have become more important for every construction project. If the project is in California, the governing laws and regulations are (a) National Environmental Policy Act (NEPA), (b) California Environmental Quality Act (CEQA), and (c) Federal and State Environmental Permits. The owner should always determine if the project is subject of NEPA and/or CEQA, and review for exemptions and complete the environmental study.

Public Relation or Outreach Transmission pipelines are usually several miles long and travel through different neighborhoods in urban and rural areas. It would be prudent to present the proposed alignment and associated structures, and explain the benefits of the project and some of the seismic resistance or upgrade features to the public, and solicit their input. Hopefully, by doing so, the project can avoid or minimize possible delays or unwanted lawsuits.

Emergency Response Planning An emergency response plan should be in-place before the earthquake to make it part of an overall cost-effective earthquake mitigation plan.

When developing an emergency response plan, the following tasks should be considered:

(1) Establish a planning team including personnel from management, operations, safety and engineering.

(2) Complete hazards assessment and vulnerability analysis.

(3) Define emergency response categories such as

a. Minor earthquake event defined as damages confined to one location and not whole region.

b. Moderate earthquake event defined as damages affecting multiple locations within some parts of a region and coordination among neighboring agencies might be necessary.

c. Major earthquake event defined as a disaster involving widespread damage to the whole region.

11 (4) Conduct condition assessment of the existing pipelines including appurtenances.

(5) Provide inventory of material for pipeline repair such as different size and material of pipes, reducers, couplings, gaskets, plates, pipe/adaptor fabrication and pipe installation/repair equipment.

(6) Conduct a survey of current staff availability.

The plan should include the following activities:

(1) Establish repair priority

(2) Develop repair strategy

(3) Set up personnel, materials and equipment requirement.

(4) Provide repair procedures.

(5) Prepare staffing and material/equipment purchasing plans.

(6) Purchase different size of pipes and reducers (or adaptors)

(7) Locate stockpile sites for material and equipment

(8) Establish schedules and procedures of emergency exercises and provide training.

(9) Provide locations to store as-built drawings

(10) Establish a pipe replacement program to replace sections of aging pipeline at regular basis.

(11) Secure long-term contracts with outside contractors for availability during a major seismic event – It might be difficult to find available contractors immediately after a major disaster.

(12) Maintain contact with other utilities such as developing a mutual aid and assistance program among utilities.

(13) Include an action item to establish a seismic upgrade program, if there is none, so that repair effort can be minimized.

12 Security Security has become a more important issue for water supply system projects. It is important for transmission pipeline seismic design to coordinate with the security work so that any security measures will not impede future repair efforts or create seismic hazards for the pipelines.

Other Special Design Issues In addition to issues discussed above, other special issues might be considered:

• Waterway crossing (river/creek/channel crossing) • Highway crossing • Bridge crossing • Potential impact due to failure of adjacent structures such as highway overpass, buildings, transmission towers, reservoirs and etc. • Hydraulic transient design.

DESIGN OF WELDED JOINTS IN STEEL TRANSMISSION PIPES

Elastic Stress Limits

If using a single lap welded pipe, the stress in the joint will be amplified over the stress in the main body of the pipe. This is caused by three reasons: the geometry of the joint will introduce net bending, which will increase the maximum longitudinal stress; the stress within the lap weld will include the factors of longitudinal axial, bending and hoop forces; the thickness of the weld material; and possibly stress concentrations at within and near the weld due to weld flaws. In the ESM method, the Guidelines make the overly simplified assumption that most single lap welded joints (outside welds) with minimum leg size equal to the minimum pipe wall thickness can sustain some localized yielding before leading to failure, so the Guidelines suggest the following acceptance criteria.

σ pipe ≤ 0.40Fy where Fy = nominal specified yield stress of the pipe. This formula implies a joint efficiency of about 35% as compared to the strength of the pipe. For cases where a single lap welded pipe is used with thinner welds, then use:

t σ pipe ≤ 0.40Fy tweld

Single lap welded steel pipes exhibit about the same strength in tension or compression. Once the compressive load reach about ±0.60 Fy in the main body of the pipe, strains

13 within the single lap weld will reach about 5% - 6% (compressive loading) or 8% to 9% (tension loading). The design longitudinal stress allowable for a single lap welded pipe (external lap weld equal to wall thickness) should not exceed about 0.60Fy in the main pipe, under maximum earthquake.

For double lap welded steel pipe with common fit up tolerances, replace 0.40 with 0.90 for tensile loading. Due to eccentricities in a double lap welded pipe at the connection, predicted tensile stresses of 0.90Fy in the main body of the pipe away from the joint will translate to about 3% strain or so within the highest strained part of the lap welded joint. Initial yielding of the double lap welded joint will occur at about 50% to 60% Fy in the main pipe. In the highly nonlinear realm, predicted tensile longitudinal strains of about 5% in the main body of the pipe will translate to about 10% strain in the highest strained part of the welded lap joint.

In compression, a double lap welded steel pipe with common dimension tolerances (D/t = 175) will buckle at a compressive load of about 0.60Fy. The pipe will continue to shorten in its buckled shape as compressive loading is maintained, albeit with load shedding and with increasing strain in the pipe. By the time the wrinkle has formed to cause about 1 inch bulging in or out, the peak strain in the male or female parts of the spigot joint will reach about 13 to 14% strain (unpressurized) or about 12% to 15% strain (pressurized to 150 psi).

For butt welded pipe, replace 0.90 with 1.00; or use nonlinear strain acceptance criteria.

If the designer opts for some nonlinear performance of the pipe, the stress checks should be replaced with tensile strain limit checks and wrinkling checks. Single lap welded pipe should generally be limited to the above elastic limits. Double lap welded or butt welded pipe can accept some strain or wrinkling, with butt welded pipe performing better than double lap welded pipe. For water pipes, some wrinkling is acceptable if the owner accepts this performance, and if pipe failure does not lead to serious impacts to nearby pipes, structures or habitat.

For pipes connected using bolted flanged joints, then the above equations are used, with the weld being that between the flange and the pipe. It is assumed that the flange and bolts will be sized based on pressure requirements, and that seismic loading from ground shaking will not control.

Wrinkling Strain Limit The theoretical onset of compressive buckling in a thin-walled cylinder (not including lap joints) is between one-third to one-fourth of the theoretical value of:

t ε = 6.0 theory R

14 where t = pipe wall thickness, and R = pipe radius. This is derived from the classical buckling stress of a perfect cylinder (Timoshenko and Gere) of:

1 tE σclassical = 31()−µ 2 R where µ is Poisson's ratio and E is Young's modulus.

A conservative estimate of the onset of local buckling in a butt welded pipe is:

t t ε = 0.175 to 0.2 onset R R

Onset of wrinkling might be a suitable design allowable for a high pressure gas pipe, or oil pipe, where wrinkling of the pipe may restrict the passage of pigs; or failure of the pipe might result in fire or other serious consequences to nearby facilities and habitat. However, for the case of water pipelines, it is rare that release of water poses serious consequences to the nearby environment, so some post-wrinkling performance may be acceptable.

Once local buckling (wrinkling) starts, there is usually a 50% to 500% increase in capability before the pipe wrinkles sufficiently to initiate a through wall crack. Recent tests of a 30" diameter, t=0.327" (D/t=92) pipe with Fy=70 ksi, (DelCol, 1998 showed that for internal pressures in the range of 0 psi to 312 psi for that pipe, the initial buckle formed at an average compressive strain of about -0.5%, which corresponds to 0.229 t/R. For an unpressurized pipe, average compressive strains over one pipe diameter length, at the wrinkle, reached 3.5%, without breach of the pressure boundary.

Once a wrinkle forms, additional shortening of the pipeline will tend to accumulate at the wrinkle.

Under wave propagation, peak longitudinal compressive strains in the pipe should be lower than the onset of significant wrinkling (equation 1). Under fault offset or other limited area PGD loading, peak compressive longitudinal strains should be kept below equation 2 (or below 5% strain within the wrinkled joint, when considering local joint geometry).

⎡ 2⎤ wave passage t ⎛ pD ⎞ εc = 0.75⎢ 0.50 ' − 0.0025 + 3000⎜ ⎟ ⎥ [eqn. ⎣ R ⎝ 2Et ⎠ ⎦ 1]

15 5.0 D R ' = 3 1 ()−− DD D min

t εPGD = 0.88 c R [eqn. 2] where D is pipe outside diameter.

Example. Assume a 96" inside diameter butt welded steel pipe with t = 0.75 inches. The nominal onset of compressive wrinkling (0.175t/R) is -0.27%. Assuming that Dmin is 95 inches (2.5 inch out of roundness), and an internal pressure of 150 psi, equation (1) gives the allowable strain at -0.37%. For fault offset, equation (2) gives the allowable strain at - 1.35%. Equation 2 allows for post-wrinkling behavior, and assumes that this is acceptable to the owner.

For compressive strains higher than -5% (when measured ignoring wrinkle geometry), tears in the pipe should be expected. For most water pipelines at moderate temperatures (over 40°F), the tear length has not been observed to propagate, with a resulting leak. Tear openings have been observed as about 0.25 inches wide x 12 inches long (36" diameter pipe with double lap weld impacted by fault creep), resulting in leak rates on the order of a 1,000 gpm to 2,000 gpm.

The above equations do not apply for single or double lap welded pipes, where the onset of wrinkling occurs at lower forces owing to the major geometric discontinuity at the joint. For double lap welded pipes, the longitudinal compressive stress in the main pipe should be kept to 0.60 Fy to prevent wrinkling; or the peak bending strain within the wrinkled joint kept below 5% when considering joint geometry.

For single lap welded pipes, the longitudinal compressive stress in the main pipe should be kept to 0.40 Fy to prevent wrinkling; or the peak bending strain within the wrinkled joint kept below 5% when considering joint geometry.

In all cases where yielding of the steel is allowed, the weld consumables, welding procedures and inspection criteria shall be suitable to ensure development of gross section yielding of the pipe section both for field girth joints and shop fabricated longitudinal spiral or straight seam joints.

Tensile Strain Limit The longitudinal strain in a butt welded steel pipe should be limited to a level to achieve the target performance level of the pipeline. For offset displacements which are defined as having about a 16% chance of exceedence given the design basis earthquake, maximum tensile longitudinal strains should be kept to about 0.25 times ultimate uniform

16 strain (strain before necking) of the steel, or about 5%. This design limit provides for some capacity to withstand larger fault offset, or to accommodate minor flaws in the pipe and girth joint.

Should double lap welded steel pipe be used, then the maximum longitudinal strain in the pipe must be kept low enough such that there is a reasonable chance of survival of the joint. Test data on double lap welded joints suggests that perhaps one quarter of the joints will break when the strain in the pipe away from the joint reaches about 8%. This suggests that the maximum allowable strain in the main body of the pipe should be kept to 2%, or perhaps no more than 4% to have a reasonable chance of maintaining the pressure boundary. At 2% strain, the reliability of a double lap welded pipe will be similar to a similar quality butt welded pipe at 5% strain.

The girth joints in single lap welded steel pipe will generally not be strong enough to allow longitudinal tensile yielding in the main pipe. Therefore, single-lap welded transmission pipe should not be used in areas prone to PGD (landslide, liquefaction or fault offset) should the owner wish a highly reliable pipeline.

DESIGN CONSIDERATIONS AT FAULT CROSSINGS

Additional design considerations specific to pipelines at fault crossing are: (1) fault types and fault zones, (2) orientation of the pipes with respect to the fault line, (3) design earthquakes and the associated magnitude of fault displacements, (4) geotechnical hazards, (5) soil-pipeline interaction, (6) joints used to accommodate fault displacements, i.e., expansion-contraction joints and flexible couplings, (7) analysis methods, and (8) design redundancy.

Section 8.2 of the ALA Guidelines provides detailed discussions on the above issues. Section 12.1 also provides additional information about using EBAA Iron ball joints at fault crossing. An example of San Francisco’s fault crossing project is described in Cheng’s paper (2001), which was presented at the 2nd Japan and U.S. Workshop.

CONCLUSIONS

This paper provides general description of the major seismic design issues that should be considered during the planning and design phases of a transmission pipeline project in moderate and high seismic regions. Detailed design procedures or specific detailed information can be found in the ALA Design Guidelines. The designer can also use this paper as a checklist for planning and reviewing a transmission pipeline project.

ACKNOWLEDGMENTS

The author wishes to thank the following individuals: Susan Yee of SFPUC/EDB for her encouragement and support; Thanh Nguyen of SFPUC/CDD and Mongkol Mahavongtrakul of SFPUC/EDB for their valuable information on ductile iron pipes and

17 corrosion control; and John Eidinger of G&E Engineering for his invaluable contribution on welded joints design and patient review of the paper and Chapter 8 of the Guidelines.

The Guidelines are written under contract to the American Lifelines Alliance, a public- private partnership between the Federal Emergency Management Agency (FEMA) and the National Institute of Building Sciences (NIBS). This report was prepared by a team representing practicing engineers in the United States water utility industry and academics.

The ALA “Design Guideline for Seismic Resistant Water Pipeline Installations” is currently available as a draft, for industry comment. The complete draft guidelines is available for download at: http://homepage.mac.com/eidinger/Menu7.html (click on link: “downloads”, click on: Seismic Guidelines.doc). Comments should be sent by January 29, 2005 to the author of this paper at [email protected] or to John Eidinger at [email protected].

References

American Lifelines Alliance (to be published in 2005), Design Guideline for Seismic Resistant Water Pipeline Installations

ASCE Manuals and Reports on Engineering Practice No.79, Steel Penstock, 1993.

American Water Works Association Manual of Water Supply Practice, Concrete Pressure Pipe, AWWA M9, 1995.

American Water Works Association Manual of Water Supply Practice, Steel Pressure Pipe, AWWA M11, .1989.

American Water Works Association Manual of Water Supply Practice, Ductile Iron Pipe and Fittings, AWWA M41, 1996.

Ballantyne, D., Taylor, C., USGS Grant Award 14-08-0001-G1526, “Earthquake Loss Estimation Modeling of the Seattle Water System”, Kennedy/Jenks/Chilton, Federal Way, Washington, 1990

California Office of Emergency Services, Emergency Planning Guidance for Public and Private Water Utilities March 1999

Cheng, L., “Seismic Design of Water Pipelines at Fault Crossing”, Proceedings of the Second Japan and U.S. Workshop on Seismic Measures for Water Supply, August 2001

Earthquake-Proof Design of Buried Pipelines, Pipe Research Laboratory of Kubota Ltd. Tokyo, Japan, 1981.

Eidinger, J., Collins, F. and Conner, M. 6th International Conference on Seismic Zonation, “Seismic Assessment of the San Diego Water System”, Earthquake Engineering Research Institute, Palm Springs, CA, 2000.

18 Timoshenko, S. P. and Gere, J. M., Mechanics of Materials, 3rd ed., Pws Pub Co, 1990.

19 S6-4

“Seismic Diagnosis of Extensive Water Distribution Network”

Presenter: Nobuhisa Suzuki

(JFE R&D Corporation, Japan)

Cover page

Title: Seismic Diagnosis of Extensive Water Distribution Networks

Authors: Nobuhisa Suzuki (Contact person) Principal Researcher JFE R&D Corporation 1-1 Minami-Watarida, Kawasaki Kawasaki, Japan 210-0855 Phone: (81)44-322-6234 Fax: (81)44-322-1512 E-Mail: [email protected]

Nobuhiro Hasegawa Pipeline Engineer Water Facilities Division JFE Engineering Corporation 2-1 Suehiro, Tsurumi Yokohama, Japan 230-8611 Phone: (81)45-505-7669 Fax: (81)45-505-8903 E-mail: [email protected]

Takahiro Yabuguchi Pipeline Engineer Waster Facilities Division JFE Engineering Corporation 2-1 Suehiro, Tsurumi Yokohama, Japan 230-8611 Phone: (81)45-505-7669 Fax: (81)45-505-8903 E-mail: [email protected]

Seismic Diagnosis of Extensive Water Distribution Network

Nobuhisa Suzuki, Nobuhiro Hasegawa and Takahiro Yabuguchi

ABSTRACT

Validation of NeEX (Network EXpress), an accurate and fast simulation program to predict seismic responses of water distribution networks, is presented in this paper. NeEX has been equipped with an extremely swift solver, so it is capable of calculating the seismic responses of the water distribution networks more than 10,000 times as fast as FE programs do. And NeEX provides almost the same accuracy as the FE programs which accuracy is sufficient to verify seismic integrity of the water distribution networks. Therefore, NeEX is applicable to seismic diagnosis of even a huge water distribution network with an intricate geometry in metropolitan areas. This kind of sophisticated seismic diagnosis had never been performed before due to a lot of expenses of FEA. Verification of NeEX is performed based on comparison with results obtained by the regression formula developed by JWRC (Japan Water Research Center) in order to predict the number of breaks of pipes and fittings. A hypothetical water distribution network is defined for the verification in this paper, which consists of ductile iron pipes connected with K-type joints and is extended to a square area with sides 5 km long. Results of the verification are summarized as follows; 1) The numbers of breaks obtained by NeEX show good agreement with those by the regression formula, 2) NeEX may yield the conservative number of breaks, whose maximum is approximately twice as large as that calculated by the regression formula.

______Nobuhisa Suzuki, Dr. Engineering, P.E., Principal Researcher, JFE R&D Corporation, 1-1 Minami-Watarida, Kawasaki-ku, Kawasaki, Japan 210-0855 ([email protected]) Nobuhiro Hasegawa, Pipeline Engineer, Water Facilities Division, JFE Engineering Corporation, 2-1 Suehiro, Tsurumi-ku, Yokohama, Japan 230- 8611 ([email protected]) Takahiro Yabuguchi, Pipeline Engineer, Water Facilities Division, JFE Engineering Corporation, 2-1 Suehiro, Tsurumi-ku, Yokohama, Japan 230- 8611 ([email protected])

1 INTRODUCTION

Seismic diagnosis and mitigation of existing water distribution networks have been recognized to be one of significant issues to ensure seismic integrity of the water distribution networks in urban areas. It is not to mention that seismic diagnosis and mitigation will also be required for distribution networks to be constructed. Water distribution networks cover wide areas and consist of relatively small diameter pipes with complicated geometries. Therefore, seismic diagnosis of the water distribution networks using FE programs had not been conducted due to a lot of expenses and their complicated geometries. While transmission lines consist of relatively large diameter pipes with simple geometries so that it would be relatively easy to perform seismic diagnosis. An extremely fast simulation method was developed to realize seismic diagnosis of extensive water distribution networks, which fast simulation method has been installed in the sophisticated simulation program NeEX. NeEX has been already applied to seismic diagnosis of several water and gas distribution networks [1]-[4]. As presented in the literatures [1]-[4], NeEX can calculate more than 10,000 times as fast as the FE programs can do and accuracy of NeEX is equivalent to FEA. Due to the outstanding performance of NeEX with respect to the accuracy and the swiftness, it has become possible to obtain sufficient information about seismic improvement methodology for the water distribution networks. Even when the water distribution network is huge and complicated, NeEX will be able to predict the number and the location of breaks of the distribution pipes and the fittings. This paper describes verification of NeEX due to comparisons with results obtained by the regression formula developed by JWRC [5], which can predict the number of breaks of distribution pipes and fittings such as tees and bends. A hypothetical water distribution network spreading over a square area with sides 5 km long was defined for the comparison. The hypothetical network was consisted of ductile iron pipes connected with K-type joints (hereinafter DIP-K) and had the total length of 150 km. Results of the verification are summarized as follows; 1) The numbers of breaks of the pipes obtained by NeEX show good agreement with those predicted by the regression formula, 2) NeEX may give a conservative number of pipe breaks, whose maximum value is approximately twice as large as that obtained by the regression formula, 3) The number of breaks due to past earthquakes sometimes become twice as many as the regression formula predicts, therefore, the prediction of NeEX will be appropriate to cover the maximum number of the actual breaks or damage.

ESTIMATION OF BREAKS OF A WATER DISTRIBUTION NETWORK

Overview of the fast simulation program NeEX

Idealization of the hypothetical network is described in this section as the theoretical basis of NeEX was presented at the 3rd US-Japan symposium [3]. Figure 1 presents a small part of the hypothetical network where seismic waves propagate along the straight pipe. The small part of the network shown in Fig. 1 should be defined as a segment, which is a fundamental unit to idealize the complicated water distribution network. The small part of the network is composed of a straight pipe, a tee and a bend. The tee and the bend should be defined as boundary elements for the segment. Other types of segments should be defined in order to idealize the water distribution networks, which have different boundary elements such as a bend with arbitrary angle and a branch perpendicular to the straight pipe.

2 Seismic waves propagating along the straight pipe in Fig. 1 are a component of shear waves in the longitudinal direction, which have been defined in the Seismic Design Codes for Water Facilities [6]. The design spectrum for response velocity of Ground Motion Level-2, whose maximum response velocity is as large as 100 kine, was chosen for the verification of NeEX. Critical extension of the K-type joints should be assumed in order to estimate the number of breaks in the water distribution networks. The critical extension of 40 mm was assumed for 100, 150 and 200 mm nominal diameter pipes and 64 mm for 300, 400 and 500 mm nominal diameter pipes.

1000

G.M. Level -2 100

G.M. Level -1

Sv (cm/sec) 10

1 0.1 1 10

ＴG (sec)

Fig. 1 A segment to be defined as a piece of a Fig. 2 Design spectrum of response velocity water distribution network, which consists for seismic design of buried pipelines of a straight pipe and a bend and a tee

Estimation of the Number of Breaks by the Regression Formula

The number of breaks of pipes can be estimated by the following regression formula, which was developed based on breaks of reported after the 1995 Kobe earthquake [5]. The regression formula was derived from the data on the breaks of straight pipes, tees and bends. The data on breaks of pipe beam bridges and air valves and hydrants was not included in the regression analysis. Therefore, the regression formula should be used to predict the number of breaks of buried pipes.

= ∗ ∗ ∗ dldpb ∗ LSCCCN (1)

Where Nb expresses the number of breaks, C p , Cd and Cl are correction factors for pipe material, pipe diameter and extent of liquefaction, respectively. And Sb represents the standard number of breaks and L is the total length of each pipe material within every mesh.

The standard number of breaks, Sb , can be calculated by the following formula.

−5 10.2 b ∗∗= S1033.6S V (2)

3 Where Sv is the maximum response velocity (kine). The formula should be applied when Sv is 110 kine and smaller.

The correction factors C p for DIP-K becomes 1.0 for pipe diameters from 100 to 150 mm and

C p becomes 0.9 for nominal diameters from 200 to 250 mm. The values of 0.51 and 0.07 should be applied for nominal diameters from 300 to 450 mm and from 500 to 600 mm, respectively. As shown in Fig. 2, when natural ground period is 0.7 sec and longer, equation (2) gives Sb of 1.00 as the maximum response velocity is equal to 100 kine. Consequently Sb becomes 0.3 for DIP-K with a nominal diameter of 150 mm.

A HYPOTHETICAL DISDRIBUTION NETWORK FOR THE VERIFICATION

The hypothetical water distribution network is presented in Fig. 3, which covers a square area with sides 5 km long and has the total length of 160 km. Two networks with different nominal pipe diameters were assumed for the verification of NeEX. Network-1 consists of DIP-K with a uniform nominal diameter of 150 mm and Network-2 consists of DIP-K with six nominal diameters of 100, 150, 200, 300, 400 and 500 mm. The hypothetical water distribution network shown in Fig. 3 was generated from a road map in Yokohama city. The six nominal diameters of DIP-K taken into account in Network-2 were roughly assigned in accordance with breadth of the roads. Borehole data in the area was used to calculate a natural period of the ground for every small square with sides 50m long. In order to discuss the effect of the natural period of the ground on the number of breaks of the buried pipes, the mean natural period for every small area was adjusted to 0.6, 0.8, 1.0 and 1.2 sec.

Fig. 3 A hypothetical water distribution network in a square area with sides 5km long

4 Relationships between the natural period of the surface ground and the total length of the pipes are presented in Figs. 4 through 7. The natural period for the segment was assigned to be the same as the mean value calculated for the small area. When a segment spanned over two or more small areas, the natural period of the small area was used for the segment where the longest portion of the segment was buried. As shown in the figures, the pipes with the nominal diameters of 150 and 200 mm accounts for more than 60% of the total length of the pipes.

50 50 φ500mm φ500mm φ400mm φ400mm 40 φ300mm 40 φ300mm φ200mm φ200mm φ150mm φ150mm 30 φ100mm 30 φ100mm

20 20 Total length (km) Total length (km) 10 10

0 0 0.0 0.5 1.0 1.5 2.0 2.5 0.0 0.5 1.0 1.5 2.0 2.5

TG (sec) TG (sec)

Fig. 4 Natural period at segments Fig. 5 Natural period at segments

(Mean natural period is 0.6 sec) (Mean natural period is 0.8 sec)

50 50 φ500mm φ500mm φ400mm φ400mm 40 φ300mm 40 φ300mm φ200mm φ200mm φ150mm φ150mm 30 φ100mm 30 φ100mm

20 20 Total length (km) Total length (km) 10 10

0 0 0.0 0.5 1.0 1.5 2.0 2.5 0.0 0.5 1.0 1.5 2.0 2.5

TG (sec) TG (sec)

Fig. 6 Natural period at segments Fig. 7 Natural period at segments (Mean natural period is 1.0 sec) (Mean natural period is 1.2 sec)

5 PREDICTION OF BREAKS IN THE HYPOTHETICAL DISTRIBUTION NETWORK

The Number of Breaks in Network-1 (Pipe diameter of 150 mm is taken into account)

Figures 8 through 11 compare the number of breaks in Network-1 calculated by NeEX and the regression formula. As presented in the figures, it is common to the both results that the number of breaks varies with natural period and presents the maximum number at the mean natural period. The regression formula tends to present the number of breaks with almost a symmetric distribution around the mean natural period. However, the numbers of breaks at the short periods are slightly larger than those at longer periods. On the other hand, NeEX tends to yield the number of breaks twice as large as those of the regression formula except at very short natural periods less than 0.5 sec.

20 20 NeEX NeEX Reg. formula Reg. formula 15 15

10 10 Breaks (N) Breaks (N) 5 5

0 0 0.0 0.5 1.0 1.5 2.0 2.5 0.0 0.5 1.0 1.5 2.0 2.5

TG (sec) TG (sec)

Fig. 8 The number of breaks of Network-1 Fig. 9 The number of breaks of Network-1

(Mean natural period is 0.6 sec) (Mean natural period is 0.8 sec)

20 20 NeEX NeEX Reg. formula Reg. formula 15 15

10 10 Breaks (N) Breaks (N) 5 5

0 0 0.00.51.01.52.02.5 0.00.51.01.52.02.5

TG (sec) TG (sec)

Fig. 10 The number of breaks of Network-1 Fig. 11 The number of breaks of Network-1 (Mean natural period is 1.0 sec) (Mean natural period is 1.2 sec)

6 Figure 12 compares the number of breaks mentioned above. Most of the numbers of breaks in the figure are between 0.4 and 0.8 breaks/km when the natural period is 0.6 sec or longer. The broken lines present the average number of breaks with respect to the mean natural period. They are 0.29, 0.41, 0.48 and 0.50 breaks/km for 0.6, 0.8, 1.0 and 1.2 sec, respectively. The total numbers of breaks are compared in Table 1, where the ratios regarding the numbers of NeEX to the regression formula range from 1.3 to 1.8. Based on the comparison of the number of breaks, it can be concluded that NeEX is conservative to predict the number of breaks by approximately 80%.

2.0 1.2sec 1.0sec 0.8sec 1.5 0.6sec

1.0

0.5 Breaks per km (N/km)

0.0 0.0 0.5 1.0 1.5 2.0 2.5

TG (sec)

Fig. 12 The number of breaks per km in Network-1

Table 1 The number of breaks in Network-1

Mean natural period (sec)

0.6 0.8 1.0 1.2 NeEX 46 66 76 80

Regression formula 36 44 43 46 Ratio(=NeEX/R.F.) 1.3 1.5 1.8 1.7

The Number of Breaks in Network-2 (Six pipe diameters are taken into account)

Figures 13 through 16 present the number of breaks in Network-2 predicted by NeEX and the regression formula. The similar tendency to Network-1 can be seen in the figures, which can be summarized as follows. The number of breaks varies with natural period and presents the maximum

7 number at the mean natural period for both Network-1 and Network-2. The regression formula tends to present almost a symmetric pattern of distribution of the number of breaks with respect to the mean natural period. However, the numbers of breaks at shorter periods are slightly larger than that at longer periods. While, NeEX shows a tendency to yield the conservative numbers of breaks except at a natural period of 0.5 sec and shorter. Table 2 presents the ratios of the number of breaks predicted by NeEX to that by the regression formula. The ratios are 1.3, 1.7, 1.9 and 2.0 when the mean natural periods are 0.6, 0.8, 1.0 and 1.2 sec, respectively. Observing the comparisons, NeEX may be conservative to predict the number of breaks, which can be twice as large as that calculated by the regression formula.

20 20 NeEX NeEX Reg. formula Reg. formula 15 15

10 10 Breaks (N) Breaks (N) 5 5

0 0 0.00.51.01.52.02.5 0.00.51.01.52.02.5

TG (sec) TG (sec)

Fig. 13 The number of breaks in Network-2 Fig. 14 The number of breaks in Network-2 (Mean natural period is 0.6 sec) (Mean natural period is 0.8 sec)

20 20 NeEX NeEX Reg. formula Reg. formula 15 15

10 10 Breaks (N) Breaks (N) 5 5

0 0 0.0 0.5 1.0 1.5 2.0 2.5 0.00.51.01.52.02.5

TG (sec) TG (sec)

Fig. 15 The number of breaks in Network-2 Fig. 16 The number of breaks in Network-2

(Mean natural period is 1.0 sec) (Mean natural period is 1.2 sec)

8 Figures 17 through 20 compare the number of breaks with respect to the pipe diameters. As seen in Fig. 17, in the case of the natural period of 0.6 sec, no breaks is observed at a natural period of 0.5 sec and shorter. At the natural periods of 0.6 and 0.7 sec, the number of breaks can be seen for the nominal pipe diameters of 150 and 200 mm. When the natural period is 0.8 sec and longer, the number of breaks of the pipes with the nominal diameter of 300 mm appears. And the number of breaks of the large diameter pipes tends to increase with natural period. The results shown in Figs. 18 through 20 are similar to Fig. 17, where the natural periods are 0.8, 1.0 and 1.2 sec, respectively. The number of breaks comes to appear when the natural period is 0.6 sec and longer. When the natural period is 0.8 sec and longer, the breaks of the pipes tend to concentrate to the pipes with nominal diameters equal to 300 mm and larger. However this tendency is reflected in the cases of the longer natural periods, it is that the number of breaks of the pipes with larger diameter tends to increase with increasing natural period.

Table 2 The number of breaks in Network-2

Mean natural period (sec.)

0.6 0.8 1.0 1.2 NeEX 31 50 60 68

Regression formula 23 29 32 34

Ratio(=NeEX/R.F.) 1.3 1.7 1.9 2.0

20 20 φ500mm φ500mm φ400mm φ400mm φ300mm φ300mm 15 15 φ200mm φ200mm φ150mm φ150mm φ100mm φ100mm 10 10 Breaks (N) Breaks Breaks (N) Breaks 5 5

0 0 0.0 0.5 1.0 1.5 2.0 2.5 0.0 0.5 1.0 1.5 2.0 2.5

TG (sec) TG (sec)

Fig. 17 The number of breaks in Network-2 Fig. 18 The number of breaks in Network-2 (Mean natural period is 0.6 sec) (Mean natural period is 0.8 sec)

20 20 φ500mm φ500mm φ400mm φ400mm φ300mm φ300mm 15 15 φ200mm φ200mm φ150mm φ150mm φ100mm φ100mm 10 10 Breaks (N) Breaks (N) Breaks 5 5

0 0 0.0 0.5 1.0 1.5 2.0 2.5 0.0 0.5 1.0 1.5 2.0 2.5

TG (sec) TG (sec)

Fig. 19 The number of breaks in Network -2 Fig. 20 The number of breaks in Network -2

(Mean natural period is 1.0 sec) (Mean natural period is 1.2 sec)

The Number of Breaks Predicted by NeEX and the Regression Formula

The regression formula has been derived from the data regarding breaks of pipes and fittings shown in Fig. 21, which data was collected after the 1995 Kobe earthquake. The figure plots the data in terms of the number of breaks per km and the maximum response velocity of the ground. The broken line (the blue line) presents the standard number of breaks per km given by the regression formula, and the solid line (the red line) presents the same value as the broken line for the DIP-K pipes multiplying the correction factor of 0.3 for the DIP pipes.

3 Damage Reg. formula DIP 2

1 Breaks per km (N/km)

0 020406080100 Maximum response velocity (kine)

Fig. 21 The number of breaks and the maximum response velocity of the ground (Data was collected after the 1995 Kobe earthquake)

10 As observed in Fig. 21, some of the collected number of breaks per km range from 1.7 to 2.0 times as large as those predicted by the regression formula. As mentioned above, NeEX may predict 1.3 to 1.8 times as large number as the regression formula predicts in the case of Network-1. And the predicted number of breaks due to NeEX may be 1.3 to 2.0 times as large as the value calculated by the regression formula in the case of Network-2. Consequently, based on the results mentioned above, NeEX tends to predict conservative values that are almost the same as the maximum number of breaks in the collected data after the 1995 Kobe earthquake.

CONCLUSION

In order to verify of the fast and accurate simulation program NeEX, the results with respect to the number of breaks of pipes were compared with those calculated by the regression formula developed by JWRC. A hypothetical water distribution network covering a square area with sides 5 km was used for the verification, in which area the DIP-K pipes with the total length of 160 km were taken into account. The results of the verification can conclude that the number of breaks predicted by NeEX will be twice as large as those calculated by the regression formula and NeEX may give the conservative number of breaks. The regression formula was developed based on the actual number of breaks occurred during the 1995 Kobe earthquake, however, the maximum numbers of breaks reported after the earthquake are about twice as large as those calculated by the regression formula. Consequently, NeEX will be able to estimate the appropriate number of damage that is very close to the reported maximum number of breaks of the pipes after the 1995 Kobe earthquake. Although it was not explained in the paper, however, it is possible for NeEX to predict locations of breaks of buried pipes, fittings and joints. This kind of information will be useful to consider countermeasures and mitigations for extensive water distribution networks. More, NeEX will be able to simulate the effects of countermeasures and mitigations, which would be the most significant work in order to ensure seismic integrity of extensive water distribution networks.

ANKOWLEDGMENTS

The accurate and fast simulation program NeEX was originally developed in order to conduct seismic diagnosis of an extensive gas distribution network in Tokyo metropolitan area. The author would like to thank Dr. Yoshihisa Shimizu, Mr. Ken-ichi Koganemaru and Mr. Naoyuki Hosokawa of Tokyo Gas Co. Ltd. and Mr. Toshiyuki Mayumi and Mr. Takeshi Mori of JFE Engineering Corporation for their invaluable comments and advices. The authors also wish to thank Mr. Izumi Kubo of Japan Industrial Testing Co. Ltd. for development of NeEX.

REFERENCES

[1] Suzuki, N., Horikawa, H., Mori, T., Mayumi, T., Shimizu, Y., Koganemaru, K., Yoshizaki, K. and Hatsuta, Y., 2002, “A Fast Analytical Method for Seismic Responses of Buried Pipeline Networks,” Proceedings of the 11th Japan Earthquake Engineering Symposium. [2] Koganemaru, K., Shimizu, Y., Yoshizaki, K., Hatsuta, Y., Suzuki, N., Horikawa, H., Mori, T. and Mayumi, T., 2002, “Development of an Evaluation Method for Earthquake Resistance of Buried Pipeline Networks,” Proceedings of the 11th Japan Earthquake Engineering Symposium.

11 [3] Suzuki, N., 2003, “A Fast Simulation Method for Predicting Seismic Responses of Extensive Water Distribution Networks,” The 3rd US-Japan Symposium on Seismic Design for Water Facilities. [4] Suzuki, N., Koganemaru, K. and Shimizu, Y., 2004, “Swift Algorithm to Predict Seismic Responses of Extensive Buried Distribution Networks,” Paper No. 500, 13WCEE. [5] JWRC, 2000, “Report on Prediction and Investigation of Damage to Water Pipes due to Warthquakes.” [6] JWWA, 1997, Seismic Design Codes for Water Facilities.

12 S6-5

“Seismic Upgrade of Prestressed Concrete Water Tanks”

Presenter: David D. Lee

(East Bay Municipal Utility District, USA)

Seismic Upgrade of Prestressed Concrete Water Tanks

David D. Lee, P.E. Eric M. Fieberling, P.E. East Bay Municipal Utility District

ABSTRACT

Most of the East Bay Municipal Utility District’s distribution water tanks are located near a major earthquake fault. Many of the tanks are above grade, prestressed concrete tanks that were built prior to the 1970’s before the onset of modern seismic codes. The potential failure modes for these tanks include sliding and uplift of the tank wall over the existing foundation, and hoop overstress of the wall prestressed steel, causing loss of contents. The retrofit design criteria and analysis methods are based on the AWWA Standard with modifications to include near fault and site-specific soil effects. Retrofit schemes for sliding include interior curbs connecting the wall to the floor slab or exterior cables connecting to a new foundation. To prevent uplift either seismic cables or rock anchors were used. Retrofit schemes for hoop overstress involve the addition of pre-stressed wire wrapping or post-tensioned tendons or bars and a cover of shotcrete. Many tanks were upgraded with the tanks in service to reduce impacts to the system and cost. Many non-seismic components of the tanks such as seals, ladders, piping, valves and telemetry were rehabilitated or replaced in conjunction with the seismic upgrade work. Remote controlled seismic isolation valves with flow meters were included at selected reservoir sites to provide a means of saving water in those tanks after a pipe break following an earthquake. The upgrades have proved cost effective compared to replacement of the tanks. The program should greatly improve the post earthquake performance of the District’s water system.

INTRODUCTION

The East Bay Municipal Utility District has over 170 water distribution reservoirs including 67 pre-stressed concrete tanks ranging in size from 200,000 gallons to 11 million gallons. Most of these tanks are above grade and were built prior to 1965. The majority of these are located near the two major earthquake faults that traverse the District’s service area and 58 of them were analyzed to have an unacceptable risk of failure in a probable earthquake (magnitude based on 10% probability of occurrence in 50 years). The District is currently in the final year of implementing a ten-year, $189 million Seismic Improvement Program to seismically strengthen its water distribution system. Over $40 million will be spent to seismically upgrade the vulnerable pre-stressed concrete tanks.

Seismic Hazards

The District’s service area covers approximately 325 square miles on the east side of San Francisco Bay. The adjacent Figure 1 shows the two major earthquake faults, the Hayward and Calaveras Faults, in the District’s service area and a third fault, the Concord Fault, which is just to the east of the service area. Most of the District’s 1.3 million customers are located in close

1 proximity to either the Hayward or Calaveras Faults.

Figure 1. Major Faults

Following the 1989 Loma Prieta earthquake, the District undertook a comprehensive study of the seismic vulnerability of District facilities. This study identified the vulnerabilities, presented a capital improvement program and established upgrade goals to maximize system performance following a major earthquake. The evaluation was system-based to model overall system performance. Each facility was evaluated for the various ways an earthquake can cause damage, including: Ground Shaking, Fault Rupture, Seismic induced Landslides, Liquefaction, Ground spreading and lurching. Scenario earthquakes with a 10% in 50-year probability of occurrence were used in the study and included magnitudes 7.0 on the Hayward fault, 6.75 on the Calaveras fault and 6.5 on the Concord fault.

Ground shaking and landslides are the primary seismic concerns for the prestressed concrete tanks. Fault rupture is rarely an issue because the tank sites typically avoid known fault locations. Liquefaction and ground spreading and lurching tend to occur in weak soils found in low-lying areas, however to provide the required pressure for water delivery, most tanks are located on higher ground with better foundation materials (typically stiff-soil or weak-rock). Each tank was prioritized based on the following factors: life safety (adjacent or “downstream” habitants), importance to the system, damage susceptibility, and current condition.

FAILURE MODES

The primary failure modes for pre-stressed concrete tanks subjected to seismic loads are sliding, hoop overstress and wall uplift causing damage and loss of contents. One or more of these modes apply to each of the vulnerable tanks. The tank aspect ratio (i.e. the height to diameter ratio) is a key index for the anticipated types of failure. Higher aspect ratios lead to uplift problems while

2 sliding is a greater concern for lower aspect ratios. Damage to tank piping can also cause failure.

Sliding

The tanks were built with little connection between the wall and ring footing or interior slab. The wall base was designed to slide outward with increased water inside the tank. The outer edge of the interior slab sits on the footing with a continuous “ring” joint between slab and the interior face of the tank wall. Figure 2 shows a typical wall section. Beginning in the early 1950’s, radial keys were cast in the ring footing (and bottom of core wall) to resist tangential in-plane wall movement but allow radial outward movement. However, the shear capacity of the keys is well below the expected earthquake forces. Excessive movement of the wall would breach the ring joint seal causing severe leakage. Pounding against the interior slab would cause excessive stress in the slab and wall resulting in damage and possible permanent deformation. Figure 2. Typical Wall Section Hoop Overstress

Hoop stress is dependent on the original design forces (with corresponding area of steel and applied stress levels), corrosion of the existing prestressing wires or bars and residual stress after long term creep losses. These factors were evaluated through a combination of field inspection and analysis. Prestressing bars were used until the early 1950’s when wire-winding techniques were developed. The older tanks with bars tend to have severely insufficient applied force and hoop capacity for the combined hydrostatic and hydrodynamic (seismic) loads. However the bars are not high strength material and have some ductility. In addition, little corrosion damage has been observed. Consequently, the expected damage is extensive wall cracking and excessive movement in the wall joints. By contrast, the newer prestressing wires are a higher strength material and were applied at higher stresses. However, they are more likely to be damaged or break if corroded. The wire-stressed tanks also have insufficient hoop capacity due to the low original design forces and would likely experience wall cracking. They could also experience local wire breakage and subsequent severe damage to the wall.

Most of the prestressed tanks have cast-in-place dome roofs. Each dome is supported at the top of the tank wall with a dome ring (tension ring) that has additional prestressing to resist dome thrust. Analysis revealed overstress in the dome rings under vertical loads and seismic forces, which can be either horizontal or vertical. This would result in cracking and potential collapse of the dome roof and damage to the walls.

3 Uplift

The hydrodynamic forces caused by seismic events include both the impulsive and convective (sloshing) components. The convective force tends to act higher up the tank wall and thus contributes to the overturning tendency. If the dead load of the wall and roof are not sufficient to resist the overturning the wall will uplift from the existing footing. This can cause damage to the ring seal as well as the wall and footing, and result in complete loss of contents. Many of the high aspect tanks were built with small ring footings that are undersized for the additional bearing pressure caused by the seismic overturning forces. This can result in footing settlement and further damage.

Valve Pit and Piping Damage

In addition to the tank damage described above, failure can result from damage to the valve pit or the inlet/outlet (I/O) piping. Most of the valve pits are cast in place concrete structures located adjacent and connected to the tank ring footing. The I/O pipe enters into the tank through the tank floor and is encased in concrete under the tank to the valve pit. Many of the pits have precast concrete roofs. The pre-cast roof panels could fall and damage the valves and piping. The I/O pipe is susceptible to damage at the transition from the concrete valve pit wall to the adjacent soil. This is a particular concern for cast iron pipes or soft soils. If the tank roof is damaged and debris falls into the tank, the I/O pipe could get clogged and valves could be damaged.

Pipeline breaks downstream from the many tanks could cause the tank to be drained within hours after a major earthquake. Refilling the reservoirs could take many days depending on the difficulty of the pipeline repairs. The danger of fires is also present in most pressure zones and numerous fires are anticipated to occur within the District Service Area in the first 24 hours after a seismic event. This is particularly hazardous in zones that are remote or difficult to access.

RETROFITS

The retrofit design criteria were established after completion of the Seismic Evaluation Study. It was determined that funding could be justified to strengthen the tanks to prevent collapse but that additional funding to retrofit or replace the tanks to a higher standard was not merited. The upgraded tanks may suffer damage, and the water may require boiling before drinking. However, the tanks should survive to provide water for fire suppression and basic needs immediately after the earthquake. For example, some tanks may develop cracks and leak after an earthquake, however, the leakage rate should be low enough to keep the tank usable.

The basis of the retrofit designs is the American Water Works Association Standard D110 (Wire- and Strand-Wound Circular Prestressed Concrete) with modifications by the District to account for the high seismic region. The modifications include the use of seismic coefficients for type and nearness of faults, and for soil profiles. Site-specific response spectra were used in the analysis of some tanks and a finite-element seismic analysis was performed on one of these tanks. The finite-element analysis gave the District a better understanding of wall stresses and validated the design. Retrofit methods were developed for each of the primary failure modes. The retrofit designs vary depending on the size, aspect ratio, location and condition of each tank.

Retrofit for Sliding

Two primary methods have been used to resist sliding. The first incorporates new interior perimeter concrete curbs with connections to the existing floor slab and the tank wall. For this method the tank must be drained. Reinforcing dowels are anchored in each vertical wall panel. The curbs are separated by gaps at each of the vertical wall joints and a slip pad separates the curb bottom from the floor slab. The floor slab connections consist of grout filled steel pipes anchored in the floor slab inside tube steel sockets cast in the curb. This allows for wall movement in the radial direction for normal hydrostatic and thermal loads but provides resistance in the tangential direction (i.e. parallel to wall) for seismic loads. The seismic sliding load is thus transferred to the slab and is ultimately resisted by friction on the soil below the slab due to the weight of the contents. Figure 3 shows construction photos of a typical interior curb.

Figure 3. Interior Sliding Restraint Curbs

The second method includes an exterior ring footing addition with diagonal seismic cables attached to the exterior wall surface and extending into the footing. The ring footing addition is typically anchored to the foundation sub-grade with drilled concrete piers. This upgrade can be done even if the tank remains in service. The cables are high strength galvanized seven-wire strands and bond to a layer of new shotcrete on the wall under the new prestressing. The cables pass through a neoprene sleeve at the wall base to allow for perpendicular wall movement. Figure 4 shows construction photos of this method.

Figure 4. Exterior Seismic Cables & Ring Footing Addition

Retrofit for Hoop Overstress

To mitigate hoop tension overstress, new circumferential wire-wound, strand-wound, or tendon prestressing is applied to the tank. The degree of new prestressing depends on the area of steel, remaining force and condition of the existing pre-stressing. Elastic shortening will also further reduce the force in the existing prestressing. On most tanks, wound wires (or strands) have been preferred due to economics and quality control. The wires or strands are high strength steel, hot dipped galvanized and applied with an automated wrapping machine that continuously monitors the applied stress. Two inches of shotcrete cover is applied over the wires (or strands) to bond and protect them. Photos of the wrapping machine and shotcrete application are shown in Figure 5. Tendon prestressing has been primarily used on smaller tanks that have lower demand (i.e. less degree of hoop overstress) or for tanks that have little clearance around the tank or difficult access for a wire or strand-wrapping machine. The tendons are typically epoxy coated 7 wire strands, greased and encased in plastic sheathing. They are applied by hand and tightened with hydraulic jacks at coupling anchors located at specific points on the tank circumference. The couplers are staggered between tendons and friction losses are monitored to ensure a uniform application of stresses on the tank. Figure 6 shows a relatively small but tall tank with tendons installed prior to the application of shotcrete. The shotcrete is required for protection against corrosion and vandalism. This method was typically not as economical as wire or strand wrapping due to the automation of the wrapping and shotcreting.

Retrofit for Uplift

Wall uplift is also mitigated by the seismic cables, ring footing and drilled piers described above for sliding. The angle of the cables can be adjusted if the required force for uplift is greater than the sliding requirement. At some of the lower aspect tanks with lesser seismic uplift demands, the new interior sliding restraint curbs attached to the base of the tank wall described above provides sufficient additional dead load to resist uplift. At a few sites with high uplift demands

6 and rock sub-grades, rock anchors consisting of grouted multiple strands, were used. The anchors were installed in grouted holes drilled tight against the tank wall & through the existing ring footing. The top of the strands were splayed out and attached to the tank wall prior to the application of shotcrete and strand wrapped prestressing. The new prestressing and shotcrete provide sufficient bond to anchor the strands.

Figure 5. Automated Wire or Strand Wrapping and Shotcrete

Figure 6. Post-tensioned Tendons prior to final shotcreting (Note: bottom of wall has new seismic cables that are encased in new shotcrete)

7 Valve Pit and Piping Retrofits

To prevent damage to valves and piping the existing precast concrete valve pit covers were secured with galvanized steel hardware and epoxy anchor bolts in the valve pit walls. To reduce the risk of damage to the Inlet/Outlet pipes, a gap was typically chipped out around the pipe circumference in the exterior valve pit wall and flexible sealant material was placed in this gap and thus eliminating the “hard spot” where the pipe transitions from soil to the concrete valve pit. At sites where the soil around the pipe outside the valve pit is much softer than the foundation sub-grade of the tank and valve pit, a flexible expansion joint was installed which allows for both rotation and longitudinal movement.

To mitigate the potential for downstream pipe breakage, remote controlled electrically actuated reservoir isolation valves are being installed at certain reservoirs (typically only one reservoir per pressure zone), to maintain some water storage for emergency use. Re-chargeable battery powered valves were chosen to allow for remote operation even if the normal power supplies are down. The installation includes a flow meter downstream of the valve, to allow the operator to monitor the flow, and a hydrant upstream of the valve, to provide water from the reservoir for emergency vehicles.

CONCLUSION

Existing aging circular pre-stressed concrete water storage tanks located in an area of high seismic risk can be effectively and economically retrofitted. Some tanks may be retrofitted while they remain in service. Seismically upgrading and rehabilitating the tanks may result in lower life cycle costs. East Bay Municipal Utility District’s tank retrofit program will greatly enhance the post earthquake performance of the District’s water distribution system.

8 S6-6

“Evaluation of Scenario Earthquakes and Examination of the Seismic Resistant Design Method of Waterworks in Hiroshima City”

Presenter: Kenji Totoki

(Waterworks Bureau, The City of Hiroshima, Japan)

Evaluation of Scenario Earthquakes and Examination of the Seismic Resistant Design Method of Waterworks in Hiroshima City

Kenji TOTOKI, Shinji KAMURA Waterworks Bureau, The City of Hiroshima

Abstract Earthquake ground motions caused by scenario earthquakes (Koi fault earthquake, Geiyo earthquake, Ogata-Oze fault earthquake) at main points in Hiroshima city were evaluated as the input earthquake motions for seismic resistant design by comparing with those obtained from “the Guideline and Explanation to Seismic Resistant Method of Construction for Waterworks Facilities”. Based on the importance, the waterworks facilities are divided into two ranks, “Rank A” and “Rank B”, in the guideline. Considering the local topographical and geological characteristics of Hiroshima city area, and the various functions of the water supply facilities, “Rank A” was classified into three categories according to the seismic resistant performance and reliability. Natural period of the structure is an essential parameter for the seismic resistant design, parametric eigen-value analyses were performed to examine the natural period of the reservoir structures by using two-dimensional and three- dimensional finite element models to check the validity of the method to obtain the natural period.

1. Introduction The water supply system of Hiroshima city was established in 1898 and the facilities of the system have been constructed and expanded with the development of the city. The present water supply capacity is about 630,000 ㎥ per day, and it supports citizen’s lives and activities. The population is 1,160,000 within the service area, i.e., Hiroshima City, Fuchu town and Saka town as shown in Fig.1. As the Hiroshima city district was attacked by severe earthquakes about ten times in the past, the waterworks bureau had been made effort to construct strong water supply system against earthquakes based on the seismic resistant design codes in those days. The 1995 Hyogo-ken Nanbu earthquake, however, changed our thinking about the feature of the earthquake, and forced to re-consider the measures against earthquakes from both sides of hardware and software, in exchange of the many victims. The Japan Water Works Association revised “the Guideline and Explanation to Seismic Resistant Method of

- 1 - Construction for Waterworks Facilities”1) (it is expressed as “Seismic Resistant Guideline” below) in 1997. Taking the opportunity, diagnoses of the seismic resistance performance of the water supply system was carried out in 1995 in Hiroshima city, and the “Seismic resistant

Fig. 1 The water supply system in Hiroshima city region project” explained in Table 1 started. The facilities which needed seismic resistant reinforcement were immediately selected and the performance of them were evaluated by dynamic finite element analyses in 1996. In the analyses, synthesized earthquake motions were made which were used as input ground motions by the Koi fault. We have recognized the importance of the topographical and geological characteristics of Hiroshima region which was formed on the Ota river delta through the process of the project. A study team was formed during the project. The goal of the study team was to establish the seismic resistant design method considering the rational earthquake input motion and necessary resistance performance of the waterworks facilities.

- 2 - Table 1 Seismic resistant project of water supply system in Hiroshima city Item Tasks 1. Seismic resistant project Renewal of deteriorated water pipes Strengthening the backup function of the water supply system Use of seismic resistant joint pipes Seismic resistant retrofitting of the water supply reservoir 2. Securing of drinking water Increase the capacity of water supply reservoir Construction of seismic resistant water tanks Installation of emergency shutoff valves Strengthening the initial stage framework in an emergency 3. Establishment of wide area Collective disaster support among twelve major city mutual-aid organization waterworks bureaus Mutual-supply of water at the time of disaster Cooperation of the emergency measures at the time of disaster Mutual support for disaster with the Japan Water Works Association This paper reports the examination results by the study team, that is, the evaluation of the scenario earthquakes, the dynamic characteristics of wall structure water supply reservoirs and the view of the seismic resistant design of water supply system in Hiroshima.

2. Evaluation of scenario earthquakes in Hiroshima city 2.1 Scenario earthquakes After the 1995 Hyogoken-nanbu earthquake, the Tottori western earthquake occurred in 2000 and the Geiyo earthquake occurred in 2001, and it can be said that the western part of Japan entered the active term of earthquake occurrence, therefore, it is very important to guarantee the seismic resistant ability of water supply facilities. In order to carry out seismic resistant design, it is necessary to determine the earthquake external force, and the earthquake motion level is usually set up according to the Seismic Resistant Guideline. The estimation of the earthquake motion has highly been progressed by virtue of the accumulation of earthquake records due to the progress of earthquake observation systems, grasp of the underground structure by physical

- 3 - investigation, and development of the theory and the analysis technique for fault rupture process and seismic wave propagation. Therefore the method of setting up the earthquake motion level using a fault model comes to use as a realistic technique in recent years. We determined the spectra of three scenario earthquakes, i.e., Koi fault, Geiyo, Ogata-

Table 2 Factors of the scenario earthquakes Eearthquake Length of the fault Epicentral distance Magnitude Expected (km) (km) maximum seismic intensity

Koi fault １０６６．５ 6 or more

Geiyo ２８５７７．２５ 6 or more

Ogata-Oze fault. ２６４７７．２ 5 or more Oze fault earthquakes, which are considered to affect Hiroshima city region. Then we calculated earthquake motion levels in many points, and compared them with the earthquake motion levels from the Seismic Resistant Guideline. The scale of the scenario earthquakes considered to affect Hiroshima city region is summarized in Table 2. The locations of the epicenters of the earthquakes are shown in Fig.2.

Fig. 2 The locations of the scenario earthquakes

- 4 - 2.2 Examination method The fault parameters were set up for three scenario earthquakes, and 15 main reservoirs in Hiroshima city were selected as the earthquake motion prediction points. The acceleration response spectra were obtained by using the ground models, which were employed to establish the local disaster prevention plan for 5% damping. Comparison was made between these spectra and that obtained from the level 2 (L2) earthquake motion based on the Seismic Resistant Guideline. The L2 earthquake motion was determined by the observed records at about 150 points of the Hanshin area at the time of the 1995 Hyogoken-nanbu earthquake, and the response spectra for each ground foundation type I, II and III were obtained for 90% and 70% non-exceeding probability as the maximum and the minimum values, respectively. In order to compare the severe case this time, the 90% non-exceeding probability spectrum was used in the comparison.

2.3 Examination result and evaluation The acceleration response spectra at the Ushita reservoir and the Koi reservoir are shown in Figs.3 and 4 as examples among 15 main reservoirs. L2 earthquake motions from the Koi fault and the Geiyo earthquake spectra are drawn in these figures. The Ogata-Oze fault earthquake was omitted on this examination, because the earthquake motion level was smaller than that of the Koi fault. As shown in Figs. 3 and 4, the spectra obtained by Koi fault earthquake at the Ushita reservoir and at the Koi reservoir exceeded L2 earthquake motion in short period renge, but are less than L2 earthquake motion in long period renge. The spectra obtained from Geiyo earthquake were less than L2 earthquake motion in all period range. There were 10 places where the scenario earthquakes exceeded the L2 earthquake motion among the 15 main reservoirs. The results are summarized in Table 3.

onse acceleration onse p Response acceleration Res

Period T (second) Period T (second)

Fig. 3 Acceleration response spectra at the Fig. 4 Acceleration response spectra at the Koi Ushita reservoir (Ground foundation type I) reservoir (Ground foundation type I)

- 5 - Table 3 Exceeding period of range of the acceleration response spectra from the scenario earthquakes than that from the L 2 earthquake motion Evaluation point Capacity of the The scenario earthquake with exceeded Note of earthquake reservoir (m3) L2 earthquake motion and its period motion range

Earthquake Period range (sec) name

Tsuboi reservoir 6,000

Higasisako 1,800 Koi fault 0.1～0.3 reservoir

Kouchi reservoir 4,500 Koi fault 0.1～0.2

Midorii reservoir 16,900 Koi fault 0.1～0.4

Ushitai reservoir 36,230 Koi fault 0.1～0.6

Ougonzan 16,400 Koi fault 0.1～0.4 reservoir

Kouyo reservoir 68,000 Koi fault 0.1～0.3

Koi reservoir 25,000 Koi fault 0.1～0.4

Tunnel reservoir(No.1) 25,000 Koi fault 0.2～0.4 *1) The tunnel

Tunnel reservoir(No.2) Koi fault 0.1～0.4 reservoir, length

Tunnel reservoir(No.3) is about 3.7km,

Tunnel reservoir(No.4) Koi fault 0.1～0.2 has seven

Tunnel reservoir(No.5) evaluation

Tunnel reservoir(No.6) points.

Tunnel reservoir(No.7 Although the return period of the Koi fault earthquake is thought to be so long as from thousands years to tens of thousand years, it is necessary to take into consideration not only the L2 earthquake motion of the Seismic Resistant Guideline but also the earthquake motion by the Koi fault in the seismic resistant design, because the earthquake motion by the Koi fault have exceeded the L2 earthquake motion. On the other hand, it is thought that it is not necessary to examine the Geiyo earthquake motions, because the Geiyo earthquake motion is less than the L2 earthquake motion in all period range. We have, however, earthquakes once every about 50 years in the past, it is necessary to take into consideration the Geiyo earthquake motion as L1 earthquake motion.

- 6 - 3. Eigen-value analyses of wall structure reservoirs 3.1 The purpose of examination Natural period of a structure is very important because the seismic resistant design of the structure is generally designed by the seismic coefficient method and the design seismic intensity is determined based on the natural period of the structure. Reinforced concrete bearing wall structures are commonly used as the reservoirs in Hiroshima city district. It is thought that the RC bearing wall structure has very high rigidity because the deformation of the structure is restricted by the short side wall, which is called as “gable wall” below. The present design, however, does not take into consideration the rigidity of the gable wall, because the calculation method of the natural period for the design generally uses 2-dimentional(2D) analysis model. Therefore, the validity of the modeling of the 2D model is examined by comparing with 3 dimensional(3D) model for the reservoir of RC bearing wall structure.

3.2 Analysis method and Condition In the seismic resistant design of reservoirs, such as a RC reservoir, the horizontal design seismic intensity is calculated by the following equation (1) in the case of L2

earthquake motion. h02

Kh2=Cs・Kh02 (1)

Where, Kh2 = Horizontal design seismic intensity (0.3 or more), Cs = Structural characteristic factor (= 0.45 for RC reservoir), Kh02 = Standard horizontal seismic intensity at the center of the horizontal seismic intensity K Fig. 5 The upper values of standard horizontal gravity of the structure. seismic intensity (L2 earthquake motion) The standard horizontal seismic intensity is determined based on the natural period of the structure and the ground condition of the site as shown in Fig.5. In this study, the Ushita reservoir (10.75 x 4.875 x 19.4m) of which ground condition is Type I, is analyzed. The cross section of the structure is shown in Fig.6. The 2D model shown in Fig. 6 was Fig. 6 Cross section view of the Ushita reservoir

- 7 - used as the basic case, and 2D and 3D models of different sizes explained below were analyzed. The height of the model was fixed as to be 5m for all models. (1) Models with the width of twice and three times long (Case A) (2D model) (2) Models with the twice and three times length both of the long side wall and the short side wall (Case B) (3D model) (3) Models with twice and three times length only of the long side (Case C) (3D model) The models were made with beam elements in 2D analyses and shell elements in 3D models. The soil foundation beneath the bottom of the structure was modeled by the soil springs, and the spring coefficients were obtained by the formula based on N-values proposed by “Specifications and Description of Highway Bridges”2).

3.3 Analysis results In this study we only discuss the natural periods which correspond to the shear deformation of the cross-section of the structure. Figure 7 shows the fundamental mode shape of Case B-1. From this figure we can see that the maximum (1) Overall view (3D) (2) Cross section view deformation appears at the center of (at the center) the long wall and deformation of the Fig. 7 Fundamental mode shape gable wall is small. Figure 8 and Table 4 summarize the obtained natural periods. From Fig.8 we can see that the natural period of 2D models are twice or three times longer than those of 3D models. This is attributed to that the rigidity of the gable wall is not considered in 2D models. On the other hand the Fig. 8 Obtained natural period natural periods in 3D models are not so different even when the length of the wall are twice or three times, i.e., 78% for twice and 75% for three times long. This means that the rigidity of the gable wall plays very important role for the natural period of the reservoir. As for corresponding seismic intensities, it is 1.0 for Case A which is the upper limit,

- 8 - but 0.7 for Case B. This means that the inertia force and the dynamic hydro-pressure

Table 4 Obtained natural periods and corresponding design seismic intensities Design Case Dimension Enlargement Natural period seismic intensity

A-1 2-dimensional 2-dimensional basic case T=0.l82 0.953

A-2 2-dimensional Twice in the direction of the short side T=0.253 1.000

A-3 2-dimensional 3 times in the direction of the side T=0.325 1.000

B-1, 3-dimensional 3-dimensional basic case T=0.0783 0.700 C-l

Twice in the both sides of the short and B-2 3-dimensional T=0.0846 0.700 long wall

B-3 3-dimensional 3 times T=0.0938 0.700

C-2 3-dimensional The direction of the long side, twice T=0.107 0.725

C-3 3-dimensional The direction of the long side, 3 times T=0.125 0.785

*The maximum value (earthquake motion level 2) of design seismic intensity is calculated by ground foundation type I obtained by 3D analyses are smaller by about 30%. The Seismic Resistant Guideline was revised in 1997, however, there still exist many problems which should be solved. To estimate the natural period accurately shown here is one of the examples. Because the seismic force strongly depend on the natural period of the structure.

4. Examination of the seismic resistant design method 4.1 Earthquake resistant performance of structures It is necessary for the waterworks facilities to keep the reliability even where earthquake occurs because the facility is very important for citizens lives. From this point of view, the following three seismic resistant performances are established in the Seismic Resistant Guidelines.

1. Earthquake resistance performance 1: Capability of maintaining the original function without any repair.

- 9 - 2. Earthquake resistance performance 2: Capability of making quick recovery of the original functions with repairs without reinforcement. 3. Earthquake resistance performance 3: Capability of keeping the entire water supply system. The earthquake resistance performances are mainly described by the capability of keeping the function of the water supply system.

4-2. Facility importance and earthquake resistance performances Water supply facilities are principally classified into two ranks according to their importance, “Rank A” and “Rank B”. “Rank A” facilities have high importance, while “Rank B” facilities have ordinary importance. “Rank A” facilities are determined by considering the following factors: 1. potential to generate serous secondary disaster, 2. located in up-stream within the water system, 3. major facilities which do not have substitute or back up, 4. major distribution pipelines supplying to socially important institutions, 5. major facilities which are difficult to repair, and 6. information centers during a disaster.

What facilities belong to “Rank A” are determined by water utilities themselves, based on the experience, local disaster prevention plan and so on.

Hiroshima city can be divided into three areas, first, the urban area which is formed on the Ota rivers’ delta, second, residential area on the hills east and west sides of the urban area, and third, the north country side. The water supply system in Hiroshima city district extends on such variety areas. Due to the geological and topographical varieties, the water supply system has relatively small reservoirs as many as 208. The number is large compared with other same size cities as Hiroshima city.

Because on this feature, “Rank A” facilities are divided into three i.e., A-1, A-2 and A- 3 depending on the earthquake resistance performance, reliability and function of the facility, and “Rank B”, total division are four i.e., A-1, A-2, A-3 and B. The division is summarized in Table 5. The earthquake resistance performances of “Rank A” facilities are shown in Table 6 and damage levels in the table are explained in Table 7.

- 10 - Table 5 Importance classification of fundamental facilities Category Facility name Remarks in earthquake Importance resistance classification

Intake Intake, intake pipe, Direct facilities for water A-1 sedimentation basin, pump supply and no alternative station facility

Aqueduct Aqueduct Direct facilities for water A-1 supply and no alternative facility

Purification Pre treatment facilities, Direct facilities for water A-1 Plant chemical injecting facilities, supply and no alternative receiving well, mixing basin, facility flocculation basin, sedimentation basin, filter, waste water treatment facilities, equipment for disinfection, clear water basin

Conveyance Conveyance Pipe, regulating Direct facilities for water A-1 well, junction well, pump station supply and no alternative facility

Distribution Reservoir Reservoirs for conveyance A-1 water supply stations

Reservoir Main reservoirs of each A-2 water supply system and the reservoir of which important facilities is included in the water supply area

Reservoir, pump station Other reservoirs and pump A-3 stations

Distribution main, distribution Distribution mains and A-3 branch distribution branches which supply water to important

institutions

Distribution branch Other distribution branches B

- 11 - Table 6 Seismic resistance performances for “Rank A” facilities Importance L1 Earthquake motion. L2 Earthquake motion Note classification Seismic resistance The grade Seismic resistance The grade of damage performance of damage performance

A-l Seismic resistance 1 No damage Seismic resistance 2 Small damage *1

A-2. Seismic resistance 1 No damage Seismic resistance 2 medium damage *2

A-3 Seismic resistance 1 No damage Seismic resistance 2 medium damage

*1 The incidental facilities which maintains the various function of water supply service take the L2 earthquake motion into consideration

*2 The incidental facilities which maintains the various function of water supply service take the L1 earthquake motion into consideration

Table 7 The concept of damage levels The degree Examples of damage, and typical restoration period of damage No damage ① Almost no damage ② Almost no leakage of water Small ① Damage which can be restored by repair within 3 days damage ② Damage that water can be supplied by repairing Medium ① Damage which maintains the minimum function damage ② Damage that only important facilities can supply with water ③ Damage which takes restoration from three days to two weeks Heavy- ① Not collapse but the function of the facility is lost damage ② Although re-use is possible if reinforcement is done, a long period (two weeks or more) is required to recover Collapse ① Damage that is considered to be collapse ② Removal or rebuilding are required

5. Seismic resistant design of basin-well type structures 5-1 Foundation of earthquake resistant design Most of the basin-well type structures such as reservoirs are RC structures and have high rigidity. They are constructed on and in the ground, and they are strongly affected by the earth pressure and the deformation of the surrounding ground. As the basin-well type structures are located in the upper-stream within the system, they belong to “Rank A” facilities, and divided into three categories as A-1, A-2, and A-3 in Hiroshima city district as

- 12 - mentioned above.

5-2 The design input earthquake motions Earthquake motions are affected by the fault rupture process, the wave propagation path, the ground condition, and so on, and the structural response is also affected by the predominant period of the earthquake motion. The input earthquake motions, therefore, must be taken into account these factors. As a result, L1 and L2 earthquake motions are determined as shown in Table 8 and Table 9, respectively.

Table 8 L1 earthquake motions corresponding the importance of the facility Importance The design input earthquake motion to be adopted classification A-1 facilities The greater one among L1 earthquake motion obtained from the Seismic Resistant Guideline or the response spectra obtained from the Geiyo earthquake. A-2 facilities Level 1 earthquake motion obtained from the Seismic Resistant A-3 facilities Guideline

Table 9 L2 earthquake motions corresponding the importance of the facility Importance The design input earthquake motion to be adopted classification A-1 facilities According to the characteristics of the structure, it is adopted according to the followings;. ①The larger one among those obtained from dynamic analyses using the waveform from the Koi fault or the Ogata-Oze fault. ② A larger one among the L2 earthquake motion obtained from Seismic Resistant Guideline and the response spectrum obtained from the Koi fault or that from the Ogata-Oze fault. A-2 facilities Level 2 earthquake motion obtained from the Seismic Resistant A-3 facilities Guideline

5-3 Target level of performance For L1 motions, “Rank A” facilities should not be damaged while “Rank B” facilities may be allowed to suffer from light damage as far as their functions are maintained, in other words, facilities may be designed by the traditional allowable stress method. For L2 motions, “Rank A” facilities should be designed so that they do not give severe influence on

- 13 - human lives, and even in case of light damage, it can maintain their primary function. Table 10 summarizes the check points in the seismic resistant design according to the importance of classification and to the earthquake motions.

Table 10 Check point of the earthquake resistance performance

Importance L1 earthquake L2 earthquake motion classifica- motion tion A-1 By the allowable Investigate the seismic response stress design characteristics of the structure by using the method, check the proper model (ex. 3D model) subjected to induced stress in national inspect earthquake motion and the element proper analysis method such as FEM. A-2 whether it Check that the bearing capacity does not A-3 exceeds the decrease and that the plastic deformation and allowable stress residual displacement are with the limit or not. values by the limit state design method.

6. Future subjects in the seismic resistant design In this study, the basic issues such as input earthquake motions, seismic resistance performance are discussed, however, investigation of the damage levels, evaluation of the characteristics of the facility members, and materials, etc. must be done for the rational seismic resistant design. The waterworks facilities are very complicated and delicate system which is composed of many components, therefore, it is very difficult to grasp the damage state of each component and that of the total function. It is, however, necessary to determine the limit state of the components, that of each facility, and also that as a total system of the water supply system. Moreover, the facility characteristics, such as facility scale, local conditions, facility practical use after suffering a disaster, and so on, and seismic resistant performances such as, non-damaged, slight-damaged, restorable-damaged, severely damaged, collapse etc., are classified and model cases should be set up according to these damaged types. At the same time, cost benefit analysis is needed to clear the benefit of the seismic resistant design and the necessary expense for the design. We need to determine appropriate earthquake resistant ranks and the database corresponding to facility type, important rank, and earthquake resistance performance should be established. Based on these, we need to determine the appropriate earthquake

- 14 - resistance performance for the system. It is also important to set up the clear decision- making process.

7. Concluding remarks Researches on the analysis and the seismic resistant improvement of structures have been energetically done after the 1995 Hyogoken-nanbu earthquake, and the seismic resistant guideline for water supply facilities was revised based on these research results. The performance requirement type seismic resistant design method will be the mainstream of the design method in the future. To realize the performance requirement type seismic resistant design, we need to evaluate the input earthquake motion precisely, to determine the damage levels and limit states of each components of the water supply system, and to grasp the performance of the total system. This requires engineers to promote their skills and to enrich the knowledge on the earthquake resistant design method. From this point of view, we will make efforts to increase the reliability as high as possible of the water supply system in Hiroshima city district, and to realize it, we continue the research on rational seismic resistant design method of water supply system and improve our skills.

Acknowledgments The authors are grateful to Prof. Fusanori Miura of the University of Yamaguchi, for his helpful opinion and precious guidance.

References: 1) Japan Water Works Association; The guideline and Explanation to Seismic Resistant Method of Construction for Waterworks Facilities, 1997. 2) Japan Road Association; Design Specifications of Highway Bridges Part IV Foundations, 1996.

- 15 - S6-7

“Seismic Upgrades of Pump Station Located on Liquefiable Soils in Portland, Oregon”

Presenter: Tim Collins

(City of Portland Water Bureau, USA)

Seismic Upgrades of Pump Station Located on Liquefiable Soils in Portland, Oregon

Tim Collins1 Tim Blackwood2

ABSTRACT

The City of Portland Water Bureau constructed a groundwater pump station and well field on the south side of the Columbia River in the early 1980’s. This facility is the main backup water source for the City of Portland system and has historically been used to augment summer flow and during turbidity events of the main surface water source. The pump station consists of a control building, pump gallery, two (2) million-gallon tank, chemical feed building, and a large electrical substation. Since its construction, the seismic hazard rating for this area has been increased (from Zone 2 to Zone 3). The Water Bureau considers this a critical, life safety facility and decided to upgrade it to withstand current seismic code requirements. The upgrades began in 2000 and include structural retrofits to existing buildings and contents, a new post-tension mat foundation for the electrical substation, ground improvements to mitigate lateral spreading, and a completely new chemical feed building constructed on a subgrade improved with stone columns. This paper describes the major improvements made at the site.

Specific attention will be given to the ground improvement work incorporating stone columns to mitigate liquefaction and lateral spreading. Geotechnical explorations performed before and after the ground improvements will be used to assess the effectiveness of the stone columns in densifying the sandy and silty soils.

______1Tim Collins, Engineer, Maintenance Engineering Group, City of Portland Water Bureau, 1120 SW 5th Avenue, Portland, OR 97204 Tel: (503) 823-5033, Fax: (503) 823-4500 [email protected] 2Tim Blackwood, Division Manager, GeoEngineers, 7504 SW Bridgeport Rd., Portland, OR 97224 Tel: (503)624-9274, Fax: (503) 620-5940 [email protected]

1 1. FACILITY DESCRIPTION

From the 1920’s through the 1970’s, the City of Portland relied exclusively on the surface water from the Bull Run watershed as its water supply. A number of events in the 1960’s and 1970’s spurred interest in developing a backup source of water. These included a flood in December 1964 which severely damaged the main conduits from the Bull Run watershed, and another flood and associated landslide in January 1972 which caused a prolonged period of highly turbid water. To provide redundancy to the overall system a groundwater resource was developed along the south shore of the Columbia River in northeastern Portland.

The groundwater facility can supply up to 420 mld (110 mgd) using 26 wells to Portland and the surrounding area as a backup and supplemental supply. The original facilities (1980) were a pump gallery building, control building, a 2 million-gallon tank, associated vaults, and an electrical substation (Figure 1). This facility is considered a critical, life-safety element of the overall City of Portland system. This paper describes improvements to the original structures and how seismic hazard mitigation was incorporated into the new improvements at this facility.

Figure 1: Site Map of Original 1980’s Facility. (Geotechnical borings locations noted)

On January 1, 1993, the Oregon Seismic Safety Policy Advisory Commission adopted Seismic Zone 3 for western Oregon 1 . In response to this code change the Water Bureau commissioned a seismic vulnerability study of the facility. This study investigated the seismic hazards present at the site, how these hazards could effect the existing structures, and what mitigation efforts should be undertaken2. This study provided the core geotechnical information upon which the site improvements were based.

2 2. OVERALL GEOLOGIC AND SUBSURFACE CONDITIONS

Seismicity

The structural geology of the project area is primarily a function of regional stresses and deformation developed as the oceanic Juan de Fuca Plate subducts beneath and translates northward relative to the continental North American Plate creating northwest trending strike-slip faulting (Figure 2) The dominant geologic structures in the project area are the Portland-Vancouver Basin and the associated Tualatin Mountains. These structures create a broad northwest trending syncline/anticline pair. The basin has filled with sediments delivered by the Columbia and Willamette River systems.

Seismicity in the Portland area can be categorized into three types of earthquakes: crustal, intraplate, and subduction zone earthquakes. Due to their proximity, crustal earthquakes present the greatest hazard to this site (Figure 3). The controlling maximum credible earthquake (MCE) used during the design of the improvements is a magnitude 6.5 shallow crustal event on the Lackamas Creek fault at an epicentral distance of 9 km. The last major crustal earthquake near the site occurred on March 25, 1993 when a M5.6 earthquake occurred 100 km from the site at 10 to 20 km beneath the ground surface.

Intraplate earthquakes occur within the Juan de Fuca Plate as it is subducted and deformed beneath the western edge of the North American Plate. The maximum magnitude for this type of earthquake source is estimated to range from 7.0 to 7.5. The earthquake focus is usually centered at a depth of 40 to 60 km. This type of earthquake has have occurred in 1949 and 1965, with the most recent occurring on February 28, 2001 (the Nisqually Earthquake). The Puget Sound area incurred significant damage from the Niqually Earthquake. While it could be felt in Portland, this area only experienced minor damage.

Subduction zone earthquakes are the most far-reaching types of earthquake expected in this region. Recent research has indicated that this type of earthquake could generate an estimated magnitude of 9.0. The most recent subduction zone event is believed to have occurred in 1700. Geologic records indicate that these large events have a recurrence interval of 350 to 700 years. The epicenter of such an earthquake would range from 100 to 150 km west of Portland.

Figure 2: Cross-sectional view of the Cascadia subduction zone at the latitude of Portland, Oregon.3 Dots indicate locations of significant crustal earthquakes.

Site Conditions

The groundwater pump station is located in the Holocene floodplain of the Columbia River. Along the south boundary of the site is a back slough of the Columbia River with the actual river 1000 m to the

3 north. A large stormwater detention pond is situated along the western boundary. The site slopes gently southward toward the slough.

Subsurface Conditions

Subsurface conditions have been evaluated through a series of four separate geotechnical evaluations. The evaluations characterized site soils and groundwater through explorations consisting of mud rotary auger borings with Standard Penetration Testing (SPT) borings, Cone Penetrometer (CPT) soundings, and laboratory tests run on collected soil samples. The evaluations were completed by Shannon and Wilson (1981 & 1982), Dames and Moore (1995), GeoEngineers (2000), and City of Portland (2004). Subsurface conditions were found to consist of a surface layer of soft to medium, clayey to sandy silt to a depth of 3.5 to 4.5 m (11 to 15 ft) with SPT blow counts (N) ranging from 1 to 5. The silt layer is clayey near the surface and becomes more sandy at depth. Below the silt is very loose to loose silty sand with N values ranging from 6 to 13. The sandy material is considered to have the highest liquefaction potential. The sand layer was observed to a depth ranging from 4.5 to 5.5 m (15 to 18 ft). Below the sand to the maximum depth explored is very dense cemented gravel (interpreted to be Troutdale Formation, Figure 4) with N greater than 50. The silt and sand are Holocene aged overbank deposits from the Columbia River. The Troutdale Formation is a Pliocene-aged cemented gravel deposit. Groundwater was observed 2.5 to 3m (8 to 10 ft) below surface grade, but was expected to occur as shallow as about 1.3 m (3 ft) during the winter months.4

Figure 3: Structural geology and crustal faults mapped in the northern Willamette Valley.5

Figure 4: Subsurface site condition, East-West Cross-section (Section B-B in Figure 1)

Liquefaction and Lateral Spreading Analysis

As noted above three seismic sources contribute to the hazards at this site. Regardless of the source, the main affect of an earthquake at this site would be the liquefaction of the sandy soils and the subsequent lateral spreading toward the slough or stormwater detention pond. Conventional liquefaction analysis (Seed & Idriss)6 indicates that the non-plastic silts and sands present have Factors of Safety ranging from 0.3 to 0.6 under MCE conditions. A finite difference computer code, FLAC, was used to build a model to further refine the anticipated liquefaction conditions. The model indicated that liquefaction occurs after 6 cycles of strong shaking with lateral spreading predicted at 15 to 30cm (6” to 12”) near the pump gallery building and 30 to 45cm (12” to 18”) near the tank. This magnitude of deflection was anticipated to result in intolerable deformation between the pump gallery/tank and the connecting piping.

3. FOUNDATION DESIGN AND STRUCTURAL IMPROVEMENTS

Electrical Substation

The original substation foundations consisted of a series of unconnected shallow mat foundations without pile support. The selected location of the new chemical feed building mandated that the substation be moved (Figure 5). The new location had estimated lateral spreading values ranging from 15 to 30cm (6” to 12”)7. During an earthquake the existing mats could move independently of each other potentially ripping the electrical elements apart. To mitigate this movement, the entire substation was placed on a single post-tensioned mat foundation (Figure 6). This type of foundation would keep the substation intact while allowing for movement during an earthquake. Settlements beneath the new foundation were estimated at 2.5-3.8 cm (1”-1.5”) static and 5-7.6 cm (2”–3”) during an earthquake event. Connections into the substation facility were incorporated in order to accommodate these anticipated levels of ground movement.

5 Existing Substation

Proposed Chemical Feed Building

New substation

Figure 5: Substation relocation, site map Figure 6: Substation foundation with post tension cables showing Pump and Control Buildings

The pump gallery and control building are two of the original structures on the site. They are tilt-up, double-T construction with pipe pile foundations driven to refusal in the Troutdale formation. As part of the FLAC analysis the stresses introduced on the pipe piles during lateral spreading was analyzed. This analysis determined that the piles could accommodate the anticipated stresses.

A structural assessment of the buildings was done, based on FEMA 273-NEHRP Guidelines for the Seismic Rehabilitation of Buildings for immediate occupancy (IO) performance objective8. This analysis showed that the roof/wall, wall/wall, and wall/footing connections were not strong enough to meet the IO criteria under MCE seismic conditions. A structural engineering company (Degenkolb Engineers) developed a series of retrofits designed to improve the strength of these connections. Stainless steel plates were used to tie the wall and the roof section together (Figure 7). Reinforced concrete tie downs were constructed to improve the wall and foundation connections (Figure 7). Steel connections were installed at the roof/wall corbells (Figure 8). Tie downs were also installed on many non-structural elements in the buildings namely related to pumps and electrical control units.

Wall/wall connection plates

Footing holddown

Figure 7: Pump building west wall showing plates Figure 8: Wall corbell/roof flange reinforcing between wall panels and wall/footing holddown

6 Chemical Feed Building

The chemical feed building was designed using the 1997 Uniform Building Code (UBC) starting in 1999. The building is a seismically reinforced concrete block structure. The subgrade soils did not have adequate bearing capacity to support the anticipated structural loads without undergoing excessive settlement. To reduce settlement within tolerable levels and mitigate against liquefaction a reinforced mat foundation was constructed on a subgrade improved with stone columns. Stone columns were installed beneath the entire footprint of the building. They were installed at grade and once in place the basement excavation was made.

Tank and Large Diameter Pipes

The tank is steel and placed on a concrete ring wall foundation that is founded approximately 3 feet below grade. Unlike the other original building on the site no piles are used in this foundation. The lateral spreading near the tank was estimated at 30 to 45cm (12” to 18”). Three rows of stone columns were installed around the perimeter of the tank. The stone column ring is designed to mitigate liquefaction settlements near the ring foundation and prevent any liquefied soils underneath the tank from moving laterally.

Several large diameter pipes are present on the site. The main pipes are a 54” concrete cylinder pipe (CCP) tank inlet, a 72” CCP tank outlet pipe, and a 54” CCP outlet from the pump gallery. These pipes were susceptible to varying magnitudes of lateral spreading 0 to 30 cm (0 to 12”). The main areas of concern with these pipes were differential movement at the interfaces with the existing pile supported buildings. Initially flexible couplings were proposed at these locations. After additional study, stone columns were used to minimize liquefaction around the tank and along the lengths of the critical sections of pipe, thereby keeping settlement within acceptable levels.

4. SOIL IMPROVEMENT

Value Engineering Studies

As an element of the chemical feed building design a Value Engineering study (VE) was completed for the project. Ground improvement proposed for the site received significant attention and some revisions were made to the original design concepts. Initially piles were recommended for installation around the tank to protect against lateral spreading. During VE the following mitigation options were considered: 1) construct a densified soil or structural ring around the tank to carry foundation loads to underlying non- liquefying soils, 2) relocate the tank to an alternate area that would first be improved with stone columns, deep soil mixing, or grouting, 3) install stone columns or grout wall around the existing tank, or 4) bypass the tank with new piping. The VE study found that stone columns were a viable and more economical alternative to piles and other methods investigated. Stone columns could be used to stabilize the existing tank foundation and also mitigate liquefaction potential for the large diameter piping. The initial soil improvement method for the new chemical building was already stone columns. This was not revised during the VE process. This allowed one ground improvement/construction technique for all the project components. The project saved $335,000 by using stone columns to mitigate the liquefaction hazard compared to the next least expensive method that still met the design criteria.

7 Stone Column Design

Stone columns were designed to have a 40”-42” diameter with an 8 foot center to center triangular spacing. This created a 15%-18% replacement ratio which was acceptable provided stone columns were installed at a 75% to 85% relative density or 38.5 to 42 degree friction angle. A minimum of a three row width was designed around the tank, resulting in a 14 feet wide reinforced zone (Figure 9). The minimum depth of the columns was 18 feet and did not require that the columns be advanced into the underlying dense gravel, although they generally were. To minimize the potential for settlement of adjacent facilities during installation, a minimum distance of 1m (3’) was specified between any stone column and the outside edges of the existing buried utilities.

This design resulted in computed factors of safety (FS) of 1.3 against lateral spreading failure and just over 1.0 against localized bearing capacity shear failure. At the tank location, total and differential settlement of up to 0.15 m (6”) was estimated from post liquefaction settlement, as the soils directly beneath the tank could not be densified. The settlement and low FS against localized shear failure were considered acceptable because of the tank’s flexibility.

72” CCP Inlet 54” CCP Outlet

54” CCP Inlet

Figure 9: Stone column layout

Stone Column Installation

Stone column installation began on September 19, 2002. Four hundred and thirty eight (438) stone columns were installed using the dry bottom feed vibro displacement method. The columns were advanced to depths ranging from 4.5 to 7.5 m (15’ to 25’) below ground surface (bgs). The vibrator had a capacity of 120 kW at 1800 rpm. The electric feed was 3-phase, 380 volts at 60 Hz. The vibrator weighed 2450 kg, was nearly 3 m long and was supported on a 100-ton crane. The feed system included a stone hopper and pressurized chamber with an 8” tremmie line that directed the rock from the hopper to the target depth (Figure 10). The diameter of the vibrator head was approximately 0.7 m wide. Rock was loaded into the hopper with a loader with a 2 cu m bucket (Figure 11). The stone column material consisted of 2” minus, well to moderately graded durable crushed rock that was compacted by the vibrator’s motion as well as raising and lowering the apparatus 0.7 to 1m during installation.

Figure 10: Stone column installation equipment. Figure 11: Showing stone hopper being filled Flying soil caused by pressurized rock/air in with front-end loader. Note spoils around hopper at shallow depths. vibrator shaft.

Approximately 20 to 32 stone columns were installed per day except for periods of breakdowns and verification testing. Average production for all days was 24.5 columns per day over the 19 days of production. Due to the number of utilities and tight site access, production in many areas proceeded at a relatively slow pace compared to the large open areas. Columns were installed first on the outside of a pattern and working inward. Also neighboring columns were not installed on the same day to allow for pore pressure dissipation. No significant difficulties were encountered while installing the columns, although some columns had to be relocated to avoid underground utilities. Most were moved within 1 to 2 m of their design location and were all approved by the engineer prior to installation.

A construction issue that developed after soil improvement was complete was soft subgrade conditions encountered when the chemical feed building basement was excavated (1.5m (5’) bgs). This condition was possibly caused or exacerbated by the stone column installation. Soft subgrade conditions required overexcavation and the placement of filter fabric and 18” of crushed rock was necessary to obtain acceptable subgrade conditions and provided some bridging between stone column locations (Figure 12).

Because the facility was to remain operational during construction, a number of measures were taken to ensure this, including the following: • The 54” CCP outlet pipe was potholed and surveyed at 4 locations near the proposed stone column locations. The survey points were placed where the pipe was considered to be most rigid and, therefore, most vulnerable to damage (elbows and entrance/exit points at concrete vault). Measurements were taken on the pipe by placing the survey rod through a plastic pipe and measuring horizontal and vertical locations periodically during stone column installation. No significant displacements of this line were measured. • Survey points were placed on the tank exterior in three equally spaced locations and surveyed periodically during installation of adjacent stone columns. No measurable displacements occurred. • The tank was drained during installation of stone columns adjacent to it. This reduced loads on the surrounding soils to further reduce settlement potential. • Water pressure in the 54” pipe was reduced during installation of adjacent columns. • A distance of 1.5m (5’) between stone columns and adjacent utilities was maintained for all columns. This was greater than the design distance of 1m (3’) and was requested by the contractor. Since this did not significantly affect design and was the nearest the contractor was willing to install columns and accept responsibility for damage to adjacent utilities, it was accepted by the owner and design team.

9 Construction Verification and Monitoring

Verification testing of stone column installation was completed as follows: • A geotechnical engineer was monitoring the stone column installation throughout the process. • The acceptance criterion during construction was based on assuring that the specified rock volume was installed into each column. • An auger boring was completed through the center of an early stone column. SPT and Dames and Moore sampling were conducted in the boring to compare density and friction angle of the installed column relative to design. N through the column was found to average 16 in the upper 10 feet (silt soils) and 40 in the lower 5 feet (sand soils). These roughly correlated with design friction angles of between 38.5 and 42 degrees. • Soil was excavated at the top of the stone columns to approximately 1.7m (5 feet) deep to verify if the design diameter was reached. It was found that the top approximately 1 m (3 feet) did not have a well-defined column, but below this depth a well-formed circular column developed that met the design diameter (Figure 13). • Volume of installed aggregate was tracked during installation to ensure that the design replacement ratio was achieved. An installed and compacted volume of 0.36 cu. m per 0.3m length was required, so a loose volume of 0.4 to 0.45 cu. m was mandated, allowing for a 10 to 20% volume reduction from densification. • Amperage was recorded during installation, but was not found to consistently increase in the fine- grained soils, so it was not considered to be a reliable indicator of achieving design criteria. Amperage was found to be an acceptable indicator of reaching the gravel layer and also consistently showed densification of the sand soils.

Figure 12: Chemical feed building basement Figure 13: Close-up of stone column. Note very excavation showing stone column and soft little mixing between rock and surrounding silty subgrade conditions. soils.

5. CONSTRUCTION COST AND TIME

The site improvements were designed and constructed in three stages; substation relocation, building seismic upgrades, and chemical feed building/ground improvements. The substation relocation occurred first in 2000, followed by the seismic upgrades and improvements to the existing buildings. The soil improvements were done in summer 2002 with the chemical feed building completed in late 2003 (Figure 13). Each of the stages was stand alone with its individual design, bidding, and construction timelines. By breaking the seismic upgrades and site improvements into smaller projects, the Water Bureau was

10 better able to manage the labor and financial impacts to the organization. Table 1 outlines the cost and construction time necessary to complete each stage of the project.

Table 1: Project Elements Construction Costs and Time Project Name Construction Costs (US $) Construction time (months) Substation Relocation $780,000 3 Pump Gallery and Control Building $600,000 6 Seismic Upgrades Chemical Feed Building construction $2,200,000 18 Facility Ground Improvements $250,000 2 Total $3,830,000 29

6. POST CONSTRUCTION CONDITIONS

The facility was in operation for 6 weeks during the summer drought of 2004. There have been no settlement issues noted during facility operation or afterwards. The most significant earthquake to recently affect the site was the Nisqually Earthquake centered beneath Olympia, Washington (Mw=6.8, February 28, 2001) approximately 175 km (110 miles) north of the site. This earthquake occurred after the substation relocation and building upgrades but prior to the stone column installation and the chemical feed building construction. No settlement or building damage was observed at the site due to this earthquake.

In an effort to determine how well the stone column installation densified the liquefiable soils, CPT soundings were advanced to measure subsurface parameters. The soundings were advanced in September 2004, 2 years after the completion of the stone column installation. CPT data obtained from areas not impacted by stone column installation were overlaid onto readings from improved areas (Figure 14). These CPT soundings indicate that the strength of the silt materials above the groundwater level and the sand layer increased. The sand layer appeared to have the greatest strength increases. However, the strength of the silty materials below the groundwater level remained unchanged or lost strength compared to those areas where stone columns were installed.

The finding that saturated silts had an unchanged or lower strength after stone column installation may be explained by the conditions observed during stone column installation and building construction. During column installation, significant volumes of wet spoils appeared around the vibrocompaction equipment at the surface, nearly a cubic meter per column (Figure 11). This is less than the volume actually displaced, but suggests that the silts were liquefying around the compaction tool and at least some were being pushed to the surface by the inserted crushed rock rather than densified. Liquefying the silt would destroy the natural soil fabric and possibly reduce its strength, even after dissipation of pore water pressure. Also slow pore water pressure dissipation would prevent densification.

This is different phenomena than those evidenced during driven pile installation where softened fine grained soils typically regain or exceed their pre-construction strength as disturbed soils reconsolidate after driving is completed. This difference may be due to the stone column installation method that more easily allows the soft silts to be displaced to the surface, especially those that are saturated. This is compared to driven piles where there is little opportunity for such an occurrence and localized ground heaving is typically the only surface disturbance noted.

As indicated by the Seed and Idriss liquefaction potential method, soils with greater amounts of fines show less susceptibility to liquefaction. The silty soils at this site have fines (% passing the No. 200

11 sieve) ranging from 72% to 99%. These are far in excess of the 35% normally assumed as the limit where soils are susceptible to liquefaction. However, the site conditions would indicate that the silty soils present could be made to liquefy at the energy levels provided by the vibrocompation probe. The liquefaction potential of silty soils is currently an area of increased study and is not well established.

El. 4.0 m 4

3 Elevation, m

0 0 20 40 60 80 100 120 140 160 180 200 Tip Resistance, Qt (tsf)

CPT1 CPT53 CPT50 CPT52 CPT54

Figure 14: CPT data from before and after the stone column installation. CPT1 and CPT53 are in areas not affected by stone column installation.

7. CONCLUSIONS

Stone columns were a key element of this project’s seismic mitigation upgrades. They provide a viable and flexible means of improving sandy soils in and around existing facilities without significant installation induced settlement. Post installation strength measurements indicate that the stone columns have increased the strength of the sandy layer and thereby reduce liquefaction potential of this soil type. However, the installation of the columns through the silty layer created some construction issues (i.e. soft subgrade, excessive spoils). A clear understanding of stone column installation impacts on silty soils are necessary to achieve effective integration into a project. These issues concerning silty soils should be taken into account during the design elements of the project.

The seismic upgrades of a facility such as this with a combination of existing and new structures create challenging design and construction issues. Tackling these issues requires a clear mandate, careful planning, innovative design and construction practices. It was paramount that these projects had the underlying mandate that this facility was to be functional as the backup water supply immediately after a significant earthquake. This mandate provided the institutional understanding necessary to address the

12 complicated array of poor subgrade conditions, existing facilities, system upgrades, and revised seismic codes that exist on this site. Using a wide variety of construction methods and ground improvement schemes, the Groundwater Pump Station is now prepared to be operational soon after a significant seismic event.

References

1 Kennedy R.E, “The development of the Portland, Oregon Building Code – 50 years of evolution, 1945-1995. A comparison of seismic events and structural aspects”, Oregon Geology, Volume 58, Number 1, January 1996

2 Crouse C., Schwarm D., Porush A., “Groundwater Pump Station Interstate Facility Seismic Vulnerability Study, Portland, Oregon”, Dames & Moore Report, September 13, 1996.

3 Wong I.G., 2000, “The Rapidly Changing Perception of Seismic Hazards in the Pacific Northwest: Implications to Dam Safety, ASDSO West Regional Conference, May 2000

4 McDevitt S., Schwarm D., Thielen D., “Geotechnical Engineering Report, Proposed Ground Water Pump Station Improvements, Portland, Oregon”, GeoEngineers Report, November 3, 1999.

5 Geomatrix Consultants, 1995, “Seismic Design Mapping for the State of Oregon”, unpublished final report prepared for the Oregon Department of Transportation.

6 Youd T.L, Idriss I.M., (2001), “Liquefaction Resistance of Soils: summary Report form the 1996 NCEER and 1998 NCEER/NSF Workshops on Evaluation of Liquefaction Resistance of Soils” Journal of Geotechnical and Geoenviromental Engineering, Vol. 127, No. 4.

7 Ballantyne D., “Portland Groundwater Pump Station Improvements Site Piping and Vault Seismic Vulnerability and Mitigation Recommendations”, EQE International, November 8, 1999.

8 Roggenkamp D.A., Thompson C.L, “Groundwater Pump Station Seismic Upgrades”, Degenkolb Engineers, January 19, 1999.

13 Session 7: Restoration after Earthquake

S7-1 “Introduction of the Disaster Information Management System of Osaka Municipal Waterworks” Presenter: Hiroaki Miyazaki (Osaka Municipal Waterworks Bureau) Japan

S7-2 “Damage to Water Supply Pipelines System due to the 2004 Niigata-ken Chuetsu Earthquake and Its Restoration Presenter: Masakatsu Miyajima (Kanazawa University) Japan

S7-1

“Introduction of the Disaster Information Management System of Osaka Municipal Waterworks”

Presenter: Hiroaki Miyazaki

(Osaka Municipal Waterworks Bureau, Japan)

Introduction of the Disaster Information Management System of Osaka Municipal Waterworks

Kazuya Yamano, Katsuhiko Eguchi and Hiroaki Miyazaki

ABSTRACT

Osaka Municipal Waterworks Bureau established the ‘Disaster Information Management System’ from 2000 to 2002, as part of its effort to ‘improve the information and communication system’s reliability,’ one of the basic measures of “Osaka Municipal Waterworks Earthquake Preparedness Improvement Plan 21” (devised in March 1996). The Disaster Information Management System, which is designed to be used in the event of natural disasters (earthquakes, storms and floods etc.) and large-scale accidents, went into full-scale operation in December 2003. The system has various functions. Such functions include one that estimates seismic damage to the distribution pipes in the city immediately after an earthquake, enabling prompt formation of pipeline restoration teams. Another major function is to collectively manage all the information received from different sources at various times, including locations of low water pressure or water suspension, establishment of emergency water supply stations, damage to water intake/purification/distribution plants and water supply lines, and estimated dates of restoration. In addition, the System has a function that keeps track of the number of staff reporting for duty at each department/office after a disaster, so that an appropriate number of backup staff can be assigned to respond to emergency situations. The Disaster Information Management System enables us to promptly initiate emergency response actions after a disaster and to conduct effective disaster emergency activities for early restoration of the damaged waterworks facilities. This paper introduces basic functions and organization of the Disaster Information Management System of Osaka Municipal Waterworks Bureau and explains how this System will be used in actual disaster emergency activities.

______Kazuya Yamano, Manager for Earthquake Preparedness, Engineering Div., Osaka Municipal Waterworks Bureau, 1-14-16 Nanko-kita, Suminoe-ku, Osaka 559-8558 JAPAN Katsuhiko Eguchi, Chief of Planning Section, Planning Dept., Engineering Div., Osaka Municipal Waterworks Bureau, 1-14-16 Nanko-kita, Suminoe-ku, Osaka 559-8558 JAPAN Hiroaki Miyazaki, Staff Officer, Planning Dept., Engineering Div., Osaka Municipal Waterworks Bureau, 1-14-16 Nanko-kita, Suminoe-ku, Osaka 559-8558 JAPAN

effort to ‘improve the information and 1. Introduction communication system’s reliability,’ one of the eight basic elements (Fig. 1) of the Bureau’s earthquake The core of the disaster restoration activities of the preparedness improvement plan. Waterworks Bureau are emergency water supply and The newly-developed system, the ‘Disaster restoration of intake/ purification/ distribution Information Management System,’ was put into plants and pipelines. Implementation of information full-scale operation in December 2003. communication and public relations activities in conjunction with these emergency measures ensures ①①ImprovingImproving earthquake earthquake resistance resistance of of key key facilities facilities effective and smooth disaster recovery. ② To minimize damage from a large-scale disaster, ② EstablishingEstablishing a a water water supply supply and and distribution distribution centercenter network network prompt, effective and well-coordinated activities in the initial stage are essential. Immediately after the ③ ③ ImprovingImproving mutual mutual compatibility compatibility among among different different distribution systems occurrence of a disaster, the Bureau will be swamped distribution systems with various types of information and will have ④ ④ Countermeasure against power failure BasicBasic Countermeasure against power failure numerous tasks requiring urgent attention1) . ElementsElements ⑤ ⑤ ExpandingExpanding emergency emergency material material stockpiling stockpiling To ensure appropriate response to such an systemsystem anticipated chaotic situation, after the Kobe ⑥ ⑥ EstablishingEstablishing stable stable water water supplying supplying routes routes to to Earthquake, the Bureau analyzed emergency man-mademan-made islands islands recovery activities and water supply activities ⑦ ⑦ ImprovingImproving the the information information and and communication communication conducted in some areas damaged by the Earthquake. system’ssystem’s reliability reliability As a result, we concluded that all the information ⑧ Improving earthquake resistance of headquarters sporadically coming into the emergency ⑧ Improving earthquake resistance of headquarters necessarynecessary for for disaster disaster relief relief and and recovery recovery activities activities headquarters piece by piece, including which sections of water supply lines are damaged by the Fig. 1. Basic Elements of Earthquake Preparedness earthquake, should be organized speedily so that Improvement Plan appropriate decisions can be made in a timely manner. This System has seismic damage simulation Osaka City owns and operates its own disaster function that estimates seismic damage to the information management system to process disaster distribution pipes in the city immediately after an information collected by each bureau and ward earthquake. In addition, the System has a function office by computer and manage all the collected that collectively manages all disaster-related information for easy information sharing. It is information, such as low water pressure/ water desirable that the Bureau’s system has suspension, installation status of temporary water interoperability with Osaka City’s disaster supply stations, damage to pipes and other water information management system. supply facilities, and estimated recovery period. In 1999 Osaka Municipal Waterworks Bureau This paper describes basic functions and commenced the development of its own disaster organization of the Disaster Information information management system to be used in the Management System of Osaka Municipal event of natural disasters (earthquakes, storms and Waterworks Bureau and how this System will be floods etc.) and large-scale accidents, as part of its used in actual disaster emergency activities.

-1 -

2. Basic Functions of the Disaster ground surface velocity of the entire city as well as damage condition of water distribution pipes in the Information Management System city are estimated and plotted on the map of the whole of Osaka City, based on seismic scale data

Fig. 2 shows basic functions of the system. An collected by seismometers installed throughout the outline of each function is provided below. city, ground surface velocity information monitored

How each function is utilized in actual emergency by the Bureau’s seismic monitoring system, and IntegratedIntegrated ManagementManagement information released by the Meteorological Agency (1)(1) SeismicSeismic damagedamage of Japan such as epicenter and ground characteristics. 1. Support for simulationsimulation functionfunction This function enables early detection of damaged pipeline restoration activities (2)(2) PipelinePipeline restorationrestoration parts of water distribution pipes and prompt information manage- information manage- formation of pipeline restoration teams. ment function function For detailed description on this function, refer to 2. Support for Emergency water Emergency water 2) emergency water supplysupply informationinformation the paper presented at the 3rd Workshop . Fig. 3 supply activities management function shows one example of computer output 3. Support for Facility restoration screen. facility restoration informationinformation activities management function (1)(1) StaffStaff attendanceattendance management function Severe damage (2) Emergency information 4. Support for (2) Emergency information Slight damage transmission/receptiontransmission/reception other activities No damage functionfunction

(3)(3) On-the-spotOn-the-spot picturespictures management function

5. Support for (1)(1) PublicPublic relationsrelations information provision functionfunction to citizens and Osaka City Disaster (2)(2) FunctionFunction ofof Countermeasures interfacinginterfacing withwith Headquarters Disaster Information Management System of Osaka CityCity Fig. 3. Results of Water Supply Lines restoration activities is described in Chapter 4. Damage Simulation

Fig. 2. Basic Functions of Disaster Information Management System (2) Pipeline restoration information management function This function is used to manage earthquake 2-1 Support for water supply lines restoration damage information, collected, registered and activities updated by the pipeline restoration teams, by GIS. Such information includes the conditions of (1) Seismic damage simulation function damaged water distribution pipes, estimated time of When an earthquake hits, the distribution of restoration, and low water pressure/ water

-2 -

suspension. condition of emergency water supply activities in Since the function interfaces with the each area where water supply is suspended. Pipeline Information Management System Also, this function enables registration of water (Mapping System), which is used for supply-related information received from citizens. managing the pipeline database in the Fig. 5 shows one example of computer output normal situations, damage locations screen. registered in the Mapping System will be automatically transmitted to the Disaster Information Management System. Fig. 4 shows one example of computer output TemporaryTemporary screen. waterwater supply stationstation

Welfare facilityfacility Fig. 4. Registration of Damaged Points of Water suspension area Hospital Low pressure area

Wide Service areaarea site site EvacuationEvacuation sitesite

Service pipe: Service pipe: DistributionDistribution pipe:pipe: Fig. 5. Registration of Emergency Damaged point Damaged point DamagedDamaged pointpoint Water Supply Stations

Water 2-3 Support for facility restoration activities supply line Water Pipelines This function is used to register and manage information on damage to facilities at water intake/ purification/ distribution plants. Damage 2-2 Support for emergency water supply information can be registered and viewed on the activities water purification process flow chart screen, which enables a prompt grasp of damage status of the entire This function is used to register and manage water system. Capability of registering/managing emergency water supply information on the GIS, emergency water quality examination data taken by such as installation of emergency water supply the water quality examination team is useful for stations (evacuation sites etc.) and temporary water assessment of function of each water purification supply equipment (temporary water tanks, feed tanks system. etc.) at each station and total volume of water In addition, this function is capable of collecting supplied. real-time information on storage volume of each This function enables effective grasping of the purification/distribution reservoir and possible

-3 -

supply time. Fig. 6 shows one example of reporting for duty at each department/office after a computer output screen. disaster, so that appropriate number of backup staff can be assigned to respond to emergency situations. Fig. 7 shows one example of computer output Water Purification Process Flow Chart screen.

Sedimentation Basins, Filter Basins, Finished Water, Intake Ozone Contact Basins, GAC Filters Service Facilities Chlorine Contact Basins Reservoirs (2) Emergency information transmission/

Damaged reception function SectionSection This function enables a specific department to transmit/receive emergency information. It is also possible to transmit images and broadcast a single mail to multiple recipients. Several templates are provided to speed creation and transmission of notification documents and Fig. 6. Registration of Damage to Water various emergency messages. When a new e-mail Intake/Purification/Distribution Plants message arrives, the popup notification box (shown in Fig. 8) will be displayed on the screen of the terminal accompanied by a loud buzzer sound, so 2-4 Support for Other Disaster Emergency that important emergency information can be Activities checked without delay and without fail.

(1) Staff attendance management function

図７

Name of departmentdepartment You’ve got one new mail. Do you want to display it now? AttendanceAttendance rate rate ofof staff staff

Fig. 8. New Mail Notification Box Displayed on the Screen Fig. 7. Staff Attendance Information

(3) On-the-spot pictures management function The Disaster Information Management System Pictures taken at disaster/accident sites are has the function to keep track of the number of staff imported as image data and transmitted via system

-4 -

terminals to the emergency headquarters and each Waterworks Bureau. department. Visual information helps the staff grasp The following are information that can be the actual conditions of damage to water supply automatically collected and tabulated with this facilities and emergency water supply more vividly. function. This function facilitates prompt decision making of Disaster information (earthquakes, storms and the headquarters and effective sharing of damage floods etc.) information among all departments. Fig. 9 shows Quality of water sources one example of computer output screen. Operation/restoration status of water purification/distribution plants Number of houses to which water supply is suspended; number of repairs to distribution pipes Number of repairs to water service pipes Number of temporary water supply stations; total volume of supplied water at temporary stations etc. In addition, to speed the creation of public relations materials, several standard expressions are registered.

(2) Function of interfacing with the Disaster Information Management System of Osaka City Fig. 9 On-The-Spot Picture of Water Leakage This function is used to import data stored in the Registered on the System Disaster Information Management System of the Osaka City Disaster Countermeasures Headquarters, into the Bureau’s Disaster Information Management 2-5 Support for information to citizens and System, in order to make full use of such data for the Osaka City Disaster Countermeasures Bureau’s disaster recovery measures. The Bureau’s Headquarters Headquarters receives damage or restoration information from the facilities in the City and (1) Public relations function various other sources; information on facility This function supports the creation of public damage and number of evacuees at each evacuation relations materials, by automatically gathering and site etc. is useful in planning emergency water tabulating information registered by each emergency supply activities, and information on fires and road response team regarding water supply cutoff, damage is useful in planning water distribution and emergency water supply and restoration status of emergency restoration measures. damaged water supply facilities. It is also possible to transmit the information Since produced materials can be output in PDF collected by the Bureau, regarding damage to water file format, they can be submitted as reports to news supply facilities, water suspension, installation of agencies and Japan Water Works Association or can emergency water supply stations etc. to the Disaster be posted on the web site of Osaka Municipal Information Management System of Osaka City.

-5 -

••SeismicSeismic damagedamage simulationsimulation InformationInformation onon lowlow HistoryHistory informationinformation pressurepressure andand waterwater displaydisplay ••WaterWater distributiondistribution informationinformation suspensionsuspension

PublicPublic relationsrelations TotalTotal waterwater storagestorage volumevolume informationinformation

••FacilityFacility damagedamage informationinformation ••DistributionDistribution pipepipe damagedamage informationinformation ••TeTe lemetry lemetry datadata ••EmergencyEmergency waterwater supplysupply informationinformation ••StaffStaff attendanceattendance informationinformation etc etc

Fig. 10 Integrated Management Screen

2-6 Integrated management of disaster-related 3. Construction of the Disaster information Information Management System The Disaster Information Management System has the function to collect/organize all the 3-1 Basic policy of construction of the System information registered through various functions (e.g. information on damaged water pipelines (Fig. 4); Osaka Municipal Waterworks Bureau has information on temporary water supply stations (Fig. installed telemeters throughout the City to 5); information on damage to intake/ purification/ collect/monitor water distribution data, etc.; 101 distribution plants (Fig. 6); information on staff water distribution telemeters that automatically attendance (Fig. 7)), as well as to show the registered measure water pressure and distributed water information on a single screen (Integrated amount and 40 water-quality telemeters that Management Screen (Fig. 10)). This function automatically measure water quality have been enables at-a-glance understanding of the restoration installed, and 11 seismometers have been installed to activity information of the entire Bureau, which record earthquakes at the water intake/ purification/ facilitates prompt and appropriate decision making distribution plants. These telemeters are installed in of the Waterworks Headquarters. locations all over the city. To use information collected from these units most effectively, it is necessary to collectively manage all information related to risk management

-6 -

by using web delivery technology. Information transmission to terminals in the Bureau To allow all relevant staff to collect necessary is performed using wireless LAN. information any time from disaster information system terminals installed at all departments, the (2) Use of the System in normal times/emergency Disaster Information Management System is The Disaster Information Management System designed to interface with other existing systems, features switchable operation modes (“practice such as the Water Distribution Information mode” or “disaster mode”), which allows the System Management System. to be used in normal times as well as in an emergency. This feature enables the Bureau staff to 3-2 Construction practice operating the equipment at normal times. When the seismometers of the Bureau detect The Disaster Information Management System is ground motion exceeding a specified value, the specifically designed to be used during a disaster. “operation practice mode” will be automatically Therefore, the servers and major communication switched to the “disaster mode.” lines are duplexed, to prevent the shutdown of the entire system. 4. Utilization of the System in (1) Hardware units Servers are key hardware units of the System. Earthquake Disaster Emergency The following three servers are used: Activities ① Disaster information processor: Manages databases stored through each function of the 4-1 Pipeline restoration System. ② Disaster information transmission unit: (1) Initial stage (Occurrence of an earthquake – 24 Processes requests received from the disaster hours after the earthquake) information system terminal installed at each Prior to on-site investigation of damage, the water department. pipeline restoration teams collect and check the ③ Image data management processor: Manages information registered in the Disaster Information various image data to be used in the System. Management System, such as water distribution In order to spread risk and load, two units each for information collected by distribution data for the above three types of equipment are installed. telemetry and pipeline damage simulation Optical fiber cables, integral part of LAN for information. disaster information, are also duplexed, and two Based on the collected information, the teams units of network equipment are installed. investigate damage to important water pipelines on Dedicated lines are used for communication lines the spot, such as transmission/distribution mains, connecting between the Bureau and each distribution mains linking to distribution plants and department/office, part of which are duplexed. distribution pipes linking to evacuation sites, to In emergency situations, it is unpredictable who locate damaged points as soon as possible. They will operate which terminal at what time; to ensure also register information gained during investigation mobility, notebook computers are used, and a activities in the System. Web-based information delivery system is employed. At the same time, they operate valves to ensure

-7 -

that a sufficient quantity of water is distributed to System. shelters and areas where fires occurred, and identify areas where water supply will be suspended due to (2) Initial stage (3 hours after the earthquake – 24 valve operation, as preparation for planned pipeline hours later) restoration work (emergency water supply Each service office keeps track of water supply suspension). status (low water pressure/ water suspension) in the Based on the predetermined order in which city. It also inspects water service equipment distribution pipes are restored and water supply installed in shelters, important facilities etc. in its suspension information collected by the Disaster service area, to collect information regarding damage Information Management System, the emergency and water supply suspension, and registers the restoration plan for distribution pipeline should be inspection result data in the Disaster Information made. Management System. Based on the water suspension information registered in the System by the pipeline (2) Restoration stage (24 hours after the occurrence restoration team, the service offices develop of an earthquake － ) emergency water supply plans; and then, in Based on the information collected in the Disaster accordance with the plan, emergency water supply Information Management System regarding low activities should be commenced. water pressure/water suspension and pipeline damage, pipeline restoration teams make the (3) Restoration stage (24 hours after the occurrence distribution pipeline restoration plan for the launch of an earthquake －) of full-fledged restoration activities, to restore Emergency water supply teams collect important water distribution routes as soon as information on low water pressure/ water suspension possible, such as transmission mains, distribution information that continues to change according to the mains linking to distribution plants, distribution progress of restoration work and valve operation, pipes to shelters, important facilities such as using the Disaster Information Management System. hospitals and welfare facilities, nearby parks. At Based on information on “shelters in areas where the same time, commence emergency restoration water supply has been suspended” and “temporary work, mainly on damaged transmission mains, water supply stations set up in city parks,” conduct distribution mains, distribution branch pipes and emergency water supply operations to ensure that distribution branch lines. sufficient amounts of water are delivered to temporary water supply stations and important facilities. 4-2 Emergency water supply Resister/manage information on “water supply status at each temporary water status” and “number (1) Initial stage (Occurrence of an earthquake – 3 of times of water supply by administrative district” hours after the earthquake) in the Disaster Information Management System; Each service office registers water suspension and based on such information, make improvements information at important facilities such as hospitals on the emergency water supply plan to ensure more and welfare facilities in its service area, as well as effective operation. Careful consideration and water service-related information received from planning is necessary to ensure that backup parties citizens, in the Disaster Information Management from other cities will be assigned to the right places.

-8 -

4-3 Restoration of intake/ purification/ station whose role is to secure close coordination distribution plants between relevant departments and to develop most appropriate emergency policies through discussion, (1) Initial stage (Occurrence of an earthquake – 24 so that emergency measures will be implemented hours after earthquake) smoothly, speedily and effectively. Inspect the operation status of monitoring/control systems and damage to constructions in intake/purification/distribution plants, and register inspection results in the Disaster Information Management System. Based on the collected information, evaluate the purification capability of each plant. At the same time, check the data on the water storage volume at each purification/service reservoir that are automatically inputted and calculated in the System, and make preparations to operate distribution pumps for water distribution control in the City. In repairing damaged pipes, first perform emergency repairs to the portions with minor damage, to restore purification capability as soon as Photo 1 Utilization of the System in the Disaster possible. Management Office

(2) Restoration stage (24 hours after earthquake － ) With reference to information about restoration status and progress of the damaged intake/ purification/distribution plants collected in the Disaster Information Management System, make a damage restoration plan for the launch of full-fledged recovery activities.

Example) Utilization of the System for Photo 2 Information collected in the System, shown drill on the large screen in the Disaster Management Office Photo 1 and 2 show the inside of the Waterworks Headquarters, photographed during the earthquake disaster drill conducted in January 2004. The Waterworks Headquarters is a key operational

-9 -

5. Conclusion

The Disaster Information Management System of Osaka Waterworks Bureau has enabled collective management of an enormous amount of information, which used to be shared only in the form of paper documents. The System has also enabled at-a-glance understanding of ever-changing damage situation and restoration work progress. In addition, owing to the System’s function of showing various disaster-related data collected in the System on integrated management screen, the Headquarters has become able to make appropriate decisions in a timely manner, and each department/office has become able to acquire information collected by other departments/offices promptly and efficiently. Proper and effective use of the System for emergency activities of the Waterworks Bureau will greatly help shorten the time required to repair damaged facilities and restore water supply. In the future, it is necessary to increase periodic operation training and simulation-type drill to ensure effective use of the System, and to continue to make efforts to enhance the functions of the System.

1) Kobe City Waterworks Bureau：Record of Kobe Earthquake Water Lines Restoration,，1996. 2) Kazuya Yamano, Katsuhiko Eguchi, and Koichi Murata ： Seismic Damage Simulation of Distribution Pipeline Based on the Monitoring Data Collected by a Seismometer Network，The 3rd Japan and U.S. Workshop on Seismic Measures for Water Supply，2003.

- 10 - S7-2

“DAMAGE TO WATER SUPPLY PIPELINE SYSTEM DUE TO THE 2004 NIIGATA-KEN CHUETSU EARTHQUAKE AND ITS RESTORATION”

Presenter: Masakatsu Miyajima

(Kanazawa University, Japan)

Cover page

Title: Damage to Water Supply Pipelines System due to the 2004 Niigata-ken Chuetsu Earthquake and Its Restoration

Author: Masakatsu Miyajima (Contact person, presenter) Professor Department of Civil Engineering Graduate School of Natural Science and Technology Kanazawa University 2-40-20, Kodatsuno Kanazawa, 920-8667 Japan Phone: (76) 234-4656 Fax: (76) 234-4644 E-mail: [email protected]

DAMAGE TO WATER SUPPLY PIPELINE SYSTEM DUE TO THE 2004 NIIGATA-KEN CHUETSU EARTHQUAKE AND ITS RESTIRATION

Masakatsu Miyajima

ABSTRACT

This paper deals with an outline of damage to water supply pipeline system and the restoration process in the 2004 Niigata-ken Chuetsu Earthquake. An earthquake with a magnitude of 6.8 occurred in Chuetsu region of Niigata Prefecture of Japan on October 23, 2004 at 17:56 on local time. The earthquake caused the loss of 40 lives and injured about 3,000 people. Many rock and soil slope failures took place, particularly in the mountainous area and along Shinano River. The permanent ground deformation as well as ground shaking caused some structural damage to residential houses, buildings, bridges, road, highways, railways and lifelines. A suspension of water supply was occurred at more than one hundred thousands houses in the stricken areas just after the earthquake. An outline of the water supply system and damage to it at Ojiya and Nagaoka City is given. The relation between the damage to water supply pipe and pipe material was investigated. Furthermore, the restoration process was discussed in the relation to a number of refugees.

______Masakatsu Miyajima, Professor, Department of Civil Engineering, Graduate School of Natural Science and Technology, Kanazawa University, 2-40-20, Kodatsuno, Kanazawa, 920-8667 Japan

INTRODUCTION

An earthquake with a magnitude of 6.8 occurred in Chuetsu region of Niigata Prefecture of Japan on October 23, 2004 at 17:56 on local time. The earthquake caused the loss of 40 lives and injured about 3,000 people. The earthquake had an unusual after shock activity and at least four large after shocks with a magnitude greater than 6.0 occurred. Many rock and soil slope failures took place, particularly in the mountainous area and along Shinano River. The permanent ground deformation as well as ground shaking caused some structural damage to residential houses, buildings, bridges, road, highways, railways and lifelines. The most heavily damaged areas were Kawaguchi Town with a seismic intensity of 7 on the Japan Meteorological Agency (JMA) seismic intensity scale, Tokamachi, Ojiya and Nagaoka City. All resident of Yamakoshi Village evacuated to Nagaoka City because of heavy damage due to slope failures induced by the earthquake. The present paper is focusing on the damage to water supply system due to the 2004 Niigata-ken Chuetsu Earthuquake. An outline of the water supply system and damage to it at Ojiya and Nagaoka City are given. The relation between the damage to water supply pipe and pipe material was investigated. Furthermore, the restoration process was discussed in the relation to a number of refugees. The material presented in this paper is the interpretation and compilation of the available materials and information released by waterworks bureaus at the stricken areas and mass media within one month after the earthquake. Further investigation should be done to clarify the causes of the damage in relation to the ground conditions.

DAMAGE TO WATER SUPPLY FACILITIES

Nagaoka City

Population of Nagaoka City is 192,322; the number of household is 67,821. More than 800 houses and buildings were totally collapsed due to the earthquake. Severe damage occurred at the mountainous areas of eastern part of the city due to slope failures (see Photos. 1 and 2). The water supply system of Nagaoka City has approximately 1,084km of service main and distribution pipelines. The percentage of pipe length to pipe type is shown in Fig.1. About 66% of the total length is made up of ductile cast iron pipe (DIP), 21% unplasticized polyvinyl chloride pipe (VP), 7% steel pipe (SP), 6% cast iron pipe (CIP) and 0.5% asbestos cement pipe (ACP). The total number of failures for service and distribution pipelines was 287. The damage rate, defined as the amount of failure per length of pipeline, was about 0.26/km. The relation between the number of occurrence of damage, pipe type and pipe diameter is listed in Table 1. The number of occurrences of damage and the damage rate related to pipe type are illustrated in Figs. 2 and 3. Fig.3 indicates that the damage rates of steel pipe and unplasticized polyvinyl chloride pipe is high. Failure mode of the unplasticized polyvinyl chloride pipe was pull-out at joint and breakage of pipe body and it of steel pipe was failure of screw joint.

Photo. 1 Damage to hydrant due to slope Photo. 2 Damage to pipeline due to slope failure (Takamachi, Nagaoka City) failure (Takamachi, Nagaoka City)

DIP CIP SP VP ACP 0% 20% 40% 60% 80% 100% Percentage

Fig. 1 Percentage of pipe length to pipe type (Nagaoka City)

Table 1 Relation between the number of occurrence of damage, pipe type and pipe diameter (Nagaoka City）

DIP CIP Pull-out Pull-out Nominal Brakeag Brakeag Brakeage of and Brakeage of and Diameter e Others Sub Total e Others Sub Total pipe body leakage pipe body leakage at joint at joint at joint at joint 800 300 0 1 1 250 0 0 200551 1 125,150 13 13 3 3 100 16 16 1 1 2 75 16 218311 5 50 0 0 40 0 0 Sub Total0500252471012 Pipeline length 710.9 66.6 (km) Damage rate 0.07 0.18 (incidents / km) SP VP Pull-out Pull-out Nominal Brakeag Brakeag Brakeage of and Brakeage of and Total Diameter e Others Sub Total e Others Sub Total pipe body leakage pipe body leakage at joint at joint at joint at joint 800 2 2 300 1 21 22 0 23 250 1 1 0 1 200 1 1 0 7 125,150 1 1 2 0 18 100 1 2 3 0 21 75 1 1 0 24 503197 2911575 73102 40235 1024514 7989 Sub Total 9 22 40 0 71 35 108 9 0 152 287 Pipeline length 73.8 227.5 1079 (km) Damage rate 0.96 0.67 0.27 (incidents / km) Notes DIP : Ductile iron pipe CIP :Cast iron pipe SP : Steel pipe VP : Unplasticized polyvinyl chloride pipe

200 1.5

150 1

100

0.5 50 Damage rate (/km)

Number of case of damage 0 0 SP VP SP VP DIP CIP DIP CIP Pipe type Pipe type

Fig. 2 Number of damage occurrences Fig. 3 Damage rate related to pipe type

related to pipe type (Nagaoka City) (Nagaoka City)

Ojiya City

Population of Ojiya City is 41,314; the number of household is 12,376. More than 600 houses and buildings were totally collapsed due to the earthquake. Shiodono purification plant was moved downwards of the slope due to large slope failure shown in Photos 3 and 4. The surrounding ground sank about 50cm and a room of valve was damaged at Funaoka-yama distribution plant (see Photos. 5 and 6). The water supply system of Ojiya City has approximately 328km of service main and distribution pipelines. The percentage of pipe length to pipe type is shown in Fig.4. About 71% of the total length is made up of ductile cast iron pipe, and 16% steel pipe, 9% unplasticized polyvinyl chloride pipe, 4% polyethylene pipe. The total number of failures for service and distribution pipelines was 102. The damage rate was about 0.31/km. The relation between the number of occurrence of damage, pipe type and pipe diameter is listed in Table 2. Many failures occurred at pipe with a relative small diameter. The number of occurrences of damage and the damage rate related to pipe type are illustrated in Figs. 5 and 6. These figures indicate that the damage rates of the steel and unplasticized polyvinyl chloride pipe are high. This tendency is similar to that of Nagaoka City. Most of failures of ductile pipe with no aseismic joint occurred at joint, and those of steel and unplasticized polyvinyl chloride pipe occurred at both of pipe body and joint.

Photo. 3 Damage to Shiodono purification Photo. 4 Damage to Shiodono purification plant (Ojiya City) plant (Ojiya City)

Photo. 5 Damage to Funaoka-yama Photo. 6 Damage to Funaoka-yama distribution plant (Ojiya City) distribution plant (Ojiya City)

DCIP SP VP PE 0% 20% 40% 60% 80% 100%

Percentage

Fig. 4 Percentage of pipe length to pipe type (Ojiya City)

50 1.5

40 1 30

20 0.5

10 Damagerate (/km)

Number of case of damage 0 0 PE PE SP VP SP VP DCIP DCIP Pipe type Pipe type

Fig. 5 Number of damage occurrences Fig. 6 Damage rate related to pipe type

related to pipe type (Ojiya City) (Ojiya City)

Table 2 Relation between the number of occurrence of damage,

pipe type and pipe diameter (Ojiya City)

DIP SP Nominal Brakeage of Pull-out Leakage Brakeage of Pull-out Leakage Diameter Others Sub Total Others Sub Total pipe body at joint at joint pipe body at joint at joint 300 2 2 0 250 1 1 0 200 4 4 0 150 9 9 2 2 100 20 20 1 1 2 75 1 1 2 2 1 3 6 50 1 1 10 12 2 4 28 40 0 2 1 3 Sub Total 0 37 1 1 39 13 17 4 7 41 Pipeline length 234.4 51.8 (km) Damage rate 0.17 0.79 (incidents / km) VP PE Nominal Brakeage of Pull-out Leakage Brakeage of Pull-out Leakage Total Diameter Others Sub Total Others Sub Total pipe body at joint at joint pipe body at joint at joint 300 0 0 2 250 0 0 1 200 0 0 4 150 0 0 11 100 3 1 4 1 1 27 75 1 1 0 9 50 5 6 2 13 1 1 43 40 2 2 0 5 Sub Total10712201010 2102 Pipeline length 29.9 12.4 328.5 (km) Damage rate 0.67 0.16 0.31 (incidents / km) Notes DIP : Ductile iron pipe SP : Steel pipe VP : Unplasticized polyvinyl chloride pipe PE : Polyethylene pipe

RESTORATION PROCESS OF SUSPENSION OF WATER SUPPLY

Restoration curves of water supply at Nagaoka City, Ojiya City, Tokamachi City, and Kawaguchi Town are shown in Figs. 7 – 10. The restoration curves indicate a percentage of recovery from a suspension of the performance. The restoration curves of gas supply are also shown in these figures. Since an effect of the suspension of water and gas on a number of refugees is investigated here, the rate of refugees in each day to the maximum number is also indicated. Since a number of severe after shock may affect refugees, the number of after shocks greater than 4 of JMA seismic intensity is also drawn in these figures. The number of refugees shows the maximum value after three or four days from the event in each city and town. The number of refugees decreased gradually in spite of high activity of the after shock after two weeks from the event. The rate of restoration of Kawaguchi Town was lowest, then Ojiya, Nagaoka, Tokamachi City. One of the reasons seems that Kawaguchi Town was delay for a request of aid to another cities and towns. At Ojiya City, only one emergency cut-off valve was installed at service reservoir and it did not work sufficiently at the earthquake. All water in the service reservoirs lost just after the earthquake because of leakage from the service main and distribution pipelines. This is seemed to be one of reasons of the low rate of restoration at Ojiya City. On the other hand, an emergency cut-off valve was installed at one of two reservoirs at each water distribution facility and they worked well at the event at Nagaoka City. An increase of the restoration rate at Nagaoka City was, therefore, greater than that of Ojiya City. The rate of refugees to the maximum number of Kawaguchi Town rapidly dropped after 12 days from the event. Electricity was restored at all area of Kawaguchi Town after 11 days from the event. The perfect recovery of electricity at Nagaoka and Tokamachi City was on three days after the event. These figures indicate that the recovery of electricity affects a decrease of refugees. Moreover, the rate of refugees decreased in about 10%, when the rate of restoration of water supply increased to about 90% in Ojiya and Tokamachi Cities. These facts suggest that recovery of lifeline performance strongly affects the number of refugees.

Gas Rate of refugee Water 100 7 90 6 80

70 5

60 4 50

40 3

30 2 20 Rate of restoration of lifeline, Number of after shocks 1 10 greater than 4 of JMA intensity

Rate of refugees to the maximum nember (%) 0 0 0 7 14 21 28 Date after the main shock

Fig. 7 Restoration curves of water and gas supply system、rate of

refugees to the maximum number and Number of after shock

greater than 4 of JMA intensity (Nagaoka City).

Gas Rate ofC refugee Water 100 7 90 6 80 70 5 60 4 50 40 3

30 2 20 Rate of restoration of Rate of lifeline, restoration Number of after shocks 1 10 greater than 4 of JMA intensity

Rate of refugees to the maximum nember (%) 0 0 0 7 14 21 28 Date after the main shock

Fig. 8 Restoration curves of water and gas supply system、rate of refugees to the maximum number and Number of after shock greater than 4 of JMA intensity (Ojiya City).

Rate of refugee Water 100 7 90 6 80 70 5

60 4 50

40 3

30 2 20 Number of after shocks Number of after shocks

Rate of restoration of lifeline, 1

10 greater than of 4 JMA intensity

0 0 Rate of refugees to the maximum nember (%) 0 7 14 21 28 Date after the main shock

Fig. 9 Restoration curves of water supply system、rate of refugees to the maximum number and Number of after shock greater than 4 of JMA intensity (Tokamachi City).

Rate of refugee 100 Gass 7 Water 90 6 80

70 5

60 4 50

3 40 30 2 20 Rate of restoration of lifeline,Rate of restoration Number of after shocks 1 10 greater than 4 of intensity JMA

Rate of refugees to the maximum nember (%) 0 0 0 7 14 21 28 Date after the main shock

Fig. 10 Restoration curves of water and gas supply system、rate of

refugees to the maximum number and Number of after shock

greater than 4 of JMA intensity (Kawaguchi Town).

CONCLUDING RENARKS

An outline of the damage to water supply pipelines from the 2004 Niigata-ken Chuetsu earthquake was presented and the restoration process of the performance of lifeline systems was discussed. The following conclusions may be drawn based on the present study. (1) The damage rates of the steel and unplasticized polyvinyl chloride pipe are high. (2) Failure mode of the unplasticized polyvinyl chloride pipe was pull-out at joint and breakage of pipe body and it of steel pipe was failure of screw joint. (3) Most of failures of ductile pipe with no aseismic joint occurred at joint. (4) The recovery of lifeline performance strongly affects the number of refugees.

ACKNKOWLEDGEMENTS

The present paper is based on data collected from an investigation by reconnaissance team of Japan Society of Civil Engineers. Many individual and organizations generously helped with this investigation. In particular, I am deeply indebted to Prof. Y. Takahashi of Niigata University, Prof. T. Komatsu of Nagaoka University of Technology, Prof. N. Araki of Nagaoka National College of Technology, and staff of waterworks bureaus of Nagaoka and Ojiya City for their help.

Final Discussion

Chairpersons

Dr. Hiroyuki Kameda (NIED) Mr. David Lee (EBMUD)

Tomioka; Ladies and gentlemen, we would like to start the final discussion. Thank you very much for your cooperation. Thank you. Anyway I have to talk about the objectives of the workshop. First of all, this workshop was started from 2000 in San Francisco, USA. Before this, in 1998, we have had a pre-workshop; it was arranged by the IWA, and sponsored by JWWA in Tokyo. There were 4 times workshop so far including this time, During those 7 years, we JWWA have a supporting committee, Chaired by Prof. Kameda, and We have the several members supporting the committees. I would like to introduce to Prof. Hamada, and Prof. Takada, he is not here, Mr. Kojima from Hachinohe Waterworks, Mr. Yamada from Tokyo Water Supply, Mr. Naito from Yokohama City, Mr.Yamano from Osaka City, Mr. Kijima from Kobe City, Mr. Kobayashi, he is not here from Hanshin waterworks supply, Mr. Toshima from Japan Ductile Association, Dr. Suzuki from Steel Pipe Association. And also Mr. Ishii from JWWA, he is not here. On behalf of him I am here, and observer, Mr. Sado from Ministry of Health, Labour and Welfare, on behalf of him, Mr. Azuma, he is here. And the secretariat, Mr. Fukuda, he is there. Those are our committee members. So, Prof Kameda should be qualified for the Chairperson of this final discussion, also Mr. Lee, he is a group leader of US. Then, I would like to ask Prof. Kameda to be chaired the final discussions. Thank you.

Kameda; OK, let’s go for a final stretch of the course. I am very happy to be here and very glad to be co-chair this discussion session with David Lee whom I have been enjoying so much of meeting together and working together. For the discussion, First, I would like to explain you about the some hand out that were prepared late at night last night and early this morning. After that we are going to have a presentation from Craig Davis, and on the previous workshop out comes also, after that, I would like to ask David Lee the discussion as they are well professional English speaker, in that way we would like to conclude this workshop. Then, first of all, I would like to go through this hand out. Well, the purpose of this discussion is general discussion and Q&A’s to presentations and we would to find out commonalities and differences. On that basis, we would like to review significance of the 4th Workshop which is this workshop, but we have continuation from the previous workshop which is the 3rd Workshop and on this basis, we would like to have some perspectives to the next workshop, 5th workshop in terms of topics, mode of discussion, etc and time-venue and those things. And Lee says request for filling opinion form, I felt to put Japanese in this line, but please see and look at last page, page 5, and please treat the following items and detach the page and give it to the secretariat when you leave the Workshop, before you leave the Workshop. What we are going to ask you to do is to better purpose for the future workshop topics, we are requesting all participants to (1) rank the following categories and (2) for any of these categories describe the most important topic as specific as possible that they would like to, you would like to learn from other counterpart. We would like to make these responses, very important basis for preparation of the next workshop. And the categories, listed here, “Risk assessment”, “Risk management”, “Seismic preparedness and readiness”, “Seismic resistant design”, “Experiences from recent earthquakes”, “Post-earthquake response and recovery”, and any other specific topics and these come from the previous workshop, in this way, we have a continuation and sociability discussion. If you have any general opinion about the workshop, please do, describe that, too, and be sure to put your name and affiliation. Thank you very much, and then returning to page 1, next chapter is the brief review of the issues addressed at the Kobe Workshop. Yesterday, on day 1, we had 3 presentations on session I, “People’s cooperation for seismic measures”. Session II, 1 presentation on “Risk assessment” actually field based assessment. Session III, “Risk management” are particularly, my impression is assessment and planning. There were 5 presentations. Session IV, “Seismic performance”, there were 2 presentations. And today, we had a “Mitigation and prevention of damage”, there were 4 presentations, And also we had a presentation from Taiwan, particularly on the Chi-Chi earthquake. Next session, Session VI, we had 7 presentations, most of the design procedures guideline development, and particularly, I personally was very much interested and impressed by US activities, and next session, you discussed and described on going activities, maybe almost finishing by the American lifeline alliance. Then, final session was on “disaster information system” and “Recent Niigata-Chuetsu Earthquake Effects”, and you still have a very clear impression about that. And next section 3; this is my personal impression or personal summary of the major points in Q&A during presentation sessions. I don’t think this is exhausted, but just for your information or to motivate your ways of thinking and I want you to further expand this and to develop this. Let me just indicated what I have felt, one major topic was “priority on disaster reduction in the US and Japan”. We have the comparative relations. “Differences in budget systems”, particularly in Japanese case, everything to be managed within regular revenue basis. That is the limitation as well as somehow stables this kind of thing. “Policy decision process”, I don’ know if putting this way, on that, my impression is that US side, US way is explicit and quantitative, Japanese way is more implicit and quantitative. I don’t know if you are agreed or not. “Implementation Strategies”, particularly, I was interested in many disaster reduction planning have been conductive in US, and I am a cureless how they are going to be implemented. And as I said while ago, ALA design guideline development, this was very interesting, and I believe that we have a lot of learn from this activity. And there are of course technical points, discuss, such as seismic upgrade technologies for pipes, water tanks, pump stations, pipes crossing active faults, etc, and also there were another aspects of the disaster management, “Mobilizing capabilities of residents, including water service monitors, emergency water management leader, by the way, this particular it was addressed by Yokohama-city, and I believe that Prof. Hamada as a citizen of Yokohama is going to raise his hand to be this leader, I am very much looking forward to seeing you in that way. Chapter 4, there was a preliminary discussion meeting between JWWA & AWWARF on a previous day for this workshop which is Jan. 25, at Hotel New Otani, it was attended by from US side, Elizabeth Kawcznski, Marilyn Miller, Craig Davis, David Lee, and from Japan side, Toru Tomioka, Ayako Hirata, Shiro Takada, Charles Scawthorn, and myself, and we discussed some perspective for the future activities, and if I summarized discussion, JWWA’s approach for the next workshop was presented which is the shown below, I will detached later, and US side responded and opinions were exchanged very frankly. At that time, it was agreed or maybe I should say endorsed at this time. “Discussion will be continued to accommodate wide topics”, “Opinions of the participants will be solicited in a written form at the end of the 4th WS”; this is what I asked you to do using the last page of this document. On this basis, both US and Japan side will discuss the subject matter and discuss and effective framework of the 5th Workshop, then, hopefully by having a preparatory meeting. For the next workshop, this is a current understanding that 5th Workshop shall be hosted by the US side within and expected period of 18-24 months. Next item, JWWA was proposal something like this, let me just read, “Recently we Japanese water supplies are interested in the emergency temporary water supply, restoration process, temporary construction, emergency water quality control, information collection, communication system, mutual cooperation network in waterworks, public relations method and so forth. It becomes important in Japan. We JWWA would like to discuss the risk management which is expanded the present discussions for sustaining this workshop. It should be combined both aspects of hardware and software in the field of disaster management (as the seismic measures).” Some background, in the early part of this document says that there are, we have an in-house, officials and engineers and also we, Japanese water services use a lot of contractors and consulting companies and major technologies are born by them. They would like to concept on this type of discussion. Anyway, that was a beginning of the discussion and we anyway decide to be reflexible about this, and we would like to continue on the discussion. Let’s next section says that “Survey Summary from the 3rd Workshop, Los Angeles. After I came here, I found that Mr. Craig Davis has made a wonderful proceeding of previous workshop which I was not able to attend and found out also that all the things I have just mentioned about have a very sorry relation with the previous workshop also. So, after this I would like to ask him to those aspects. I am not a right person to talk about this, so Craig,

Davis; Thank you, Prof. Kameda. I would like to start by adding one clarification on page 5. This comes from experiences in Los Angeles, the item, experiences from recent earthquakes. Please consider that as actual experiences, for example, such as Prof. Miyajima just presented for Niigata area. For post-earthquake response and recovery, please consider that as something that would be either planned, for example for Osaka today or something that was actually experienced, for example, we saw from Mr. Lund for Northridge and Kobe. I noticed in Los Angeles, the United States saw everything one way, Japanese saw everything the other way, so we need clarification. Ok, so, I want to start with the picture, University of Southern California, American college football team, they are the national champions in the United States, but in 2003, we were the international champions for the 3rd US-Japan Workshop. This will be a summary of results for the workshop which are very consistent with the request of Japan Waterworks Association during the meeting we had on Tuesday. I titled this “Future Directions for Improvement in Water System Seismic Practices Results of the Workshop”. However, I want to clarify that no matter what terms I am using here, they are consistent with the Japanese requests. The United States and Japanese have the same interests to focus the workshop in the same manner. So, the purpose of the workshops, and reason for the results summarized here are to discuss water system seismic issues, what we don’t know, to learn from each other, and to understand state-of-the-practice. This is something that the United States colleagues with American Waterworks Association Research Foundation discussed before the 3rd Workshop. We determined that we must further develop issues needed for a broad based seismic improvement program from the learning that is going on in the workshops; it shouldn’t just go home with one person. And to further identify topics in need of discussion and development. The background for the information that I am going to present to you comes from the 3rd Workshop for identifying needs for future focus. They are derived from the presentations, discussions, and survey results. Some of you who attended the 3rd Workshop may recall getting a survey I identified with the title “Important Aspects to Include in a Water System Seismic Improvement Program”. The results summarized are in terms of goals for the type of program and tools needed for the type of program. And again, my phrases, don’t misunderstand, this is the way I summarize and what we tried to do at the workshop. Everybody falls within this category. Ok, so the goals which are summarized, are also in the handout from Prof. Kameda, are “to provide adequate post-earthquake water supply throughout service area;” “to reduce earthquake damage to facilities;” to ensure minimum level system functionality and rapid system recovery;” “to achieve a rapid emergency response;” to accomplish a well planned, cost-effective, and publicly responsible seismic improvement program to ensure public safety;” and “to continually develop and improve earthquake disaster prevention capabilities.” These goals are my summary of a very large amount of information. They are obtained from discussion and presentations in the 3rd Workshop and from the previous Workshops as well. In addition, the discussion identified tools needed for water system seismic practices. There are including “Management strategies;” “Seismic resistant pipe;” “Mitigate component damage effects on system functionality;” ”Establish seismic performance criteria;” “Water supply;” “Education.” Everybody falls in this category. I would suggest most all of the categories of topics we had during this workshop also fit in these needs as well. Now, I am going to move faster so that I can leave here and we can just talk. Prof. Kameda asked me to leave the detail but I want to move faster through the detail. Ok, “Management Strategies” have been identified in the past and need “community involvement” and “planning and prioritization.” For what the Japan Waterworks Association is interested in, I want to point out “other tools also deal with Earthquake Disaster Management.” Seismic resistant pipe:, “Design guidelines and standards”, you see the ALA Guideline, “Manufacturing capabilities”, for example, Japanese S-joint is not available in U.S and I believe that there is some aspects where the technology exists in the United States and Japanese are also interested in it. “Replacement/improvement of old and bad pipe” and “Replacement prioritization schemes”. “Damage Effects on System Functionality”---“System redundancy,” “Block distribution system”, “Isolation capabilities”. “Seismic Performance Criteria”---“What are acceptable system outage times,” “How should a water system perform with blank exposure”, we need to fill in the blank, “Should criteria be established on National, State, or Local level.” For water supply, “Cost effectiveness of adding water supply for seismic purposes or distributed supply adding redundancy”, “Means of providing water in any manner following a major disaster.” It is interesting that Japanese and United States approaches to this were very different. And “Education”; this consists of owners, engineers, field crews, and so on. All sides in every country need to be better educated on seismic issues, why seismic issues are important to consider, what should be considered, how to implement and use. OK, so conclusion, the 3rd US-Japan Workshop—identified several items needing further discussion and development. 4th Japan-US Workshop—we are requested to further develop topics in need of discussion and development, we will be requesting your opinion for desired focus. In Japan, we have a Japanese professional team, Seibu Lions which were national champions, and I suggest we are now International Champions at the 4th Japan-US Workshop. Thank you.

Kameda; Thank you very much. Now I think that there are many things already there in a previous workshop, and this time, we have been doing our discussion in a very sociable way, and much more propounded ways of discussion, so after this, please give us your inputs about your thoughts of workshop and do your perspective for the next step. That would be, I hope you reflect when your write of requesting form. Then, I would like to ask David Lee to the discussion.

Lee: First of all, please ask the general questions to anybody, any topics and we would like to continue to exchange with Japanese about presentations.

Kameda: Excuse me, one comment to the Japanese, please

Prof. Kameda asked Japanese attendees to have any comments or questions in Japanese language in this discussion.

Kameda: I am encouraged them to speak out either English or Japanese language.

Lee: Let me ask to this question from Japanese, what topics you would like to discuss the most in the future on the last page? Anybody, speak up first.

Naito: I am Naito from Yokohama Waterworks. Now, we have met 4 times for the workshop. I think that we have less covered the major topics so we feel that collaboration with citizens is going to be important topic from now on. I say because the local government, I think we have worked very much improving hardware to that and I think we have almost limited. So what we would like to pay attention to from now on is improving on the software side which means say that collaboration with citizens and community. This is going to be main topic. So, Yokohama waterworks would like to request the main topic on the future meetings.

Miller: Marilyn Miller from EBMUD. Many water utilities have developed sophisticated tools to access seismic vulnerabilities and identify needed seismic upgrades. I think that maybe opportunities to use those tools to help focus emergency respond planning to identify in advance where what part of system will need repairs and use as the way to focus immediate post-earthquake activity. I would like to explore that further.

Lee: So, that category, which one? Seismic preparedness and readiness?

Miller: Yes.

Land: Le Val Land from Los Angeles. I support the idea of the tools for emergency respond but I also I think we need to recognize that many more small water utilities or the large utilities, those types tools need to be simple and available to non technical personal.

Shinozuka: Shinozuka from University of California, Irvine. I have a couple of comments. First of all, I can agree more to the opinions that have been already expressed. I would like to add first of following. Prof. Kameda, you asked to us to fill out this last page, that is good and I will do that. On the other hand, all these things, “Risk assessment”, “Risk management” and et cetera, these are very much highly colligated and it is not really correct to see that each item individually. For example, although “Seismic resistant design” is important for a technical component of the profession, but as same time, that is very important part of risk assessment depending on they are design procedures and design guidelines, risk is quite different. So, next time, I think someone should, next workshop, maybe will invite someone to talk about system integration and we get sometimes blind about some other factors when concentrate one thing and this is an important issue for us to be aware. Second, there has been no mentioned seismic resilience concept that is when system is hit by earthquake, the performance goes down immediately, and then you have to start to repair and restore the system. More time it takes, it will be involved more significant impact and economics involved and societal destruction. And this resilience concept is now considered for community, that is much more difficult issue, but we could start out with water system first, so what is resilience of water system, limiting our service to the issues making sure water is supplied within shortest time period depending on type of earthquake, intensive earthquake when you introduced at past to become probabilistic. That resilience is an important concept. Then, I would like to say that also there is not much discussion about real-time damage identification not only that but dissimulation of information, broadcasting information so that not just government agencies or utilities know about it, but even citizens should know what is happening. This is not necessarily easy but for us the part of what I have just said that is damage identification in real time, this is something I am very much interesting it in, not just for water system but transportation system, power system and I would, you would say that that research, I think the research should not be excluded, expense of more immediate utility of what we are trying to do, I think university of professors are willing to research if you identify some things we should do with cost are very effectively, I think that something I would to impress upon on you. You said that I can say anything?

Lee: Yes.

Kameda: With responsibility or without?

Lee: Without, without. I totally agree with you because I know how much I paid when I was in university. I think you are correct, and university, you can do research very effectively.

Kameda: We would like to hear more from most of Japanese side including the requests. Mr. Naito earlier from Yokohama has given a very clear idea of what they would like to focus on the future workshops and exchanges. If I meant summarized what he had requested it extremely, may I take it that Mr. Naito, you have said we have gone through the various measures, countermeasures is to reinforce the hardware aspects. What we can do has been somehow done already, although the level might not been high enough, therefore what you are suggesting is to try to focus on the exchanges on how we are able to have collaborations with residents. Is this miss my understanding or correct that what I would like to confirm with Mr. Naito? And for the other Japanese delegates, may I ask this kind of opinion is going to be the voices of the other city representatives as well? As pointed by Prof. Shinozuka and Ms. Miller, they are must be differences about the responses in the measures taken by the large scale utilities and the small scale utilities. What about the voices from Hachinohe, of course Hachinohe is a fulfilled waterworks utility, but in terms of scope of the city, is not categories one million population city, therefore we would like to hear from voices from Japanese people.

Kojima: I think I seem to be nominated on behalf of smaller scale cites. This is the 4 times I am attending this workshop, and I am very much grateful this occasion was given to me and so far. I have had informative experiences with you. Through the workshops, I was able to understand the differences in the United States and Japan. I don’t know whether this is going to be the voices of other utilizes of Japan, but if I may share my personal view points, in our case of Hachinohe, water utility is taken whole responsibility in preparing programs for countermeasures, for reduction of disasters, and that program whatever you do is communicating to the citizens, but before that we try to confirm whether such programs or countermeasures should take place by the water utility. After getting the confirmation by the citizens within given to us to save the budget plan, for example, for 4 to 5 years we tried to improvement whatever we can do as the countermeasures for the disaster of countermeasures. Suppose during this budget period, the program period, if our top management of our utility says we are not going to raise the charge, then without raising the charge, we try to implement countermeasures as much as possible. But as citizen side, it seems that we water utility did not have the opportunity to present us to the citizens. How old countermeasures are prepared? How old countermeasures are to be implemented? I think this kind of explanation so far lack on that part. Because the citizens know that countermeasures somehow have been taken by water utilities and they are sure 100%, but there is nothing we can sure to extend of 100% was presented by Mr. Matsushita of Kobe, at the time we had 3rd Workshop on Risk Management. Therefore I am personally interesting in hearing from the workshop not only Hachinohe utility but personally we would like to know more about what portion can be implemented among those countermeasures to develop 100%. And any the upper limitations which we can only do some extend or there are any portions of the countermeasures we are not able to respond. Those are areas I want to focus coming for our future workshop. Thank you.

Kimura; I am Kimura from the utility of Osaka prefecture. The biggest problem we face is not the seismic measures but rather I have to say the aging of equipment and hardware we have. After the great Kobe earthquake, we have come up with the seismic measures, but I think we realize that it is necessary to review what we need to do. We have to focus in our area is how we are able to respond to aging facilities. We might to expect some droughts or we might to expect some incidents; therefore we should never only concentrate on the seismic areas but rather we should broaden this scope of our perspectives so that we will be able to review our systems as the whole comprehensively. That is why we have a plan to introduce the large capacity transmission line. The purpose of the introducing such new transmission line was not seismic measures along. We want to upgrade or rejuvenate the current system with new pipe. Of course, as the result or as the effect, this had also contributed it to responding the seismic measures as well. As I presented yesterday, I believe hardware measures have to be look at from very comprehensive, integrated manner, not only in Osaka prefecture but also the other municipalities. I believe they share the same view points in the future we expect that water demand goes down and we will suffer physically from revenues, from charges of the types. Therefore we have to run on business very efficiently; therefore we should never focus only seismic measures. We believe in cooperative software aspects, and we need to have the detail to researches and studies. To this effect, I think it is necessary to get real time information especially from citizens so that we will be able to respond quickly to be requested citizens even in any kind of disasters. Therefore somebody says earlier that collaboration with residents is going to be very important for our utility including over all software measures.

Furuya; I am Furuya from Yokosuka city. It seems that because the Yokosuka city is not an urban, large city, there are different view points I can share with you. I had presented earlier as living in a smaller city surrounding the large cities. In the past, we were rather depended on initiatives by the urban, metropolitan cities whenever any incidents happened. But, since the Hanshin-Awaji Great Earthquake, we have changed for our thoughts. There seems that cannot be done by the utility alone as was pointed by Mr. Naito. I believe the necessary is to focus on the collaboration with the residents so that other types of disasters we will be able to respond and overcome the difficulties in collaboration with residents. In another point, we realized from experience of the Hanshin-Awaji Great Earthquake, before that time we had not expected or we had not made assumption that they would be a lot of support from other cities therefore this is another area we need to discuss how we are going to coordinate our activities with this support given from other cities or other utilizes in case of disaster. Two more comments, in the emergency water supply I hope that we will be able to have the opportunity to discuss water quality matter as well. Is it necessary to maintain the quality level as the ordinary case, of course for drinking water probably we need to maintain the similar level? But it is really necessary to maintain the water quality just like the ordinary cases for the other applications as well. That is the area I think the further discussion is needed in the water quality. And also for the use of drinking water at that time of emergency, I would like to know how that use can be reconciled with water to be used for the fire fighting. They should be further discussions because that area is rather blurred or ambiguous so far. Those are the points of discussion I am very much expected in the future.

Matsushita: I am Matsushita from Kobe, from the form theory of the workshops to come, I think it is rather pretty much sure to focus one particular discussion area, for example, collaboration with residents when it comes to the papers to be presented and discussed in the workshop. When we prepared the papers, I think we should broaden the scope of the areas to be presented and to be discussed, and I myself, I am interested in getting more information from different areas. But having said that, probably it is necessary to discuss one particular area and focus on the particular are such as collaboration with residents. Therefore, for those kinds of specific topics, may I suggest to allocate one half day to discuss theory on the subjects, on top of the series of the papers to be presented in general terms? I have been attending the meeting in the past 3 times, and I came to realize a little by little some of the differences between two countries. In order to understand mutually, the situation and majors to be taken, I believe it is the necessary to understand the social background, common practices and common senses we each have. So, it looks like to try to Japanese side try to make a presentation, prepared the paper based on their own practices, common senses which makes American delegates are difficult to understand sometimes, and US side presents based on their own common senses to us and sometimes Japanese people have difficult understanding at the situation. Therefore, in order to understand, extend of understanding, it is necessary to know the background with this presentation and situation are happened. That is one suggestion. And Mr. Kimura from Osaka said, those points should discuss very comprehensively from different perspective and in that integrated discussion and presentation are the seismic measure should be reflected as one of the discussion items. I suggested similar thing when I made representation on the need of rebuilding system in the 3rd workshop. Unfortunately, there was less, a very small response from the US side, but those are some of the request I have.

Yamano: My name is Yamano from Osaka city. As how we can proceed the future workshop, let me have one comment. I think all water utilities need to proceed aggressively with seismic measures. In those measures, not only the software but also the hardware aspects should be incorporated. Therefore, I still think that hardware aspects are going to continue to be our important areas for the discussions, the upgrading of the pipelines or the equipment of facilities. I also think risk management, risk assessment and collaboration with residence. They are all important. Therefore, I do not think this is necessary to focus on one particular topic whenever we have those workshops, but having said that, I regret that all the time we have a very small time for the discussion on each topic, therefore in order to get understanding both sides, it is necessary to have ample time of the discussion on each topic. That is my request.

Lee; I totally agree with you.

Nukui; I am Nukui from Yokohama City. I am not an engineer and I am attending here for the first time that I was given to present a paper, and I am on the administration side of the work. It looks like the discussion focused mostly engineering or technology side. That is OK with me, but when it comes to the measures should be taken by Yokohama-city with a very limited amount of budget we have taken the various measures for seismic programs. But in order to improvement effective programs, time and money is necessary. Whenever the disaster or incidents are happened, we have to work very hard to respond quickly and to provide water supply on the emergency cases. We do have the underground tanks available in Yokohama, and also somebody from Yokosuka said one of the two tanks can be available for the fire fighters to be utilized so much so that we have made of those efforts to take various measures through the various programs. But the question has be asked who are going to operate on those facilities equipment at the time of disaster, it must be very difficult to have all the employees of the utility present to give the support, I apologize if I am wrong, I believe even the Hanshin-Awaji Earthquake, they are only little more than 60 percent employees of the utilities were able to attend in the emergency response works. Therefore, it is necessary to have the support not only from the employees but also from residents at the time of disaster. That is why I think the collaboration with residents is very important, and, of course, it is possible even for the residents if they are trained they are able to operate on the water supply through the underground tanks. Therefore I would like to continue to focus not only for the hardware but also for the software areas for the future discussion. Thank you.

Lee; I get the permission from 2 bosses here, Tomioka-san and Prof, Kameda. They told me time is not important, but personally I think we should finish around 5:15 so just we give everybody’s idea. So, if you have questions, please ask. Prof. Kamade says this is unusual it has so many Japanese speakers. I said that they are making up for the last workshop. That is even.

Kameda; I think the interpreter has used a strong word for the interpretation word, “unusual”. I have said the way of Japanese delegates have been acting is little bit different from the ordinary cases, and I am very much grateful to see that. That is what I meant.

Land; I would like to add for the discussion we talked about seismic mitigation hardware. We talked about software for emergency planning and response, but I think it is necessary for water utility to think about recovery of community. One of the items of discussion we don’t seem to talk a lot of recovery of the community related to water supply. I think that is very important because of the need for continued emergency but also for business, manufacturing, schools and hospitals and other activities, more permanent basis you put back the normal operations. So, I would like to ask the continuous or consideration of recovery as one of the items discussion.

Scawthorn; Just some thoughts on the next workshop. First of all, when you look at the history of both the Western United States and Japan, in terms of the huge fires that occur every year in Western United States, the two largest urban fires that ever occurred in San Francisco, 1906 and Tokyo in 1923 and so on and so forth, the topic of fire following earthquakes, might be of some valuable. I always think how absent fire people are from these meetings and how absent water people are from fire department meetings. Now, on the American side, John Eidinger and Bill Heubach and others have worked on a recent publication for the American Society of Civil Engineers. So, that might be interesting for the Japanese side. That is one topic. Second topic is Prof. Shinozuka has done a lot work on aging infrastructure and I was recently at a meeting on the topic where the whole world of micro-sensors is just opening up, and I wonder if that might be something a year from now that might be interesting not only for ordinary operations but also for proposed earthquake operations, as Prof. Shinozuka has identified. The third point is money, as Bill Heubach presented. How programs are financed and innovative new ways of financing, perhaps combining ordinary bond measures in the United States or perhaps insurance, on the Japanese side perhaps there is less bonding and so on. But they might be interested in how the US does things. Both sides are interested in money. I always think the true cost presented to citizens has always led to their support of mitigation, and I think the true cost of the loss of a water system is not known. Mr. Heubach did a good job, but I think we still don’ t know the true cost of the lost water for a period of days or weeks, as Prof. Miyajima pointed out in Nagaoka. So perhaps involving some economists from the various larger water utilities here might help someone. These are my thoughts.

Lee; I would like to comment as the current exchange information. Please do not hesitate to ask us if you have any interesting material that you want. We will send you the material, and of course you will translate it by yourself.

Hamada; We are now discussing how we are going to proceed the workshop in the future. Looks like listen to the voices from the other participants, we are now coming closer to the conclusion. What I would like to suggest is that it would be possible to narrow down one or a few particular theme for the workshop. Take it one example of the “Seismic resistant design”. I believe this is the theme which was put in this paper for the first time. And in the United States had started this kind of design and we are expected they will be more and more interaction on this subject between Japan and the United States. I expect that they will be a lot of inputs in this area. I would like to also comment that we should never stick only the presentation. We have to have a lot of lively discussion following presentations. Therefore what we can suggest is to take one or a few important, critical themes which reflect the overall presentations or the papers of the workshop, and allocate half a day or full day for the discussion on the important themes. For example, I believe for this particular workshop in advance we were told that collaboration with residence is one of the themes to be discussed. If this is a sight to us, I think some of Japanese delegates refer to the subject and I know that the US delegates are working on this subject as well. If there is such an assignment in advance to us, we should be fully prepared to discuss on that subject when we come to the forum.

Lee; Anybody else? OK, last 3 speakers.

Yamada; I am Yamada from Tokyo metropolitan utility. I think our utility is the largest scale in Japan. But size of the utility does not correspond to the scope of my answer. Having said that, if I may share my observations, probably we do not have to focus one particular theme for the papers and the discussions in the workshops to come. But still we need to focus on some of the priority areas for the discussion to understand mutually. I myself, personally, are interested in hearing more about the measures of the responses to be taken after the earthquake from US side. I have been attending this meeting since the previous meeting, and I came to realize that there are differences between the two countries when it comes to the measures and responses taken after the earthquake. They seem to not be done by the US side because of charges or because of the need to convince residents for the things on Japanese side doing routine matters. Therefore the prevention side is also important but I am very much interesting in hearing from after the quake measures and responses. Those are areas which were not fully touched by the US side, therefore I am very much expecting to hear the measures and responses after the incidence from US side. The reason why is that I am interested in this post-earthquake measures and response is when comes prevention, probably we can set this down criteria so much and so forth. But when comes the post-earthquake, we cannot just discuss this based on standard, looks like each utility can come up with various and unique measures and responses. That is something very much I would like to hear.

Ando; I am Ando from Kobe City. I was given by the only last 3 speakers. This workshop is named as the Japan and US Workshop that is the very important significance, and of course, we are very much grateful with the Taiwan delegates in this workshop. I think through the discussions between 3 countries, we are able to focus on the common subjects. It is natural that we broaden the scope of the discussion from the earthquake and to the other incidents. But the measures in the programs, we have been implementing for the earthquake program can apply to the other incidents and disasters as well. Although we have been doing a lot of upgrading rehabilitation works. That is not for upgrading purpose, but rather by doing that we will be able to contribute to prevent the occurrence of other incidents. For example, in the case of Kobe, we had recently stricken and terrible typhoon. There was the outage of the power in long period of time. At that time, some parts of water pipes were flooded. Unless we have been taken various measures after Hanshin-Awaji Earthquake, I believe the damage or the failure could have been worth what we experienced earlier. Therefore what we have been done through the countermeasures against the earthquake has to going to be a lot for the various incidents. It is advisable for all of us to discuss on those various themes coming workshop. In our scope of the works of the ministry of Health, Labour and Welfare has prepared a guideline. That guideline, there are various indicators by which we are to be evaluated and we are in the position of making the variation on ourselves, especially for the safety and security areas. Therefore we expect that we will be a lot of discussion in risk management as one of the important things because valuation is going to be important for all of us. That is the area I am very much willing to hear. Last thing, I hope that we will be able to bring this workshop further to prosperity so that we can be called as the internationally top leader’s workshop between Japan and US.

Lee: It is very proper that US leader will make up the final conclusion.

Shinozuka: Thank you very much. This is Shinozuka from UC, Irvine. I think the comment was made from Japanese side that they would like to know how US side thinks about the post-disaster strategies. I would like to say my own opinion about that. First of all, I would like to say that UC, Irvine has a huge research grant from NSF (National Sanitation Foundation) to deal with post-disaster rescue issues we called rescue projects. It is 15 million dollars grant from NSF. They are fortunately or unfortunately, principal investigator is the IT expert, he is a computer scientist, but it is fortunate because we can look at this post-disaster issue without really asking industry to pay. This is one of the points I wanted to make. Many researches are being done without your knowing. And they are interested in using micro-censors, and they are interested in knowing what happened immediately after earthquake. You have to realize that this takes reliability out of equation; I should say probability out of equation. Now, you are dealing with what happened and important point is that you have to know what happened immediately. That is where this advanced technology come in. There are the ones who proposed that we should ask every citizen to become human censor as long as we have the means of transmission of the information from each citizen; let’s say central control organization. Then everyone, censor, you don’t have to put all these expensive censors anymore, and only thing is the advanced technology that everyone is to become censor and communicate. These are very advanced concept. You see, we don’ ask anybody about money except for tax payers. I would like you to think about this concept because it is very important if you know what happened right away and if you then accumulate our knowledge going through these simulations of disaster and so forth. You should choose their closest scenario of disaster to what happened. Probably you would know what we should do. I think this is very important research contributions. We are making sort of this behind the scenes and I am quite sure satellites image. During pay for that development we can take advantage of that technology very well. Right now, seismic urban damage assessment, this is my opinion including how important it is to carry out research and if you can give me input what you need. Please use your imagination and then I can work with these people and come up some technologies. This is something I would like to say. Thank you.

Lee: I would like you to remind that Taiwanese colleagues that they had raise a hand and noticed that. That would be a last speaker.

Shih; Prof. Shih; Ban-Jwu Shih from Taiwan. I would like to thank everybody here to invite us, the Taiwan side, to participate in this successful and fruitful workshop. Believe or not, near 99% of the population in Taiwan use tap water, but we do not have any official seismic design code for water pipelines. From this workshop, we learned a lot about seismic designs and soft countermeasures like emergency response plans from practicing engineers as well as professors from Japan and the US. Especially we can download ALA reports from Mr. Eidinger’s website, which should be very helpful. Earthquake mitigations and emergency plans for water supply systems are very important for Taiwan, but we do not have much experience on them. Therefore we asked people from Tokyo or Kobe water bureau to give us copies of their response plans, and thanks to them they already consented. In Taiwan, according to law, our government asks all utility companies, not only water supply companies, gas and fuel companies, as well as electric power companies to prepare emergency response and recovery plans, especially for earthquakes and typhoons. So, I would like to ask people from both Japanese and US sides that if you have any documentations or reports regarding to lifeline seismic codes or emergency plans, please let us know. Because of this successful and joyful workshop, we hope that we can again attend in the next US-JAPAN workshop. Thank you very much.

Kameda; You know as much about what I say. We had capable interpreters in this workshop so that we were able to communicate this meeting smoothly. Thank you very much.

Interpreter; Thank you for your compliments.

Kameda; I think that we had a very fruitful time for the discussion. Time was extended half an hour more than original planned. I think we can return to both sides and elaborate this discussion and think about strategies for next one and meet again to talk about future strategies. I think this workshop was very successful. Thank you very much for your cooperation. Now, I would like to return the micro-phone to the real boss of this workshop.

4th Japan-US Workshop on Seismic Measures for Water Supply Survey Results for Seismic Improvement Program Topics (アンケート調査結果） Ranking the Requested Categories for the Future Workshop (The Survey responses are received 9 from US, 3 from Taiwan and 20 from Japan) （回答結果：米国９名、台湾３名、日本２０名） Experiences from Seismic Seimic Post-Earthquake Risk Assesment Recent Risk Management Preparedness and Resistant Response and Other Topics Organization (リスク予測評 Earthquakes （リスクマネジメント） Readiness Recovery 価） Design （最近の地震災害（他の課題）（地震災害への備え）（震災対応・復旧）（耐震設計）の経験） US Marilyn Miller, 1 52 1 6 3 4 7 EBMUD Tim Collins, 2 22 2 1 3 City of Portland John Eidinger, 3 G&E Engineering 34 5 1 6 2 7 Systems Inc. Don Ballantyne, 4 24 7 3 5 6 1 ABS Consulting Bill Heubach, 5 54 6 1 3 2 Seattle Public Utilities Michael Conner, 6 12 3 City of San Diego Luke Cheng, 7 21 1 1 1 2 San Francisco PUC Steve Welch, 8 Contra Costa Water 6 5 4123 District 9 Nameless 11 Taiwan Pei-Chung Hsu, 1 44 1 3 2 2 Taipei Water Ban-Jwu Shih, National Taipei 2 12 6 5 3 4 7 University of Technology Wei-Sen Li, National S&T Center for 3 Disaster Reduction, 34 2 1 5 3 2 Taiwan Experiences from Seismic Seimic Post-Earthquake Risk Assesment Recent Risk Management Preparedness and Resistant Response and Other Topics Organization (リスク予測評 Earthquakes （リスクマネジメント） Readiness Recovery 価） Design （最近の地震災害（他の課題）（地震災害への備え）（震災対応・復旧）（耐震設計）の経験） Japan Masanori Hamada, 1 3421 Waseda University Charles Scawthorn, 2 21 6 4 3 5 Kyoto University Kenetsu Kojima, Hachinohe Regional 3 51 2 4 6 3 7 Water Supply Authority Hiroshi Yamada, Bureau of 4 Waterworks, Tokyo 34 2 5 6 1 7 Metropolitan Government Kiyoshi Naito, 5 Yokohama 64 5 7 2 3 1 Waterworks Bureau Hironori Nukui, 6 Yokohama 21 3 5 6 4 7 Waterworks Bureau Kazuya Yamano, 7 Osaka Municipal 41 1 5 7 1 6 Waterworks Bureau Hiroaki Miyazaki, 8 Osaka Municipal 21 2 2 2 2 Waterworks Bureau Tetsuro Kijima, 9 Kobe Waterworks 1312 Bureau Toshio Toshima, 10 Japan Ductile Iron 21 5 6 3 4 Pipe Association Shogo Kaneko, 11 Japan Ductile Iron 56 1 2 3 4 7 Pipe Association Experiences from Seismic Post-Earthquake Risk Assesment Seimic Resistant Recent Risk Management Preparedness and Response and Other Topics Organization (リスク予測評 Design Earthquakes （リスクマネジメント） Readiness Recovery 価）（耐震設計）（最近の地震災害（他の課題）（地震災害への備え）（震災対応・復旧）の経験） Nobuhisa Suzuki, 12 32 1 1 5 4 6 JFE Takahiro Yabuguchi, 13 32 1 4 6 5 7 JFE Tatsumi Fujishiro, 14 Japan Water 52 4 6 1 3 7 Resarch Center Eizo Seki, 15 Japan Water 1 Resarch Center Masahiro Kimura, 16 Osaka Prefectural 43 2 7 6 5 1 Waterworks Kenji Totoki, Waterworks Bureau 17 65 1 2 3 4 The City of Hiroshima Shinji Kamura, Waterworks Bureau 18 31 5 2 6 7 4 The City of Hiroshima Takashi Furuya, Yokosuka City 19 65 4 7 1 2 3 Waterworks and Sewerage Bureau Makoto Matsushita, 20 Kobe City 71 4 5 6 3 2 Waterworks Bureau Closing Address

Toru Tomioka Japan Water Works Association

Thank you very much for Prof. Kameda and Mr. David. I have to say something for our closing. Thank you very much for attending this workshop and I would like to close this session but tomorrow we have a tour in Kobe city. First of all, I would like to thank for people who engaged this workshop in Kobe city and all of you. I hope you enjoyed this workshop and also enjoyed the active discussion. Especially Japanese participants could discuss each other with the American people. We, all of us agreed to be held the next workshop, maybe in the United States. So, I am sure we will meet you again in the United States. Thank you very much, and I would like to ask to Elizabeth.

Elizabeth Kawczynski American Water Works Association Research Foundation

I want to add my thank you to all of the presenters. From the United States, we are very much appreciating to the Japanese being willing to make presentations in English. That was very clear and understandable. I know we couldn’t do the same in Japanese so very much appreciated it. Thank you, Thank you. Especially thank you to all of the utilities that made presentations. Because we know that you all have full-time job and are very busy. It is gratifying to have you taking so much effort to make presentations and exchange information. The discussion of the end was fabulous. So, I think we are really a champion team and we certainly look forward to hosting next time in the United States, again. Thank you very much. Technical Tour

Kobe Municipal Waterworks Bureau Overview of Large Capacity Transmission Main

１．Introduction

Kazuhiko Mizuguchi ２．Overview of Damages PullingPulling outout ofof MechanicalMechanical JointJoint （Rokko Bridge －φ700 SP）

Bridge-attached aqueduct pipe to be broken down Damage of Raw Water Conveyance Tunnel Damage of Conveyance pipe by typhoon Tank of Other Water Supply Utility Citizen waiting for water supply from the tank wagon Emergency Water Supply in the Distribution Station Collapsed Building blocking up Road Collapsed Houses, Highway viaduct blocking up Road Daily Traffic Jam in the Main Road

３．Restoration Plan Based on the lessons of the 1995 Hanshin-Awaji Disaster

Limit to tank Emergency Reliable Emergency truck supply Water Supply Water Supply Wide Area system Service Interruption

Pipe Network Disaster Resistant of ion Water System uct s Upgrade Red eak e br pip

Extended Large Capacity Easy to repair recovery term Transmission water system Securing Main source water Five main aims in the Basic Plan

(1) Temporary restoration within 4 weeks (2) Emergency water supply (3) To secure water at the disaster prevention centers (4) Geographically fair restoration (5) Stability of a social welfare Main Projects of Restoration

(1) Emergency Storage System (2) Large Capacity Transmission Main (3) Seismic Pipe Network System (4) Seismic Upgrade of Principal Facilities (5) Monitoring and Controlling Facilities ４．Water Sources Water Sources in Kobe

20 Daily Water Supply

Daily Water Supply Capacity Daily Water Demand (as of April 2000) (2003)

Hyogo Prefecture Nunobiki Reservoir Hyogo Prefecture Others Water Supply Project 3 Water Supply Project 20,000m Others Dondo Dam 23,371m3 Dondo Dam 3 Aono Dam [2.2%] 33,000m Aono Dam [4.2%] 3 Karasuhara Reservoir [3.7%] 3 19,152m 28,000m Sengari Reservoir 3 [3.1%] [3.5%] 28,000m 3 [3.1%] 40,972m [7.4%]

Sengari Reservoir Hanshin Water Hanshin Water 119,000m3 Authority City-owned Authority City-owned Lake Biwa Water sources Lake Biwa [13.2%] Water sources Yodo River 3 Yodo River 3 3 64,343m 3 200,000m 672,000m [11.6%] 471,169m [22.3%] [74.7%] [84.9%]

Daily water supply Daily water demand 900,000m3 554,664m3 ５．５．LargeLarge CapacityCapacity TransmissionTransmission MainMain

１）Main functions

1. For renewal of existing tunnels 2. For the inspection of existing tunnels 3. Diffusion of risk of transmission main 4. Emergency water supply 5. Quick restoration ２）Route Ashibedani Junction well

Uegahara Purification Plant Third Tunnel (Plan)

Okuhirano Purification Plant Ashibedani Junction well ~ Sumiyoshi River Shaft

Ashibedani ( In Service ) Junction well Uegahara Purification Plant

Ashiya Border line

Sumiyoshi River Shaft Motoyama Shaft Sumiyoshi River Shaft ~ Nunobiki Shaft ( Under construction )

Sumiyoshi River Shaft Ishiyagawa Shaft

Oji Shaft Nunobiki Shaft Nunobiki Shaft ~ Okuhirano Shaft

( Under a Plan ) Nunobiki Shaft

Okuhirano Shaft ３）Illustration

ls nne g tu istin Junction well Ex

B ra nc h L ine Distribution n reservoir ai M n sio smis an Tr LargeLarge CapacityCapacity TransmissionTransmission MainMain

Two existing tunnels

LCTM

Shaft(water supply station) Emergency water supply Fire Fighting

Link to Distribution Pipe ４）Cross Section

Transmission Main Foam mortar

Segument

ain f M r o ete am Di

Segument outer Diameter Table. Overview of Large Capacity Transmission Main (ASHIYA Boundary ~ OKUHIRANO Shaft)

Extension 13．7km

Diameter 2．4m

Conveyance Capacity 400,000 m3／day

Investment 45 Billion yen

Construction Period 1996 ～ 2010

Storage Capacity 59,000 m3 ５）Emergency Water Supply Facilities

Water Emergency supply to water supply water via temporary wagons and water taps fire trucks Connecting pipe Motoyama shaft

Elementary school Emergency water supply station ６）Drilling of emergency water supply Drilling to enable to practice Firefighting & Water Supply ６．Method of Construction

１）Construction of ISHIYAGAWA Shaft SMW 立坑築造施工順序３立坑築造施工順序５立坑築造施工順序７立坑築造施工順序９ Building of Body

立坑築造工写真 Excavation and pressing Base Concrete ２）Shield Tunnel Method Assembly of Shield Machine Departure from Shaft Excavation

Back-fill injection

シールド掘進状況３）Slurry Shield Machine Inside of Shield Tunnel ４）Transmission Main ５）Inside of Shaft

Branch Line

Air Valve

Transmission Main Inside of Shaft

Branch of Transmission Main

Pressure reducing valve ７．The Usage of Large Depth Underground

送水管トンネル

地下河川トンネル ▲東京都内地上風景 0m 10m 20m 鉄道トンネル 30m 40m 50m 60m 70m 80m 90m

地中送電線トンネル下水道トンネル地中ガストンネル Definition of Large Depth Underground

40m

Pile Fundamentals

Support Layer

10m Large Depth Underground Large Depth Underground Adoption of Law

Nunobiki Shaft

Usage of the Law

Okuhirano Shaft ８．Total Plan １）Nankai and To-Nankai Earthquake

ＫｏｂｅＣｉｔｙ２）Total Plan of Large Capacity Transmission Main

Ashibedani Junction well

Myodani Pump Okuhirano Ashiya Boundary Purification Plant