<<

EERI Special Earthquake Report — April 2012

Learning from Earthquakes

The Mw 7.1 Erciş-Van, Earthquake of October 23, 2011

Introduction as northwest-southeast right-lateral The following EERI reconnais- and northeast-southwest left-lateral sance report on the 23 October At 1:41 pm local time on Sunday, translational fault zones (METU, 2011 Erciş, Van, Turkey earth- October 23rd, 2011, a Mw 7.1 2011a) (see Figure 1). quake has been prepared by earthquake struck Van Province in Rafael Alaluf of EQRM Interna- eastern Turkey (USGS, 2011). The The main shock is believed to have tional Inc., Ricardo Hernandez of earthquake claimed 604 lives. The been on a WSW-ENE reverse Degenkolb Engineers, Cemalettin town of Erciş, with a population of fault with north-dipping fault plane Dönmez of İzmir Institute of 77,000 (GOVP, 2011), was hit hard- (METU, 2011a) (Figures 2 and 3); Technology, and Ayhan İrfanoğlu est. Located about 40 km NNW of the the fault was not identified on the of Purdue University. Funding for estimated epicenter, it had 191 build- active fault map of Turkey (MTA, travel for these team members ings collapse totally or partially, killing 1992). The November quake is was provided by the individual more than 600 people. Van, the larger believed to have occurred on a authors, EERI, the İzmir Institute provincial capital located about 15 km strike-slip fault also previously of Technology, and the George SSW of the estimated epicenter, was unidentified. E. Brown, Jr. Network for Earth- mostly spared, with only six build- quake Engineering Simulation ings collapsed. On November 9th at The largest earthquake recorded Operations Center, NEEScomm. 9:23 pm local time, a Mw 5.6 earth- in the region during the last cen- In preparing this report, as ref- quake struck about 10 km SW of Van tury was the 1976 M7.2 Çaldıran erenced in the text, the writers (AFAD, 2011). This event claimed 40 earthquake (Gulkan et al., 1978), benefited from the work of several lives and caused further damage in although the 1945 M5.8 Çatak, of their colleagues who visited the city of 332,000, including collapse 1972 M5.2 Van, and the 1977 M5.1 the disaster area and generously of 25 buildings, most of which had Erciş were significant as well. shared information. In conducting been condemned following the earlier the field work, the authors also earthquake. The strong ground motion station in collaborated with the Earthquake Van did not record the ground shak- Engineering Research Center Seismotectonics ing from the October main shock of the Department of Civil Engi- due to a malfunction. The closest neering at Middle East Technical The region around Van has a complex station that recorded motion was University, and the Van Yüzüncü seismic setting due to the interaction the station (38.99011N, Yıl University. See acknowledge- between the Arabian and the Eur- 43.76302E), approximately 40 km ments for a complete listing. asian tectonic plates. The area has NNE of the epicenter. Unprocessed east-west thrust fault zones, as well records (Figure 4) indicate peak

Figure 1. Above: Van and Erciş are located in eastern Turkey by (source: BBC, 2011). Right: fault zones around Lake Van (source: METU, 2011a).

1 EERI Special Earthquake Report — April 2012

buildings; 19 of the 81 provinces were chosen as pilot districts to implement the 2001 law, but Van was not one of them. Only on January 1, 2011, did the Build- ing Control Law go into effect in Van province, along with the rest of the country. It is believed that the 2001 Building Control Law will help improve the quality of construction as it estab- Figure 2. Fault model for the Mw 7.1 earthquake lishes control authorities showing the hypocenters of the main shock and and prohibits unchecked aftershocks (source: Hayes, 2011). construction. There are ground accelerations were 178 gal reports that 89 of the (N-S), 170 gal (E-W), and 80 gal the top two highest seismic hazard buildings in Van province that were (U-D). Acceleration response spec- zones in Turkish earthquake-resistant built under the Building Control Law tra are shown in design. The northern and southern performed very well (Eyidogan, Figure 5. It is believed that Erciş thirds of the province are considered 2011). had ground shaking stronger than as seismic Zone 1, with effective that in Muradiye. The station in Van PGA for design calculations equal to Building stock and construction recorded the November ground or higher than 0.4 g. Erciş is in the type. The dominant types in the motion with peak accelerations northern third of the province. The region are one- to nine-story rein- (unprocessed) were 148 gal (N-W), middle third of the prov- 246 gal (E-W), and 151 gal (U-D). ince, which includes Van, The original and processed records is considered as a seis- from the October main shock can mic Zone 2, with effective be found on http://eerc.metu.edu. PGA in the range of 0.3g tr/, and in METU, (2011b). to 0.4g. In 2000, following the Performance of Structures devastating 1999 Mar- mara earthquakes, The seismic hazard map of the Turkish state authorities province (Figure 6) comprises issued an ordinance on control of private build- ing construction. However, the ordinance was stricken down by the Constitutional Court within ten months, a setback that hindered estab- lishment of independent and effective control mechanisms. In 2001, the Turk- Figure 3. Distribution of slip over the estimated ish Government fault plane. The arrows indicate the direction passed a Build- Figure 4. Main shock uncorrected ground accel- of motion of the hanging wall relative to the ing Control Law eration records from the Muradiye station (source: footwall (source: Hayes, 2011). for privately built METU, 2011b).

2 EERI Special Earthquake Report — April 2012

forced concrete (RC) buildings with infill walls and moment-frame or moment-frame and structural walls, and one- to two-story bearing-wall buildings. Buildings constructed before 2000 tend to be one to four-stories high. The majority of the buildings have full or half-base- ments. With the population growth in the provincial capital (Figure 7) and the economic boom in the region over the last decade, many taller buildings of eight to 12 stories have been put up in Erciş and Van in recent years. In Erciş, where nearly 200 buildings collapsed totally or partially, most of the casualties occurred in the newer and taller buildings. We learned from local engineers that ready- mix concrete became available in the last 3-4 years, and that there has been widespread use of poorly washed sand and gravel from local creeks. Inspecting the debris of col- lapsed structures showed a number of deficiencies: insufficient confine- ment reinforcement and 90-degree hooks in transverse rebars, use of smooth rebars, improper splicing of column longitudinal reinforcing Figure 5. Response spectra for the N-S and E-W records for the Muradiye station motion (unprocessed) recorded during the main shock (source: bars, poorly graded concrete mix

METU, 2011b). design, and substandard concrete casting (Figure 8). Apparently, the

new RC structures were designed by engineers, but they had no qual-

ity control during the actual con- Zone 1 struction of the building. Our local Zone 2 Zone 3 contacts noted that the construction Zone 4 process had become so informal Zone 5 that land owners would hire sepa- rate local specialty workers (for

formwork, rebar placement, con- crete casting) and not even hire a

contractor to organize the process. Few, if any, inspections or quality

control checks were done. October 23rd Event

November 9th Event The widespread poor material use and construction quality do not

explain the variation in damage in the building stock. We observed

that a majority of the damaged buildings suffered from obvious

fundamental design errors such as inadequate lateral-load resisting Figure 6. Seismic hazard map for Van Province (MPWS, 1996) with systems, soft-story at ground level epicenters of the October and November earthquakes.

3 EERI Special Earthquake Report — April 2012

In most of these buildings, mez- zanine floors are constructed to provide additional office space. Typ- ically, the mezzanine floor slabs are wrapped around the elevator core, which typically consists of structural walls. The resulting extreme differ- ence in the stiffnesses rendered these structural walls vulnerable. Figure 11 shows a commercial building under construction in Erciş that sustained heavy damage in its core wall at ground story (including mezzanine level) and second story.

Residential buildings not on main streets are typically free of commer- Figure 7. Population growth in Van and Erciş, 1935-2008. cial units. These residential build- ings performed generally better (tall and open-front ground story for cial units at the first (ground) story and during the earthquake. Newer mass commercial use), mezzanine level by residential units at upper stories housing buildings north of Van built construction resulting in dispropor- (Figure 9), particularly along the main mainly by TOKI, the Housing Devel- tionate stiffness distribution and streets. In older mixed-use buildings, loading of structural elements at the lower stories are stripped of their ground story, floor torsion, flexible partition walls because of commercial block-infill joist floors, strong-beam/ activity. In newer mixed-use buildings, weak-column, pounding, and cap- lower stories are taller than upper sto- tive columns. ries and again lacking partition walls, particularly along the street side of The majority of the commercial the buildings. Such a structure can be activity in the region takes place in seen in Figure 10. mixed-use buildings. These struc- tures are occupied by the commer-

Figure 8. Examples of deficient reinforcement detailing, poorly graded con- Figure 9. Typical mixed-use build- crete mix, substandard concrete casting (photos: Alaluf, Dönmez, İrfanoğlu, ings in Van (top) and Erciş (bottom) METU group). (photos: Dönmez, İrfanoğlu).

4 EERI Special Earthquake Report — April 2012

opment Administration of Turkey, sustained no damage. Figure 10. Heavily damaged RC structure with tall ground story Even though there are no ground (photos: Dönmez, İrfanoğlu). motion records from Erciş or Van for the October event, it is believed that the levels of ground shak- ing in these towns were such that architectural infill walls, in general, assisted the actual lateral load- resisting structural systems in the buildings. However, in certain cases, these supposedly nonstruc- tural walls interacted with the struc- tural systems to the detriment of the latter by forming captive column conditions either from the beginning of (Figure 12) or during the dynamic response of the building (Figure 13). In adjacent buildings with insufficient separation, pounding damage was observed (Figure 14). Figure 11. Six-story + basement RC moment-frame and structural wall build- ing in Erciş. The relatively taller ground story has a mezzanine level wrapped The typical floor structural systems around the elevator core structural walls. The core structural walls parallel to in the RC structures are two-way the building front sustained heavy damage, but there was no damage in the slab systems and infill joist floor columns (photo: İrfanoğlu). systems (Figure 15), the latter being found mainly in buildings built in the last five years.

Figure 12. Erciş inter- There is a tendency to have struc- city bus terminal. Partial tural walls in five-story or taller height infill walls along buildings constructed recently. the side of the building at However, we observed new or the ground story caused under-construction buildings with a captive column condi- improperly positioned structural tion. The captive columns walls that resulted in floor torsion. failed and the building One nearly completed building was put out of operation (Figure 16) had a single large (photo: İrfanoğlu). structural wall on the back that cre- ated floor torsion, and block-infill

5 EERI Special Earthquake Report — April 2012

of the structure was determined to be around 30MPa with a Schmitt Hammer. Even though deformed bars were found in the columns, almost no confinement reinforce- ment was observed at the plastic hinge regions.

There are few buildings with structural walls. Except in some of the school buildings, which will be discussed below, structural walls were usually not adequately propor- tioned. Still, walls around elevator shafts did tend to reduce damage to the rest of the structure though they Figure 13. Hollow concrete masonry units failed during the earthquake and sustained considerable damage caused a captive column condition. The column experienced brittle failure in themselves. shear (photos: İrfanoğlu and Erdil). In Gedikbulak, a large village with 250 buildings located about 15 km NW of the October earthquake epicenter, a three-story primary school building with structural walls collapsed during the earthquake (Figure 18). The building, con- structed in 1988, had an approxi- mately 14.4m x 21.4m footprint. It had two 7.2m x 0.3m structural walls along its plan long direction built on one end of the building and only columns, some of which were captive, at the other end. Two 4.6m x 0.3 m structural walls bordering the stairwell, which was near the centerline, acted as the main lateral load-resisting elements in the plan short direction. The seven 0.5m x 0.3m columns around the building Pounding damage in Figure 14. perimeter were the sole vertical Van (photo: İrfanoğlu). structural elements along one half of the building front, although four joist floors which were too flexible 0.3m x 0.3m interior columns were compared to the columns. There also present. The collapse configu- was concrete crushing and/or shear ration evidenced floor torsion as it failure at lower ends of almost appeared that the building twisted every column at the ground story, in plan as it slumped. and flexural failure at the ends of the perimeter beam installed Buildings in low-income districts in along one edge of the building. the region are typically constructed Furthermore, cracks developed either with concrete masonry units Figure 15. Joist-floor system at column-joist floor interfaces at in buildings under construction as bearing walls for single-story every floor level, and a web of in Erciş. The infill used in the buildings, or with reinforced con- cracks spanning between column floor system in the top photo is crete frames with masonry infill for lines was observed in the joist floor concrete masonry block, on the up to two-story buildings. These slabs at every floor level (Figures bottom polystyrene foam (photo: structures commonly have light 16 and 17). The concrete strength İrfanoğlu). timber roofs with light gage metal

6 EERI Special Earthquake Report — April 2012

covers. Most of these structures survived the earthquake with no damage, an indication that their elastic limits were not exceeded. Buildings in villages are constructed typically as bearing-wall structures using concrete masonry units, bricks, or rocks. The typical fail- ure mode in these structures was out-of-plane wall failure due to the absence of diaphragms tying the walls together (Figures 19 and 20).

Seismically strengthened build- ings. There were several seismi- cally strengthened buildings in the area, three of which were visited by team members The Ziraat Bank building in Erciş survived the earth- Figure 16. Six-story commercial building under construction, column fail- quake with minor damage (Figure ures at the base on the first (ground) floor (photo: Dönmez). 21). The Kazim Karabekir Primary School in Erciş is a four-story rein- forced concrete frame/structural wall building with hollow clay tile walls used as partitions; it had been retro- fitted by replacing some of the hollow clay tile infill walls with reinforced concrete structural walls (Figure 22). A survey of the structure revealed that the material quality and the reinforcement detailing in the Figure 17. Damage in the perimeter beam, and the joist floors seen from original framing was inferior. During bottom and top of floor (photos: Dönmez and İrfanoğlu). the October earthquake, there was widespread partition wall damage in the structure. Beams supporting the stairs exhibited damage as well,

Figure 18. The three-story primary school in Gedikbulak village after collapse (photos: Erdil, İrfanoğlu).

7 EERI Special Earthquake Report — April 2012

ment was anchored. The damage pat- tern indicates that the added shear walls did not engage with the exist- ing structural frame right away.

The Van Figure 19. Typical out-of-plane failure of concrete masonry unit bearing Merkez İskele walls (photos: Alaluf and Bedirhanoğlu). District Boarding School classroom building is similar to the school in Figure 22. The two-story RC build- ing had diagonal nets of hairline cracks in some of its original struc- tural walls. There were also signs of separation between the added structural walls and the perimeter framing, particularly at the wall- upper beam interface.

Reportedly, the Van branch build- ing of the Turkish Central Bank had been reviewed for possible seismic strengthening, but that had been deemed unnecessary (Figure 23). This very irregular five-story build- Figure 20. Failure of mud brick bearing walls in Gedikbulak village ing has a three-story atrium and (photo: İrfanoğlu). wing walls at the perimeter as the primary structural elements of the system. No structural failure was seen in the system except for a crack on one of the beams; how- ever, extensive partition wall failure in the building disrupted services.

Government buildings. Structural systems in government buildings varied depending on the age of the structure. The spectrum stretches from stone bearing-wall systems in old structures to RC frames in relatively new structures, but RC frames predominate. They are sub- Figure 21. Local branch of Ziraat Bank which was seismically strength- ject to special regulations depend- ened before the recent earthquakes (photo: Alaluf). ing on the ministry with which they are associated. The Erciş Palace of Justice is an example of the old but it was due to discontinuity in the to the third story. There was sepa- bearing-wall systems. As shown flexural reinforcement. The exist- ration at the interface between the in Figure 24, cracks formed in the ing (original) structural walls of the added structural walls and the perim- walls and went through concrete structure in the N-S direction had a eter framing, particularly the beam to lintels within the walls. diagonal web of hairline cracks up which the added structural reinforce-

8 EERI Special Earthquake Report — April 2012

The Erciş District ’s Office, a campus of two- to five-story plus basement modern RC buildings with frame and structural wall systems, sustained no structural damage and only minor damage in nonstructural walls. The main problems were failed water tanks in the basement, collapsed archive support systems (Figure 25), and ceiling tile failure in the conference hall on the top floor.

Primary/secondary educa- tion buildings. Primary schools inspected in both Van and Erciş had Figure 22. Kazim Karabekir Primary School in Erciş. Interface crack in-plane and out-of-plane damage between the added RC structural wall and the original beam (photo: to the hollow clay tile infill walls and İrfanoğlu). minor structural damage in some of them. Many schools had cracking at the frame/wall interface, diagonal cracks, through-cracks, and crush- ing of the infill walls. Out-of-plane gable wall failure was typical in all of the schools that were observed throughout the area.

At the Cumhuriyet Primary School (Figure 26), a conglomeration of four-story buildings located approximately 0.5 km southeast Figure 23. The Van branch building of the collapsed building (“Sevgi of the Turkish Central Bank (photos: Apartmanı”) that claimed 50 lives in Dönmez). the city center of Erciş, there was only minor damage to the structural system. Except for a corridor in which the infill walls were crushed, minor to moderate damage to the infill walls with cracking at the wall/ frame interface.

At the Atatürk Primary School, approximately 0.5 km northwest of the “Sevgi Apartmanı” building, a three-story plus basement school building had out-of-plane gable wall failure, but the building was in a good enough condition for the military to use it for response opera- tions.

Near the Van pier, a three-story building used as a teachers’ resi- dence had out-of-plane wall failure.

◄ Figure 24. Severe damage to walls of Erciş Palace of Justice (photos: METU group).

9 EERI Special Earthquake Report — April 2012

damage at the stairs and some beams in the structure (Figure 28). A few buildings were deemed to require further investigation, while others had only a little nonstruc- tural damage with some cracking at the wall/frame interface.

The Medical School of the univer- sity is not on campus but in Van, with seven main blocks intercon- nected to each other. Four of these blocks were built 60 years ago. Figure 25. Archives in Erciş District The other blocks were built in the Governor’s Office (photos: İrfanoğlu). last 15 years. All of the blocks are RC structures with two-way slab detailing, sustained the most damage systems. Even though the over- Above the third-floor ceiling slab, and is currently out of commission. all construction quality was not the structure consisted of wood The damage appeared to consist good, no structural damage was post/beam framing supporting the primarily of an out-of-plane collapse observed in the buildings after the roof with a story height equivalent to at a gable wall and the exterior wall October earthquake. However, the floors below. The wood-framed at the exit stair well, heavy structural portion had wood diagonals placed systematically to provide stability. On two sides of the structure at the uppermost story, out-of-plane wall failure was observed (Figure 27).

Universities. Van Yüzüncü Yıl University, established in 1982 with a population of approximately 18,500 students, is located approxi- mately 20 km from the epicenter of the October earthquake. Overall, campus buildings performed well from a structural point of view, with most buildings having only minor Figure 27. Teachers’ housing in Van near the pier (photo: Hernandez). to moderate damage to the hollow clay tile infill walls. The library, a five-story building with an irregu- lar footprint and obvious inferior concrete quality and reinforcement

Figure 26. Cushed infill walls in Figure 28. Library building at Van Cumhuriyet Primary School in Erciş Yüzüncü Yıl University (photos: (photo: İrfanoğlu). Dönmez).

10 EERI Special Earthquake Report — April 2012

due to medium to heavy end ornaments, and toppling of the nonstructural damage, minaret just above the transition the university hospital segment. Figure 32 shows that the was out of service. The toppled minaret punched through nonstructural elements the elevated concrete slab above and contents that the ablution room. rendered the hospital inoperable were partition At the Salihiye Mosque on the north walls, molded gypsum side of Erciş, one of the minarets ornaments at the ceil- failed just above the transition zone ings, wall panels on top and toppled southward, damag- of partition walls, and ing the dome of the prayer hall. laboratory glassware The other minaret showed signs of and hardware. The uni- incipient failure well above the tran- Figure 29. Mosque under construction in Erciş versity was closed until sition zone. In Gedikbulak, the local (photo: İrfanoğlu) February 2012. Bülbül Mosque collapsed totally (Figure 33). Mosques. Construc- tion of the mosques we Hospitals and health-care observed generally con- facilities. There is a wide range sisted of concrete frame of hospital buildings in both Erciş with hollow clay tile infill and Van. The Erciş State Hospi- in the prayer hall and tal, constructed in 1962 with five stone cladding in the buildings, has a current capacity minarets (see Figures 29 of approximately 134 beds. It is and 30). located approximately 100 m from the Sevgi Apartmanı. Most buildings The prayer halls of most on the campus are two- to three- mosques in Erciş and story reinforced concrete build- Van had little damage, ings. Discussions with the security largely minor diagonal personnel at the site revealed that cracks and/or cracks the building was red-tagged despite at the interface of the only minor damage. The hospital infill/frame, although could have played a significant part the Kara Yusuf Paşa in the emergency response had it mosque in Erciş sus- been open. It received a green tag tained heavy damage in approximately one week after the Figure 30. Typical Turkish minaret (source: its main structure (Figure earthquake, and was scheduled to Doğangün et al., 2006). 31). Minarets had dam- resume normal operations shortly ages that ranged from thereafter. minor cracks in the shaft, toppled The Van State Children’s Hospi- tal, located in the heart of Van and 22 km southwest of the epicenter of the Octo- ber earthquake, consisted of several three-story buildings constructed in the 1960s. Most of the buildings had minor to moderate cracking at the frame/wall interface, with some local damage due to pound- ing and infill damage. Like its counterpart in Erciş, the entire campus was closed to patients. Figure 31. Kara Yusuf Paşa Mosque in Erciş, built in 1974 (photos: METU group)

11 EERI Special Earthquake Report — April 2012

those bins were filled to capacity, while the others were not. The col- lapsed bins also appeared to have a cross-bracing design that did not continue to the base of the verti- cal legs, possibly causing them to buckle under the seismic movement (Figure 34).

A concrete ready-mix plant, located just down the street from the flour manufacturing plant, also had a storage bin collapse. The bin was situated atop a small concrete structure, and the vertical legs, and Figure 32. Toppled minaret at Vanyolu Mosque (photo: Hernandez). possibly the cross-bracing, buckled. The only two large industrial facili- our discussion with hospital ties in Van Province are the Van personnel, there was only Cement Plant in Edremit and Erciş minor nonstructural damage, Sugar Plant. The cement plant had such as some interior wall no damage. The Sugar Plant--a cracks and a partial collapse complex of buildings consisting of of a small portion of the concrete frames with hollow clay exterior façade. There was tile infill walls, steel braced frames, no interruption in service and prefabricated concrete frame and the hospital cared for buildings--sustained damage only injured immediately after the to its administration building. A earthquake. three-story concrete frame structure with a daylighting basement had Figure 33. The Bülbül Mosque in Gedikbulak Industrial facilities. Indus- some unreinforced hollow clay tile village (photo: Alaruf). trial facilities in Van and facades fall off at several locations Erciş had little to no damage and moderate infill wall damage. to the primary Neither of the plants had direct structures. The business interruption. The cement industrial area plant was going through its regular of Van is located annual maintenance, and was not in approximately operation. The sugar plant stopped 20 km SW of production, though it was operable, the epicenter because it was selected as the of the October warehouse/center for aid distribu- earthquake. The tion to earthquake victims. building inventory includes precast concrete buildings, Lifelines concrete frame

with hollow clay Highway roads and bridges. Figure 34. Collapsed bin at flour plant in provincial Except for the road surface defor- tile infill walls, and capital Van (photo: Hernandez). mation near the epicentral region steel-framed build- on the main Van-Erciş highway, no The Van State Hospital is a state- ings. Damage appeared to be limited impact on the roads was observed. of-the-art facility with several build- to a few non-building structures such According to reports from the ings that opened approximately as storage bins. one year ago. Located 30 km METU-EERC and civil engineering teams (METU, 2011a), damage in southwest of the epicenter of the At a flour manufacturing plant, two the 14 reinforced concrete highway October quake, it has deep foun- out of six small grain bins constructed bridges inspected by the teams was dations supporting the reinforced of light steel framing collapsed. Dis- minimal. Most of the bridges were concrete frame building. Based on cussion with the owner revealed that made of reinforced concrete piers

12 EERI Special Earthquake Report — April 2012

some glazing failure in the tower, so the operation was moved to the fire brigade building and func- tioned without disrup- tion. At the time of the earthquake, the main terminal building was being dismantled to move the airport oper- ations to the adjacent new terminal building (Figure 35). Services Figure 35. The old and the new terminal buildings at Van were provided from airport (photo: İrfanoğlu). the old building until the November earthquake, which caused struc- tural damage to it (Cebeci, 2011).

Lake Van port. The port is con- structed of deep foundations sup- porting a 84-m long by 14-m wide concrete pier (Figure 36) structure with a capacity to dock one large cargo ship on one side of the pier and several small to medium-sized boats on the other side of the pier. From our preliminary observations, there were two transverse cracks in the pier, the first near the begin- ning and the other approximately 20 m from the end. The crack near Figure 36. Pier at Van (photo: Hernandez). the beginning of the pier did not continue through the width of the supporting prestressed reinforced pier, but the crack near the end concrete I-beams, multi-span was a through-width crack. At the with 17.5m typical span length, time of our observation, the pier supporting the deck. Permanent was in use and the damage did not transverse offsets on the order of appear to affect its operation. 1-2 cm were observed at expan- sion joints, as were spalling and A prefabricated one-story structure cracking in concrete transverse used to house passengers had shear keys. Flexural cracks were minor cracking due to the spread- observed in the piers of the Ben- ing of soil in the vicinity. Small dimahi bridge near Erciş. sand boils and spreading cracks approximately 2.5 to 3 cm wide . No damage was Railroad system were visible. reported in the railroad system, including in the railroad tunnel Electrical transmission sta- in the epicenter region (METU, tions. Observations were made 2011a). at two local electrical substations in the affected area near the city Airport. In the wake of the Octo- Van. The 154-kilovolt transformer ber quake, Van airport was put center approximately 22 km south- back into full service after a quick west of the October earthquake inspection of the runway and the epicenter, which receives power Figure 37. Transverse crack at navigation equipment. There was the Van pier (photo: Hernandez). from the local dam, had no

13 EERI Special Earthquake Report — April 2012

damage although service to Van winters are harsh in the disaster Norway, Pakistan, Qatar, Roma- was interrupted for approximately region, over 336,000 blankets, nia, Russia, South Korea, Spain, 3 hours due to downed power 7,192 sleeping bags, and 27,500 Sweden, Switzerland, Syria, lines in the city. A new 380/154/33 heaters were distributed. Approxi- Tunisia, Turkish Republic of kilovolt capacity electrical substa- mately 20,000 people were living Northern Cyprus, Ukraine, United tion located approximately 20 km in 14 tent cities built in Van and Kingdom, the USA, and the UN southwest of the October epicenter Erciş. Nearly 3,500 tents were HCR and OCHA. was under construction and sched- distributed in the countryside to uled to open in November 2011. shelter animals. Over 132,000 Economic Impact and The substation reportedly had no meals were being distributed daily Long-term Recovery damage except some transformers in three central cafeterias and 35 that shifted during the earthquake. schools in the region. At the end Final damage surveys were The opening of the substation was of January, 19,500 inhabitable carried out by 1,136 technical delayed until the transformers were containers were providing shelter personnel between November repositioned. to around 100,000 people (Uras, 15th and December 22nd in Van, 2012). In Erciş, eight soup kitch- Erciş, and several other districts Emergency Response ens served 81,000 meals a day; in and villages in the province. As Van, three soup kitchens served noted in Table 1, of the 103,478 In the October quake, 604 people 60,000 meals each day. Addition- buildings inspected, 33,016 were killed, about 2,600 were ally, 72,000 meals were purchased buildings (31.9%) were either injured, and 222 were rescued from daily by the state from private heavily damaged or collapsed, collapsed structures during emer- providers. 4,755 buildings (4.6%) sustained gency response operations. During medium-level damage, 35,545 the November quake, 40 people The domestic and international buildings (34.4%) sustained light lost their lives and 30 were rescued monetary aid is estimated to be damage, and 30,162 building from collapsed structures near Van. TL 411 million (approximately US$ (29.1%) had no damage. 220 million). Aid was received from According to the Turkish Disas- Algeria, , Australia, Azer- Of the 14,904 commercial units ter and Emergency Management baijan, Bahrain, Belgium, Bulgaria, inspected, 2,440 (16.4%) were Directorate (AFAD), within hours Canada, Egypt, Finland, France, either heavily damaged or col- of the October earthquake, nearly Germany, , Ireland, Israel, Italy, lapsed, 2,748 (18.4%) sustained 500 search and rescue, emer- Japan, Jordan, Kazakhstan, Kyr- medium damage, 6,307 (42.3%) gency medical and Red Crescent gyzstan, Malaysia, Netherlands, sustained light damage, and 3,409 personnel arrived at the disaster region. They were mobilized from Table 1. Damage distribution in inspected buildings (VSG, 2011) 48 provinces and 39 agencies. As of mid-December, 5,267 search and rescue and 2,976 medical person- Damage Level Number of Buildings Fraction of Total nel from state, provincial, non- Heavy/Collapse 33,016 31.90% governmental, and Turkish Armed Medium 4,755 4.60% Forces had been to the disaster area. Nearly 200 ambulances, Light 35,545 34.40% including 18 medevac helicopter, None 30,162 29.10% supported the emergency medical TOTAL 103,478 teams. Six field hospitals were con- structed. Nearly 200 social work- ers, sociologists, psychiatrists and Table 2. Damage distribution in inspected commercial units (VSG, 2011) consultants participated in providing psycho-social support. Damage Level Number of Buildings Fraction of Total Heavy/Collapse 2,440 16.40% As of mid-December, 73,679 tents (29,222 from abroad), 260 pre- Medium 2,748 18.40% fabricated housing units, 10,155 Light 6,307 42.30% inhabitable containers, and 3,794 None 3,409 22.90% specially designed prefabricated TOTAL 14,904 houses were sent to the region. As

14 EERI Special Earthquake Report — April 2012

(22.9%) were not damaged (Table walls are engaged in the structural should be a higher level of licens- 2). response even though, at least ing required for essential facilities on paper, they are supposed to such as hospitals and schools. The total cost of this earthquake, be nonstructural elements. While Those facilities should be designed including financial losses and these infill walls may sustain little only by experienced professional reconstruction efforts, is expected damage at low-intensity shaking, engineers specialized in structural to be US$ 500 million to 1 bil- during stronger shaking they are and earthquake engineering. lion. To help with the recovery, damaged, typically in the form of Industrial losses also were avoid- over 2,250 permanent houses widespread cracking and even able. Many of the losses were are expected to be completed by crushing. Crushing of the infill walls caused by equipment (tanks, ves- August 2012. Additionally, founda- can cause development of cap- sels, etc.) not properly anchored tions of 3,000 new housing units tive column condition which often or braced to resist seismic loads. were laid with initial work started results in premature and brittle Such damage is well understood on additional 12,000 housing units, failure of the affected columns. by professionals and can be miti- all supported by the government. gated at a modest cost. Distribution of these residences is Given the observed high frequency expected to start in August 2012. of destructive earthquakes in Post-earthquake damage Turkey, it would be worthwhile to assessment. Given the frequency Recommendations revisit the code-specified drift ratio of large earthquakes affecting limit considered for the design level urban areas in Turkey, it behooves Without doubt, Turkey has the earthquake. It is possible that low- the Turkish government to orga- structural engineers, seismic ering this drift limit would provide nize teams of experts that could design code, and construction long-term benefits that outweigh be called to duty on short notice know-how to avoid earthquake the short-term costs. to assist with the rapid inspection disasters of this scale, but the of the buildings. Such teams could damage clearly demonstrated that Several buildings inspected by the be supported by professional civil improvement is needed in pre- EERI team had improper designs. engineering associations. It is not earthquake mitigation. Almost all It could be that the code-enforce- reasonable to expect that local of the damage and deaths were ment bodies, which are the local engineers, who are very likely to caused by the collapse of inad- municipalities in the case of private be in crisis themselves, could bear equately designed and constructed buildings, are unable to provide the burden of inspecting buildings buildings, particularly buildings built competent reviews and/or enforce in their hometowns. Hospitals, during the last decade. Modern the seismic design code. If it is the emergency operations, and water buildings should have had light latter, the state has to provide the and power distribution networks damage, given that the shaking necessary checks and balances of need continuity of operations after intensities were moderate, but enforcement. However, if it is the an earthquake, so these facilities they were not properly designed former, the technical skills of the should be surveyed and tagged and/or not properly constructed. engineers reviewing the design right away. Instead, the inspec- Obviously, an advanced building need to be improved and a mini- tion process was quite slow and seismic design code does not guar- mum level needs to be established. apparently resulted in much confu- antee good performance of build- A licensure process in engineering sion. We visited several structures ings and their contents. should be introduced to provide where the residents claimed that a formal and objective means to some people without official or Design and construction. The assess the competency of design technical credentials inspected building designs in seismic regions engineers. This would result in their buildings, made statements, such as Van region should incor- establishment of engineering and gave ratings regarding the porate more and better distributed cadres that could protect the safety, damage state and soundness of structural walls. Building lateral well-being and other interests of the the structure. No documentation load resisting structural systems general public. Professional engi- was left about those statements and the infill walls should be neering certification ensures that and ratings. designed considering their possible practicing engineers meet require- interaction during earthquakes. In ments for competency in their A post-earthquake building tag- the current construction style the specific disciplines, and that engi- ging system should be developed, unreinforced infill walls are wedged neering designs are reviewed and and engineers trained in “rapid between and flush with the struc- accepted by qualified practicing visual screening methods” should tural elements. As a result, the infill engineers. Such a licensure pro- be dispatched to the field immedi- cess is being developed, but there ately after the earthquake to inform

15 EERI Special Earthquake Report — April 2012

building occupants whether it is East Technical University (Center gov.tr/sarbis/Shared/WebBelge. necessary to evacuate their build- Director Prof. Sinan Akkar and aspx?param=103 ing or it is safe to go back inside. Department Head Prof. Güney If the occupants of a damaged Ozcebe), the Van Yüzüncü Yıl Uni- Doğangün, et al., 2006. “Tradi- building are evacuated, it is also versity (Rector Prof. Peyami Battal), tional Turkish Minarets on the necessary to decide if and when and the Turkish Gendarmerie Force Basis of Architectural and Engi- emergency workers can enter the (Lt. Gen. Yıldırım Güvenç, Col. neering Concepts”, Proceedings building for search-and-rescue or Rıza Yaşar, Cpt. Emre Güneri, Mr. of the 1st Int. Conf. on Restora- temporary shoring missions. Such Kemal Balyemez, and Mr. Cenk tion of Heritage Masonry Struc- a system is already in place in the Soysal). The EERI team mem- tures, Cairo, Egypt, April 24-27 USA, and has been used after bers are thankful to Profs. Ayşegül many earthquakes with great suc- Askan, İdris Bedirhanoğlu, Halis Eyidoğan, H., 2011. Interview cess. Usually, the local government Bettemir, Barış Binici, Alp Caner, (in Turkish). http://www.t24. body responsible for enforcing the Altuğ Erberik, Nuretdin Kaymakçı, com.tr/vandaki-felaket-herke- building safety code examines Arda Özaçar, Afşin Sarıtaş, Ahmet sin-gozu-onunde-insa-edildi/ the affected structures and tags Yakut, Tolga Yılmaz, Dr. Mehmet haber/184080.aspx. Accessed them as appropriate, but often they Çelebi (USGS), Dr. Hasan Korucu on November 29 receive volunteer support from pro- (Cpt., Turkish Naval Force), and fessional structural and earthquake graduate students Vesile Hatun Gedikbulak Primary School, engineering associations. Akansel, Alper Aldemir, Özer Ay, 2011. http://okulweb.meb.gov. Ozan Demirel, Barış Erdil, Özkan tr/65/01/245329/contact.htm . Hazard mitigation. To develop a Kale, Sercan Kıran, Öner Mete, Governor’s Office of Van Prov- program for seismic risk mitiga- Dilek Okuyucu, Ramazan Özçelik, ince, 2011. http://www.van.gov.tr/ tion, the potential risks need to be and Cihat Yıldırım for their help and default_B0.aspx?content=1006 . quantified as accurately as pos- generosity. sible. Compared to other earth- Gülkan, P., Gürpinar A., Çelebi, quake vulnerabilities, buildings This report was edited by Sarah M., Arpat E., and Gençoğlu S. pose the largest risk to life, limb, Nathe, EERI Newsletter Insert (1978), Engineering Report on property, and economic welfare. Editor. the Muradiye-Caldiran, Turkey, To determine vulnerable buildings, Earthquake of 24 November a countrywide building inventory References 1976. Published by the National needs to be prepared, starting from Academcy of Sciences, Wash- the heavily populated earthquake- British Broadcasting Corporation, ington D.C., 59 pp prone city centers. This inventory 2011. “Turkey earthquake: Survi- should be prepared based on rapid vors outdoors on freezing night.” Hayes G., 2011. “Finite Fault engineering assessments for every http://www.bbc.co.uk/news/world- Model, updated result of the building, so mitigation priorities can europe-15437014 Oct 23, 2011 Mw 7.1 eastern be set. Afterwards, an earthquake Turkey earthquake,” http://earth- mitigation plan for the entire coun- Cebeci, Uğur, 2011 Hürriyet News- quake.usgs.gov/earthquakes/ try should be prepared. paper: http://www.hurriyet.com.tr/ eqinthenews/2011/usb0006bqc/ yazarlar/19229796.asp finite_fault.php, Acknowledgments Disaster and Emergency Man- Middle East Technical Univer- The EERI team members are agement Directorate (AFAD), 9 sity Earthquake Engineering thankful for the support provided November 2011 Earthquake Bulletin Research Center, 2011a. “The by the EERI Learning from Earth- (in Turkish). http://www.deprem. 23 October 2011 Mw 7.2 Van quakes program, İzmir Institute gov.tr/sarbis/Shared/WebBelge. earthquake seismic and struc- of Technology, and the George E. aspx?param=103 . Accessed 29 tural damage field observations Brown, Jr. Network for Earthquake November 2011 (in Turkish),” Report # METU/ Engineering Simulation Operations EERC 2011-04, http://www.eerc. Center, NEEScomm. They are Disaster and Emergency Man- metu.edu.tr/sites/default/files/ also grateful for the logistical and agement Directorate (AFAD), Van_ODTU_DMAM_Rapor.pdf technical support they received 2011. “Press Bulletin on the 23 from the Earthquake Engineering October and 9 November 2011 Middle East Technical Univer- Research Center of the Depart- Van earthquakes (in Turkish),” 24 sity Earthquake Engineering ment of Civil Engineering at Middle November. http://www.deprem. Research Center, 2011b. “Pro-

16 EERI Special Earthquake Report — April 2012 cessed Mainshock Accelerometric Recordings of the 23 October 2011 Van Earthquake,” Report # METU/ EERC 2011-02, http://www.eerc. metu.edu.tr/sites/default/files/ Report_Van_EQ_2011.pdf

Mineral Research and Explora- tion of Turkey (MTA), 1992. “Active Fault Map of Turkey,” prepared by F. Saroglu, O. Emre, and I. Kuscu. http://www.mta.gov.tr/mta_web/ dirifay1.asp

Ministry of Public Works and Settle- ment (MPWS), 1996. “Seismic Hazard Map,” http://www.deprem. gov.tr/sarbis/shared/DepremHarital- ari.aspx

United States Geological Survey (USGS), 2011. “Bulletin for the 23 October 2011 magnitude 7.1 eastern Turkey earthquake,” http:// earthquake.usgs.gov/earthquakes/ recenteqsww/Quakes/usb0006bqc. php

Uras, G., 2012. Article on economic impacts of Van earthquake, http:// ekonomi.milliyet.com.tr/8-290- vanli-cadirda-isi-gunduz-eksi-5- gece-eksi-14-derece/ekonomi/ ekonomiyazardetay/03.02.2012/1497350/ default.htm,

Van Provincial Office of the State Religious Affairs, 2011. http://www. vanilmuftulugu.gov.tr/cami_koy.htm

Van Sesi Gazetesi (VSG), 2011. Article on building damage in the Van earthquake, http://vans- esigazetesi.com/haber_detay. asp?haberID=1193,

17