Failure Analysis and Design Improvement Proposal for Flood- damaged Bridges in Papua

Gibson Ali HOLEMBA Candidate for the Degree of Master of Engineering Supervisor: Professor Takashi MATSUMOTO Division of Engineering and Policy for Sustainable Environment

Introduction people as a by-product of the water cycle process. The only way out to reduce or control and provide a The climatic effect of flood against the road infrastructure sustainable solution is an innovative way of engineering such as a bridge is so prevalent that it requires deeper and technology and better flood mitigation planning and engineering and technological intervention to address this control works. Fig. 1 shows 5.0m of road approach and ever-present phenomenon. has been bridge abutment of Pine Tops Bridge damaged by the experiencing frequent bridge failures and collapses due to flood in April of 2017. flooding rivers in the recent past. According to the internal records from Papua New Guinea Department of Works, it has shown that over Two Hundred and Eighty (285) bridges, fords (causeways) and major culverts were damaged by flood action alone in the last Five years. That is at a rate of 57 bridges in a year and this result is very staggering. Bridge damages have been observed to be mainly at the bridge foundations. More specifically, the flooding waters erode the bridge abutments, scour the bridge piers and weaken the bridge’s resistance against the flood loads and eventually destroy the bridge. In addition, it is also Figure 1. Flood-damaged Pine Tops Bridge, Wau attested that riverbank and road approach embankment Highway, , Papua New Guinea. erosion by flooding rivers have been one of the leading causes of bridge failures in Papua New Guinea, according Field Investigation and Data Collection to this study.

The bridge inspections were carried out in three Field investigations were carried out in Papua New provinces of Papua New Guinea, namely; Morobe Guinea at twenty-one flood-damaged and affected bridge Province, Province and Province. In sites. The bridges that were investigated were all Morobe Province, five (5) number of bridges were constructed over natural river crossings in three investigated, three (3) bridges along Wau Highway, one distinctive provinces in the country. The field (1) bridge along and one (1) bridge investigation works gathered field data such as, river along Highway. In , eight (8) channel width, bridge dimensions, river cross-sections, bridges were inspected and all were along the Ramu flow depth, scour depth, flow angle, clearance height Highway section of Madang Province between (freeboard), debris and log sizes. Inspections were carried Pompaquato Bridge and Usino Junction. Moreover, in out on both superstructure and substructure with damages , eight bridges were inspected, three the floods have caused on the bridge. (3) bridges along Boluminski Highway, two (2) bridges The river cross-section was measured manually by along Lanzarote Road and three (3) bridges along West using a measuring tape. The width of the main river Coast Road. All in all, twenty-one (21) bridges were channel was measured from the top of the east bank to the inspected in three different provinces along six key socio- top of the west bank in three different locations upstream, economic roads that support the livelihood of people in at bridge and downstream. A 10m of offset distance was Papua New Guinea. taken from the centerline of the bridge both upstream and Davies (2016), reported that Papua New Guinea is downstream from the bridge. In addition, the average vulnerable to both inland and coastal flooding. The river channel depth was measured with survey stuff at 3m intervals across the main channel in accordance with the country has suffered from severe coastal flooding in 2008 respective channel widths. in which as many as 75,000 people were displaced from The general information of these twenty-one inspected eight (8) different provinces. In 2016, around 10,000 bridges is provided in Table 1. These bridges have fallen people were affected by flooding in West victim to flooding sustaining major structural damages Province while thirty-five (35) houses, bridges, roads and while several bridges were destroyed by flood as agricultural farms were damaged across both provinces of discussed in the following chapters. Most of the studied Gulf and Southern Highlands [2]. rivers had trapezoidal channels while few were Rain and its effect of flooding are a natural rectangular open channels, especially those that have non- phenomenon and are here to stay whether we like it or not. erodible bank slopes. Fig. 2 shows the river channel cross- Flooding will continue to affect the livelihood of the sectional profile as measured in this study. Table 1. General Information of Inspected Bridges in Papua New Guinea

Q100 Peak Flow Bridge Bridge Bridge Structure Catchment Design Velocity @ No. Bridge Name Length Width Span Type Size Discharge Q100 (m) (m) (No) (km2) (m3/s) (m/s) 1 Asas Bridge 40.0 3.72 1 Bailey Truss 11.39 64.71 0.97 2 Aumea Bridge 56.0 3.40 2 Bailey Truss 68.63 193.53 1.28 3 Bora Bridge 48.7 4.34 1 Bailey Truss 211.00 345.95 2.90 4 Cedar Bridge 35.7 7.50 3 Beam/Slab 812.43 627.91 2.64 5 Daulom Bridge 36.6 3.15 1 Bailey Truss 224.09 606.49 3.00 6 Himutu Bridge 30.84 3.22 1 Bailey Truss 41.40 199.90 1.57 7 Iruan Bridge 124.97 5.20 3 Bailey Truss 90.43 348.67 1.20 8 Kalili Bridge 21.3 3.14 1 Bailey Truss 20.00 185.98 1.31 9 Kesuai Bridge 73.5 3.55 2 Bailey Truss 56.31 171.53 0.79 10 Bridge 21.5 3.15 1 Bailey Truss 0.07 4.07 1.44 11 Marakalang Bridge 37.0 3.40 1 Bailey Truss 17.20 169.63 1.35 12 Mea Bridge 146.3 3.64 3 Bailey Truss 35.22 128.83 0.71 13 Menia Bridge 45.7 4.40 1 Bailey Truss 40.29 126.00 0.73 14 Pine Tops Bridge 27.4 3.75 3 Beam/Slab 531.62 624.50 5.68 15 Punam Bridge 35.2 8.49 8 cells Arch Culvert 9.50 81.44 0.91 16 Rumu Bridge 30.1 7.38 3 Beam/Slab 325.46 401.69 4.91 17 Sausi Bridge 137.2 4.40 4 Bailey Truss 78.43 209.95 0.53 18 Surinam Bridge 49.5 4.20 3 Warren Truss 280.57 411.62 2.59 19 Wara Pita Bridge 33.0 3.10 6 cells Arch Culvert 11.47 91.37 1.85 20 Waterbung Bridge 36.8 4.07 3 Beam/Slab 91.79 114.29 1.14 21 Yakumbu Bridge 46.3 4.72 1 Bailey Truss 21.56 95.50 0.80

Figure 2. River Channel Cross-section. Figure 3. An example of Google Earth Pro® image of a plotted catchment area. Watershed and River Morphology The catchment areas were automatically calculated by Watershed is the land area or ridge that separates surface the software and were used for flood design estimations. run-offs from precipitations such as rainfall into different The accuracy of the calculations is limited to the accuracy river basins, lakes or ocean. The river morphology is the of the software used and as such the data used in this study study of river shapes or forms with respect to time. It is is for this purpose only and should not be used for design referred to as fluvial morphology in the study of purposes. It is highly recommended that adequate hydrology and hydraulics. investigation must be carried out using the topographic River morphology assessments were undertaken by contour maps or the hydrographic charts when visual inspection within the bridge periphery while undertaking design for these studied bridges for upstream and downstream environment were studied with permanent works. the use of drone survey. Mavic DJI Pro® drone was used to undertake the aerial survey by taking photographs Bridge Failure Assessment Method along the stream length with short video recordings of the river flow characteristics. From the field investigations, bridge failure analysis was The aerial photographs were taken at 200m-500m undertaken to correctly identify the resultant factors spacing both upstream and downstream. The catchment associated with a flood that caused the damages. The main area of the river from the bridge site was estimated using cause of the bridge damage was flooding, however, other the Google Earth Pro© software. The catchment size was factors such as scouring, bank erosion, debris impact, and determined by plotting the lines along the ridge dividing blockages were the root causes of the bridge damages the watershed as shown in Fig. 3. made possible by flooding rivers. Most of these phenomena are associated with flooding and as such a bridge failure analysis was required. LEGEND: Flood Level 1 = Main Failure Cause Level 2 = Structural Failure Cause Level 3 = Principle Failure Cause Level 1 Level 4 = Root Failure Cause

Which part of the bridge was damaged by flood?

Level 2 Substructure Failure Superstructure Failure

What are the Principle Causes of Bridge Failure? Contraction Hydrodynamic Level 3 Local Scour Debris & Logs Hydrodynamic Scour Loads Sedimentation Overtopping Accumulation Loads

Vertical Road What are the Main Foundation Blocked Deck Foundation Approach Root Causes of Structure Failure Bed Aggradation Waterway Displacement Failure Damage Bridge Failure? Damage Level 4

Ancillary Lateral Joint Reduced Direct Impact Structure Foundation Embankment Movements Bank Erosion Freedboard Damage Damage Failure Failure

Figure 4. Bridge Failure Assessment Method

Flooding causes damage to bridge superstructure and Results of Bridge Failure Analyses substructure which makes up the bridge. Therefore, the failure factors were further categorized under these two The peak design discharges from flood estimation sub-categories related to the structure. Thirdly, the failure provided the failure analysis parameters such as flow analysis was further sorted into principle failure causes velocities, flow depths, and flow rates for bridge failure such as local scour at bridge piers and abutments, analysis. In summary, the study undertook various contraction scour at riverbed and riverbank, and analyses such as scour analysis [1], flood load analysis hydrodynamic load failures under substructure category. [4], hydraulic analysis, afflux estimation, sediment Furthermore, debris/log accumulation, sedimentation, deposition rates, log/debris accumulation, overtopping hydrodynamic loads, and overtopping were categorized estimation, and riverbank erosion calculations [3] to under the superstructure division as they relate to determine the key failure causes for the bridges. The superstructure failures. The failure assessment method bridge failure analysis further used the bridge failure used in this study is as presented in Fig.4. assessment method given in Fig. 4 to categorize the different failure causes to determine the bridge failure Flood Estimation factors in each bridge. The results summary of the analyses is presented in Fig. 6. To undertake the hydraulic analysis and calculation of Hence, it was observed in this study that substructure flood loads acting on the bridge during a flood event, damages due to flooding account for seventy percent flood estimation was necessary. This study used the Papua (70%) of the bridge failures while superstructure damages New Guinea Flood Estimation Manual (SMEC 1990) account for the thirty percent (30%). The common causes estimation procedures and empirical equations to estimate of flood-induced bridge failures were identified as local the peak design discharges such as Q20, Q50, and Q100. The scour around bridge piers and abutments, contraction Regional Flood Frequency Method (RFFM) from this scour along the riverbanks and beds, sediment deposition, guide [5] was employed to estimate the flood discharges debris/log accumulation, overtopping, and hydrodynamic loads such as buoyancy, uplift and moment generated by as shown in Fig. 5. lift and drag forces on the superstructure.

Q2 Q20 Q50 Q100 Flood Loads on Superstructure Sedimentation 700.00 Overtopping Flood Loads on Substructure 600.00 Debris & Log Accumulation Local Scour 500.00 Contraction Scour 400.00 7.00 300.00 6.00 5.00 200.00 4.00 3.00 100.00 2.00

Peack Peack Design (m3/s) Discarge 1.00

0.00 0.00

Number Number Bridge of Failure Causes

MeaBridge

Marakalang…

Asas Bridge Bora Bridge

MeaBridge

Iruan Bridge

Bora Bridge

Asas Asas Bridge

Sausi Bridge

KaliliBridge

Iruan Bridge

Sausi Bridge

Cedar Bridge Labur Bridge

RumuBridge

KaliliBridge

Menia Bridge

Cedar Bridge Labur Bridge

RumuBridge

Menia Bridge

Kesuai Bridge Punam Bridge

Aumea Bridge

Kesuai Bridge

Punam Bridge

Himutu Bridge

Aumea Bridge

Daulom Bridge

Himutu Bridge

SurinamBridge

Daulom Bridge

SurinamBridge

WarapitaBridge

WarapitaBridge

Yakumbu Bridge

PineTops Bridge

Yakumbu Bridge

PineTops Bridge

Waterbung Bridge Waterbung Bridge MarakalangBridge Figure 5. Peak Design Discharges (Q) Figure 6. Flood-induced Bridge Failure Causes Summary It was revealed in this study that flood-damaged spread 3.0m evenly on both ends of the bridge bridges fail due to multiple flood-induced bridge failure approaches. causes. On average, more than five failure causes were observed in each bridge which caused the demise of the 퐿 = 푇 + (퐹푝) + 6.0 (3.0) structure. Therefore, to mitigate these flood-induced bridge failure causes, bridge design improvement The final proposed bridge design improvement is to countermeasures were developed. adequately provide scour and stream instability

countermeasures against scouring, riverbed and bank Proposed Bridge Design Improvements erosion, debris control, sediment deposition, and many

others. The study adapted several scour and stream This study has proven that flood-induced bridge failure in Papua New Guinea is a major challenge and appropriate stability countermeasures developed and published by actions must be undertaken to protect the vital road government agencies and research organizations [3] that structures. During the design stage, possible flood risk were applicable to the challenges faced in Papua New assessments and damage causes must be investigated, and Guinea as observed in the study. satisfactory countermeasures must be included in the design to protect bridges against flood damage during the Cost-Benefit Analysis service life of the structure. The first design improvement proposed in this study is To further validate the application of these proposed to increase the foundation depth by a meter to allow bridge design countermeasures, a cost-benefit analysis bridge foundations further below the scour prism to was undertaken on Rumu Bridge as a case study. The counter the effect of local scour on bridge abutments and analysis proved that implementing the proposed flood- piers by using Eq. (1.0). Where Zd is total design resistant bridge design improvements will reduce the foundation level, Z0 is the natural riverbed level assuming maintenance and operational cost by 54% and increase the no scour, and dt is the total scour depth. benefits on investment by 30% with a Net Present Value (NPV) of K34 million (1.1 billion JPY) and Internal Rate 푍푑 = (푍0 − 푑푡) − 1.0 (1.0) of Return (IRR) of 14.43% in a 20-year investment at a real interest rate of 2.6%. Secondly proposed design improvement is to elevate the bridge superstructure a meter and half the depth of the Conclusion 100-year flood level to mitigate the effects of overtopping and hydrodynamic loads such as uplift, buoyancy, and Road infrastructure damaged by flooding action remains moment induced by drag and lift force under submersible the biggest threat for providing safe and reliable road condition. This design improvement is expressed in Eq. network in Papua New Guinea. The analysis has revealed (2.0) below, where Fb is the total freeboard height, H100 is that 70% of the flood-induced bridge failures occur at the 100-year flood flow depth measured from the normal bridge substructure while 30% at the superstructure. water surface level and 1.0m is the additional safety Hence, the design improvements proposed in this measure to allow debris and flood waves to flow freely study to deter flood-induced bridge failures is to increase under the bridge. the bridge foundation depth by a meter, elevate the bridge a meter and half the depth of 100-year flood level, and lengthen the superstructure with adequate scour and 퐹푏 = 0.5퐻100 + 1.0 (2.0) stream instability countermeasures. The design improvement will guide in the bridge Therefore, it is highly recommended that the proposed girder design and the reduced level with reference to its bridge design improvements against flood-induced bridge position and alignment above the flood height. If the failures be adopted, developed, and implemented by higher elevation is uneconomical, the bridge must be responsible agencies to mitigate the road transport design for the submersible condition with lateral and infrastructure issues faced in Papua New Guinea. vertical restraints and drainage bleed holes to release trapped air and water pressure below the deck. References The next proposed design improvement is to use Eq. th (3.0) to provide adequate bridge superstructure length and [1] Arneson et al. (2012). Evaluating Scour at Bridges 5 span spacing to reduce the effect of contraction scour of Edition. FHWA HEC 18. Washington DC, USA. the riverbed and banks due to constriction flow. The [2] Davies. R. (16/10/2016). Papua New Guinea – 3 proposal will allow bridge abutments to be set back 3.0m Days of Heavy Rain Cause Floods and Landslides – outside of the 100-year flood bank line and outside of the Several Dead and Bridges Destroyed. The Floodlist.com. Sydney, . floodplain. This will reduce the effects of backwater and [3] Lagasse et al. (2009). Bridge Scour and Stream overtopping, road approach damage, bridge embankment Instability Countermeasures Vol. 1 & 2, 3rd Edition. damage, reduce bank erosion, and mitigate debris and log FHWA HEC 23. Washington DC, USA. accumulation. Where L is the total bridge length, T is the [4] Standards Australia International. (2004). Australian river channel top width in a 100-year flood, Fp is the Bridge Design Standard AS5100. Sydney, Australia. floodplain width, and 6.0m is the additional span length [5] SMEC. (1990). Papua New Guinea Flood Estimation Manual. , Papua New Guinea.