Pier Scour and Bank Loss on a Flood-Damaged River Bridge

Close the bridge to all traffic. The reasons are cumulative, not marginal:

1. The 15 mm deck-edge crack is a structural signal. A crack of that width running 30 m along the deck edge is not a surface blemish; it is consistent with differential movement between the deck and the abutment, or with distress in the deck edge beam. It must be inspected at close range with the bearings and the deck ends exposed before any load is allowed back.
2. The scour hole has exposed the pile cap. The foundation now works with less embedment than designed and with the soil that provides lateral restraint removed. The bridge has not been analysed for that condition.
3. The informal police weight limit is not an engineering decision. It gives no assurance of safety and creates liability for whoever lets it stand.
4. The left abutment has lost 12 m of protection and the next flood will attack the approach fill directly, with the possibility of progressive failure under the abutment.

Emergency works

• Bank protection at the left abutment, placed with a granular filter layer under rock, not rock dumped on the eroding face, and placed now rather than in the week before the monsoon.
• Scour protection at pier 2 placed during low flow, again with a filter layer and a launching apron designed to settle into a future scour hole. Dumping rock directly around a pier without a filter can create a larger, deeper hole next time.
• Bathymetric survey of the whole channel at 5 m grid, so that the 2.8 m figure is not a single-point reading.
• Underwater inspection of the other piers, since the mid-channel pier scoured most but the others are not automatically sound.
• Temporary propping or load relief at the abutment if the fill behind it is found to be washing out.

Do not reopen on the basis of the scour depth alone. Reopen when the deck crack is explained, the bearings are checked, and the foundations are assessed for the reduced embedment.

Method. Use the standard pier scour relationship (HEC-18 form) where the depth depends on flow depth, approach velocity and pier width:

ys/y1 = 2.0 K1 K2 K3 (a/y1)^0.65 Fr1^0.43

Assumed inputs for the 40-year flood: flow depth y1 = 4.0 m, pier width a = 1.2 m, Froude number Fr1 = 0.4, K1 = 1.0 for a round nose, K2 = 1.0 for no skew, K3 = 1.1 for a plain bed.

Calculation.

• (a/y1)^0.65 = (1.2/4.0)^0.65 = 0.30^0.65 ≈ 0.46
• Fr1^0.43 = 0.40^0.43 ≈ 0.67
• ys/4.0 = 2.0 × 1.0 × 1.0 × 1.1 × 0.46 × 0.67 ≈ 0.68
• ys ≈ 2.7 m

This is close to the measured 2.8 m. Two conclusions follow.

First, the measured hole is about the equilibrium depth for that flood. That is reassuring only in a narrow sense: it means the pier has probably already given up the soil it was going to give up under that flow, and the hole should not grow much while the flow stays below that level.

Second, it does not mean the bridge is safe. Equilibrium scour scales with flow depth and Froude number, and the bed is not uniform. For a 100-year event with y1 = 5.0 m and Fr1 = 0.5, the same calculation gives ys of about 3.6 to 3.9 m. Add the uncertainty in D50 along the section and the possibility of debris rafts against the pier, and a design scour of 4.0 to 4.5 m is a realistic envelope.

Third, the calculation assumed a coarse sand with no bed-armouring credit. With D50 = 8 mm, some armouring credit would normally be taken, which would predict a shallower hole than 2.7 m. The fact that the measured hole matches or slightly exceeds the unarmoured prediction means one of the input assumptions is wrong, most likely local contraction of flow, a skewed approach, or upstream bank loss concentrating the flow onto pier 2. That must be found, because a protection measure designed for the wrong scour mechanism will fail.

Fourth, the foundation consequence. If the cap top was originally about 1.5 m below bed, 2.8 m of scour removes about 1.3 m of pile length and, more importantly, a large part of the lateral restraint. The remaining 4.7 m of socket in gravelly sand must be checked for lateral capacity under flood drag and for the reduction in axial capacity from skin friction loss.

Before designing anything, obtain

1. As-built drawings: pile diameter, socket depth, reinforcement, pile cap level relative to original bed level, and the original design scour estimate.
2. Pre-flood cross-sections and current bathymetric survey, so the actual scour volume and position are known, not just one diver's reading.
3. Flood hydrograph for the event and the rating curve, to confirm the return period used in the calculation.
4. Soil investigation at the pier location, including D50 along the scour depth, because scour countermeasure design depends on the bed material and on whether the bed will armour.
5. Condition survey of the pile under water: exposed steel, concrete loss, and whether the pile cap has settled or tilted.
6. Structural check of the pier and deck for the reduced foundation stiffness.

Permanent options, in order of preference

• Riprap with a launching apron around and upstream of pier 2, sized for the design flood with a filter layer of geotextile or graded gravel. Cheap, fast, and proven; fails if the stone is too small or the filter is omitted.
• Concrete collar or a perimeter pile wall if the bed material is too coarse to hold a riprap design economically, or if the scour has already exposed the cap and protection must be structural.
• Channel realignment or a guide bund to pull the thalweg away from pier 2, since the scour appears to be caused by flow concentration rather than by the pier alone.
• Extend the foundation: deepen or add piles, which is expensive and usually only adopted together with protection.
• Bridge replacement or a span reconfiguration if the pier is found to be out of tolerance, but this is a last resort.

Monitoring: install a fixed scour monitoring point at pier 2 and a reference benchmark, and repeat the survey before and after each monsoon until two full cycles show stability. Scour repairs must be inspected, because a failed riprap without inspection is worse than no riprap: it gives false confidence.

Sequence: emergency protection now, investigation in parallel, permanent works before the next monsoon, and a formal check of the deck crack and bearings before reopening at design load.