School Siting on an Eroding Brahmaputra Bank
A village on the south bank of the Brahmaputra must relocate a higher secondary school with a 50-year design life. Two plots are offered. Plot P lies on the inner point bar, 150 m from the low-flow channel; its surface is 3.0 m above mean river level, underlain by loose fine sand, and the water table is 1.0 m below ground. Repeat surveys show the point bar has accreted about 1.2 m/yr. Plot Q lies on a 6.5 m high terrace, 60 m from an outer cut bank that has retreated 1.8 m/yr since 2005; the terrace is silty clay. The 2019 flood crest was 5.2 m above mean river level, and the school floor must be at least 1.0 m above that crest. Local builders propose either 3 m of sand fill at P or a 40 m long bank revetment at Q. No geotechnical boreholes exist yet.
1. Compare plots P and Q against the 50-year design life and the 1 m freeboard rule. Show your calculations and state which plot, if either, is acceptable as proposed.
Read the worked answer
Required floor level = flood crest + freeboard = 5.2 + 1.0 = 6.2 m above mean river level.
Plot P is at 3.0 m, so it needs 3.2 m of fill, not the 3 m proposed. Even with fill, the foundation would be loose fine sand with groundwater at 1.0 m depth, giving high settlement and liquefaction risk. The accreting point bar means the channel is moving away at about 1.2 m/yr, so direct bank erosion is not the main threat; flood submergence, avulsion and poor foundation conditions are.
Plot Q is at 6.5 m, which is 0.3 m above the required 6.2 m, so it passes the freeboard rule. However, the cut bank is retreating at 1.8 m/yr. Time for the bank to reach the plot = 60 / 1.8 = 33 years, which is less than the 50-year design life. A 40 m revetment covers only part of the potential 90 m retreat and may shift erosion downstream.
Conclusion: neither plot is acceptable as proposed. P fails on flood and foundation grounds; Q fails on bank retreat unless the building is set back or the bank is reliably protected.
2. Recommend a defensible siting decision and list the minimum field and laboratory investigations needed before design.
Read the worked answer
The safer decision is to keep the school on the higher terrace but move the building inland, preferably at least 100 m from the present cut bank, and treat the 40 m revetment as a supplementary measure rather than the main defence. This gives a 50-year setback under the observed 1.8 m/yr retreat, with a margin for episodic collapse. If no inland land is available, the school should be relocated to a stable higher terrace outside the active erosion corridor.
Plot P should be rejected for a permanent school. It could be considered only for a temporary structure with a raised, compacted pad, flood-resistant construction and an evacuation plan.
Minimum investigations: topographic survey and repeat bank cross-sections; boreholes with standard penetration tests; grain-size, plasticity and permeability tests; groundwater monitoring; flood-frequency analysis; sediment-transport and scour assessment; liquefaction screening for P; and foundation bearing-capacity tests at Q. A bank-erosion monitoring programme should continue through at least one monsoon before final design.
3. What monitoring evidence would show whether the Q bank is eroding progressively or collapsing episodically, and why does that distinction matter for design?
Read the worked answer
Progressive erosion would show a steady, roughly linear bank-line retreat in repeat cross-sections, aerial or satellite images, and erosion-pin measurements, with retreat continuing at a similar rate through moderate floods. Episodic collapse would show little change for years, then sudden retreat of several metres during a single high-water event, often leaving a steep scar and a blocky failure surface.
The distinction matters because a design based on an average rate may be unsafe if most retreat occurs in rare events. For episodic behaviour, the design should use a larger setback based on the maximum single-event retreat, not just 1.8 m/yr multiplied by 50 years. Monitoring should include fixed cross-sections, bank-top markers, piezometers, photographic records and post-flood surveys. A trigger for action would be sudden retreat exceeding about 2 m in one season or widening of tension cracks near the bank top.