Case Study 5 - Forty-Eight Hours on the Embankment
A district magistrate in north Bihar holds an emergency meeting at 6 p.m. The gauge at the embankment headworks reads 47.90 m against a danger level of 47.50 m, and the water has been rising 15 cm a day. The catchment forecast is for heavy rain for three more days. Twenty-six villages with 41,000 people lie in three clusters. Cluster C: six villages on the diara land between the embankment and the river, 11,000 people, to be under about 1.2 m of water within 48 hours. Cluster A: nine villages inside the embankment on low land, 12,000 people, whose only link road is already under 30 cm of water and will be cut within a day. Cluster B: eleven villages on ground 1.5 m above the danger level, 18,000 people, which will be cut off from district headquarters. Two NDRF teams with four boats are available; each boat carries 25 people per trip and takes 90 minutes for the round trip. There are three relief camps in colleges at headquarters and no helicopter.
1. Set out an evacuation and relief sequence for the three clusters, with the reason for each decision.
Read the worked answer
First: Cluster C - evacuate, immediately. Highest inundation depth (1.2 m), earliest onset (48 hours), and no road alternative: boats are the only means, so the resource must go where it is irreplaceable. Within C, prioritise the old, the sick, pregnant women and children.
Second: Cluster A - begin evacuation at once, but by staged boat movement along the road alignment. They face the second-highest risk because if the embankment breaches, low land inside it floods with no warning. The road is still usable for a day, so move people and livestock toward the embankment crown and the road while it is passable, then by boat.
Third: Cluster B - do not evacuate; supply in place. At 1.5 m above danger level they are unlikely to drown. Moving 18,000 people across water is the most expensive option available and would consume the boats needed by C. Their risk is isolation - drinking water, medicines, fodder - so the response is stockpiling, a satellite phone or VHF link, and a medical team, not boats.
Principle to state: allocate the scarce asset (boats) to the cluster with the highest depth × imminence and the fewest alternatives, not to the cluster with the largest population.
2. Using the boat data, test whether Cluster C can be fully evacuated in 48 hours, and revise the plan accordingly.
Read the worked answer
Requirement: 11,000 people ÷ 25 per trip = 440 boat trips.
Capacity: four boats, each round trip 90 minutes → 48 hours ÷ 1.5 hours = 32 trips per boat → 4 × 32 = 128 trips → 128 × 25 = 3,200 people in 48 hours.
Gap: 11,000 − 3,200 = 7,800 people cannot be moved in the time available; the full evacuation would need about 165 boat-hours per boat, roughly seven days.
Revised plan:
- Shelter vertically, then evacuate. Move C's population to the highest ground within the cluster - embankment crown, raised school and panchayat buildings, and any concrete plinth - and only then ferry people to camps. Vertical evacuation is what the boats cannot do.
- Requisition private country boats. Dozens typically exist locally; commandeering them under the disaster management law multiplies capacity faster than waiting for more NDRF teams.
- Prioritise the ferry for those who cannot shelter safely - the bedridden, newborns, pregnant women - and for resupply of the shelters.
- Pre-position at the camps water, ORS, dry food and fodder for the numbers actually arriving, not the numbers planned.
Report honestly upward: a plan that promises full evacuation and delivers 3,200 people is worse than a plan that shelters 7,800 on high ground and says so.
3. In the review meeting three weeks later, the district is asked why north Bihar still floods despite decades of embankments. Give the reasoning the district officer should place on record.
Read the worked answer
The embankment strategy addresses the symptom, not the cause.
1. Silt, not just water. The Himalayan feeders - Kosi, Bagmati, Gandak, Burhi Gandak, Kamla - carry an enormous silt load. Inside an embankment the current slows and silt drops, so the bed rises between the levees. The same discharge then spills at a higher level each decade, and the embankment has to be raised again. Embankments buy time and transfer risk downstream.
2. Confinement turns a spread into a breach. When the bed rises, a levee failure releases water at speed and depth. The Kosi's breach at Kusaha in Nepal on 18 August 2008 shifted the river into a new channel - the reason the Kosi is called the 'sorrow of Bihar'. A river that has historically wandered across its own megafan cannot be permanently fixed by a single line of earth.
3. Drainage congestion. North Bihar's tals and chaurs cannot drain while the Ganga runs high downstream; ponding lasts weeks even after the rain stops.
4. What complements embankments: basin-level planning with Nepal, spurs and revetments at attack points, drainage sluices, flood-plain zoning so new construction stays off the most hazardous land, early warning from upstream gauges, and cropping choices that assume a flood season rather than resist it.
The point to record: the district's job is flood management within a structural reality it cannot change in one season.