Cyclone Shelter Ration and Evacuation Planning
A Category-3 cyclone is forecast to make landfall in 30 hours near a coastal union of 12,000 people. The union has two concrete cyclone shelters: Shelter A holds 600 people and is on a road that floods when the tidal surge exceeds 1.2 m; Shelter B holds 450 people and sits on higher ground. The forecast surge is 1.5 m. Roughly 40% of residents are expected to seek shelter, and 1,800 more people from a neighbouring union with no shelter will also arrive. Local officials have 30 hours to move people, and the main road to Shelter B is usable for only 18 of those hours before it is cut. Describe the numbers, the shelter capacities and the timing constraint.
1. Estimate total shelter demand and determine whether the two shelters can absorb it. What should officials do about the overflow?
Read the worked answer
Demand. Local seekers = 40% of 12,000 = 4,800. Add 1,800 from the neighbouring union = 6,600 total seekers.
Capacity. Shelter A = 600, Shelter B = 450, total = 1,050. But Shelter A's access road floods at a surge above 1.2 m, and the forecast surge is 1.5 m, so Shelter A may become unreachable or unsafe for late arrivals. Realistically usable capacity is close to 450 (Shelter B) plus whatever can enter A before the road cuts.
Gap. Even counting both shelters fully, 6,600 − 1,050 = 5,550 people have no shelter space. The gap is far too large to close by squeezing people in.
Judgement. Officials must not simply direct everyone to the two shelters. They should: (1) identify additional safe structures (schools, mosques, multi-storey buildings) on high ground; (2) begin moving the neighbouring-union group early, since they have the longest journey; (3) use the 18-hour window on the Shelter B road first, because it closes before landfall; and (4) plan vertical evacuation (moving to upper floors of sturdy buildings) for those who cannot reach a shelter. The core reasoning is that capacity is the binding constraint, not willingness to evacuate.
2. How should the 30-hour window and the 18-hour road limit change the order of operations?
Read the worked answer
Work backwards from the deadlines.
- T−30 to T−18 (first 12 hours): the Shelter B road is open only for 18 hours, so the highest-priority movement is the 1,800 incoming people and any locals assigned to Shelter B. Move them first, while the road exists.
- T−18 to T−0: Shelter B is now cut off to vehicles. Remaining movement must go to Shelter A, but only while the surge stays below 1.2 m. Since the forecast surge is 1.5 m, assume A's road may flood before landfall, so do not plan late mass movement to A.
- Last 6–8 hours: stop road movement and switch to vertical evacuation and shelter-in-place advisories.
Reasoning. The binding constraint is not the 30-hour forecast but the 18-hour road life and the 1.2 m surge threshold. Sequencing by when access closes, not by distance, is the correct judgement. Officials should also announce the plan early so people do not wait and then find roads cut.
3. A local leader argues that because only 40% are expected to seek shelter, there is no real problem. Evaluate this reasoning.
Read the worked answer
The 40% figure is an assumption about behaviour, not a guarantee, and it is being applied to a population that includes 1,800 people with no shelter at all in their own union.
- 40% of 12,000 = 4,800 locals, plus 1,800 outsiders = 6,600.
- Capacity is about 1,050, and effectively less because Shelter A's road floods at 1.5 m surge.
So even accepting the 40% assumption, demand is roughly six times usable capacity. The leader's conclusion inverts the logic: a low turnout estimate makes the shortfall less visible but does not make it smaller, and turnout often rises sharply once people see the cyclone approaching.
Judgement. Planning should use the 40% figure as a baseline but prepare for higher turnout, because the cost of under-preparing (people trapped on flooded roads) is far worse than the cost of over-preparing. The correct response is to expand capacity and start early, not to rely on the assumption.