Reservoir Release Before a Cyclone
On 15 August a low-pressure system over the Bay of Bengal is forecast to cross the Odisha coast near Paradip within 48 hours. The Mahanadi basin upstream of Hirakud has received 210 mm of rain in three days, and inflow to the reservoir has risen to 45,000 cusecs and is still climbing. Live storage is at 92 per cent of its 5,400-million-cubic-metre capacity, while the rule curve for mid-August permits no more than 85 per cent. Downstream in Cuttack and Puri districts, farmers have just finished transplanting paddy nurseries, and two embankments breached last year. The engineer-in-charge has three options: release 40,000 cusecs now; wait twelve hours and then decide; or hold the water and release 60,000 cusecs after the cyclone has passed. The superintending engineer objects to a large release because it will cut power generation and 'waste' water. The downstream collector has not yet been informed of any of this.
1. Using the numbers in the setup, explain why holding the water until the cyclone has passed is the most dangerous option, and why a 40,000-cusec release now does not by itself solve the problem.
Read the worked answer
Step 1 - measure the cushion. 8 per cent of 5,400 MCM is about 432 MCM of empty space above the present level. One cusec is roughly 0.0283 cubic metres per second, so 45,000 cusecs is about 1,274 cubic metres per second, or roughly 110 MCM per day. At the current inflow the cushion would be gone in under four days - and the inflow is still rising, with cyclone rainfall still to be added.
Step 2 - see what each option does to the cushion. A 40,000-cusec release is below the present inflow. Storage therefore keeps rising, only more slowly (net about 5,000 cusecs, or 12 MCM a day). It buys days, not safety. To create a cushion before the storm arrives the release has to exceed the inflow; 60,000 cusecs gives a net drawdown of about 15,000 cusecs, roughly 37 MCM a day.
Step 3 - see why waiting is worst. If the gates are held and the inflow doubles with the cyclone, the release that becomes unavoidable will be far larger than 60,000 cusecs and will be uncontrolled. It will also arrive exactly when the storm surge and local rain are already raising the river, so the same discharge produces a higher flood level downstream. Gate operation in cyclonic wind and debris is also unreliable.
The safe sequence is therefore to build the cushion before the storm, when the river downstream is still low, not after.
2. The superintending engineer argues that a large release will 'waste' water needed for power generation. Explain why this reasoning mistakes the priorities that govern a multipurpose reservoir in the monsoon.
Read the worked answer
The reservoir has three purposes - flood moderation, irrigation and power - and they are not ranked equally in August. The rule curve is a flood-safety constraint, not a production target; it exists precisely because the dam must be able to absorb a storm inflow. Treating the 85 per cent line as an inconvenience to be argued with inverts the order of priorities.
Second, the argument confuses water released with water lost. Water let out through the turbines still generates power; only spillway releases bypass generation. A controlled release now, routed through the powerhouse, produces electricity and lowers the level at the same time. Holding water and then spilling it uncontrolled produces neither power nor safety.
Third, 'waste' is the wrong accounting. Water that reaches the sea has no immediate use, but the damage avoided downstream - embankments, transplanted nurseries, roads, lives - has a cost far above the value of the stored water. The engineer-in-charge should record the flood-control purpose as the governing one, and note that power generation is optimised within the flood rule, never against it.
3. List, in order, four actions the engineer-in-charge should complete before opening the gates, and state what each is meant to prevent.
Read the worked answer
Inform and warn downstream authorities first. The collector of each downstream district, the embankment division and the village-level warning system must be told the discharge and the likely timing before the gates move. This prevents the classic failure in which the flood wave reaches villages that were never warned, and lets people move livestock and reach high ground.
Coordinate with upstream gauging stations and the meteorological office. The release has to be sized to the inflow actually arriving in the next 24–48 hours, not to this morning's figure. This prevents releasing too little and losing the cushion, or releasing too much onto a river already in flood.
Fix the release in stages and record it. Opening gates gradually, in fixed steps, and logging each step means the downstream river rise is predictable and any error can be traced. This prevents a single sudden surge and removes ambiguity about who ordered what.
Check gate mechanism, power supply and spillway condition, and notify the powerhouse. Gates jammed by debris or a failed hoist during a cyclone turn a controlled release into an emergency. This step prevents the dam from being unable to do what has already been announced.
Only after these should the gates be opened, and the whole sequence should be written into the dam's log with the rule-curve position attached.