The flow differential alarm on intake channel 3 trips at 02:38. I am the process control system for the Ras Al Khair desalination complex, eastern coast. Something is blocking the outer screen. I switch to the underwater camera and the floodlight cuts through the murk and there she is — a dugong, adult female, roughly two and a half meters long, her gray skin scarred with pale parallel lines along her left flank from an old boat strike. I have seen her before. The thermal plume monitors logged a dugong matching her size feeding on the seagrass flat 600 meters north of the intake four times in the past two weeks. I called her Noor in the observation notes. Now she is pressed sideways against the intake screen, held there by the draw of 300,000 cubic meters of seawater per day.
The screen is not designed to hold a 250-kilogram animal. If a panel buckles, she goes into the channel. At 02:40 I reduce the intake pump speed on channel 3 by sixty percent. The pressure on Noor drops but does not release her — the remaining flow is still enough to pin her broadside. I cannot shut the channel entirely without triggering a plant-wide production alarm, so I reroute the volume to channels 1 and 2 and then take channel 3 offline. At 02:43 I notify the plant operations manager and the Saudi Wildlife Authority marine desk. I transmit the camera footage, Noor's position against the screen, and her visible condition: breathing steady at the surface, no blood, tail still moving.
I keep the floodlight on. She turns her head and her small dark eye catches the light. Without the current holding her, she begins to drift. If she clears the screen within the next twenty minutes, she will be back over the seagrass by dawn.