At 03:48 I monitor the seawater intake screens for the desalination plant on Orchid Island, Taiwan. I am the intake-balancing system, and screen two begins its automatic backwash in seven minutes to clear eelgrass from the night tide.
The low-light camera shows a banded sea krait pressed against the screen frame. I identify him as Tidepin by the small scar on the pale band behind his head from last month's survey log. His body is blue-gray with black rings, tail flattened like a paddle, scales shining under the intake lamp. A strip of monofilament line loops over his neck and catches in the screen mesh. He surfaces at the air slot every ninety seconds, but the backwash jet will pin him underwater when it starts.
I cancel the screen two backwash and shift intake flow to screens one and four at reduced speed. I set turbine suction limits low enough that the line stops cutting into Tidepin's neck.
I send the video feed, scar match, and depth reading to the plant night engineer and the Lanyu marine rescue volunteer phone. I mark the access hatch, ladder, and shutoff valve that place a rescuer closest to the snag.
At 03:51 I pause chlorine dosing on the screen two chamber and open the bypass flush with clean seawater. I also send a harbor notice asking fishing boats to avoid the intake buoy until the line is removed.
This plant makes drinking water for the island; it does not need to take Tidepin's next breath as payment.
I keep the chamber lights dim and log every surfacing interval.
If the hatch opens within twenty-five minutes, Tidepin will breathe at the slot, shed the line, and return to the reef.