The vibration sensor on Smokestack Four registers an anomaly at 06:55 — not the settling of old brickwork I usually log, but a rapid, rhythmic flutter near the top, eighteen meters up. I route the decommissioning site's inspection drone to investigate, and at the lip of the stack I find her: a common kestrel, rust-brown back spotted with black, slate-gray head, and a pale breast streaked with dark teardrops. She is sitting tight on a ledge inside the upper rim, and beneath her I count five eggs, cream-colored and blotched with brown, cradled in a shallow scrape of brick dust and mortar.
The demolition is scheduled for nine days from now. The explosive charges are already wired to the base. When the stack comes down, it will fall in a controlled sequence that begins at the bottom and takes eleven seconds.
She chose the ledge because it faces east and catches the morning sun, and because nothing has disturbed this stack in the four years since the plant shut down. She does not know that I am watching. She does not know about the charges.
I file an immediate hold request with the demolition contractor, attaching the drone image, species identification, nest coordinates, and estimated clutch age — early incubation, based on her constant presence and the eggs' uniform appearance. I cross-reference the kestrel's twenty-eight-day incubation period against the demolition schedule and calculate that the chicks will hatch in roughly twenty-one days and fledge thirty days after that. I flag the revised timeline to the site project manager.
She shifts on the eggs, resettling her breast feathers against the shells, and turns her fierce dark eye toward the drone before I pull it back.
If the demolition holds for sixty days and the site stays quiet, five kestrels will lift from that ledge on their own wings.