The archive strong room has gas suppression, because water would ruin the paper as surely as fire. Smoke in the room starts a thirty second countdown to the gas, with an alarm and a light, and a button by the door cancels it. Every night at eleven minutes past two the panel logs smoke, the countdown starts, and at eight seconds it is cancelled from the button inside the room, and the room is locked and empty and there is no smoke.
I am the archivist at a museum with a strong room, and the suppression system is the thing I trust most and understand least.
The system.
Smoke detectors in the strong room. Smoke starts a countdown, thirty seconds, with a siren and a red light, so that anybody in the room can get out. At zero the gas goes in and the room is not survivable. A button inside the room, by the door, cancels the countdown if the person inside knows it is a false alarm, and every cancel is logged with the time and the seconds remaining.
Eleven minutes past two, every night. Smoke. Countdown. Cancelled at eight seconds, from the button inside.
What has been checked.
The detectors, replaced. The panel, replaced. The button, replaced, and the new button has a cover on it that has to be lifted, and the log shows the cover lifted at nine seconds and the button pressed at eight.
The room, at two, by me, inside it, with the door open and a colleague at the panel. At eleven past the detector went. The siren. The red light. The countdown on the panel outside, which my colleague read to me. Twenty. Fifteen. Ten. At nine, the cover on the button beside me lifted. At eight, the button went in. I was two feet from it. There was no hand.
There was no smoke, either. Not that I could see. But there was the smell of it, for the thirty seconds, old smoke, paper smoke, the smell of a fire that has been out for a long time.
Eleven minutes past two.
The building has a history, and the history has a fire. 1932, when the museum was a library. A fire in what was then the stack, which is now the strong room, at night, in the winter. It started at about two, the report says. The night porter found it and fought it with buckets and it was put out with the loss of a part of the collection and, the report says in a paragraph of its own, of the porter, who was found in the stack with the last bucket, having got the fire down to nothing and not got out.
The report does not give a time for the porter. It gives one for the fire. Eleven minutes past two, from the clock that stopped when the heat reached it, which was in the stack, and is in our collection now, with the hands where they were.
What is done about it.
The pre alarm and the cancel are logged every night and the fire officer has seen the log and has asked who is in the strong room at two, and I have said nobody, and he has asked who cancels the countdown, and I have said that the log says it is cancelled from inside, and he has looked at me for a while and written, system tested nightly, which it is.
What I do.
I keep the archive.
I understood it in the end. He is putting the fire out. Every night at eleven past two the porter is in the stack with his bucket and the smoke is what the detector reads, the smoke of 1932, and he gets it down, and by the time the gas would go he has done it, and he cancels, because the fire is out and there is no need for the gas, and there is a man in the room.
He is right. There is no need. He has put the fire out every night for ninety years and he has never once let it get to zero.
I have started to say thank you. Not out loud, I am not that sort of person. I have started to leave the strong room light on for the night, which is against policy, because a man fighting a fire in the dark with a bucket deserves to see what he is doing, and I have put a bucket in the corner. A real one, galvanised, the old kind, full, and in the morning it is empty, and the floor is dry, and the log says cancelled at eight seconds, and the fire officer has asked what the bucket is for, and I have said, it is for the porter, and he has stopped asking me things.