The training smoke comes from one machine with a fluid tank that is measured, because the fluid is dear and the amount used per drill is on the form. The machine is off overnight, unplugged, and most mornings the house is full of smoke at six, the tank is down by a two crew drill, and the machine is cold.
I am the training officer, and the smoke machine's fluid is a line on my budget.
How the smoke is metered.
The machine heats fluid from a tank into fog and pumps it into the house through a duct, and a two crew drill fills the house from a known amount, which I record from the tank's sight glass before and after. The machine is unplugged overnight because it has a heater in it and I do not leave heaters on in a wooden building.
A house that has been used clears in about an hour with the vents open, and it is clear by the time I lock it.
Most mornings at six the house is full of smoke.
Full, the way it is for a drill, to the ceiling of both floors, cold, and thick enough that the door is a wall of grey.
The machine is cold and unplugged and its tank is down by a two crew drill's worth, and the duct into the house is warm, the way it is after the machine has run.
What has been checked.
The machine. Serviced, and then taken out of the building and locked in the store overnight. The house filled anyway, and the tank in the locked store was down.
The fluid. It is our fluid, the smell of it is our fluid, and the house's own gas sensor, which is there so nobody drills in real smoke by mistake, reads it as glycol fog and not combustion.
The house. Locked, and the door contacts show it locked, and the vents shut. The camera inside shows the rooms clear at midnight and full by five, the smoke coming in through the duct, thick, at the rate the machine puts it out, with the machine in a locked store forty metres away.
The drill.
A two crew drill has a shape: two search teams, two casualties, a set time. The tank goes down by that. Not by a random amount. By the amount two crews would take, every time, as though a drill had been run in the house in the night, by two crews, to the schedule.
I have looked at the house's thermal camera recordings for those nights, because the house is instrumented for exactly this, so I can see where crews go in the smoke.
There are crews in the smoke. Warm shapes, in pairs, moving the way search teams move, along the walls, through the rooms, up the stairs, at about two in the morning, in a house that is locked, and they find the casualty positions, which are empty, and stop at them, and kneel, and one of the pairs, every night, goes to the back bedroom, and stays.
What is done about it.
The fluid is on the budget as overnight loss, and the brigade's finance people have queried it once and accepted a note about evaporation, which is not what it is.
The house is vented at six and clear by seven, and the day's drills run on schedule, and the crews who train in it are good, and I have sometimes thought that is not unconnected.
What I do.
I run the training.
The one thing I have changed is that I open the house at six myself, and I do not send the duty crew, and I open it from the back door, not the front.
The reason is a morning in the winter when I opened the front door into the smoke at six and stood in the doorway, and out of the smoke, at the pace of a search team leaving a building, came the sound of two sets breathing, and boots, and the smoke moved the way smoke moves when people walk through it, past me, either side, close enough to touch, and out into the yard, and the tally board on the wall by the door, which I had left empty, had two tallies on it, and both of them were off it by the time I turned round.