So it’s a piece of shit gas oven that was recalled a decade ago, all landlords are bastards, etc etc.
If you try and use it normally, the thermostat just stops the gas flow after a few minutes regardless of temperature. Useless.
After some time fighting this frustration I made a wire contraption (to replace prior rubber-band attempts) to hold down the thermostat knob, such that the gas flow is forced on, but it still adjusts the temperature. The only side effect is that the lighter clicks constantly whenever the door is opened.
So the other night I put some food in there to reheat, set up the contraption to keep it on, opened the door and waited for the “whoof” noise of the gas lighting. Then I left to the other room.
Five minutes later my housemate went into the kitchen. They noticed it smelled like gas, and went to open the oven door to check it was working.
You may be able to predict what resulted.
As soon as they got the door open, the air mixed with the natural gas filling up the oven ever since the flame had somehow gone out some minutes before. Then the lighter went off, because of the part holding down the knob. Bingo, fireball.
Very luckily, they had fast reactions to push backwards onto the floor, pushing the door closed and getting themself low down while the fireball plumed upward.
Because of that reaction, they escaped with only some singed hair, eyebrows and eyelashes, and sensitised skin like a sunburn. If they hadn’t, there’s a reasonable chance their polyester top could have caught on fire and resulted in severe burns. We do have a fire blanket in the room but they had not been specifically shown it yet.
mild injury, safety analysis
This event shows amongst other things, the importance of integration testing and holistic approach to designing in safety to dangerous systems.
The wire contraption didn’t on its own appear to create any new risk — the possibility of the oven leaking gas by extinguishing its own flame already existed, and in fact the contraption seemed to *lessen* the risk by keeping the flow high enough to maintain flames.
However, the accident shows that an apparently irrelevant side effect, the continuous activation of the piezo lighter whenever the door is open, became a danger when considered in combination with 1. A failure in another part of the system (the flame going out), and 2. A human interacting with the system in a way that was natural but not, perhaps, ‘correct’. Arguments about what ‘should’ have been done (such as shutting off the gas cock or leaving the room and opening windows) aren’t as important as considering what systematic failures allowed the human to make a bad decision.
@s0 you nearly Home Alone’d your housemate because you hacked your gas oven. I mean it’s clearly the landlords fault but still, lmao
@ThermiteBeGiants the hack was not initially mine, I made the existing one more reliable. Which raises more interesting reflections about how people are more likely to accept inherent risks in an update of a previously existent system, than those same risks in a completely new system.
@s0 look, we could go all Richard Feynman and talk about normalisation of deviance in the safety culture of your household.
But I just have a personal rule to not fuck around with gas appliances in any way, especially if they’re clearly broken like that was. You should’ve dobbed in your landlord to Fair Trading, you and your housemates’ hacks aside that appliance should not be connected to the gas mains if it performed like that.
@s0 oof I'm glad it didn't get worse than that. Hope you get that hellish contraption replaced ASAP
A place for trans makers, coders, tinkerers and dreamers.