This may be completely off the wall. After a friend's mother had a bout of CO poisoning, and I had episodes of drowsiness, the friend suggested I got a CO alarm myself. Since I don't combust anything, fully or partially, except the odd candle, I got the sort of spot detector that goes through a colour change. I put one in my car, one by my bed, one in the living room, and one on my fridge-freezer.

A little while ago, shortly before the use by expiry date, and during a spell of temperature inversion, the colour changed on the two in the living room and the kitchen, in the case of the fridge-freezer very strongly. Thinking that I had had a leak in from the neigbours below, the flue not working because of the weather, I asked them if they had a detector, which they did not, and alerted other neighbours to the possibility of leaks between properties. And opened the windows.

Today I had a visit from a fridge engineer, as the fridge wasn't working properly, and he told me that there was a leak of refrigerant - he did not tell me what it was, though I've seen the word pentane about somewhere. He said it was one which needed to be removed and sequestered as in the case of CFCs. He said pentane was the insulator. He also said the refrigerant would not affect the CO detector, which is now pale again, and a new one has shown no effect.

The time scale is about right for when the fridge went wrong. Is it possible that leaked pentane (and I'm pretty sure that the refrigerant is a flammable gas, because I remember feeling some concern when I bought the thing) could partially oxidise while hanging around in the air? I know methane oxidises in the atmosphere, but I don't know the steps it passes through on the way to CO2 and H2O.

Or could escaped refrigerant discolour a detector?

I really don't like the idea of escaped hydrocarbons round my kitchen.

Penny