[benjamir]

I recently bought one of the best CO_2 meters available here (and had a CO alarm since before that): the low temperature boiler (natural gas, 1991) in our kitchen contributes close to nothing to the CO_2 level despite getting its air from inside. Letting it run on automatic a whole workday contributes less than 200ppm in our flat. But an adult and two kids can bring the level from below 500ppm to above 1000ppm in around one hour -- additionally: doesn't really matter if the doors to other rooms are open or not. Ppm fluctuation is (way) less than 100ppm between rooms.

(All time low 380ppm -- 10% above the world average during my birth :-( , highest 1620ppm)

[mindcrime]

Yeah, under ordinary circumstances your boiler or whatever should not contribute enough CO to cause problems. It's when something weird happens that you have to worry (usually). Like if the jet is clogged or something else happens that can cause incomplete combustion. Incomplete combustion is your enemy when it comes to CO[1].

[1]: https://www.abe.iastate.edu/extension-and-outreach/carbon-mo...

[benjamir]

* I wrote about CO2.

* The boiler's flame is visible (through a slit).

* The system doesn't have a separate supply stream -- it takes oxygen from the flat.

An CO sensor seems obligatory to me: after the chemney sweeper (obligatory in Germany) measured 10,000ppm CO in the exhaust stream (defect, not normal operation), I didn't want to depend on a 25y+ boiler's electronics alone.

CO kills relatively fast. Even if rare, I don't want to die caused by a rare event that could have been avoided by spending ~50€ (here)