Previous Entry Share Next Entry
Car Talk-based geekery.
A caller on Car Talk this week asked the guys to confirm something her co-worker had told her. As a person involved in emergency medicine (I didn't catch her exact involvement) she often sees people brought in suffering from CO poisoning from boating. Again, I didn't catch exactly how come fishers/boaters get exposed to CO to the point of needing medical care. One of her coworkers commented that the patients could have gotten the same effect by simply sitting in their car in a closed garage for an hour or so and bypassed the sunburns, boats, etc. Another of her coworkers (a resident) countered that modern car's emission systems are so good that one could run your tank dry in a closed garage and there would be hardly any CO buildup.

The Car Talk guys sided with the resident, noting that nowadays the needles on the emissions testers barely budge when testing a car, as compared with the old days when no emissions tester was needed to tell that lots of bad stuff was coming out the tailpipe. They suggested that you would be more likely to die from oxygen deprivation then CO poisoning.

Below the cut are my calculations and thoughts on the matter...

Some starting assumptions...

We are talking about an air-tight 1-car garage which is 3m tall, 5m wide and 5m deep (roughly 16'x16'x10'). At standard temperature and pressure there is 1mol of gas per 22.7 liters of volume. In the garage there are 75000 liters, enough room for 3300 mols of gas. The partial pressure of oxygen in the atmosphere is about 0.20 atmospheres, so about 0.20 of the gas in the room is is O2, or 660mols of O2. There are 32g/mol for O2, so a little over 21kg of O2. There's about 73kg of nitrogen in the room, for a total of about 94kg of gasses total.

I am assuming that the gasoline is roughly an isomer of heptane, so it'll be roughly C7H16 in chemical formula. The chemical equation balances with C7H16 + 11O2 ==> 7CO2 + 8H2O, so burning one mol (100g) of gas with 11 mols (352g) of O2 yields 7 mols (308g) of CO2 and 8 mols (144g) of water.

A liter of heptane masses about 680g, so for every liter (6.8mol) of gas burned, about 2.4kg (74.8mol) of O2 is consumed. That's about a ninth of the oxygen in the room. There are 585mols of O2 left in the room, so the partial pressure is about 18%. Still breathable. However, there is now an additional 48mols, about 2kg, of CO2 in the air, and an additional 1kg of water vapor, so the CO2 concentration is about 2kg/95.6kg, or 2%. That's enough to make you drowsy.

After the 2nd liter is burned, there's 510mols of O2 left, the partial pressure of O2 is 15%, which is enough to start impairing your performance and consciousness. There's an additional 2kg of CO2, pushing the concentration to 4% or so.

It seems a race to see which kills you first, the falling O2 levels or rising CO2 levels. The O2 seems to be falling by about 2%/liter, and the CO2 seems to be rising by the same amount. After about 5 liters burned (a bit more than 1 gallon), the O2 concentration and CO2 concentrations are both about 10%, and either of these levels alone will cause unconsciousness -- with more extremes both leading to death.

But I'd also be concerned about the engine of the car. After 5 liters of gas burned, the O2 levels in the air have fallen by half, equivalent to (say) the Mount Everest base camp. Would a standard car made for US domestic use be able to run well at that altitude? Would it overtax the ability of the emissions control system to eliminate partial combustion products? At what point would the engine be so starved of O2 that it would quit? More morbidly, which would die first, the car or the human?

Should I send it to the Car Guys?

  • 1
burning gasoline with air produces both CO (carbon monoxide, which is toxic) and CO2 (carbon dioxide, which can cause suffocation, and which at high blood levels monkeys with the breathing reflex). in the classic car-exhaust death, it's CO poisoning, which will kill you much faster than just suffocating from lack of O2 in the air. basically, CO binds to hemoglobin even more strongly than plain oxygen does, so you wind up dying from lack of oxygen, even if there's still plenty of O2 in the air as well as the CO.

i have a hunch that the high temperature and pressure inside the combustion chambers of a modern engine have an effect on the CO to CO2 ratio of the exhaust. (i think that's also what causes the formation of oxides of nitrogen.)

I understand how CO poisoning works, the question asked of the Tappet Brothers was whether modern emission control systems are effective enough in controlling the combustion process within the engine and catalytic converter to basically turn the "classic car-exhaust death" into a thing of the past -- that new cars put out so little CO that it isn't enough to kill you.

Their opinion was, yes, modern emission control systems effectively control the emission of CO such that the "classic car-exhaust death" shouldn't happen with a properly-running modern car.

My concern was, assuming that was the case, that CO was not a problem, could you really run the gas tank dry without killing the driver? I think the answer is "no", as there isn't enough oxygen in a small garage to do the job.

and garages aren't airtight anyway...

That was my thought there will be some diffusion

"diffusion", my worn-out driver's seat! every garage i've ever been in has been a stellar example of the definition of the word "drafty". (there are quite a few people who use their garages as auxiliary refrigerators in the wintertime.) no way could a car use up all the oxygen inside a garage, because more air is always coming in.

The car talk people are also wrong about another issue- emissions tests are not for CO2 levels primarily. They test a variety of different substances. If I'm not mistaken they look mainly for particulate and ozone (but I could be wrong). They definitely aren't searching for CO2 though (MO maybe?).

Click 'n' Clack never said that emissions testing tested for CO2, which I admit would be pretty stupid. They implied that emissions testing tested for CO, among other things, which would make sense. I'd expect emissions testing would test for CO, NOx, unburned hydrocarbons, soot, and other products of incomplete combustion.

BTW, welcome to reading the backlog of my journal.

  • 1

Log in

No account? Create an account