Household gas: rises or falls?

February 13, 2006 5:00 PM

Does household natural gas pool in low places or rise to high places?

We have an occasional "gas" smell at the base of our internal stairs (we live at the top of them). I'd thought perhaps our appliance pilot lights and or ignition delays were leaking a little and it collected down there. The gas man came by and said that wouldn't make sense because the gas would rise. Now my desire to give the repairman specific hints to reduce his time/cost, and my scientific and practical curiosity are tickled.

Hints on identifying the leak source will earn extra points, but please be aware we know enough to have the gas turned off already, we'll be hiring a professional to do any actual work, and we'd just like to gather a little additional data before taking further and costly steps.

posted by freebird to home & garden (19 comments total)

Google thinks natural gas is lighter than air.
posted by mbrubeck at 5:10 PM on February 13, 2006

Yup, natural gas is lighter than air, butane (which you could get from gas bottles) is heavier.
posted by fvw at 5:19 PM on February 13, 2006

Yep, I've just had enough professionals tell me it would pool in low spots that I thought it was worth checking for some practical "gotcha" about how it's generally delivered, or temperature differentials, or something, that would explain it seeming to collect in the lowest part of the house if it was leaking from appliances in upper areas.
posted by freebird at 5:23 PM on February 13, 2006

Natural gas should rise to high places, assuming that it isn't much colder than its surrounding air.

Here you can find the composition of natural gas. There's also a quick primer on gas densities here. Finally, you can learn about the composition of air here.

You can see from the above that methane is the main component of natural gas, and that nitrogen is the main component of the air we (normally) (should) breathe. Great! Now you can geek out with a periodic table of elements and add up the molecular weights of the molecules in question, as this quick guide on calculating molecular weights describes. Or you can just go to this javascript-based molecular weight calculator.

Ok! So if you did all that, you'll see that the molecular weight for methane (CH4) is 16.043 g/mol, and that nitrogen (N) comes out to 14.0067 g/mol. Methane, in other words, is denser than nitrogen, assuming constant volume and number of molecules. But it's more complex than that. You need to calculate the specific gravities of the gases in question, maybe by using this free online calculator. You could also just look up the overall specific gravities of natural gas and air in this table of commonly-used specific gravities. Anything with a specific gravity of less than 1.0 means that it'll rise in air, assuming it's a gas.

Note that mercaptan (CH4SH), which is added to natural gas to make it stink, is a bigger molecule. I'll leave it up to you to figure out its molecular weight and specific gravity, but you may discover that that's what you're smelling.

Finally, you may want to consider that the human nose quick acclimates to stable environmental odors. Thus, you might not notice the smell after a short period of time. Similarly, if you farted in a closet, you'd get used to it, assuming the fart gases didn't break down or otherwise change over the time of smelling.
posted by herrdoktor at 5:32 PM on February 13, 2006

To find a natural gas leak, try rubbing some a mixture of 5 parts water, 1 part dishsoap on any pipes or tubes or sockets that gas might pass through. I've applied the mix successfully with a toothbrush. If natural gas is present, it should hiss and bubble...it will be very obvious.
posted by charmston at 5:45 PM on February 13, 2006

By herrdoktor's logic, all the oxygen should be down around the floorboards, and those of us walking around with our air intakes at the lofty height of 5 feet or more should be choking to death on pure nitrogen.

That doesn't happen.

Of course, gases mix; the process by which this occurs is known as diffusion. The rate of diffusion is inversely proportional to the square of the molecular weight (I think; high school chem was a long time ago). Methane, being relatively light, diffuses very rapidly in air, and thus it is difficult (not impossible) to create a large 'pool' of methane gas anywhere that is open to the out-of-doors.
posted by ikkyu2 at 6:20 PM on February 13, 2006

Molecular nitrogen is N2 so the weight would be 28 grams per mol.
posted by 517 at 6:49 PM on February 13, 2006

Yes, herrdoktor is totally wrong (and ikkyu2 is right) because he doesn't take into account molecular diffusion. Two "ideal" gasses will expand to take up all of the available space, and mix with each other. Air is a heterogeneous solution.

In order to calculate a density gradient, you'd need to factor in diffusion speed, compared to the effect of gravity. You could get a general idea of how much separation there would be by comparing the X velocity of an individual molecule of substance A (like methane) to the velocity of a molecule of substance B (like N2 and O2) and their gravitational acceleration –9.8 ms².

So just compare the effect of gravitational acceleration on the partials per unit of kinetic energy to the diffusion speed.

Anyway, you'll probably need to use differential calculus.
posted by delmoi at 7:44 PM on February 13, 2006

(Well, actualy you might be able to get by without diff eq, I have no idea)
posted by delmoi at 7:46 PM on February 13, 2006

Let me see if I can say this more clearly.

Another way to think about it: one kilogram of methane at 72&deg has the same amount of kinetic energy as one kilogram of N₂ at 72°. But, because the velocity component of that kinetic energy is different, one will be more effected by gravity.

Similarly, because the velocity will be different, the diffusion speed will also be different.
posted by delmoi at 7:58 PM on February 13, 2006

delmoi: Essentially, you're talking about deviations from the Ideal Gas Law, which states, among other things, that at standard temperature and pressure a mole of Ideal Gas will fully and evenly occupy a 22.7 liter volume.

Obviously, gases aren't ideal. But for the purposes of anyone but serious molecular-chemistry geeks, the Ideal Gas Law describes the behavior of gases - they expand in predictable ways to fill the volume available to them.

If you are in a source-sink environment, like a building with radon seeping from the wallboards, or a closed room with a methane-gas leak, you also need to know about diffusion speed to understand how the gas is going to behave. But, for your first approximation, you do not need to worry about deviations from the Ideal Gas Law, such as the fact that non-ideal gases are composed of massy particles; these deviations don't change the behavior of gases in ways that non-quantitative estimators care about.
posted by ikkyu2 at 8:08 PM on February 13, 2006

Whoops! I had assumed that the spec gravs listed for "air" and "natural gas" took that into consideration, as as measurement of the components of the gases mixed together. My apologies for supplying the wrong answer.
posted by herrdoktor at 8:58 PM on February 13, 2006

We've digressed a bit from the original question. Short answer: yes methane is lighter than air. You won't get pooling at the bottom of your stairwell. Because of the way residential units are built, it's not likely to become a risk. The major risk with methane is explosion. Suffocation is a risk, but not likely int he situation you describe.

There's no simple way to detect methane. The cheapest is a small tube you draw air through to watch a coulour change (the most common being a Draeger tube). There are a variety of explosion meters (usually packaged as a combination "four gas" meter) which can be calibrated for methane. Finally, the best methods involve portable mass spectrometers, expensive, but widely used by the oil and gas industries for pipeline work.

Your best bet is your nose, smelling for that rotten egg mercaptan smell. You become desensitized to mercaptan in a couple of minutes however, so you'll need to take lots of breaks. Give yourself at least fifteen minutes.

If it's most noticable at the bottom of the stairs, it's most likely coming up from below. Check the basement first, furnace, water heater or dryers, then the unit on the lower floor. Visually check that the pilots are all lit. You should be able to see a blue flame in each of the units.
posted by bonehead at 7:25 AM on February 14, 2006

Essentially, you're talking about deviations from the Ideal Gas Law

No he's not. He's arguing from classical ideal gas kinematics. As long as you only consider energy and momentum conservation as he's doing, you're talking ideal behaviour. Ideal gases are point masses of non-interacting particles. A no-mass gas model is called a vacuum.

Non-ideality happens when molecular interations have to be considered. Size, in the form of a collision-cross section, is the simplest deviation from non-ideality. Shape is an added wrinkle. Intermolecular forces, coulomb, magnetic, force-potentials, etc..., are harder. Quantum dynamical statistical mechanics is (so far) the farthest anyone has been able to push gas behaviour.
posted by bonehead at 7:30 AM on February 14, 2006

Anyway, you'll probably need to use differential calculus.

You would, exact people already have and gotten the answer. P=P₀e^-mgh/RT, where P is the partial pressure of the gas at the altitude you're calculating, P₀ is the partial pressure of the gas at a reference altitude, R is the universal gas constant, T is absolute temperature (although things get messier if temperature isn't constant over the altitude difference), m is the molecular weight of the gas, g the acceleration due to gravity, and h the height difference between the two altitudes.

So at the top of a 1000-foot building at 25°C (298K), there is 96.7% as much nitrogen as at the building's base, but only 96.2% as much oxygen as at the building's base. If there were methane in the same building, there would be 98.1% as much at the top as at the bottom. Methyl mercaptan, 94.4% (This assumes the air is still and the building is airtight; ventilation in an real-life building is going to mix things up more.) But yeah, over the height of a typical house the difference is negligible.

As a practical matter, not only do you have ventilation in your house (which will mix things up more than if the air in the house was still), it's not airtight. If you have a gas leak, you're going to get a diffusion gradient such that it's highest closest to the leak, and decreasing as you get farther away from it; since the house is not sealed, some of the gas escapes from the house. Or, to put it in simpler terms, you smell gas near the base of the stairs because that's near the gas leak, not because the gas is leaking from somewhere else and collecting there.

P.S. I think what ikkyu2 is thinking of as the ideal gas law is the traditional PV=nRT taught to first-year chemistry students. That equation, in and of itself, assumes the height difference across the volume in which a gas is contained is not large enough to be significantly effected by gravity. But bonehead is correct in that a) ideal gas kinematics emphatically do not assume gas molecules are massless (if it did, it couldn't say anything about temperature, since zero mass = zero kinetic energy, and kinetic energy is related to temperature), and b) the same assumptions from which PV=nRT is derived also predicts the dependence of gas density on height and gravity, without having to take into account any of the non-ideal behaviors bonehead lists.
posted by DevilsAdvocate at 8:18 AM on February 14, 2006

You may be smelling sewer gas. Do you have any drains near the place you notice the smell? The drains may have lost the water in their traps allowing sewer smells to enter.
posted by MonkeySaltedNuts at 6:38 PM on February 14, 2006

Great stuff all, thanks. I'm really enjoying the chemistry and physics. As a practical update: it turns out that our gas lines were installed to run gas lights, that's how old they are. So it was in fact gas at the base of the stairs and no, it wasn't just trickling down from an appliance.
posted by freebird at 11:45 AM on February 15, 2006

(up! up! not down!)
posted by bonehead at 7:10 PM on February 15, 2006

DevilsAdvocate, you're right, that's what I was thinking of. Enjoyed the follow ups; it never occurred to me to consider the effects of gravity on an ideal gas.
posted by ikkyu2 at 8:18 PM on February 15, 2006