Pressure testing my tanks

In late January I filled my three (330 gal) tanks to 35 psi of air. Within a week and a half the tanks dropped to 25 psi. For over a month now it has been holding at 25 psi. My brother says the reason is likely temperature of the air stabilizing. I could not find a leak by soaping the joints. Today I pumped it up to 40 psi. Any suggestions?

The ratio of the two pressures should equal the ratio of the two temperatures -- all in absolute units. Adding 14.67 psi to convert from psig to psia, the ratio of the two pressures is about 1.25. We add 459.67 degF to convert from Fahrenheit to Rankine. Just for perspective, assume that the tanks cooled to 32 Fahrenheit, or about 492 Rankine. To account for a 20% loss in pressure the temperature at 35 psig would need to have been 615 Rankine, or 155 Fahrenheit.

It does not seem likely that the temperature at 35 psig was as high as 155 Fahrenheit, nor that the temperature now is as low as 32 Fahrenheit, so it's very hard to account for all the pressure loss as being due to temperature loss.

Maybe it only leaks at pressures above 25 psig, in which case if you test with detergent solution at 40 psig you should find any leaks.

If you can get the tanks full of water with no air at all, you can connect a piece of 3/4" or 1/2" inch pipe that is also full of water, with a grease fitting on the end, and take the pressure way up over 100 psig with a grease gun. Any water leaks at high pressure would make it impossible to maintain a steady pressure, and if the water is cold you should be able to find moisture where any leaks may be.

But to me, passing a soap bubble test at 30-80 psig is all a guy needs to be convinced the subsystem is A-OK.

--ewd

EWD, what is your profession that you would know something like that? I have never heard of Rankine?
Stihlwoody

I agree with ewdudley. Pump it up to 60 psi and do your soap and water check. If there is a leak it must be small to hold 25psi. I figure it should hold twice the working pressure to be safe. Temps can definitly change the pressure.
When I was checking my 500 gallon tanks, I capped one off in the morning at zero pressure with cool morning air. When I took the cap off in the afternoon after it had been sitting in the sun it had a suprising amount of pressure in it.

Back in the day we learned ideal gas stuff and temperature scales both for merit badge work and high school science class. Granddad used to build his own SW radios from scratch after quitting school in the eight grade.

Rankine is a temperature scale that starts at absolute zero (which is as cold as it gets in theory) using fahrenheit units. It makes all the algebra come out more simply when calculating stuff like this. Kelvin is the scale from absolute zero using celsius units. Either one works the same as long as you're consistent with the other units in the calculations.
I wonder if the TV weather guys, who seem to have converted all their winter temperature statements into wind chill factors to pump up the excitement, shouldn't start using Rankine in the summer heat wave warnings. Now that would sound dramatic.

I just want to add a warning (not in Rankine units) about pressure testing tanks.

Propane tanks can take any pressure that the typical air compressor can pump into them. But I don't think that such high pressure is appropriate for homemade tanks (or a manufactured tank that is not a pressure vessel) if you're using only air. I think homemade tanks are safer tested almost full of water and then pressurizing the little air that is in them. A tank filled with water might 'pop' if it bursts, and make a mess, but a tank filled with only air can explode dangerously.

OK EW you did not answer the ? Mr smartypants LOL I have wondered the same thing every time i read a post by you.
Rob

Hey, I resemble that remark!

I do what they call 'embedded programming', which often involves measuring physical properties and controlling processes and events through time. The one and only time I ever ran across Rankine since high school was in implementing a fail-safe open-loop compressor discharge temperature control strategy for use in the event of a discharge temperature sensor failure.

The engineer had developed a model for predicting discharge temperature according to a bunch of factors, some of which were temperatures, all expressed in degrees Rankine, of all things. And the guy was Russian, which made it all the more baffling.

You wouldn't happen to have ball valves installed on those tanks already would you? I just pressure tested a quick plumbing change I made to my system and couldn't get it to hold 28PSI but it would leak down to about 20PSI and hold. Turns out it was the packing nuts on ALL of the ball valves in that leg. They were very, very slight leaks and would hold after the pressure dropped.

I had to tighten up the nuts quite a bit but I got them all to stop. These were Watts and Webster valves.

K

I also had a couple isolation valves that I had to tighten the nut on. same brand. I thought I was going to have to replace one of them, but I finally got it to stop leaking.