6

I've recently learned that old air compressor tanks can catastrophically fail. I can't find anything about the average service life of a tank, but it appears the ones that do explode are typically fairly old (they appear to be around 40 years old).

What is the simplest way to test an air compressor tank? How frequently should this test be performed?

From what I understand, the only option is to hydrostatic test the tank to about double the working pressure. The options are to use a pressure washer (with water), a grease gun (which may be difficult to use water), or a hydraulic pump modified to use water, each of which require about $100 of equipment.

  • 1
    Have you considered taking the air compressor tank to a place that does pressure vessel testing on a professional basis, such as a welding-gas shop or a dive shop? Even if they don't have the gear for a tank of that size themselves (and many shops that deal in industrial gasses will I reckon), they'd be able to inform you much better than most of us blokes here ever could :) – ThreePhaseEel Jun 18 '17 at 14:55
  • If this is an expensive, professional-grade compressor, it would probably be worth the cost of testing. If it's old, however, the money might better be spent just replacing the tank instead of testing, which you might end up doing anyway after spending money on the test. If it's an inexpensive, homeowner-grade compressor, don't even bother testing. just replace the whole thing with a new one when they go on sale. – fixer1234 Jun 18 '17 at 18:55
  • In a pinch, if water drains clear it's unlikely to have a serious corrosion issue. Tapping lightly along the bottom will usually reveal the tonal difference of corroded spots compared to the sound of tapping the top where corrosion rarely occurs. I kept 2 of 3 based on this. – James Olson Jun 24 '17 at 20:38
  • Yes @ThreePhaseEel. The ones I called were clueless about testing air compressor tanks. Unfortunately, mine has drained with rusty-colored water since it was practically new James Olson. – Alex Lauerman Jun 27 '17 at 2:41
8

Take it to a shop that either does hydrostatic testing, or has an arrangement with a place that does. It will cost far less than $100, barring a major ripoff. Any place that deals in compressed gasses (welding supply or diving supply) needs to do this on a regular basis (5 years per tank), as do fire extinguisher service shops (which might actually be the place your dive shop sends a tank to for testing.)

The formal tests (in the USA) are based on Department Of Transportation regulations for portable tanks - fixed tanks are not required to be tested (at least by the DOT), but can be. Twice working pressure may be more than is correct - I'm more familiar with dive tanks, and there it may be as little as 5/4 rated pressure - failure is not "burst only"; a tank can fail the test because it permanently deforms (the tank is submerged in a tank of water while being tested - the expansion of the tank when it is filled causes a measured about of water to leave that tank. If the tank being tested does not shrink adequately after pressure is released, the tank fails testing.)

Most fixed tanks fail from internal rust due to lack of proper maintenance (draining the tank regularly.) If your tank is 40-50 years old, it might be worth simply replacing it with a new one, particularly if you have difficulty getting it tested at a reasonable price.

  • Thanks. I searched around on the internet and called a few places locally, and I'm having trouble finding a place that will do it. One welding gas place said they ship their tanks to another major city to have them tested. Your answer sounds like exactly what I need, but it doesn't seem like many places offer general consumer hydrostatic testing. – Alex Lauerman Jun 27 '17 at 1:07
  • @AlexLauerman -- yeah, it's usually done in the context of DOT cylinders...you may have to follow up with that welding gas place and ask them if they can give you a referral :) – ThreePhaseEel Jun 27 '17 at 2:50
3

There are several other ways to test a pressure vessel.

Also, a hydro is not necessarily double, it's usually 125-150% of the safety valve pop, depending on the situation.

A popular method today is ultrasound. This exploits the fact that the tank usually sits at one orientation, so accumulated tank water rests in one area, and that is where corrosion happens. You use ultrasound to measure the tank (metal) thickness in the bottom 1/3 or so of the tank and look for deviations.

There is a method for drilling holes to proscribed depth and position in the tank, to trigger tank failure in a particular way when it fails, and to effectively "rip-stop" a rapidly expanding crack. Not surprisingly, the holes go in the same area where water acumulates and where the metal is likely to corrode.

Another technique is using a borescope to look in the tank. That doesn't replace other testing techniques, but it tells you where to test.

  • Drilling holes in the tank is sort of counter-intuitive for a high-pressure vessel. How do you make a pressure-proof, corrosion-proof seal on the hole afterwards? And wouldn't testing that put you back at square one? – fixer1234 Jun 18 '17 at 18:42
  • It's not a test. It's a safing method. It assures that corrosion pinholes one of the holes first. Or if a tear in the metal starts to happen, it snags up on the holes. Asymmetry in pressure vessels is a very good ripstop. – Harper Jun 18 '17 at 20:51
  • @fixer1234 The way I understood that comment: you don't drill all the way through the tank wall, just make a little divot on the surface. And I assume that wouldn't be a DIY operation, but something they do at the factory. – Mike Baranczak Jun 18 '17 at 20:51
  • If a tank without holes can rust through (in one place) and fail catastrophically, instead of just leaking in one place, I wonder why the holes would help, or how effective they would be. If they do work, that sounds like a perfect solution. I would love to know how deep to drill those holes. Getting the depth right seems easy using a drill and a micrometer. – Alex Lauerman Jun 23 '17 at 14:38
  • That's what happens, the corrosion thins the metal, and a tear starts and advances quickly enough to grenade the thing. Drilling takes a bit of engineering, the trick is selecting the right depth, getting a pattern that will work, and a drill stop of some kind, because if one is too deep, yer done. – Harper Jun 23 '17 at 15:38

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.