Horizontal vs Vertical Compressor and Tank

[hermetic]

From the video, which unfortunately doesnt give us a look inside the tank, the metal has not failed at a weld, but may have been weakened by water/rust along the bottom, failed there, and opened up the side. The first thing to notice is the very thin metal the tank is made from! My 1947 tank is 3/8 thick, tested to 300psi and WP is 150psi. I doubt this tank is even 3/16 thick! With a thin tank, it will swell as the compressor fills it, and contract as it empties. Could this, over time, work harden the steel? Supposing the tank has had water in it, and as we know, the rust takes place where the air meets the water, the failure point is very low down on the tank, so it can't have had a large amount of water in it. so must have been drained at least sporadically.

 

[spumco]

Could this, over time, work harden the steel?

Very unlikely. Fatigue, yes - work-harden no. There's not enough carbon or alloying elements in the mild steel, and I don't think that tank was made of something interesting.

Total guess is that ID surface corrosion reduced shell thickness adjacent to the drain bung weld (among other places). Compressor tanks are, by definition, in cyclic service and the heat affected zone adjacent to the bung weld probably experienced a fatigue crack. The crack probably initiated at a weld defect (root undercut, incomplete fusion) which was otherwise fine, but was not once the thickness was reduced by corrosion.

See 2:09 in the video. Looks like the edge of the bung weld is at the crack face, and that's the downhill end of the tank where water would collect.

The crack propogated along the long axis first as the hoop stresses are greater than longitudinal stress. Then the crack did a right turn up the shell for no obvious reason - perhaps some discontinuity in the steel from the mill.

I thought I heard the compressor running in the video. So we don't know if the relief valve stuck and the tank was subject to an overpressure event. However, the shell and heads are not distorted, so I'm guess it unzipped below MAWP.

 

[matthewsx]

I just looked inside my compressor tank (80 gal vertical) with an endoscope and didn't like what I saw. I'll unscrew one of the big bungs on the side so I can look better but I may be looking for a new tank. Oh well, I traded it for a couple of hours of computer consulting so other than the few bucks I have into parts I don't have anything into it. I have a good local compressor service company so I'll see what they can do for a new tank but I'm not going to run it anymore. No sense in having a bomb in the garage....

John

 

[forhire]

If you search google images for compressor explosions most of the images are of horizontals. It would appear the longer drain path weakens the belly leading to failure. Verticals generally start leaking around the drain plug rather than being filleted open. That said, all recip compressors can coke up the supply line which can lead to internal oil vapor ignition. The dinky pop-off valve cannot handle the rapid overpressure and the vessel will explode catastrophically. Use of correct compressor oil and inspect the supply side for coking.

 

[mksj]

If you search google images for compressor explosions most of the images are of horizontals.

True to a point, but misses the reason. The vast majority of the images/records are small 20-30 gallon cheap compressors that have been poorly maintained, and are of questionable manufactured standards. In many cases people do not drain the water or do it infrequently at best, the tanks/compressors are old. Not saying it doesn't happen with mainstream manufacture compressors, but can't say I have seen any photos of say Quincy, Champion, IR, etc. 80-120 gallon tanks posted be it vertical or horizontal. In most cases there is often an additional co-factor of misuse such as no safety valves, over pressurization, drains not working, etc. Looking at the smaller tanks, often you will see the end bells blown off or the rupture originating around the drain weld. According to other posts on the subject the vast majority of failures are more pinhole leaks as opposed to catastrophic failure.

So the bottom line, buy a decent brand compressor/tank with a track record, if it is an old tank/compressor have it inspected. There is probably a reason it is in the junk pile or you got it so cheap.

Metropolitan Engineering Consulting and Forensics - CAUSE OF EXPLOSION OF AIR COMPRESSOR TANKS

CAUSE OF EXPLOSION OF AIR COMPRESSOR TANKS

sites.google.com

air compressor explosion pictures - Google Search

: