Explosion hazards

[Tom Bender]

When working at a factory I saw a natural gas odorizing station in a separate building. Good so far but the building was built of concrete blocks, about 12 ft x 20 ft x 10 ft. No relief panels, guess they thought it would contain it. That's a lot of potential explosive! Had I been given the assignment it would have been a sheetmetal building. Just consider whether you want flying sheetmetal or flying concrete blocks.

So how about a shop air compressor? If you are concerned about a tank explosion and decide to put it into a separate room, it makes sense to have the stronger wall on the shop side.

[roger wiegand]

We had a high pressure lab at one of the places I worked. It was built with very thick, heavily reinforced concrete walls and a very weak roof-- the idea was to direct any explosive event upward, away from the occupied spaces around it. Fortunately it was never tested during my tenure there.

[Eric Arnsdorff]

There are different ways to mitigate such a risk.
There is little risk in an explosion from the compressor itself imposed on the tank. However, one of the easiest mitigations for this is a safety relief valve. Your assembly may have this already installed. There are other ways as well.
Basically, you need something that can relieve the pressure faster than it can build above the capability of your tank.

I assume the explosion you are concerned with is something like a brittle PVC failure with the air propelling the PVC pieces as shrapnel. The typical compressor tank is made from ductile steel and the associated failure from compressing the tank would be an opening created in the tank which would relieve the pressure. These tanks should meet basic weld integrity requirements and such. Even without that, the failure of a weld seam as it is being compressed would lead to an opening first prior to a complete tank end coming off.

My advice would be to install a safety relief valve if your tank isn't already equipped with one. Other methods are rupture disks and such. A rupture disk is one level above a relief valve in that it will always fail safely.

I would use mitigations such as that as opposed to building or requiring a containment structure.

[Bill Dufour]

The dynamite plant near were I grew up the buildings were gone but the rows of trees were still there. Not the normal way to plant rows. staggered so there is no horizontal sight line from the building. I understood it was the same idea weak roof and walls but the trees would keep the blast on their property.
Rather then waste money on a containment building I would spend 1/10 of that on plumbing in quality multiple relief valves. Consider having the tank in the basement for containment.

[Alex Zeller]

When working at a factory I saw a natural gas odorizing station in a separate building. Good so far but the building was built of concrete blocks, about 12 ft x 20 ft x 10 ft. No relief panels, guess they thought it would contain it. That's a lot of potential explosive! Had I been given the assignment it would have been a sheetmetal building. Just consider whether you want flying sheetmetal or flying concrete blocks.

So how about a shop air compressor? If you are concerned about a tank explosion and decide to put it into a separate room, it makes sense to have the stronger wall on the shop side.

Usually buildings like that are designed to blow the roof off of the building, not the walls outward. Much less chance if injury and damage. If I was going to design a room for an air compressor explosion it most likely would be cement (to contain any metal from the tank) with some place for the air to go when released that would do the least damage.

[Malcolm McLeod]

T...
My advice would be to install a safety relief valve if your tank isn't already equipped with one. Other methods are rupture disks and such. A rupture disk is one level above a relief valve in that it will always fail safely.

I would use mitigations such as that as opposed to building or requiring a containment structure.

I have dealt with automation systems for everything from bombs and rockets to high pressure n.gas, and Mr. Arnsdorff is dead on here (no pun intended). Relieve any over-pressure to a safe location/area/direction - - long before any associated vessel undergoes a catastrophic failure. Since Mr. Bender is concerned with gases, any flammable gas goes to a flare, air can be vented into the same room (or outside if noise* of venting or even room pressure is a concern).

Blow-out panels and concrete walls are for explosives which cannot be 'vented'.

You can get a hand pump to validate the actual relief pressure and function of a relief valve on your average shop air compressor. And if the unit was sold in the USA, it WILL have one - - unless the system was cobbled together by a... uuuhhmm, well ... someone who's IQ will fit in a shoe box. (It is a good idea to test them yearly. Everyone does that. Right?)

Most industrial air compressor shops should also be able to test a relief valve for you.

*- The noise of a 6" PRV letting go at >500psig when you are 40ft away, will get your attention. Velocity is well above Mach 1.

[George Yetka]

I've worked in a few places that have been designed in a way to let them explode and very easily be cleaned up and get back to operational. Generally they were designed to guide the blast to easily replaceable sections of building. Block is great but can only direct the blast not contain it. So like mentioned above a sheetmetal roof actually works well. It generally doesnt blast away but blasts like a soda can with a firecracker in it(holds together just blasts out)

[Steve Rozmiarek]

The typical compressor tank is made from ductile steel and the associated failure from compressing the tank would be an opening created in the tank which would relieve the pressure. These tanks should meet basic weld integrity requirements and such. Even without that, the failure of a weld seam as it is being compressed would lead to an opening first prior to a complete tank end coming off.

Unless the tank is not maintained properly, water is not drained and the rust that results creates undesigned weaknesses. That's what happened with the one that catastrophically failed and hurt my dad. It ruptured in a long tear, turned itself partially inside out and apparently took to flight. It was a larger portable size horizontal tank, ruined a concrete block wall near it and my dad's leg when some part of the mess hit him. He was lucky. I got to help with the shop cleanup, it looked like a bomb went off in there. Obviously the tank was faulty, it was an auction purchase, you could see that the tank had lost significant thickness to rust along the bottom.

[Charlie Jones]

Rust can weaken a pressure tank. I worked in the Truck Maintenance field for several years. We always had large pressure tanks that were inspected annually along with the relief valves. We never had a problem. However, a sister company was testing a 10,000 gallon fuel tank with compressed air. The entire end blew off at 3 PSI. It knocked down a concrete wall. Even at 3 PSI it had a huge volume. Air compressor tanks should have drains or dryers installed to keep moisture buildup from forming.

[Malcolm McLeod]

...rust...

Rust ...

The particular service (steam, air, etc.) and operating conditions that a pressure vessel is used for can vary wildly, so equally broadly speaking, a vessel should be inspected by a licensed inspector at least every 5yrs as part of its ASME rating. This would identify the rust, cracks, embrittlement, or other issues that compromise the vessel's rating, and so prompt its removal from commercial service (or de-rating, including required mods to any PRV/SRV devices).

I'd suspect ANY commercial operation would be driven to this standard by insurance carriers. For home or small commercial shop use, consider why that auctioned compressor is such a screaming deal. ...Maybe it flunked its last inspection??

[Jerome Stanek]

I worked for a crop service that had a spray truck with dye markers that used a small air tank and about 5 pounds of pressure it had a quick release cap on it and one day the operator flipped the release with out draining the tank and the cap must have gone 100 feet in the air. He was lucky not to be in the path.

[Bruce Wrenn]

Think about this, steam engines ran boiler pressure between 250 and 300 PSI. This was in a boiler that was mounted to a frame that was running down the rails at as much as 100 MPH. No welded ribbon rail, just bolted, which meant there was a joint about every 39 feet on either rail. Some of the boilers were close to 100 feet long. As boiler was brought up to pressure and temperature, it grew both in length and diameter. I have no clue as to why they actually worked as good as they did. Remember when a boiler blew, the pressure on the water was instantly reduced, allowing the remaining water to boil even more rapidly.

[Tom Bender]

a vessel should be inspected by a licensed inspector at least every 5yrs as part of its ASME rating. This would identify the rust, cracks, embrittlement,

Anyone have such an inspection done on a small tank? Didn't think so.

Were I the inspector I would never pass one because it would not be possible to know the condition at any reasonable cost.

[Malcolm McLeod]

Anyone have such an inspection done on a small tank? Didn't think so.

Were I the inspector I would never pass one because it would not be possible to know the condition at any reasonable cost.

I suspect owners of small pressure tanks in a (home) shop environment don't inspect them for 2, maybe 3 reasons: owner considers it a disposable (I'll replace it when it breaks); owner doesn't know how or who can perform this; and maybe a few owners just unaware/uncaring (it just sits in the corner and compresses air, ergo tank is fine too).

If the condition of the tank is a concern, drain the condensate and look at it. Is it rusty? Are there flakes or just colored water? Large flakes are bad. Tap on the tank. Does the bottom sound different than the top/sides? On small units, roll it side-to-side; there shouldn't be anything sliding around in the receiver. Likely a rat didn't crawl in and expire, so if there is sliding, it could be that time. Or better, simply pull the drain port or even the top 'charge' port and look. For $21 you can get a borescope, or just use a flashlight and a Mark I eyeball. If you can see flaking, especially large 'spalling' flakes, then it might be time for a new receiver. All of these are an 'inspection', inform the owner of condition, and I'd argue the cost is reasonable. YMMV

I am not an ASME licensed inspector, but I inspect my receiver. A little common sense goes a long way.

Lots of companies perform non-destructive testing (search "NDT"). You be surprised how affordable it is, and informative, if the above shade-tree-methods are inadequate.

[Steve Mathews]

I don't know if this will be effective against a tank rupture but both of my air compressors will be located in a shop cutout (old horse stall) that is open to the outside. It's also lined on all three sides with concrete blocks. The reason for locating them there initially was for sound abatement but it's good to know that it might help if a rupture occurs. I'm attentive to draining the tanks of condensate but will look at the other preventative measures mentioned here for additional safety precautions. This is a good thread. I'm glad it was posted.