Air compressor safety

[Sonic EscapeParticipant @sonicescape38234]

Since I didn’t have enough parallel projects, I thought of starting another one. Last Sunday I visited a friend who has several compressors and he wanted to sell them. I never really needed one. Except a short sandblasting experience. But one model caught my attention. The compressor itself is fine, but the electric motor is almost dead. It is smoking and can’t rise the pressure above 2 bar. Also it has a dubious makeshift single phase conversion with a capacitor.

But this is fine, I have a few motors that fits. Including the pulley that is matching the existing belt. It would be a nice repair project. So I bought it without thinking too much. But then I realized that I’m going to share the workshop with a 100-liter bomb. I started to read about the air compressors. An then to watch some videos with some scary explosions. And then I read some more. The main reason for catastrophic tank failure seems to be internal rust. Except some cases where they welded the tank 🙂

Now the problem with a second hand tank is that you can’t be sure how often was drained the water. And you can’t check if it is rusty or not. Would you use such a tank? I’m thinking to resell it. I don’t think it is worth the risk. What do you think?

 

 

 

[Nigel Graham 2Participant @nigelgraham2]

It would need to be tested hydraulically to be sure, or as sure as you can be.

The motor problem is a separate matter and your description suggests due to a previous owner’s bodge.

[Nigel Graham 2]

On Nigel Graham 2 Said:

It would need to be tested hydraulically to be sure, or as sure as you can be.

The motor problem is a separate matter and your description suggests due to a previous owner’s bodge.

That means to fill it with water and then to raise the pressure to … 150% of the tank limit? That would be 20 bar. For how long I should hold it to that pressure?

[Charles LamontParticipant @charleslamont71117]

I would first check it tapping the bottom part with a hammer, and if that does not indicate anything amiss, I would want to give it a hydraulic test. A presumed date stamp of 8-11-74 does not inspire confidence.

Considering the age you could de-rate it. You are unlikely to need a working pressure above 100 PSI.

You mention reselling it. You appear to understand the risk, but you are thinking of handing it over to someone who may not?

[Sonic EscapeParticipant @sonicescape38234]

So it is half a century old?! I saw that is was repainted but I didn’t realize is so old.

[Paul LousickParticipant @paullousick59116]

Air compressor receivers should be inspected regularly. Here in Australia all commercial air receivers have to be certified by a qualified inspector.

https://www.asseteng.com.au/blog/inspection-requirements-air-compressors-receivers/

Those that are used for private use should also be inspected but is not mandatory. BUT use at your own risk.

[Sonic EscapeParticipant @sonicescape38234]

I opened the drain and a lot of water with oil came out. I had a look inside and it is very dirty. It doesn’t inspire too much confidence. I don’t think it make sense to try a hydraulic test. It seems that for the first time I will bring something to the junk yard, not the other way around.

 

[Speedy Builder5Participant @speedybuilder5]

Does anyone know if these cylinders fail by leaking or rupture and explosion ?

There must be thousands of these around, anyone seen / had experience of a failure ?

Bob

[larry phelan 1Participant @larryphelan1]

When you opened the drain, how did you manage to look inside the tank ??

[Speedy Builder5]

On Speedy Builder5 Said:

Does anyone know if these cylinders fail by leaking or rupture and explosion ?

There must be thousands of these around, anyone seen / had experience of a failure ?

Bob

In my understanding it is very very rare for a compressed air tank to explode. But when it does it is very spectacular.

There were also fatal accidents. What I found interesting was one case when the tank explosion was not caused by rust or leaks but by wrong type of oil. It created a combustible mixture that ignited inside the tank.

On larry phelan 1 Said:

When you opened the drain, how did you manage to look inside the tank ??

My tank has at least 5 holes. Just put a flash light on one and watch trough the other. Also the USB endoscopic cameras are dirty cheap these days.

 

 

[Clive FosterParticipant @clivefoster55965]

The pressure driven explosion risk of these small compressed air tanks with an appropriately small compressor is vastly overstated. There just isn’t enough compressed air, and therefore energy, inside to blow it apart with bits flying everywhere.

internal oil – air mix explosions are a different matter and hugely dangerous. Fortunately ignition is very difficult to arrange and compressor oils are formulated to reduce the risk.

Usual failure is a small rust pin hole or partial seam split so the tank can’t hold air and the compressor will no longer bring it up to pressure. On anything where there is a weakness geometrically capable of creating a serious size hole or seam split the initial damage will be localised so air starts to be lost. As the damage spreads the pressure continues to fall rapidly until there isn’t enough left to drive the hole out or split larger.

Different story with a bigger tank holding lots more air as the pressure will hold up for much longer. In a practical world once the tank gets much bigger than the large propane / butane cylinders the amount of compressed air and energy inside is probably enough to blow small bits out before the drops the pressure fast enough. Say penny or 2 p piece size. The bigger the tank the worse it gets.

Personally I’d be more worried about the threaded socket for the gauge or air connector giving out and blowing the gauge, connector or plug across the room. Some are made looser than I’d ideally like.

Cheap regulator stems can be worryingly weak. Way back I bought  cheap a “customer returned faulty” compressor said to have a broken regulator. Turned out the wall thickness of the threaded stem where it emerged from the body casting was basically slightly more than nothing! I suspect it was broken off when the original purchaser screwed the regulator on by hand before reaching for the spanner.

Eeek!

That was SIP vertical, so called quiet, compressor that sits in the corner of the garage for tyre inflation and air over hydraulic bike ramp duties. No way on God earth would I use it for real workshop duties. Its loud, really loud, but for brief pump up maybe 10 times year it will do.

Less now I’ve got one of the Makita battery tyre pumper-uppers. Really good if you already use Makita and have the batteries.

But 4,000 psi plus really scares me. No short cutting there.

Clive

[Clive Foster]

On Clive Foster Said:

 

Usual failure is a small rust pin hole or partial seam split so the tank can’t hold air and the compressor will no longer bring it up to pressure. On anything where there is a weakness geometrically capable of creating a serious size hole or seam split the initial damage will be localised so air starts to be lost.

That is how my first cheap compressor failed. I heard air escaping and investigated. There was a pinhole in the bottom of the tank. It was about ten years old. I mentioned this to a chap I know who has been a (real world) boiler inspector. He said that was about how long they last. My new compressor is bigger and much quieter, is used less and gets drained every time.

[Chris GunnParticipant @chrisgunn36534]

Sonic, you could always try this.

Chris

[Martin KyteParticipant @martinkyte99762]

I had a Lab Bambi compressor tank fail a few years ago very unspectacularly. It just went fffftttt and decompressed. Failed on a bracket weld. OK it was only 60 psi and a small tank but air is nothing like the danger of steam. All that said I would tend to treat compressors with a degree of respect and replace if I have doubts. Not sure I would bother with hydraulic tests but the endoscopy gives a good feel for condition without too much fuss. Compressors are pretty cheap new when all said and done.

regards Martin

[Martin KyteParticipant @martinkyte99762]

Incidentally small steam boilers below a certain bar litre size are considered intrinsically safe. It should be possible to use this as a basis for a don’t worry about it classification for compressed air tanks below a certain size based on stored energy.

Anyone fancy doing some numbers?

Martin

[Chris Gunn]

On Chris Gunn Said:

Sonic, you could always try this.

Chris

No, I have a better idea. I could take it to a nearby forest, pour some water inside and then 0.5kg of calcium carbide. Run and ignite everything electrically from 100m away.

[Speedy Builder5]

On Speedy Builder5 Said:

Does anyone know if these cylinders fail by leaking or rupture and explosion ?

There must be thousands of these around, anyone seen / had experience of a failure ?

Bob

Plenty of examples of air-compressor explosions on the web:

As can be seen from the example above, ancient reservoirs are far more likely to fail than new ones.

A common cause is failing to keep the tank dry by blowing off every so often.   A puddle of water left in the bottom creates a rust line running around the tank at surface level.   After ‘n’ years rusting digs a notch into the steel, which flexes every time the tank is pressurised or discharged.   Instead of  distributing pressure uniformly over the whole tank, force concentrates on the notch, gradually levering it open, and encouraging more corrosion, until one day the steel tears open along the notch, concentrating enough force at the ends to rip open good steel.

Dings in the tank are also potential stress-raisers.

Difficult to predict how long a particular tank will last before failing catastrophically, because much depends on maintenance and how often the tank is used – metal fatigue.  In normal use, where the owner replaces the whole compressor as soon as anything major like the motor fails, they are safe.  Very unwise to repair tanks by any means.  The problem arises when Mr Fixer keeps an ancient tank in service for decades by constantly repairing it without checking what the insides are like!    Or by Mr Innocent only using a compressor once in a blue moon and leaving the tank with water inside, giving it plenty of time to corrode a notch.

How dangerous depends on how big the tank is, the pressure when it fails (usually high), being in a confined space, and whether or not the operator is hit by flying objects.   Although a fair amount of energy is released, the pressures are relatively low, so mostly survivable;  loud bang, probably not enough to rupture an ear-drum.   Although fatalities seem fairly rare, I wouldn’t risk using an old compressor.  I see old kit as being “clapped out” rather than “tried and tested” until proven otherwise!

Dave

[Martin Kyte]

On Martin Kyte Said:

Incidentally small steam boilers below a certain bar litre size are considered intrinsically safe. It should be possible to use this as a basis for a don’t worry about it classification for compressed air tanks below a certain size based on stored energy.

Anyone fancy doing some numbers?

Martin

How about a 100litre tank at 7bar, which is a pretty weedy but common size compressor?

 

Having seen the damage tyre and truck air tank failures can do, plus surviving personal stupidity of not depressurising a Cessna Caravan tyre before undoing the split-rim bolts, hoping that an old tank is going to just leak rather than fail catastrophically isn’t good enough for me. My 30 year old compressor needs new rings, starter switch, regulator and release valve. It’s only fit for the bin.

[Martin KyteParticipant @martinkyte99762]

Valid comments on tyres Nick but the energy stored in the rubber has to be a factor in making tyre explosions more hazardous I would think.

regards Martin

[Howard LewisParticipant @howardlewis46836]

When I first bought my comoressor, the supplier advised blowing down after every use, and leaving the drain valve open. (They knew that being a small compressor, it would be used intermittently).

I have always followed their advice, and had no problems.

The tank was supposed to be coated internally with plastic.

A working colleague lost three fingers when a larger reservoir exploded, due to oil carry over, so care is very necessary

In a previous job, with a big twin cyliunder belt driven compressor with a 200 gallon reservoir, we blew it down and drained it every week.

Howard

[MikeKParticipant @mikek40713]

For what it’s worth, I added extended drains to both of my compressors.  It gives the condensation somewhere to go instead of sitting in the tank causing rust, but also makes it easier to regularly drain it.

Mike

[Robert Atkinson 2Participant @robertatkinson2]

250 Bar/litre is the limit for air receiver certification. So for 7 Bar (105 psi) the maximum receiver size considered “safe” based on stored energy is 35 litres.

That receiver looks pretty badly pitted inside. Personally I would not trust it in the workshop.

Robert.

[duncan webster 1Participant @duncanwebster1]

If I had an inexhaustible supply of round tuits, I’d fit a solenoid valve to my compressor tank so that when I unplug it it opens and drains. I always try to remember to open the manual valve, but no-one is perfect

[John Doe 2Participant @johndoe2]

What happens to the stored pressure though when you switch the unit off and the solenoid valve opens ? Would it be OK for all the pressure to suddenly be released ?

If so, you could more simply engineer the electricity supply switch for the compressor motor to be moved on and off by a linkage to the lever that opens a drain valve – like that shown in MikeK’s picture above.

That way, turning off the electrical supply would simultaneously open the drain valve.

But, would that be OK without first bleeding off the pressure ?

[Martin KyteParticipant @martinkyte99762]

Robert Ackinson’s comment on 35 litres being the limit of intrinsically safe is helpful.

<li style=”text-align: left;”>Its easy to get misled by comparisons with Steam Boilers. When a Steam boiler fractures the pressure begins to fall and so does the boiling point. So huge amounts of steam is released from the water which holds the pressure up and rapidly enlarges the fracture. The process is highly non linear which is why boilers explode. Air receivers merely decompress and so need a much larger volume of compressed gas to create an explosive failure. I looks like 250 bar litres is a good guide to how much to worry. I suspect most of us will have small compressors well under this size. All compressed air needs respect and good workshop practice even if our small compressors are not ‘scary’.regards Martin

[Author]

Posts