Compressor filter melting

[Wookie]

I aqm following another thread in the DNY section of the board with great interest, regarding the melting down of a compressor filter. The thread deals with the operator and if they should have notified, shouldn't have notified, etc.

I want to discuss the mechanics of a filter meltdown. I have seen compressors explode, final seperators explode, crankcases explode, filters installed incorrectly explode, filters get so saturated with oil or water that they break through, filters get so hot the hopcalite releases the CO back into the air, and heard stories of oil getting so hot it releases toluene into the breathing air supply. I've removed filters so old they fall apart when you pull the top cap off. Note to self - if that ever happens again, please turn the entire canister upside down to remove the filter element, otherwise you get filter junk throughout the air system. I have seen the wrong filter installed, the right filter installed incorrectly, and filters manufactured out of tolerance, resulting in mechanical failure of the filter. I have never seen a filter melt down, but I haven't seen it all yet. I can imagine a fire in the filter chamber, but that compressor won't come back online until many new parts are ordered.

Talk has been made of the high ambient temperatures in the fill shack of this particular facility, which I have never visited, but it can't be any higher than on any liveaboard that operates in the Caribbean. My compressor room is at ambient 120 degrees during the dive day, and we've never melted a filter.

Can anyone explain the mechanics of a melted filter?

 

[Cave Diver]

Can anyone explain the mechanics of a melted filter?

I was a bit curious about that myself.

---

---

A ScubaBoard Staff Message...

** Note to posters: This thread was originally in the Mishap Analysis forum, which means that responses won't show until approved my a moderator. It's been moved to Compressors and blending systems for better participation**

---

 

[cool_hardware52]

Can anyone explain the mechanics of a melted filter?

Pure speculation on my part, but I'd be looking for any restriction on the inlet side of the compressor.

The compression ratio impacts the output temps of any compressor.

For example pumping from 1 ata to 207 ata (~3000 psi) is a 207:1 ratio

If the inlet is restricted and the gas entering the pump is at 0.75 ata the compression ratio jumps to 207/.75 = 276:1

This link provides a detailed explanation of the effects of restricted inlets on output temps. The Deco Stop

Tobin

 

[Cave Diver]

Good info Tobin, thanks!

 

[swamp diver]

After a quick read through of the thread in question and reviewing the photo of the compressor shack at the facility it is likely there are real issues with inadequate compressor cooling.

One poster also correctly indicates that the risk of overheating is far greater when the compressor is run continuously during the hottest part of the day rather than having a set of storage banks such that the compressor can be run in the early morning and after the sun goes down in order to minimize the ambient temperature in the compressor room. Several posters mentioned the problem started around 2 pm which would be the hottest part of the day.

Typically the compressor is installed too close to the wall such that the intake fan cannot get the cool air it needs to maintain stage temps slightly above ambient. A heat loop is set up where the fan intake air temp continues to increase as it pulls increasingly hot exhaust air.

Filling high pressure tanks also leads to higher stage temps and risk of overheating.

Probably multiple issues involved here but poor cooling is root cause. Issues to review:

1. No restriction on compressor inlet as per Tobin's post.

2. Bauer limits its maximum compressor operational ambient temp to 113 F but in reality even this is pushing the limits of some compressor designs and oils. The correction factor to the filter hour capacity at this temperature is about 0.2 if I recall correctly. Best to hang a thermometer and not operate if the room temp is above 100 F else you are just going to moisture saturate the filters early and risk driving the stored volatiles off the activated charcoal bed into the breathing air or produce carbon monoxide via pyrolysis of the compressor oil.

3. Install storage banks so compressor can be run in the coolest part of the day rather than the hottest.

4. Open up back of shack and install large industrial fans to remove heat generated by the compressor(s). Don't run multiple compressor at same time without large amounts of mechanical ventilation.

5. Ensure compressor has required amount of space between inlet of cooling fan and on other three sides as per the manual. Some Bauer manuals now state a minimum of 36 inches between the fan inlet and the wall.

Some plastics melt at 300 F which is a temperature easily achieved if the ambient room temp is > 100 F and the compressor is poorly installed and ventilated. We have measured external stage temps on compressors installed to spec at 300 F which means the actual air stream temp will be much higher.

Heat is a four letter word in the breathing air compressor business.

The operation should also think about installing an Analox CO Clear monitor on the compressors. It likely would have picked up the incident early on and prevented the mishap from involving multiple tanks.

 

[iain/hsm]

I aqm following another thread in the DNY section (EDITED)

Can anyone explain the mechanics of a melted filter?

Frank can you direct me to this DNY section before I post fully I would like to see the info and photos first just to be sure.

IMHO it's not heat entirely but a plasma fire in the chemical zeolite. Rare but not uncommon in Europe with Bauer compressors my guess its its a Bauer.

Heat isn't the cause its only a contributing factor.

Long chain hydrocarbon, oil, synthetic diester, yada yada cling to the external surface of the molecular sieve. the active ingredient is zeolite.

A plasma fire in an air compressor filter is caused by two factors I can think of.

The oil is coating and forming a "seal" on the zeolite chemical ball keeping it away from the water vapour and sealing it in fresh.

When you have a auto dump on a compressor if this massive pressure change is allowed to affect the filter such as these cheaper filter/separator units you see this change in pressure on the zeolite that is saturated with water condensate expands and freezes as the pressure decreases in effect turning that bit of the zeolite to powder. The powder is pretty useless and ineffective and the adsorbing and absorbing "front" moves along the filter chemical length.

But the still potent "sealed fresh" balls of the original zeolite are left behind until a water flush removes the oil from them and it is the huge exothermic release reaction that is the cause of the plasma fire, oil heat etc only help it along together with the fresh air from the compressor feeding the fuel with the oxygen it needs.

The heat generated will melt plastic easy, on the plasma fires I have seen it also removed the anodising off the Aluminium.

Needless to say the critical point is for the filter shells to be destroyed take a hammer and chisel to the threads and get a photo published as I know full well some filters have been known after fires to find there way on E-Bay and the like.

The other cause for these plasma fires is failure of the auto dump and liquid water hitting the zeolite.

Just for grins take a small styrofoam cup and fill it with zeolite (molecular sieve) then an identical cup filled with cold water at room temperature.

Stick a thermometer into the chemical and pour the cold water over it the exothermic reaction should reach around boiling point ( I would have said 100 degrees C but I know you guys wouldn't understand if that was hot or cold LOL )

The plastic will also melt so be careful.

Its a long subject and too big for a single post. Hope it helps.

Oh yes one other is that we have high back pressure setting in Europe 200 bar 3000 psi this gives us longer filter life in hotter climates but the trade off is hotter compressor working harder wears out faster but the air is super dry. Again a dump failure allowing liquid water and hot oil carry over into the filter and the plasma effect is much more of a risk. Its a massive subject sorry I cant do more.
Iain Middlebrook

 

[Cave Diver]

Frank can you direct me to this DNY section before I post fully I would like to see the info and photos first just to be sure.

I believe this is the thread that Frank was referring to, but I can't recall any pictures in it. Just discussion of the event and the response by the dive op.

http://www.scubaboard.com/forums/dive-ny/338048-contaminated-air-fills-dutch-5-30-5-31-a.html

 

[Wookie]

Frank can you direct me to this DNY section before I post fully I would like to see the info and photos first just to be sure.

IMHO it's not heat entirely but a plasma fire in the chemical zeolite. Rare but not uncommon in Europe with Bauer compressors my guess its its a Bauer.

Heat isn't the cause its only a contributing factor.

Long chain hydrocarbon, oil, synthetic diester, yada yada cling to the external surface of the molecular sieve. the active ingredient is zeolite.

A plasma fire in an air compressor filter is caused by two factors I can think of.

The oil is coating and forming a "seal" on the zeolite chemical ball keeping it away from the water vapour and sealing it in fresh.

When you have a auto dump on a compressor if this massive pressure change is allowed to affect the filter such as these cheaper filter/separator units you see this change in pressure on the zeolite that is saturated with water condensate expands and freezes as the pressure decreases in effect turning that bit of the zeolite to powder. The powder is pretty useless and ineffective and the adsorbing and absorbing "front" moves along the filter chemical length.

But the still potent "sealed fresh" balls of the original zeolite are left behind until a water flush removes the oil from them and it is the huge exothermic release reaction that is the cause of the plasma fire, oil heat etc only help it along together with the fresh air from the compressor feeding the fuel with the oxygen it needs.

The heat generated will melt plastic easy, on the plasma fires I have seen it also removed the anodising off the Aluminium.

Needless to say the critical point is for the filter shells to be destroyed take a hammer and chisel to the threads and get a photo published as I know full well some filters have been known after fires to find there way on E-Bay and the like.

The other cause for these plasma fires is failure of the auto dump and liquid water hitting the zeolite.

Just for grins take a small styrofoam cup and fill it with zeolite (molecular sieve) then an identical cup filled with cold water at room temperature.

Stick a thermometer into the chemical and pour the cold water over it the exothermic reaction should reach around boiling point ( I would have said 100 degrees C but I know you guys wouldn't understand if that was hot or cold LOL )

The plastic will also melt so be careful.

Its a long subject and too big for a single post. Hope it helps.

Oh yes one other is that we have high back pressure setting in Europe 200 bar 3000 psi this gives us longer filter life in hotter climates but the trade off is hotter compressor working harder wears out faster but the air is super dry. Again a dump failure allowing liquid water and hot oil carry over into the filter and the plasma effect is much more of a risk. Its a massive subject sorry I cant do more.
Iain Middlebrook

Holy cow, Iain, I'd call that complete. And most of us speak metric as a second language. 100C is the freezing point of water, right? Is Zeolite the same as Hopcalite? Hopcalite adsorbs CO and turns it to CO2. I have found through trial and error that keeping a high back pressure makes the filters last much longer, although I maintain mine at 1700 PSI (115 Bar). Turning the pressure up isn't a biggie. We do that specifically to save the Hopcalite, but I wasn't sure why. I think that this is a place where many fill stations fall down, keeping their BPV (Priority Valve) set and operating correctly. We also changed to a triesther oil, which is keeping things much happier in the compressor, according to my maintenance bills.

Cave diver indeed has the correct link, and I thank you for your explanation.

 

[swamp diver]

Iain,

I had considered a purifier fire and have seen several these start in the 4th stage due to autoignition of the compressor oil with the fire propagating through the aftercooler tubing typically coated with oil/carbon and then into the purifier. Of course made much worse if one is continuous blending nitrox in an air compressor. There is mention of several compressors in the shack and on the facility's web site the compressor shack is very small with no cross ventilation.
http://dutchsprings.com/imageairfill.html

The thread in question is in the dive club 'dive NY' forum and while there are no photos of the compressor or filters in question there was a verbal description of the said filters (3) seen as "badly melted". No mention is made of any indication that there was a fire in the purifier although the hypothesis is certainly plausible.

Given the number of rental tanks we see returned from outside compressor fills with a an inch of water in the bottom (typically condensate) one would expect to hear about melted plastic filter shells more often if the molecular sieve (MS) was generating sufficient heat when flooded. Typically though the MS is gradually water-saturated over time such that there would be very little exothermic reaction if a flood occurs near the end of filter service life. In order to generate the heat required one really would have to expose dry MS to a sudden flood of condensate. It would be interesting to know if the filters were recently changed with dry MS or near the end of service life.

It is not uncommon with some of these poor compressor installations to measure final stage or aftercooler temps easily pushing 400 F (205 C) after running at full load for hours at an ambient of 120 F (49 C) in the compressor shed.

A high air temperature alarm with auto shutdown might be another option to mitigate the risk of compressor overheating. Bauer sets these to shutdown at 270 F or 132 C with the sensor about a half inch in the body of the final stage.

Bauer over here now routinely sets their PM valves at 2800 psi.

Peter

 

[cool_hardware52]

Is Zeolite the same as Hopcalite?

Zeolite = molecular sieve, i.e. the primary desiccant

Tobin