Carbon monoxide from compressors?

[captain]

It is possible for a malfunctioning, overheated compressor to reach a high enough temperature to cause the lubricating oil to decompose and produce CO. This problem was greater before synthetic oils were in use. Synthetic oil can withstand much higher temperatures than the older mineral oil compressor lubricants before starting to decompose. I have run a small compressors for 40 years and when I switched to synthetics the carbon formation I would find on the valves disappeared. The carbon was an indication that the mineral oil was decomposing.

 

[oxyhacker]

Well yes and no. One of the more common sources of CO is a cookoff of carbon in the compressor - carbon, formed by overheating of the oil, can build up over time inside the compressor, especially on the valves. If the temperature gets too high (some put the threshhold at as low as low as 500F) the carbon can start to smolder or burn, producing CO. The hotter the compressor runs, the more chance of carbon forming, and the more chance of it igniting.

Fortunately, this seems to be a rare event. The problem is, such cookoffs tend to be isolated and transient events, so occasional monitoring may very well miss them - sort of like playing russian roulette, you end up with one dead guy and five other guys who wonder what the fuss is about. So continuous monitoring is the best bet for avoiding it. New electronic monitors like the CO Experts which has been mentioned here, are making this cost effective for the first time, both for the compressor operator, and the individual diver.

However, as you say, engine exhaust is probably by far the greater problem. Filtration has hardly any effect on CO. Hopcalite will convert some of it into CO2 and moisture, but, in the quantities used in scuba cartridges, inadequate for dealing with potentially lethal quantities..

CO is not a product of compression, though it is extremely dangerous to divers. Even a little bit is dangerous and the impact is magnified with depth. It's not a common problem but, when it is a problem, it can easily cause serious permanent injury or be fatal. Though responsible shops will test their gas regularly, I am not aware of any mandatory reporting requirements which means there isn't a reliable single source for statistics regarding CO issues in SCUBA.

The source for CO in a scuba cylinder is usually a gas driven engine powering the compressor. Great care needs to be taken to keep the compressor intake away (and upstream) from the exhaust of any internal combustion engine.

 

[pescador775]

Hacker, you have mentioned the carbon hypothesis before. It could explain some things.

I've resisted the idea that compressor temperatures could be high enough to produce CO. However, I've changed my mind after a long discussion with "hopcalite" (see below):

http://www.bauer-kompressoren.de/forum/en/viewtopic.php?t=8

After "hopcalite" provided an overview of the Montreal Report, I'm fairly sure that "carbon" is not necessary to explain such events, however rare. For example, one such offending compressor was serviced but, after three hours of continuous running, the compressor spontaneously initiated CO production. My opinions are evolving as I hear from others and review the data. At this point, I'm thinking that offgassing of methane and toluene are sufficient to explain the CO phenomenon. However improbable, it does seem that the cooking (pyrolysis) of gases and water are a source. However, it is still possible that carbon is a cause. It cannot be ruled out because the servicing of the offending compressor may have missed some particulates in the cooling lines. Still, the amounts and general pattern of CO suggest a self regenerating source such as gaseous products boiling off from the oil. I don't know if the oil which was used in the Montreal compressor(s) was petroleum or synthetic stock. One might presume that mineral oils were involved.

 

[captain]

It is not the carbon itself but the formation of carbon that to me indicates that pyrolysis is taking place. CO is the product of incomplete combustion. So is the formation of carbon on compressor internal parts. If complete combustion took place the result would be CO 2 and water and no CO.

 

[pescador775]

Captain, I'm thinking that carbon formation in a compressor is different from incomplete combustion in an engine or retort. Carbon appears to be a product of "oxidation". Think of it as analogous to rust. The heat and pressure of a compressor cylinder lead to conditions which accellerate the process. The other argument would be that carbon is formed similarly to that of the old charcoal retort, eg the stuff is heated in a low O2 environment. However, compressors have high PP of oxygen.

 

[John C. Ratliff]

In the 1980s I investigated a scuba fatality that involved carbon monoxide. What happened with this case, and can happen whenever carbon monoxide is involved, is that it triggered a heart attack. So the results of even a little CO can be fatal. CO almost caused the death of Jacques Cousteau in the 1950s (read about their dives into that freshwater spring in France, where Frederick Dumas was pulled semiconscious out of the water, and Cousteau was also not fully there--from N2 narcosis at depth and CO poisoning).

There are a number of ways of controlling CO in the compressed breathing air. First, use a compressor with an electric motor, not gasoline or diesel. Second, move the intake hose high above the ground (one source would be automobile or truck exhaust). Third, periodically test the air to meet the requirements of Grade D Breathing Air, under the OSHA guidelines:

Supply air that meets or exceeds OSHA air purity standards as follows:

GRADE 'D' AIR *
Maximum Contaminant Parts Per Million

Carbon Monoxide 20

Carbon Dioxide 1,000

Oil 5

(see: http://www.dot.state.ny.us/progs/safety/sb-94-5a.html)

For the health effects on land of CO, look here:

http://www.cdc.gov/niosh/pel88/630-08.html

SeaRat

 

[John C. Ratliff]

Captain, I'm thinking that carbon formation in a compressor is different from incomplete combustion in an engine or retort. Carbon appears to be a product of "oxidation". Think of it as analogous to rust. The heat and pressure of a compressor cylinder lead to conditions which accellerate the process. The other argument would be that carbon is formed similarly to that of the old charcoal retort, eg the stuff is heated in a low O2 environment. However, compressors have high PP of oxygen.

Actually, carbon is an element; it is carbon dioxide (from respiration or complete combustion) and carbon monoxide (from incomplete combustion) that is the problem. Inside a compressor, if it is using oil for lubrication, the oil can break down under heat and pressure to also form CO (carbon monoxide). It is not at all analogous to rust, which is the oxidation of iron (Fe++). For iron, see:

http://periodic.lanl.gov/elements/26.html

For carbon, see:

http://periodic.lanl.gov/elements/6.html

For the Periodic Table of Elements, see:

http://periodic.lanl.gov/

SeaRat

 

[captain]

Pescadore775, I just look at compressor as an engine spinning without fuel. In a diesel engine the heat of compression is high enough to ignite the fuel. Because all the carbon in the fuel is not completely reacted with the o2 into co2 and h2o, co and carbon formation is the result. I base this strickly on my experience with both compressors and engines. Most of the carbon I have found on compressor valves is in the first stage where the pp o2 would be similar to that in an engine.

 

[pescador775]

John, pure carbon forms underground. In fact, thousands, no millions of tons are mined each year. It is formed from complex sugars, cellulose and other substances, the result commonly called "coal". I know there is significant heat involved, but I doubt that it compares to levels in a compressor much less an engine. However, the heat is produced by slow decomposition. We see that in the peat bogs. Increasing levels of pressure are needed to convert peat, brown coal and soft coal to anthracite. This kind of pressure cycle does occur in a compressor. Well "rust" is too picturesque and not scientific, I agree. Perhaps, "metamorphism"? Heh, heh. With coal, it looks like the entrapped oxygen in organic matter runs off as carbon dioxide, and hydrogen comes off in methane, etc. I don't know if this has any equivalent in a compressor but actual pyrolysis (heated oil) causes some gaseous stuff like methane and toluene. Normally, this product would be trapped in the filter. However, CO, when it does occur, is not easily stopped. The small pouches of hopcalite common to many filters have limited capability.

 

[pescador775]

While I'm at it I might as well recount the temperatures which Bauer lists for their compressor, the Capitano. I have others but this seems like a good example since this mod is so common. Temps for cylinder heads (all three), 140F. Temp for #1 outlet, steady 280F. Temp for #2, 220-280F. Temp for #3, 220-280F. These are from memory but are pretty close. Ambient temp for the above was 20C (68F). The variation of temps for #2 and # 3 have to do with the varying pressures in these cylinders. Number 2 cylinder pressure varies from 300-600 psi for some reason. Obviously, at higher ambients and with long run times something is happening to some compressors which causes these numbers to rise significantly. The flash points for most compressor oils (synthetic) run the range of 450-550F. The highest is for Chemlube 800, a triester oil.