Wet Gas Screw compressor 6

[mucour]

Hello,

We are considering a wet gas screw compressor for vapour recovery application. The gas is wet, water saturated and a heavy gas. Could anybody advise from the experience in their plant where they have this wet screw compressor for similar service the manufacturer that they will recommend based on the equipment's performance in terms of availability and reliability.

I know many gas screw compressor manufacturers but I want one recommended from operating experience in terms of performance.

Thank you for your help.

 

[zdas04]

You'll probably get as many preferred screw manufacturers as there are active members of this forum. Any of them will work for this application. The process derivative guys will say you have to have a (more expensive) process screw. The air derivative guys (like me) will say that the extra cost is not justified (assuming that this is a "normal" VRU that has a discharge pressure under 50 psig).

Two important points for any manufacturer's screw are: (1) make certain that the oil you choose is compatible with the gas you are compressing (very many heavies will probably force you toward a more expensive full synthetic); and (2) make sure that the package has convenient and flexible equipment to manage oil temperature. All of the screw oils are hydroscopic to some extent and will suck humidity from the gas. If you don't have enough heat in the system to cook the water out of the oil it rapidly deteriorates in its coalescing properties, its lubricity, and its viscosity. It is crucial to manage the temperature of the oil in the coalescing filter.

David Simpson, PE
MuleShoe Engineering

http://www.muleshoe-eng.com

Please see FAQ731-376 for tips on how to make the best use of Eng-Tips Fora.

The harder I work, the luckier I seem

 

[pmover]

mucour,

just curious, have you considered a thermo-compressor (ejector) for the application? without knowing the process conditioins, these units are being used for such applications.

as far as experience in similar applications, request the compressor mfg's to furnish you a reference list, including application and contact data.

good luck!
-pmover

 

[dcasto]

I've done several screw applications with wet gas, both water and condensationof hydrocarbons going through the unit. And at some point you cannot win and do the whole compression job with a screw compressor.

The last unit was taking 1500 BTU/CF gas from 20 inches vacuum to 160 psig. The amount of water and HC liquid caused us to limit the screw to 35 to 40 psig and then over to recip compressor. The screw packager we used recognized the amount of condensed liquids and would not go beyond the limits which I really respected them for.

 

[zdas04]

pmover,
I've tried VRU's with ejectors and they do ok with about 2 compression ratios per stage so if I needed 6 psia suction then I could get to 12 psia in the first stage and 24 psia in the second stage. Atmospheric pressure was 12 psia so basically with a bunch of gas pressure being expended I could only boost to 12 psig. Sometimes that could go back to a compressor suction, but sometimes it couldn't and I had to put a screw sucking on the ejector. It can work, but the limiting factor is often exhaust pressure.

David

 

[mucour]

Thank you all.

The eductor was not accepted because the amount of pressure and volume required of the motive fluid (liquid or gas) is very high to boost 1250 m3/h gas from near atmospheric to 60psig.

The wet screw compressor was considered advantageous because of the high compression ratio possible, greater than 12 is achievable. As a result, we are considering increasing the discharge pressure from 60 to 160psig. A single casing wet screw compressor will be used.

Zdas04 was correct in its thread. The challenge for wet and rich gas is how to ensure that the generated heat of compression releases the water from the lubeoil. The gas is also to be scrubbed upstream the compressor with a high efficiency scrubber having mist mat/pad plus any other means that can reduce H-C liquid carryover to the compressor.

Zdas04, pls explain what you meant by "It is crucial to manage the temperature of the oil in the coalescing filter". Are you suggesting a lagging or additional heating by an electric heater of the oil?

Thanks

 

[zdas04]

Not really, mostly you don't want to over cool the oil.

When I first started messing with flooded screws, no one had much experience with them in Oil & Gas upstream service (the first batch I bought were from Gardner Denver, and their video called "Screw Compressors in Oil & Gas" showed a bunch of screws working and all of them were mine). There were some packagers with very good knowledge, but not many operational go-bys for rough field use. After 6 months we looked hard at costs and found that oil was our largest single line item by far (bigger than the next 6 combined) and thought we had made a bad choice going with them. We found a guy that seemed to know what he was doing and he said we were not managing oil we were just writing checks. At that point we started actively managing oil temp.

Water vapor is molecular water and the "particles" are so small that there is simply no way to economically remove it with mechanical separation. The water quantity is pretty big (in your case, the vapor load will be on the order of 25 bbl/MMCF of water vapor). You need to keep enough heat in the oil to cook the water out of the oil.

We used temperature guns (there were no installed thermometers, who knew) to check the temps, changed out the bullets in 3-way valves, covered cooler sections, etc. After 6 more months Oil was out #7 cost and the machines were running trouble free. Three-way valves were installed on the oil side to bypass the cooler with part of the stream to keep temperature up, these valves were ok, but not adjustable and always sent a substantial stream through the cooler. Not a real convenient arrangement, but it worked ok.

I recently saw a lash-up that has the potential to allow temp management to be automated. This particular skid used a plate-and-tube heat exchanger with glycol on one side and oil on the other. The glycol loop included the fin-fan cooler. There was a control valve on the glycol side that was driven by the PLC based on the temperature out of the compressor. This combination allowed the temperature to stay very close to the 205F that is optimum to prevent coaking the oil while maximizing the water that is cooked off.

Bottom line is that if you don't get rid of the water, the droplets will not coalesce right and a lot of oil will go down the line as aerosol, your oil will fail samples and you'll have to dump it, lubricity will be poor and you'll wipe bearings, viscosity will be high and nothing will work right.

David

 

[Montemayor]

David Simpson has, in my opinion, given a first-class seminar on what has been arcane and experienced vital information on how to succeed in applying flooded screw compressors. It's valuable information such as this that truly justifies all the effort he made in having this Compressor Forum instituted in the first place. Kudos and gratitude should go to David.

As he has described, we all waited and waited for screw compressors to reasonably challenge the many reciprocating compressor applications where the recips were simply not "cutting it" economically or in practicality. During the 1960's and through the 70's the need became so great that dry screws were being applied even when it was known that they just couldn't perform in the long haul. But with persistance and experiment, flooded screws started to make a difference. It was a vital change because the screw needs the sealing performance given by the oil. This allowed the introduction of screws into the market place with a feeling of confidence in performance. However, we all knew we had to expect a trade off if we were to successfully replace a segment of the reciprocating applications. The resultant trade off has been well explained by David. It is not enough to just add oil to a screw application. It just isn't that simple. You have to add the vital process temperature and flow controls that allow the operator to have some leverage with respect to effective compression and oil management and effective oil management as well. Water is a contaminant to lubricating oil and it must be accounted for. It will accumulate in the oil if the process conditions are favorable for that to happen. If you don't give importance to what David has described in detail, you won't succeed in using the flooded screw at the competitive level that is expected of it. I think David's comments ought to be forged into a FAQ on this Forum.

Thanks David.

 

[dcasto]

The last screw I mentioned above had a two stage cooler. The screw outlet went through a cooler section and its outlet temperature was controlled to a fixed temperature in the oil separator to minimize oil loss and assure water would not flood the separator. Then gas was cooled in a second section to as low as possible before going to the PD compressor.

 

[mucour]

David, you could not have described it better. It was a wonderful piece. I could have given more stars but I am entitled to only 1 vote.

Thank you.

 

[crjones]

The valve David is talking about is called a thermal bypass valve usually produced by AMOT or FPE. The setting was set at 140 degF To hold the oil temperature above the pressure condensation point for water. Synthetic lubricants have made driving off the water much easier.

Extensive running unloaded cause more emulsion issues than running flat out.

 

[zdas04]

crjones,
You can get bullets for other temperatures for both diverting valves, one of the first things we did in remediation was to replace the 140F bullets with 180F bullets. This is the right oil-inlet temp for about 10 ratios, but it is really risky if you go to 15 ratios you'll be over 215F on the outlet and can wreck the oil.

The synthetic oils are pretty cool, but not always needed. I've found that with proper oil temperature management you can get years of life from mineral oils. The synthetic oils will maintain coalescing properties long after absorbing enough water vapor to ruin a mineral oil, but continued saturation with water will still result in oil loss and poor lubrication.

Even with synthetic oils you have to add 6,500 BTU per lbm of water vapor to cook the water out. That number is pretty nearly invariant and I've seen it used for this application and many reboiler applications. You can calculate the BTU's available to cook water by Q=mc delta T using the theoretical adiabatic temperature rise for the delta T (even though you don't get anything like that kind of dT, the BTU's are still there) and the gas c(p). That number of BTU's will tell you if you have enough energy to cook the water from the oil. Generally if you can maintain your screw-outlet temp between 205-210F then you have enough BTUs. If the temperature won't go there then you don't have enough energy to supply the required latent heat of vaporization.

David Simpson, PE
MuleShoe Engineering

http://www.muleshoe-eng.com

Please see FAQ731-376 for tips on how to make the best use of Eng-Tips Fora.

The harder I work, the luckier I seem

 

[crjones]

The goal had been to keep the water in its vapor state so it would move on out of the machine with the gas. The problem I have seen over the years is that raising the bypass temperature on a mineral oil you get too close to the suppressed autoignition temperature, particularly when there is "dirt" in the system suppressing it even further.

As far as synthetics, I had not had any problems with the PAO's or diesters with water. The glycols were another story, which I believe why I-R runs their units so hot.

The other side of running oil too hot is reduced film strength in the bearings.

 

[tgmcg]

If you read between the lines of what's been said above, you'll need to manage the heck out of an oil-wetted screw machine for this application. Whereas a properly designed recip does not require so much "chemistry" and TLC.

Forgive me for being so blunt, but Wild Horses couldn't drag me to support the use of an oil-wetted screw compressor for wet gas application.

 

[zdas04]

tgmcg,
Funny, last week I visited 8 recips in low-pressure wellsite service. The sites were selected based on access to the route I was planning for a tour, no effort was made to cherry-pick broken machines. The observations were interesting:

- Two 2-stage machines, one with first-stage discharge higher than second-stage discharge, the other one doing 1.01 ratios in the first stage and 4.5 ratios in the second.
- Three single stage machines not compressing any gas because the valves were busted out, but the machines were going round and round (wellhead pressures had increased enough to sell a tiny bit of gas).
- Three single-stage machines with discharge temperature MUCH higher than their ratios would calculate to (one was doing 1.8 ratios and had a temperature rise of 240F instead of the 74F that it should have been). This is the result of partial fillage due to valves being too stiff. The actual minimum pressure in the cylinder is much lower than the external suction pressure, resulting in more compression ratios than the pressures at the skid edge would indicate. In the case with 1.8 ratios the discharge was 100 psia and the suction was 56 psia. The temperature indicates that the actual minimum pressure was about 18.5 psia.

Zero out of the 8 machines were anywhere close to optimized for the conditions they were running in. Looks to me like the people who think that recips are "set and forget" machines are just plain wrong. Recips have their place, and they are often the best machine for a given task. But small hp recips need all the care that mainline machines need, and they just plain don't get it on wellsite use (I've seen 1,000 hp recips set-up just as poorly on wellsites, people have it in their minds that "wellsites are easy").

Bottom line is that screws have many fewer parts to break, much lower on-skid temperatures, and require substantially less maintenance than recips. The only finicky bit is oil temperature and competence in packaging can make that painless.

David

 

[tgmcg]

In theory, fewer moving parts means lower maintenance. But couple that simple design with gas constituents that tend to accumulate in the oil, degrade viscosity and reduce hydrodynamic oil film thickness between the screws, and that simple design becomes a potentially troublesome machine.

I'd never suggest recips were "set and forget" machines...I'm still looking for that machine. Perform regular PM on a properly design recip in a properly-designed installation and they will provide excellent service. Miss doing either and reliability will suffer.

That being said, neglect ANY compressor at your own peril.

Best regards,

Tom McGuinness, PE
Turbosystems Engineering

http://www.turbosynthesis.com

 

[djack77494]

Excellent thread. Thanks all.

 

[GSTeng]

zdas04,

I think you might be giving recips a bad name. On a well site Screws are great but when you get onto the mainline they are not so good, or even into a larger gathering systems. I don't want to argue the finer points of Screw packages vs recips, since I am a man than knows when I am in over my head. But I have seen a lot a recips that have none of the problems you listed and when they are well maintained they smash gas all day and night with out even a hiccup. There are even wellhead recips that will work as well as a screw if they are well maintained, but for that service I would go for the screw over the recip.

As far as maintenance goes, yes recips need more TLC but if you don't know how to run a screw it will kick your butt, a recip is more forgiving to incompetence. With the age of the average switcher and pumper these days incompetence is not so uncommon.

That is my two cents, just had to stick up for the poor, hard working, unappreciated Reciprocating compressor.

By the way great info on the finer workings of a screw compressor, very educational, thanks for that.

 

[zdas04]

GSTEng,
If I give the impression that recips don't have a major place in Oil & Gas then SHAME ON ME. I have set many more recip frames than screw frames and I expect that to continue.

The absolute key is knowing what parameter is most likely to change. If you are in a blood-guts-and-feathers low pressure gas wellsite, then suction pressure will change far too often and far too broadly for a multi-stage recip to ever be the first choice. Below a target suction of about 40 psig, reasonably small wellhead pressure changes can be too big a percent change to be acceptable to the valves in even a single-stage recip.

On the other hand if I can expect suction pressures to be pretty constant (either accept a dP across a suction controller or set the recip downstream of other compressors) then a recip has much better energy-consumed-per-MCF numbers than a screw and can tolerate quite a bit of variability in discharge pressure.

To me it is important to know the strengths and weaknesses of all of the technology in your toolbox and apply the equipment that maximizes the net of the strengths minus the weaknesses.

Hell, I'm sure that sometime in my life I'm going to recommend a centrifugal compressor for a specific application even though I rarely work in operations where the key parameter is energy density or oil-free service.

I cover a lot of this in faq1036-911 . Folks have told me that it is better than the FAQ-score would indicate.

David

 

[tgmcg]

zdas04,

We use slow-speed API 618 recips in flare gas service at near-atmospheric suction pressure and widely varying gas MW with no unusual problems. In fact, I'm presently buying a recip for dual 0 psig & 15 psig suction pressure service wth MW varying from 9 to 32. Hoerbiger is supplying the valves for this service....their standard ported-plate design. The folks at Cook Manley claimed their moppet valves would also handle this duty.

I don't wish to contradict you, but I'm not able to support your statement that "Below a target suction of about 40 psig, reasonably small wellhead pressure changes can be too big a percent change to be acceptable to the valves in even a single-stage recip."

Best regards,

Tom McGuinness, PE
Turbosystems Engineering

http://www.turbosynthesis.com