Frick screws for air compression?

[MacMcMacmac]

We are currently searching for a pair of ~4500cfm compressors to replace 1950's vintage Worthington BDC reciprocating compressors that are becoming somewhat expensive to keep running. These are 1250hp, 14" stroke 3 stage units that provide air for our research facilities. In that capacity, they have been somewhat troublesome, since oil is injected into the cylinders under normal operating conditions, which of course migrates to various places where it inevitably causes trouble. Given the lack of alternate compressor designs at the time, and the pressure requirements (300psi), they were likely the best choice. However, they are becoming more trouble than they are worth with maintenance and overhaul costs going well into six figures over the past 5 years.

My supervisor has asked me to look around for surplus or refurbished machines that could fill the role of these Worthingtons, with the only caveat being that we do not replace them with another pair of reciprocating compressors. Recips of this power would require a stationary engineer to operate them by law. As the current recips are located in a central heating plant manned round the clock with stationary engineers, this is not an issue presently, but they have stated that they would like to remove themselves from this role to concentrate on the operation and maintenance of the steam system. The compressed air supply from that building is more of a legacy of their being the only building supplying the whole campus from back in the 1940s and 50s.

We are looking at surplus equipment to keep the costs down. We sourced a killer deal on a used Cooper turbocompressor about 13 years ago that saved us hundreds of thousands of dollars. Since these units are run on a cost recovery system, this means the cost of the machine plus running costs gets mixed in with the billing passed on to our clients. This is one reason the Worthingtons have been kept in service so long; the machines themselves were paid for decades ago. The rising repair costs are now erasing that advantage.

Our building is far more concentrated on supplying high pressure air to several facilities (30,000hp worth of centrifugal compressors), so it is only natural that any replacement machines would be installed here, so as to be included under our maintenance and operations regime. As none of us have our stationary engineer tickets, we cannot do this if we went the recip route. We would also prefer the lower maintenance and reliability of a centrifugal or screw compressor. These compressors are used intermittently, though usually at full capacity when running.

I have been looking for surplus equipment, but the 300psi requirements means that the centrifugal pickings are quite slim. I would also like a more packaged unit than the few centrifugals I have seen, which are usually nothing more than bare air ends out in a yard somewhere. I have investigated the possibility of purchasing used/uninstalled oilfield equipment, given the severe downturn in the fortunes of oil and gas out west, and I am finding that for our flow and pressure requirements, other than recips, the Frick brand keeps coming up, the TDSH355 screw in particular. I am wondering if this machine is suitable for compressing ambient air, or if anyone has ever packaged one for air compression. I see that these are permanently attached to the gas source, so I am not sure how modulating the output would be achieved, such as with suction throttling and blow off like smaller plant screws. I see the slide valve in the screw set, and understand how is lowers output pressure, so theoretically I suppose it could just pump continuously and blow off excess air to atmosphere like our centrifugals, but this is all guesswork at the moment. There seems to be little avenue for contacting Frick itself, since it comes under the Johnson Controls umbrella and has no individual website, so I was just seeing if anyone on the board could answer these general questions.

*I do realize a centrifugal compressor is the better machine for this application*

Thank you.

 

[zdas04]

I wouldn't agree that a centrifugal is a better machine for this application. If you are "blowing off excess air to atmosphere" with your other machines then you are actually blowing off excess money to atmosphere. The controls on the Frick package can significantly reduce that waste of power.

Frick compressors are know as "process screws" and generally have a higher pressure rating than "air screws". Using a process screw in air service is not a very big deal, just slightly more expensive and slightly higher pressure rating.

The way I like to set up that machine is:
[ul]
[li]Variable speed. While between min speed and max speed keep the slide valve fully shut. Vary the speed to keep discharge pressure constant[/li]
[li]Controlled slide valve. At either max or min speed use the slide valve to keep from overloading the driver.[/li]
[li]Leave the variable VI at a fixed position based on design ratios, and don't allow it as part of your control scheme.[/li]
[/ul]

For an air application you need to be very careful about temperature control. The good news is that you do about the same compression ratios all the time, so you can determine a supply temperature that will allow water to be cooked out of the air and not taken up by the oil. Not managing the oil temperature will result in excessive oil in your delivery air and unhappy clients.

David Simpson, PE
MuleShoe Engineering

In questions of science, the authority of a thousand is not worth the humble reasoning of a single individual. Galileo Galilei, Italian Physicist

 

[MacMcMacmac]

I agree that blowing off excess air to atmosphere is wasteful, and I am trying to champion a system to re-use surplus air from a dehydrator-blower system that dumps pretty expensive filtered, dried and compressed air to atmosphere through a silencer. My aim is to re-introduce that air back into the inlet louvers of the first dehydrator to lessen the moisture load of the air it takes in from atmosphere, the trick being to cool it off from 250F to a temperature that doesn't overload the chiller loop in the first stage of the dehydration process without incurring other energy use penalties to cool it, but that is a project for another time.

The centrifugals are used to supply high pressure heated air to combustion test cells. Our clients are testing new combustor designs for power plant turbines. Sometimes the setpoints need to be held within a very narrow band, so manual control of the IGVs and blowoff valves are required to maintain flow rates within tight limits. Sometimes we just can't throttle back far enough with IGV's and need to dump the excess flow. At max flow, the recips from the other building are brought online. They can pose problems with their load/unload cycle, since flows can fluctuate a fair bit. Add in the oil carryover, which halted a fuel injector test in another lab awhile ago due to an improperly adjusted lubricator, and you can see they are far from ideal machines for our uses. Wear in a second stage cylinder has revealed porosity in the casting which will require a $50k overhaul when combined with a rebuild of a worn crankshaft journal.

As far as capacity control, am I right in assuming I could use the slide valve as a partial unloader mechanism and blowoff excess air if need be? I know that is a pretty crude arrangement, but the likelihood of a VFD being purchased with the compressor is pretty slim. I think we will be limited to full speed operation and capacity control such as the recips are using. If the slide valve limits the compression capability of the screw to the pressure in the system, would the temperature in the screw set increase dramatically as air stays in the screw and gets re-compressed over and over, or is the injection oil sufficient to keep the temperature from increasing out of control? Forgive my rudimentary questions, but I come from the air compressor world and have no direct experience with these slide valves, except on one set of Sullair screws I serviced about 15 years ago. Most capacity control I am familiar with is either a modulated or open/shut throttle valve with blowoff through a muffler.

I think oil temperature control could be easily maintained with some intrumentation and cooling water throttling.

As an aside, we used to use the rule of thumb that when converting from electric to internal combustion drive, you needed double the horsepower rating for the ICE to get the same performance as the electric drive. If these screws are running with a big Cat or Waukesha, would that rule still hold?

Thanks.

 

[georgeverghese]

If this air has to be dehydrated after compression, it will be a tall order to get trouble free operation with oil flooded screws, regardless of what brochures tell you. Adding a carbon bed adsorber on the wet compressed air outlet to take out vaporised lube - cooling oil prior to dehydration also has its limitations.

Centrifugal air compressor vendors may have a better idea to find you a used machine that some other plant may consider parting with ?

 

[MacMcMacmac]

The high pressure air is for the combustion cells, so they can tolerate some oil carryover, but not big clots of gunk all at once that wrecked the injector test. The dehydrated air is actually for another system that is used in conjunction with a single stage blower and turboexpander to produce climatic conditions in an engine test altitude cell which is being evacuated by another centrifugal exhauster. Sorry to confuse the two.

I realize an oil free solution would be best, but proper separator maintenance along with downstream filtration should do the job. I have rectified several problems that had made the issue worse. There is a 10,000cfm Balston filter in the line which had no elements in it, and there is another 10,000cfm filter in our basement which had instrumentation grade elements in it that had dropped the cfm rating down to 3200cfm, which then naturally ballooned out and ruptured at full flow. There was also an improperly installed autodrain on that filter which drained the vessel before the elements, not after, where the actual coalesced oil was collecting. I have also received a 10,000cfm Xebec XWS cyclonic moisture separator to install first thing in the line before any filters. I think this will be enough to handle any oil carryover.

 

[georgeverghese]

Well then, if you can manage with high dewpoint water wet air with oil vapor in it, then it would come down to a comparison of the power consumption at full load and at some part load between the oil flooded screw and a centrifugal, with some ongoing incremental costs for the higher maintenance for the screw. That is, provided you've got no low point pockets in all your distribution piping that would otherwise collect oily water pools that would once in a while blow out into the consumer lines. To counter this, you could add on low point piping autodrainers. As you say, adding filter coalescers to knock out oil - water mist on the compressor discharge also helps.

Assume these oil vapors dont result in any flow readout errors on flow transmitter elements at air-fuel ratio control loops at these client combustion test cells ?

 

[rconnor]

Some good points brought up by georgeverghese and zdas04. I'll add that zdas04's point of reducing the amount of wasted blow off air should be taken seriously. At this volume and pressure, the paybacks of incorporating capacity controls or selecting multiple units to match changes in demand per shift, as to reduce blow off to atmosphere, are very quick. They are even quicker if you take advantage of your local utilities incentive programs.

 

[MacMcMacmac]

Can a Frick screw be completely unloaded with the slide valve, or does it still need a suction throttle valve? I am a little vague on that. I assume you would need a system check valve, a minimum pressure valve, plus possibly a secondary gear pump for low pressure operation. Also, for shutdown wouldn't there also be a need for an oil stop valve as well? I see all the packaged screws in the oilfield with piping connected direct to the source (naturally) and I wonder how this would have to be modified for a simple air compression application.