Compressor on full recycle 2

[aikmeam]

It's a general question.

Anyone with experience operating a compressor (recips or centrifugal machines) on full recycle when the inlet to the compressor is blocked? What are the pros and cons, and also the things that we have to take note of?

Thanks in advance.

 

[unclesyd]

Neither a restricted suction or full recycle is desirable from operation and mechanical point of view. There a lot of properties of both the gas stream and compressor to consider.

Come back with a little more information, type compressor, product being compressed, temps and pressure.

 

[pmover]

is the duration of operating the compressor/driver short-term (i.e. a few hours) or long-term (i.e. days)?

short-term, maintain constant observation of operating temps and other essential parameters.

long-term, consider overall impact to process, fuel consumption, and machinery operation.

i would not be surprised if the compressed fluid is exotic or expensive, or the system is a refrigeration system, thus opting for maintaining compressor operation.

perhaps the most important factor is adequate cooling of the gas being compressed and any impact to the process.

for centrifugals, having a reliable and repeatable measuring device for surge control is essential. this would be an optimum time to verify surge system and perhaps conduct a little compressor testing - obtain P, T, + fluid properties at constant speed line at various flows.

for recips, monitoring process temps and having adequate cooling is desireable.

as far as the driver is concerned, should not be a problem, except wasting fuel/electricity (if not steam turbine driven).

good luck!
-pmover

 

[tkdhwjd]

My opinion...

Besides wasting energy to drive the compressor, I do not see any negative impacts... as long as the compressor is not surging, inlet/outlet temperatures are maintained properly. It is interesting, though, since it is rare to keep the compressor on total kickback/recycle. Not sure what you are trying to achieve....

 

[aikmeam]

I'm postulating a scenario actually.

Assuming it's a product compressor which delivers to the B.L. Due to circumstances, the upstream unit prior to the compressor has to be blocked in momentarily for a few minutes or at most an hour.

I was thinking that in case that the compressor can be run on full recycle via the kick-back valve, this will facilitate immediate start-up of the compressor when upstream unit resumes.

For me, it's not necessary to trip the compressor immediately whenever the suction is blocked in. If ever running full-recycle doesn't create any problem besides temperature rise, this could very well reduce plant downtime. Reason: compressor systems, whether recip or centrifugal, are temperature trip equipped which will eventually shut down the machine when discharge temperature gets out of hand. And whether it will last as long as we required above, it will not be an easy answer which depends largely on machine built and properties of fluid compressed as pointed out by unclesyd.

What do you guys out there think?

The next thing which comes to my mind is how do we establish the control system during such period? Do we just rely on anti-surge (centrifugal) and/or pressure control (recip & centrifugal) to do the job, or some special attention to other control parameters is required.

Up till now, I've ignore the impact to OSBL, especially if our product is a feedstock to others. Will it still be worth it when all OSBL plants had already been shutdown due to loss of flow and requires a longer start-up time. Which means that we may end up quick-starting a compressor that will eventually tripped again due to discharge pressure rise (no consumer) or sending the product flare.

 

[tkdhwjd]

My opinion...

Due to heat of compression, whether it is positive recip or centrifugal, the compressor should have interstate & discharge coolers. If so, I am not sure why you are concerned about the high temp trip scenario unless your kick-back location is upstream of the coolers.

Once again, the compressor won't care if the gas is from kickback or indegenous feed as long as the conditions are the same.

 

[Montemayor]

aikmeam:

I've had the experience of running reciprocating process compressors on what you call full recycle. I've called it "100% capacity control". What I mean to stress here is that what you've described as your scope is truly an attempt to do capacity control on a reciprocating unit. I've done this with the following caveats:

1) You must have a full-capacity designed discharge after-cooler that ensures that the amount of recycle is cooled down to a reasonable suction temperature prior to it being introduced into the recycle control valve;
2) If you don't recycle 100% of the compressor's full capacity (and I urge you to try to achieve this), then use a 100% clearance bottle. This will be difficult or expensive to do if you have a multi-staged reciprocating machine (this is one of the many information data you've left out). This feature is much-desired (& considered optimum) in controlling the capacity of a reciprocating machine because it reduces the consumed HP down to essentially zero - if done properly.
3) If you recycle 100% capacity (or less) back to suction, make sure that all included (or produced) liquids are removed prior to introducing the recyle stream into the suction.
4) Be sure to take care of the existing suction gas stream that normally enters the compressor suction; in other words, if the compressor is to be kept moving and pressurized with a constant, recirculating quantity of gas, you must dispose or re-direct the normal feed stream that no longer is entering the compressor. The feed stream can be either routed to waste disposal (a flare, etc.) or to another process. You can also shut down the feed stream if need be.
5) If your process gas is flammable or explosive, be sure to maintain a positive gage pressure at the suction valves' inlet - around 5 psig or better - and monitor it and set an alarm to detect if the pressure drops down.
6) You should have a 100% capacity PRV directly upstream of the discharge block valve.

I don't understand your remark that it's not necessary to trip the compressor immediately whenever the suction is blocked in. If you mean this while the unit is under process conditions, then you are wrong if your gas is flammable or explosive. The reason for this is that the moment the suction is blocked, a reciprocating unit will suck a vacuum which will mix air (oxygen) with the explosive process gas in the subsequent compression cylinders and result in an explosion. I hope this is not what you meant.

I agree with you and my operations proved that running the reciprocating units with 100% capacity control well reduced plant downtime because I was able to put the compressors "back on line" quickly and safely, without purging, and creating emmissions. There is nothing wrong with this technique, as long as you take the safe caveats above into consideration.

Depending on your procedures to put the compressor back on line, you should not require any more instrumentation on a reciprocating compressor. I personally prefer to have reciprocating compressors, under these conditions, be put back into production conditions under a detailed and methodical manual procedure. I have found this to be far simpler and safer for the operators and the operation.

I hope this experience helps.

Art Montemayor
Spring, TX

 

[BenThayer]

throttling on the suction of a centrifugal compressor is extremely common if you dealing with a constant speed driver.

however, you also need minimal flow for surge protection. full recycle seems to be a little extreme.

 

[augiedog]

Not a problem if the recycle gas is adequately cooled and the recycle valve is controlled by the suction pressure controller.
If its wet natural gas then ensure the JT cooling across the recycle valve doesn't cool the stream below the hydrate temp - may have to modulate the temperature with hot gas from upstream of the cooler to stay out of hydrates.

 

[aikmeam]

Well, thanks for all.

It's nice to have a mixed reaction from everyone.. some agree and some don't to the proposal. At least it's worth the discussion.

tkdhwkd: Correct that you pointed out coolers are installed on all compressors. However, most installation requires manual regulating of the cool stream to achive the required temperature. So, just hope that the operators could react on time before the compressor trips.

montemayor: You're the man. Thanks for the 1-2-3 guide. 100% clearance bottles? How does it work? When I say "it's not necessary to trip the compressor immediately whenever the suction is blocked in" it's simply to indicate flow loss. I think avoiding sucking vacuum is one of the reason recirculation valves are present in compressors.

benthayer: Can you elaborate further? I don't quite get your concerns about being "extreme".

augiedod: A very good consideration as well. Taken note on that.

So far, montemayor provided a pretty detail note and experiences on recips machine. Anyone else with centrifugal compressor?

 

[Montemayor]

aikmeam:

Clearance bottles (pockets) are vessels that are attached to, or built into, the cylinders on reciprocating compressors. These vessels (or compartments can be "fixed" or variable in the amount of volume they offer to the gases that originate in the compressor cylinders. Like clearance volume, they represent a volumetric inefficiency of the compressor. These volumes are normally segregated from the cylinder's working volume by either a manual or automatic valve that joins both volumes when opened.

As you can appreciate, if the clearance bottle (or chamber) is of the same volumetric size as the compressor's cylinder and the two are communicated then the piston can do no work (no compression takes place) and the volumetric efficiency is zero. The gas simply is pushed out of the working cylinder into the clearance pocket and back. There are no moving parts and the energy consumed at 100% unloaded capacity is simply that of keeping the machine moving through friction and mechanical inefficiency - sometimes as low as 5% of the design BHP. The compressor starts to do work as soon as the clearance pockets are closed.

Look at the following reciprocating compressor paper:

http://www.gmrc.org/gmrc/techpaperspdf/2002/basics_of_compression.pdf

Go to page 13 of 78 and read up on Volumetric Efficiency there and on following pages. You will appreciate, I am sure, the simplicity of controlling your reciprocating compressor at zero capacity with such a simple system. HOWEVER, there is a catch! (Nothing in engineering is that simple to achieve) You must have the cylinders bored and designed for this feature - or have the latent capability already in the design. Also, with double acting cylinders this becomes a physically bulky installation. This feature works best when the original compressor manufacturer is formally specified to furnish such a feature. It can become hairy on some retrofits.

You're exactly right; avoiding sucking vacuum is one of the reason recirculation control valves are present in the discharge of reciprocating compressors. There is no way to specify or maintain a constant speed reciprocating compressor at the exact required volumetric capacity under normal production. The machine has normal, variable valve inefficiencies, wear & tear, ring leakage, clearance increases, V-belt and driver slippage, etc. The only way to overcome these troublesome features is to oversize the required capacity by a contingent amount and normally make up the overcapacity by re-circulation from discharge-to-suction (capacity control). This is never taught in thermodynamics or in compressor text books, but it is the common sense portion of working with this type of machine. While centrifugals are also designed with a certain degree of over-capacity, they can be throttled or speed-varied within limits; recips, being from the positive-displacement tribe, will not tolerate throttling and so have to be capacity-controlled through other means. Unless you have a synchronous electric motor drive, you have to settle for a constant speed recip.

Regards,


Art Montemayor
Spring, TX