Antisurge control for compressor - location of PT

[phex]

I have got a question concerning the location of a pressure transmitter in the discharge line of a turbo-(centrifugal)-compressor. This transmitter is connected to the antisurge control of this compressor. I'll have a try at ascii arts to visualize my problem:

C----------------------
| |
| PT

in this "drawing", C symbolizes the compressor, the "---" line is the discharge line, the "|"-line is the antisurge bypass (which is closed in normal operation) and the PT is the location of the pressure transmitter. The socalled speciallist of the anti-surge control tells us, that this PT has to be located before the antisurge bypass line, because it has to measure the pressure at the compressor discharge. In my opinion, the location would only matter if this transmitter actually measures only the dynamic pressure or the velocity of the fluid (it doesn't do this - just the overpressure relative to the surrounding). The pressure is the same in all the system (we are talking about perhaps five meters of pipe length difference, and this is a 10" line, so pressure drop should be negligible), even if the antisurge bypass is opened due to a surge. The only thing changing is the velocity of the fluid (due to the now bigger cross sectional area with the opened bypass) and this is not measured. The flowrate through the compressor is measured on the suction side. Temperature measurements are in place. The only point we're discussing about is this PT.

Am I overlooking something here? Is there really a difference in the pressure reading due to this bypass? My colleagues are thinking in my lines and I tend to trust them on this, but my experience shows that one can get good and sound advice from these fora. Especially, since there might be people here who have more experience on this topic than I have.

Thanks for your time,
Chris

 

[phex]

In addition to the things said above (How I would love to be able to edit my posts), the bypass never accomodates the complete flow, always just a tiny fraction of the flow because in case of complete flow through bypass, the plant would shut down completely, including a total shutdown of the compressor. No more surging possible.

Chris

 

[hacksaw]

you are using a variation of surge control that recycles flow according to a given surge (mfg supplied) characteristic. xmtr location makes some but not a lot of difference. most controls are based on d/p across your equipment or require two abs press xmtrs, but single xmtr frequently adequate with single stage low pressure machines.

xmtr location is critical where you are detecting the on-set of surge and adjusting the recycle flow accordingly.

good luck

 

[phex]

Thank you for the reply, hacksaw. If I understand your words correctly, the 5 meters of additional distance in our case really make a difference for the detection of an onsetting surge? If so, why?

Thanks in advance,
Chris

 

[hacksaw]

yes and no, it all depends on your objectives, doesn't it.

 

[phex]

Heh. Are you in any way related to my boss? He gives quite the same elusive answers as that one j/k

Afaik, the objective of this antisurge controller is to take measures to prevent surging of the compressor as fast as possible, which means it should be able to react as fast as possible. The environment this compressor is running (yes, it is a single stage compressor) is as recycle gas compressor in a HDS unit. This means it is running at about 62 bar suction pressure and 82 bar discharge pressure. We have just one PT at the discharge side in use for the antisurge controller, although two are there (but located at the same distance from compressor discharge).

Mr. Montemayor, your advice in this context also would be very appreciated.

Thanks again,
Chris

 

[hacksaw]

obsurity not intended. you have a common problem. true surge control is extremely expensive and often replaced by a cheap variation.

you need the d/p across your compressor. you can measure it as d/p or calculate it. you also need the absolute pressures.

speed of response is important where true surge control is being performed, not only with the transmitter, but also its installation, location, and the valve/actuator speed. with only a single transmitter designed in to the system it appears that the designer had something else in mind, hence a cautious response.

good luck

 

[phex]

Thanks alot, hacksaw. Things are getting clearer. The system we have installed is not a real antisurge system as it should be and as such is probably not fast enough anyways. So the location of the PT doesn't make things significantly worse than they are anyways.

The final decision will be met by others, but at least I know enough to not make a fool of myself.

Bye for now,
Chris

 

[jsummerfield]

The line will be flowing if installed upstream of the surge valve and may be non-flowing downstream of the surge valve. I wonder if that might affect the transmitter readings.

John

 

[tgmcg]

Is there by any chance a check valve between the anti-surge recycle take-off and the PT? That would be a problem. There should be a check valve not too far downstream of the compressor discharge to minimize volume between the discharge flange and the anti-surge valve, such that opening the surge valve can quickly unload the compressor.

Otherwise, I don't see a problem with the present location of the PT as described.

Tom