Basis of sizing compressor Surge control valve

[cubicsupport]

I am trying to understand on what basis is a centrifugal compressor surge control valve sized. So far I have seen the following basis:
a) 1.8 to 2.2 times the CV at surge control line
b) 0.9 to 1.1 times CV calculated at intersection of horizontal line from design point with 105 % speed.
c) Valve sized for design control point

As all of the above provide different CV values, is there an industry standard? or are there more rules of thumb than I have looked at.

 

[pmover]

the criteria you've stated comes from a reliable source of many years of practical experience. there are cases, although infrequent, in which this criteria may be in excess - from my experience. however, for the majority of cases, the sizing criteria you've stated is adequate and meets the control system requirements. some factors to consider are location of surge control valve, pipe diameters, pressure/temperature/other properties of gas, and surge control valve size. as far as an industry standard, i am not aware of one.

perhaps you should make inquiries to other surge control system mfg's for their recommendations. also, different valve mfg's provide Cv values for their valves. Oh! other considerations are actual operating modes for the compressor (i.e. start-up, normal mode, process upset, min/max flow/pressure conditions, shutdown, predicted frequencies for any previously stated conditions, etc.). the case may be where major process upset condition may frequently(?) be expected/predicted, thus the user/operator may need a larger valve (Cv) to adequately meet those requirements.

sorry, i am not familiar with other industry stds, but i hope the aforemention information is helpful. each application is unique and sometimes requires careful and meticulous thought.

feel free to respond and good luck!
-pmover

 

[1969grad]

The most important criteria for a surge valve is the stroking time. Since the transients in the region close to the surge line can be very fast, the antisurge valve's speed of response must also be very fast - two seconds or less from completely closed to full open. It is not necessary to meet this same time criteria for closing.

A system with an undersized valve is never capable of preventing surge and could lead to damage. However a greatly oversized valve (bigger is never better) could result in a choked flow condition at the very worst or an unstable operation in general.

The most common approach I think is to use 1.8 to 2.2 times the calculated Cg calculated at the point on the operating curve where the surge line intersects the max speed performance curve.

BTW, the speed of response is also important in the selection of the controller and process transmitters. For the later we never use the Rosemount smart transmitters in this application.