Suction throttling valve on centrifugal compressor: sizing criterion?

[JustSomeRoark]

Dear gents,

What is the most common sizing criterion to design a centrifugal compressor suction throttling valve? On one hand, I can allow in the STV a DP equivalent to some percentage of the system dynamic pressure drop of the suction side (some 20-25%). On the other hand, some people argue that it is sufficient to consider a butterfly valve with the size of the inlet line to minimize the pressure drop and therefore the energy consumption.

I think that in the second scenario (i.e. line size valve) the controllability would be affected as there would be no margin on flow available, but I have not seen a definitive consensus.

What are your thoughts/experience?

Regards.

 

[LittleInch]

Choose ONE forum and please delete the second post before someone else does it for you....

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[1503-44]

You can't size it without consideration of the flow rate range that you expect. On the suction side that often means max flow at lower suction pressure range and lower flows at higher suction pressures range, but it can sometimes mean the opposite.
The most critical case is usually getting a target flow rate into the compressor at the lowest suction pressure while at the highest suction temperature range.

Operation is most economical at higher suction pressures, as you get a lot more gas into the compressor and it usually requires the least amount of differential pressure that the compressor must provide, so that's the best suction side operating point. But the critical case is getting maximum flow rate into the compressor at the lowest suction pressure and highest suction temperature. So, the valve must be able to pass the highest expected flow rate at the lowest expected suction pressure & highest suction temperature. A max Open of 80% position should typically correspond to that case.

Once you have the valve sized for that condition, check the lowest flow rate at the highest pressure and lowest temperature. That's the pressure where you usually should have the valve at a minimum open position. Generally not lower than 20% Open.

--Einstein gave the same test to students every year. When asked why he would do something like that, "Because the answers had changed."

 

[georgeverghese]

Is this compressor fixed speed or variable speed ? Can you post a PFD and narrative on compressor suction controls and the upstream source pressure controls also ?
Is this control valve normally full open or is it throttling ?

 

[JustSomeRoark]

@georgeverghese:

I am not at my desk now and I cant draft a sketch, but for the sake of simplicity think of the default one stage, fixed speed, centrifugal compressor, with suction throttling and the pressure upstream of the STV being controlled by the compressor performance controller.

Something like this with the pressure of the point marked with the arrow controlled by the compressor control system.

Regards.

 

[JustSomeRoark]

Oh, and regarding your last question whether it is fully open or throttling, I think it is at the very crux of the issue. If the valve is throttling at the rated point, as I would expect, there is an energy penalty in the compressor. If it is fully open, I would expect the controllability to suffer, same as would occur with a control valve having no pressure drop in a pumped system.

 

[1503-44]

At the compressor operating point, the valve could be 80% to 100% Open. You probably do not need to throttle at the precise rated operating point. Let the suction pressure float with the valve fully open. The valve could be 100% Open as far as the compressor is concerned. Why do you have to control, when the compressor is very near its duty point? It should be entirely stable there.

When the compressor reaches the end of operating range, its max limit caused by rising high suction pressure, then the valve should be at its 70-80% Open range. If suction pressure continues to rise, then close the valve more and more. If you reach 20% closure and the suction pressure is still rising, you should probably be looking to close the valve entirely, shut down the compressor/process and figure out what's wrong. The valve should not go into action at all until the compressor needs it to. That should be only when it is moving to outside normal high operating conditions, typically 10 to 15% above the compressors duty normal point. Inside the normal operating range, the compressor should be fine running with a floating suction pressure. Is there some need we do not know about for active suction pressure control when suction pressure is inside the usual operating range?

If the valve is throttling at the rated point, change the set point, or deaden the PID sensitivity, so that the valve is near 100% Open at compressor's rated operating point and reaches 70-80% Open when suction pressure reaches the high condition. Then you will have the valve in a good position to begin actively controlling suction pressure.

--Einstein gave the same test to students every year. When asked why he would do something like that, "Because the answers had changed."

 

[pierreick]

Hi,
Consider this design to support your work.

Pierre

 

[georgeverghese]

You've indicated 2x that the compressor performance controller picks up a pressure signal from UPSTREAM of this throttle control valve, so you've made it clear enough. But I've yet to see such a scheme in operation. In all such cases, the variable speed compressor has a normal operating pressure controller (sensing pressure DOWNSTREAM of the suction throttle valve) to reset the speed controller ( part of performance controller) operating the compressor. The suction throttle valve operates from another HIGH set PIC also sensing pressure DOWNSTREAM of this throttle valve.
How is this scheme you've shown supposed to work ??

 

[1503-44]

Isn't the valve just keeping suction pressure below some max compressor inlet pressure?

--Einstein gave the same test to students every year. When asked why he would do something like that, "Because the answers had changed."

 

[danschwind]

Guys,

The capacity control IS the throttle valve. There is no variable speed here.

JustSomeRoark,

Do you already have the compressor selected? If so then it is easier because you already know what suction pressure the compressor must have to develop the discharge pressure that I assume is defined by another thing (other compressor, control valve, etc).

If you don't have it and are doing everything from scratch, then it is pretty arbitrary. I've seen designs with 50 kPa and 100 kPa pressure drops @ duty flowrate. Usually aim to a ~70% open valve.



Daniel
Process Engineer
Rio de Janeiro - Brazil

 

[1503-44]

The STV is also indirectly controlling compressor inlet conditions through its pressure drop, and it is the pressure upstream that is actually being controlled, despite what the typical method is, ie. controlling pressure downstream of the STV.

JustSomeRoark, maybe you should read this now,
See Controlling Compresors

https://rotatingmachinery.com/wp-content/uploads/2023/07/Controlling-Centrifugal-Compressors.pdf

--Einstein gave the same test to students every year. When asked why he would do something like that, "Because the answers had changed."

 

[danschwind]

It really depends on the overall process. I assumed that as he is controlling upstream pressure, that it is controlling a vessel pressure (HP separator, free water KO, amine or glycol contactor, etc).

Daniel
Process Engineer
Rio de Janeiro - Brazil

 

[1503-44]

We don't really know that. As it is, it's just floating on suction pressure minus STV pressure drop. We don't know whether a 20% drop is appropriate for the compressor, or if it should be at 80%. But in any case, the valve still needs only to be selected in the usual manner. Suction pressure being whatever is needed to supply either the min or max flow, the valve still must pass max flow at 70-80% max open and min flow at 20-30% max open. He will have to live with whatever valve size(s) accommodates that range with a pressure drop less than the available suction pressure at the time. Maybe two different sized parallel valves, one for high suction pressure and the other for low suction pressure might work even better for minimising pressure drop, but supposedly there is a valve cost vs minimizing pressure drop cost limit too. All of those are beyond our scope of knowledge and control. At least now he knows what to do about it.

--Einstein gave the same test to students every year. When asked why he would do something like that, "Because the answers had changed."