Anti Surge, Hot Gas Bypass And Check Valves 1

[Rha257]

Hello All,

I have two questions regarding the discharge valves of a multistage centrifugal compressor:

1. In the case in which a dynamic simulation shows that a hot gas bypass valve is required on the compressor discharge, I have usually seen two dissimilar check valves right after the hot gas bypass take-off line. However, in case a hot gas bypass valve is not required, do we still need to have the check valves at the discharge of the compressor even though we have the check valve downstream the anti surge valve take-off line? If yes, why?

2. A compressor vendor once told me that, in case the discharge piping volume from the compressor discharge is too large, the anti surge valve size has to increase accordingly which will lead to a significant reduction in the backpressure in the recycle line during flow recycle operation which will impact the gas forward flow. I did not really understand this concept so if someone could please elaborate about this issue of increasing the ASV size due to increase discharge volume.

Thank you in advance

Regards,

Rami

 

[pmover]

1. a schematic would be helpful
2. the principle is to decrease the resistance across the compressor so positive forward flow is maintained. so, the more pipe volume between compressor and anti-surge valve implies an anti-surge valve that is capable of reducing the resistance across the compressor w/in a reasonable time frame.

 

[Rha257]

pmover, thank you for your response.

I currently do not have a schematic but the main question is basically whether two dissimilar check valves are required on the discharge of the compressor nozzle (i.e. on the upstream of the ASV take-off line) in addition to another check valve on the downstream of the ASV takeoff line.

With regards to the second point, what are the disadvantages of going to a large ASV? Aside from price and stroking time since we know that there are some manufacturers who are capable to manufacture 48" ASVs that have a stroking time of less than 1 second. The compressor vendor was stating that the main issue is mainly on reduced back pressure on the ASV recycle line which does not make sense to me.

 

[georgeverghese]

In all cases where I have come across hot gas bypass on compressor discharge, there is only one check valve downstream of HGB takeoff. In the case where there is no hot gas bypass, it is customary practice to install a check valve almost immediately after compressor stage discharge. This is typically a special non slam check valve.

It is also customary practice to install a check almost immediately after ASV take off piping branch. Hence, it doesnt matter if there is a large gas volume after the ASV takeoff, and there is no need to use a larger ASV to account for the perceived delay in discharge side pressure decay during a pressure upset event.

 

[Rha257]

Thank George for your response. I appreciate your response on this subject especially that I consider you as a legend in this forum.

With regards to the first point, what is the purpose of installing a check valve immediately after the stage? Is it to limit the volume on the discharge side of the compressor in such a way to reduce the volume of fluid that could reverse flow back to the compressor during a surge? Also, why non-slam type check valve is preferred? (does it prevent the creation of shock waves on the upstream side of the valve during shutoff or is it just because of increased reliability of non-slam check valves)

With regards to the second point, I did not mean the volume on the downstream of the ASV take-off line instead I meant the volume between the compressor discharge nozzle and the check valve on the downstream of the ASV take-off line. The OEM said that this volume must be limited to 2 x effective volume flow rate of the compressor otherwise the ASV will be large and this will affect the backpressure in the recycle line. I do not really understand how the size of the ASV (i.e. being larger) could affect the backpressure in the recycle line.

Regards,
Rami

 

[georgeverghese]

Yes, the check valve on the compressor discharge is to prevent transient reverse flow into the compressor during a potential surge event. I dont remember why this is typically a non slam check valve, but it may have to do with reliability also.

Okay, this is the first time I come across this thumb rule. Concept wise, it make sense. I can see that there will be a delay in depressurisation of the compressor discharge if there is a large volume between the compressor exit nozzle and the check valve downstream of the ASV takeoff. And hence this recommendation is to limit this hold up volume, else a larger ASV may be required to correct for the delay in depressurisation and also reduced rate of depressurisation. But the downside to a larger ASV is control instability during ASV mode recyling. This control instability can also be seen as transient reduced pressure upstream of the ASV. In most cases, there is only a small volume between compressor exit and the ASV take off branch check valve, which consists only of piping into and out of the discharge cooler. Would be a good idea to check that this volume holdup recommendation is complied with.

 

[Rha257]

I do not think that this recommendation really holds up in most cases since I have seen two different projects with the same OEM and the volume between the discharge nozzle and the inlet of the anti-surge valve is more 3 or 4 times the effective volumetric flow. However, it is definitely an interesting recommendation.

With reference to your second paragraph, what do you really mean with control instability and transient reduced pressure upstream of the ASV? Why does a larger ASV cause a reduced pressure on the upstream of the ASV?

 

[georgeverghese]

Because a larger ASV means poorer turndown. It cannot operate at low recycle flow in a stable manner - typically the lower turndown limit of a control valve is not better than 30:1. When the control valve is unstable at low flow due to oversizing, then you get pressure fluctuations upstream of the ASV.

 

[Rha257]

So if I understand correctly, having a larger control valve can lead to the constant fluctuation of the stem position due to the poor controllability at low recycle flows. So to some extent, the valve is trying to find the appropriate set point to meet the desired process targets. Is that correct?

This is very interesting to know since I was discussing with a major valve OEM about the issue of having an ASV of more than 18" and they told me that they had installed a 48" ASV in a gas processing plant and that this valve had a stroking time of 1 second. I don't know if you have come across such an ASV size before but it seems very interesting to me.

Since we are discussing control valves, I had a question related to the suction throttling valve for a fixed speed compressor. Assuming that the compressor must be always running at 20 Bara at the suction flange and 150 MMSCFD flow, however, due to possible process changes the suction pressure rises from 20 Bara to 23 Bara, this valve must throttle the pressure from 23 Bara to 20 Bara. My question is, can the STV kill 3 Bara or possibly higher while keeping the flow at 150 MMSCFD? I was discussing this with an experienced engineer and he told me that with a globe or bitterly type STV this is possible only if the line size is slightly oversized in such a way that even if the valve closes slightly to throttle the pressure from 23 (or possibly higher) to 20 Bara, the flow will not be impacted at all and will stay constant at 150 MMSCFD. What do you think about this subject?

 

[georgeverghese]

That is a huge ASV at 48inch, the recycle line must be even bigger. At these sizes, would guess it may be better to split the ASV into a low range and high range ASV, so that recycle turndown is improved, especially if the ASV is in combination capacity control service also. On the other hand, it may be better to avoid capacity control on such large ASVs' - use a separate capacity control valve.

Yes, that should be possible to kill 3bar across this STV if you are limited to fixed speed operation. An equal percentage trim would then keep pressure drop minimal when the valve is wide open when throttling is not required. Run a few sizing calcs and choose one that meets all your operational requirements.

 

[Rha257]

So an equal percentage trim has a lower pressure drop across the valve compared to linear or quick opening trim? What is the reason for this?

 

[georgeverghese]

Sorry , got it muddled up, a quick opening trim would be better in this service, since the ASV is fully closed initially ( steeper CV rise for each increment in valve stroke as it opens).