PSV sizing blocked compressor outlet (Max flow through centrifugal compressor)

[Nansen]

Hi everyone!

I am validating a PSV downstream a 1st stage recompressor for blocked outlet. The compressor is normally operating with a suction pressire of 1.5 barg and MW = 43. I have the performance curves for this case and they are showing a maximum flow of 6000 Am³/h which corresponds to 20 400 kg/h (density = 3.4 kg/m³).

Now, I want to find the maximum flow through the FIXED SPEED compressor in case the outlet is blocked, assuming PSV set pressure of 6 barg upstream and MAAP 12.1 barg downstream.

Given that this is a fixed speed compressor I am thinking it cannot provide more than 6000 Am³/h in any case, but I think it will be conservative to assume it can deliver this rate for the blocked outlet case.

I also have the curve for pressure ratio vs flow but this is also made for the operating case so I don't think it is valid for the higher inlet pressure case?

Are these correct assumtions?

Are there any other way to find the maximum flow for the blocked outlet case without requiring new compressor curves?

All comments appreciated.

Best regards

 

[vt2012]

I must be confused... how can there be any flow at all in the blocked outlet case?

 

[TD2K]

< how can there be any flow at all in the blocked outlet case? >

The flow goes through the PSV to avoid over-pressuring the system.

Look at the flow curve for the compressor. Calculate the head required to compress the inlet gas from its inlet pressure to the relieving pressure. Use the compressor curve to read off the resulting flow. That is the flow the PSV has to pass.

 

[georgeverghese]

Adding to the comments from TD2K, it is not standard pactice to use the upstream PSV settpoint as the suction pressure for this PSV flow case - use the PSHH setting for the system upstream (unless there is no PSHH upstream). Beware of difference between polytropic and isentropic eff - always use poly eff for centrifugals

What is MAAP?

 

[LittleInch]

Am I missing something here? Do you not need / have a ASV anti surge line on this compressor?

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

 

[Nansen]

Thank you for your replies!

@TD2K: Unfortunately I do not have the differential head curve but I do have pressure ratio curve which I think will correspond to the head curve. If I understand you correctly you are indirectly saying that the pressure ratio curve is also valid for the increased suction pressure, not just the operating case of 1.5 barg? If so great! That would solve my problem.

@georgeverhese: Thanks! MAAP = maximum allowable accumulated pressure, the PSV allowable relieving pressure.

@Littleinch: Yes, there is an antisurge valve but this will not be credited in the scenario.

 

[georgeverghese]

Be careful of using these pressure ratio curves ; it is better to go back to basics, use trial and error and work out where you are on the Q - poly H curve for this compressor for this higher suction press.

 

[pmover]

um, how about using the pipe material, diameter, etc. to determine max allowed by the pipe?

&/or,

use highest possible compressor suction pressure and the compressor predicted performance curves to determine outlet conditions (P, T, & Q) at surge point at max RPM. if this value is greater than allowed by the pipe, then be sure to have PSV setpoint at a lower value. the PSV capacity would be slightly greater than the Q of the compressor at the outlet conditions.

no need to consider anti-surge valve for PSV analysis.

good luck!

 

[georgeverghese]

If you have a simulator with you, it should be possible to take the pain out this exercise by using an ADJUST operation in Hysis or a CONTROLLER operation in Pro II.

Input the flow, poly head and poly E for some points on the compressor curve. Define the input stream at max normal suction press( i.e PSHH setting upstream).

Vary the compressor flow to meet the target discharge press.

 

[Nansen]

Thank you all for your good suggestions, apprecitate it!
As I do not have the diff head curves I will put in the discharge pressure curves in hysys and have it generate new curves for the higher suction pressure. If the results looks reasonable compared to the existing curves this should be an appropriate solution.

 

[apetri]

I would start with some basic introduction to centrifugal copressors equations, see for example

"

http://files.engineering.com/downlo...Curves_to_Evaluate_Behavior_of_Cent_Comp.pdf"

if you are compressing a pure fluid a mollier chart can help,
for mixtures things are more complex and you may need a software to simulate the operating (basically you need to solve polytropic stages)
I solve quickly these problems in Excel with Prode Properties (Prode simulates the polytropic stage(s) and Excel solver allows to solve for unknows as the ADJUST or CONTROLLER operations in other simulators (see the post of georgeverghese).

Of course, given some operating points you can use these to predict new operating points, but take care to obtain realistic values,
interpolation and approximation curves can easily generate wrong estimates...

 

[TD2K]

< Unfortunately I do not have the differential head curve but I do have pressure ratio curve which I think will correspond to the head curve. If I understand you correctly you are indirectly saying that the pressure ratio curve is also valid for the increased suction pressure, not just the operating case of 1.5 barg? >

I'd want to talk to the compressor vendor to ensure I was correctly predicting the compressor's performance at the higher suction pressures.

 

[chicopee]

Wouldn't the maximum flow be at free delivery and be calculated based on RPM, piston diameter and stroke length which would be the relieving capacity of this intermediate PSV? The set pressure would be based on the design pressure of that first stage.