Surge and stonewall of centrifugal compressors

[Yarik]

Hi everyone,

I would like to ask about some advise in regards to these two unstable conditions: surge and a stonewall.

One of my multistage compressor experienced a malfunction o a blow-off valve, therefore the compressor protection did not kick in when the compressor reached the control line. The compressor entered surge (flow fluctuated to 0 and up, vibration picked up). After few minutes the blow-off valve opened, suddenly and fully. As a result system air pressure dropped and IGV of the compressor opened fully. Condition of the compressor did not change (flow fluctuations and erratic vibrations continued). Even after the blow-off valve closed automatically the state did not change. Eventually the nightmare finished when the compressor tripped on high discharge temperatures (by the way, inlet temperature increased from 20 to 150 deg C)
After analysing the data I came across few questions, which I hope someone would be able to assist me with.
1. Why did not the condition stabilised after the blow-off valve opened?
2. Thrust bearing vibration did not change much (change of maybe 0.002 mm, no erratic fluctuation).
3. How will a stonewall look in terms of vibration, discharge temperatures, amps? In my understanding, vibration would not change, discharge temperatures would increase, amps will jump sky-high.
At this stage I cannot understand whether it was a combination of a stonewall and surge, or only a severe surge. I am also concerned about the level of the temperatures, that the compressor faced.
thank you

 

[rotw]

I would think of the following aspects :

1/- There could be certain volume between the compressor discharge flange and the blow off valve (depends where the valve is located). So when the blow off valve opens its effect as a relief of the system pressure downstream the compressor could not be that immediate.

2/- In my view a high discharge temperature at the inlet may be due to the reverse flow phenomenon and mixture of flow between both sides of the compressor, but I am not sure - never theoretically faced such issue.

3/- If the control valve went fully open and stayed locked in that position then flow could not be modulated so the only way to recover the system pressure was IGV to go full open. But that reasoning is on the basis of a "normal " operating map.

A question we can ask ourselves is what would happen to the design compressor map or curve if you change drastically the inlet temperature conditions...(quoted from 20 to 150) so back to point 2/ above. I wonder if this scenario wil not bring the compressor to the point of no return in view of out the control system standard algorithms ?

By the way you mention that the IGV went fully open but you did not say you if the angles are on the upper side of the map (my guess is that this could be the case to recover system pressure) or on lower side ?
The discharge temperature is the highest when IGV opens towards the upper side of the map combined with flow at control line.

I think stonewall is not a problem for integrity of the machine (vibration). It is certainly a problem for poor performance and potentially issue on axial thrust. But you indicate no abnormal indication on axial thrust.

The system tripped according to the trip setting implemented by design, why it did not trip on vibrations that is a matter to be discussed. Few minutes of surge seems to me something serious.

 

[jbattershell]

the control system may have been working properly and attempted to open blow off valve when the compressor reached the surge line, but the valve may not have opened due to being stuck. you would have to look at the history data (if available) to see if the signal was given to the valve at the point which the compressor hit the surge line.

a compressor in surge will often cause a rise in temperature on the inlet sensors due to flow reversal. hot gas surges back through the machine and heats up inlet temperature sensors.

i don't know the history of the machine but would ask if the blow off valve is large enough. is this the first time the blow off valve was called into action? do you have history of the blow off valve opening and the machine getting out of surge? if so, then why not this time. if there is no history, then you may have found out for the first time that the blow off valve is ineadequate. after the blow off valve opened, did the pressure drop to get out of surge?

is the blow off valve modulating or just an open/close valve?

stonewall is generally not bad for a compressor but certain compressor types are sensitive to aero thrust balancing and if the machine is in stonewall, the balancing of thrust is not correct and could result in overthrust. so most operators will try to avoid stonewall.

surge may not hurt a compressor, or it may. it depends on the compressor type, and vendor. generally, increased vibration is noticed with a surge and if allowed to continue, the compressor will often trip on an inlet temperature detector going high. intergrally geared machines will have inlet temp sensors in each stage normally, but a multistage barrel compressor will only have sensors at the inlet and discharge of each section. but some good surging of a barrel compressor could still get a temperature rise on the inlet to the compressor from the flow reversals.