surge 2

[142846]

It's a general question.

we know about the existence possibility of surge in centrifugal compressors. Is there any possibility of surge in positive displacement compressors? Is s there any possibility of surge in centrifugal pumps or positive displacement pumps?

Thanks in advance.

 

[Lou Scannon]

A centrifugal pump that is pumping liquid can stall, which is part of surge, but without the consequences of a stored volume of compressible fluid downstream of the impeller.
I can't imagine surge, as such, in a positive displacement machine.

"Schiefgehen wird, was schiefgehen kann" - das Murphygesetz

 

[LittleInch]

"Surge" in this context is low or no flow. You need to be careful how you use it as "surge" on liquid pipelines usually refers to pressure rises or spikes due to sudden valve closure or sudden loss of flow due to pump trips.

So PD pumps or compressors, if you restrict the flow or shut the flow off can very quickly increase the outlet pressure until the motor stalls. Hence ALL PD type pumps and compressors need to be fitted with what I guess you could call surge valves or relief valves either back to suction, back to the tank or out out to the open.

Centrifugal pumps in liquid service can withstand a certain amount of time dead head, but over time which varies by pump size and pressure, will gradually or maybe quickly overheat as the liquid churns around inside the casing with no flow through the pump. This is generally regarded as a BAD thing and usually you either add temperature trips to the casing or low flow alarms and trips to prevent damage to the pump. Some people like to put minimum flow bypasses or "spillbacks" [I hate that terminology], but this is used more for utility pumps or where flow starts and stops on a regular basis and you don't want to keep turning the pump on and off many times an hour.

NOTE: This does not apply for some reason to fire water pumps where for some god forsaken reason, the NFPA codes often seem to require a dead head test every few weeks for up to 30 minutes (!!) How these pumps survive is a bit of a mystery to me....

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

 

[1503-44]

Surge in a centrifugal pump can also occur during a pump trip when high pressure fluid in the downstream pipe may backflow into the pump. Under normal running conditions, pumped liquids do not have large increases in volume due to relatively small pressure reduction variations, so surge as experienced with backflow of gas in a compressor does not occur. Generally positive displacement machines do not allow any significant backflow.

 

[JJPellin]

This is very dependent on the definition of surge. I use the word surge to describe the stall and reversal of flow in a running centrifugal compressor. Using that definition, it could not occur any any pump or positive displacement machine. I have witnessed compressor surge many times. It can be violent and destructive. I have never witnessed anything like that in a PD machine or pump.

Pressure spikes in liquid piping systems are more commonly referred to as water hammer of liquid hammer.

The back flow of liquid into a pump when it is shut off is just backflow. I have never heard this referred to as surge.

Johnny Pellin

 

[LittleInch]

term should be "compressor surge".

"surge" is most definitely used on pipelines to describe other events, not just "water hammer" [never liked that term] which also occurs in other fluids.... It always implies a sharp bang or spike which isn't always the case.

You often also get it in two or three phase fluids when you get surges or slugging, often during normal operation as well as start / stop or upset condition.

You don't tend to notice it in a PD pump or compressor because the requirement for a relief / recycle valve is so basic, but if it didn't have one or was blocked it can also be pretty violent if you try and dead head a running PD pump or compressor... something gives - either the motor starts smoking or the coupling gives way or the pressure blows something up.

Compressor surge is pretty violent alright.

But otherwise I agree with JJ - Get your definition right.



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

 

[142846]

Hi guys. Thanks for the all great replies.
First of all I think there is a basic difference between the water hammer in liquid piping services because of fast closing the valves or pumps trip and surging that occurs because of the backflow of gas caused by upset in process conditions. In this thread I do not mean water hammer.

Lou Scannon (Automotive)
A centrifugal pump that is pumping liquid can stall, which is part of surge

As far as I know stall and surge are two different phenomena and stall is not part of surge.

1503-44 (Petroleum)
so surge as experienced with backflow of gas in a compressor does not occur.

Brief and useful

JJPellin (Mechanical)
Using that definition, it could not occur any any pump or positive displacement machine.

Great. In your opinion what is the technical reason that surge does not happen in positive displacement machine?

LittleInch (Petroleum)
term should be "compressor surge".

You are completely right. I don't know is there any way to correct the title of the thread?


Finally as far as I know pressure relief valves on the outlet of positive displacement pumps and compressors are installed because of prevention of overpressure and it has nothing to do with surge or stall.
Again I appreciate for all comments.

 

[1503-44]

Compressor surge initiates when the compressor does not have enough power, or "stalls", which is when a significant amount of excess gas volume created by the resulting pressure reduction backflows into the blades. The same stalling in a pump does not create a huge excess volume, hence no backflow into the pump impeller. Kind of the same thing, but with different results.

I think LittleInch was talking about the positive displacement machines response to overpressure created at discharge, which is to blow excess pressure through the reliefs. If they don't, something, rods, shafts, etc could break, as otherwise they just keep on increasing pressure to try to move more fluid to discharge until the motor runs up to max power output, "stalls" and burns, or the pipe pops, whichever comes first.

 

[LittleInch]

We're all getting wrapped in misunderstanding of words.

For what is described as compressor surge should really, IMHO, be called compressor stall. The impact of this is violent surgING of flow, sometimes several times per second as the compressor blades stall and then recover then stall and ....

"surge" is normally used to describe a single event when something, gas or liquid, surges in the system creating low or high pressure.

So I guess your question really was, "At low or no flow, do PD pumps and compressors and centrifugal pumps, exhibit similar characteristics to a compressor stall event (violent changes in flow)?"

If that was the question, then the answer is no.

But that's not to say you don't need pressure relief recycle systems (PD pums / compressors) or minimum flow recycles (centrifugal pumps) for other reasons...

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

 

[SwinnyGG]

I agree that confusion in terms is getting us off track.

The most simple definition of 'compressor surge' is rapid oscillations in fluid velocity in the axial direction of the compressor.

This can happen both with and without negative flow rates - ie backflow into the compressor is not a prerequisite for the compressor operating in a surge state.

Backflow (negative mass flow) does mean there is more energy being dissipated in the surge cycle the compressor is experiencing, which means surge events with periods of negative mass flow are much more destructive, but the negative mass flow itself is not the root cause of damage; it's the rapid changes in pressure due to rapid velocity changes in the flow.

 

[Lou Scannon]

As far as I know stall and surge are two different phenomena and stall is not part of surge.

As far as you know.

"Schiefgehen wird, was schiefgehen kann" - das Murphygesetz

 

[142846]

A centrifugal pump that is pumping liquid can stall, which is part of surge, but without the consequences of a stored volume of compressible fluid downstream of the impeller.

I would be grateful to know the viewpoints of others on this issue that stall is part of the surge.
Based on my understanding surge generally occurs at low mass flow rate in opposition of stall occurs at high mass flow rate. In addition surge compared with stall has very bad impact for compressors. Therefor in previous post I said that "As far as I know stall is not part of surge." If I am wrong please correct me.

thanks a lot.

 

[SwinnyGG]

By definition, a compressor which is not stalled can be operating in surge.

 

[Lou Scannon]

Surge is by definition a transient phenomenon. The compressor operating point can be forced momentarily across its surge line, but it cannot stay there (in most realistic situations). The airfoils stall (at least some of them), and pressure and flow drop precipitously, crossing back into a non-stalled, viable operating area. If no external adjustment is made to the boundary conditions, the operating point may follow the same trajectory repeatedly, i.e., increasing in pressure, crossing the surge line, stalling, and so on.
It is also possible for a compressor operating point to ride right on the surge line, but then it is just executing the cycle describe above on a small scale, where the situation is quasi-stable.
I can imagine an experiment with special apparatus that forces a compressor to be continuously on the wrong side of the surge line, but then it is continuously stalled (at least part of it) also.
If you look at a compressor map you will see that the surge line delineates an area of relatively high pressure ratio vs mass flow, from the stable operating area.

"Schiefgehen wird, was schiefgehen kann" - das Murphygesetz

 

[142846]

As far as I know stall and surge are two different phenomena and stall is not part of surge.

I apologize to everyone because the two terms "stall" and "stone wall" have been confused by me. Specially thanks from Lou Scannon to aware me.

Any way as I understood the surge phenomena (I mean surge caused by insufficient flow rate) is occurred/happened only on centrifugal compressors and surge (because of insufficient flow) is not happened in positive displacement compressors. In the other hand positive displacement compressors do not need ASV (anti surge valve) for recycling the flow to inlet. Was I right?

 

[LittleInch]

More or less correct.

PD compressors need pressure relief instead of ASV to avoid high pressure during low flow.

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

 

[compguy22]

Surge in the context of centrifugal compressors has already been clarified. It is reversal of flow that occurs because the compressor cannot generate enough pressure (or head) that would keep the flow moving forward and overcome the resistance of the system downstream. This can happen for instance when the flow is reduced and approaches towards the section of the head curve whereby the curve gradient is not steep enough or when machine decelerates (example emergency trip) but the inertia of the downstream system is such that system pressure is not relived quick enough, in both instances the flow tends to reverse towards the compressor. As the system pressure decreases due to flow reversal, the differential of pressure between compressor and downstream system equalizes and becomes again positive so the flow moves forward again, but the same cycle repeats. These cycle of flow reversals often lead to high vibrations, potential mechanical damages and may lead to loss of containment, all tough this is very exceptional. The solution is to include an anti-surge detection and control system to protect the compressor. The anti-surge system should be sized so to react promptly, have enough flow capacity in terms of satisfying capacity required at stable operating point (so called control line) plus provision dynamic response. It should not be sized too big however so not to put the compressor in choke, so a compromise is desirable.

While surge occurs in the lower flows region of the operating curve, stall however is the result of adverse pressure gradient that normally occurs in the vanned diffuser passage, typically at the inlet. Adverse pressure gradient means abrupt loss of head, similar to when a wing of an airplane angle of attack exceeds the critical angle thereby causing the lift coefficient to drop abruptly. Rotating stall is the process whereby - at an existing impeller station - a single cell exhibiting stall would move from one diffuser vane to the next in a direction opposite to the direction of rotation of the shaft. There is vibration signature typical to this phenomenon. Rotating stall is more common on compressors with low flow coefficients whereby blades with narrow widths are applied; so that also corresponds to high pressure machines, but it is not limited too. Rotating stall can occur at any area of the overall compressor map even towards choke region, and it does not necessarily mean surge, but is often a precursor of it. Rotating stall is generally predicted by means of Kobayashi and Senoo criterions (which involves calculations on velocity triangles / blade angles at exit of the impeller and entry of diffuser, flow coefficients, etc).

 

[georgeverghese]

"In the other hand positive displacement compressors do not need ASV (anti surge valve) for recycling the flow to inlet."

Yes, but you still need a capacity / pressure recycle valve (PIC - PCV) for normal operations. It need not be as fast acting as an ASV in a centrifugal compressor application. The PSV is also required when there is failure of or poor response from the PIC-PCV.