CT Compressor Surging 3

[Uncleburt]

We have a CT Compressor surging problem with one of our CTs. I can attribute the problem to excessive diaphragm clearances and compressor blade deposits. However, surging only occurs above ambient temperatures above 95 degrees F. I cannot find a reason why this would occur.
Any ideas?
Thanks!

 

[25362]

Just a thought: consider the effect temperature may have on the density and/or molecular weight of the gas.
BTW, what is the gas composition ?

 

[Uncleburt]

The working fluid is ambient air, which of course fluctuates in composition. The unit is primarily operated during summer months (using natural gas as fuel), therefore, capability is capped. I failed to mention that the surging only occurs just as the unit is reaching full capacity and ambient temperature is 90-95 deg F.
I considered temperature effects on density, however, I do not see how it could lead to surging, however, I am relatively new to CT's.

 

[rmw]

Where is that one problem CT located in relation to the other units and the wind rose? With relation to the other units?

rmw

 

[toothless]

By "CT" do you mean combustion turbine?

 

[gtsim]

How do you know the compressor surged. Compressor surge normally results in severe damage to compressor and can even destroy the gas turbine.

 

[BigInch]

Density variations can and do often cause surging and surging does not necessarily result in failure. Is this turbine identical to the others? If not, in what way does it differ? See rmw above.

http://virtualpipeline.spaces.msn.com

 

[gtsim]

BigInch,

In GT performance, density does not pay a big role. Other parameters are more significant such as rapid compressor inlet temperature changes. Inlet pressure changes are not as severe as Mach no is largely independent of pressure (e.g. Mach no= Velocity / (Gamma x R x T) ^0.5). Also, the original question refers to surge at high ambient temperature only and makes not reference to ambient pressure or wind. Please note with gas turbines, we normally work with non-dimensional parameters or Mach no for flows and speeds and density is allowed for in these parameters. Indeed, we present compressor and turbine characteristics by plotting pressure ratio VS non-dimensional flow for a series of non-dimensional speeds.

If the compressor has been persistently subjected to surge at high ambient temperature as suggested by the question, then the likelihood of damaged is significantly increased.

Regards,
gtsim

 

[BigInch]

Does the inlet temperature change NOT cause a change in density that affects your turbines? If not, how does inlet temperature change affect your turbines.

I do not understand the sentence, "Please note with gas turbines, we normally work with non-dimensional parameters or Mach no for flows and speeds and density is allowed for in these parameters." What do you mean with "density is allowed"? Density can not be, "allowed". Do you mean, "and density is included in these relations". Please explain.

I would not expect inlet temperature changes to be as severe as mach number variations. And note that I did not say that surge does not damage turbines. I was contradicting the previous statement that implied turbines always are damaged by surge, to which you apparently agree.

Why should it matter if p or p/P, is plotted against q/Q or just Q ,or RPS or rps/RPS? I would use non-dimensional plots anytime I wanted to make characteristic curves for a series of aerodynamically identical turbines, or pumps or water turbines or most similar machines that would be constructed to different scales. I don't exactly see what that has to do with surge limits per say, except possibly that is what you are trying to indicate by such plots. Please explain why this makes these GTs different from other rotating devices.

And, lastly, are you saying that high ambient temperatures could not cause surge? What is your opinion in relation to the original question?

http://virtualpipeline.spaces.msn.com

 

[gtsim]

In gas turbines, non-dimensional parameters are not those implied in your reply. Also, please note fan laws do no apply for gas turbine compressors. Compressor and turbines characteristics are therefore plotted on a Mach number basis because they are valid for all pressure and temperatures encountered in gas turbines provided the flow is turbulent, which is the normal case for industrial units. The non-dimensional groups for flow are W*((ZRT/k) ^0.5)/AP and for speed ND/ (ZkRT) ^0.5 and know as non-dimensional mass flow and speed respectively.

W = Inlet mass flow rate kg/s
Z = Inlet compressibility factor
R = gas constant (kJ/kgK)
T = Inlet temperature (K)
P = Pressure (Pa or Bars)
k = cp/cv at inlet
D = rotor diameter (m)
A = Inlet area (m^2)

These non-dimensional parameters are in fact Mach numbers at inlet to the compressor or turbine. You can refer to GT Theory and GT Performance for more details on representation of gas turbine component on a non-dimensional basis. The presence of P, T R and Z now allow for density effects. That’s what I meant that density effects ARE allowed in these parameters.

Consider a rapid (transient) increase in compressor inlet temperature (T). The compressor non-dimensional speed (ND/ (ZkRT) ^0.5 decreases for a given mechanical speed N (i.e. the aerodynamic speed decreases, cause in the compressor inlet mass flow to decrease rapidly and can result in surge). The change in pressure does not affect the compressor non-dimensional speed a great deal because Z, k and R are nor affected a great deal at normal working pressure and temperatures that occur in gas turbines. Thus transient pressure changes are unlikely to cause surge. Also, these parameters are often omitted and the resultant parameters are called the quasi non-dimensional parameters (please see GT Performance). In the steady state gas turbines can and do operate over a wide range of ambient temperatures without any problems including the ambient range stated in the question

Compressor characteristics based on these parameters clearly define the stable part of the compressor for all pressures, temperatures and speeds. Therefore, regions where we cannot operate due to surge are clearly defined by such maps.

My point was if surge was a problem at the operating conditions stated in the question then damage would result particularly in gas turbines. This is because the pressure ratios and thus temperature ratios are quite large in gas turbine compressors. Therefore during surge the compressor discharge temperature increases significantly and can cause severe damage, where compressor blades can even melt. That’s why I inquired ‘how do we know that compressor surge has occurred’.

It is hard to say much about this problem on the information so far provided. However, if the compressor has deteriorated in performance then surge line will drift further into the stable operating range. This is particularly so if the clearance between the rotors and casing has increased due to rubs. Also, the operating point on the compressor map will now move towards surge due to the loss in compressor efficiency. These factors have to be quite severe before the problem described occurs (i.e. a significant loss is compressor performance has to occur).

I trust the above helps!

Regards,
gtsim

 

[BigInch]

gtsim, Thank you very much for your time in responding to my questions. Your answers are concise and very informative. You deserve a star. Thank you.

http://virtualpipeline.spaces.msn.com

 

[gtsim]

BigInch, You are very welcome and thank you very much for your star. It is very much appreciated.

Best wishes,
gtsim

 

[toothless]

Surge can occur due to low mass flow. When the air is hot, it is less dense therefore the compressor will have a lower flow rate.

If the inlet guide vanes are out of calibration and are throttled more than expected, the air flow will be lower than design and could cause a surge.

Also, aerodynamic issues with the stator and/or rotor blades can negatively affect the air flow and result in surge.

Also, if there is some pluggage at the outlet of the compressor that would reduce the air flow and cause a surge. This would be easily detected by looking at the compressor discharge pressure.

 

[gtsim]

1) Surge can occur due to low mass flow. When the air is hot, it is less dense therefore the compressor will have a lower flow rate.

Yes, but the compressor pressure ratio also falls due to component matching (i.e. matching of the compressor, combustor and turbine). Therefore the gas turbine can operate surge free at high ambient temperatures. But during a abient temperature transient this can be a problem as discussed above.

2) If the inlet guide vanes are out of calibration and are throttled more than expected, the air flow will be lower than design and could cause a surge.

Yes, again the reduction in flow also reduces the compressor pressure ratio as discussed in 1) above. Also, the closure of the VIGV increases the region of surge free compressor operation and one the reasons we use them.

3) Also, aerodynamic issues with the stator and/or rotor blades can negatively affect the air flow and result in surge.

The proper design of the compressor should ensure satisfactory operation over the range the compressor was designed to operate.

4) Also, if there is some pluggage at the outlet of the compressor that would reduce the air flow and cause a surge. This would be easily detected by looking at the compressor discharge pressure.

Sure, but if this was the cause of the problem then we should also observe the compressor surge problem at low ambient temperatures.

Regards,
gtsim

 

[toothless]

Regarding #3 above, obviously the compressor is designed to operate over a given range. But unless that compressor is brand new it will not operate at design. As it becomes fouled or the gemometries change due to wear (erosion, corrosion, etc.) the flow characteristics will also change.

I do not agree with your statements in #1 and #2. If you cut the air flow back by increasing the temperature significantly or throttling the inlet flow, you will approach surge. Don't believe me? Go out and do that on your centrifugal air compressor - see what happens.

 

[gtsim]

Regarding #3 agreed, but the design of the engine would allow for such deterioration.

Regarding #1, #2, only in a rapid transient would you experience surge. But in steady state operation no, i.e. in near steady state conditions or slow transients the pressure ratio falls sufficiently to prevent compressor surge. The change in ambient temperature and pressure changes do not change sufficiently rapidly to cause the problem. Also, in a closed cycle GT we reduce the system/compressor suction pressure (effectively throttling it) to change the load rather than decrease the specific work with no surge problems.

 

[gtsim]

toothless,

If you wish, go can go to this website

http://www.gpal.co.uk/tryit.htm

and download the simulators and subject them ambient temperature and pressure transients and see how the operating point on the compressor map changes in the steady state. Also you can subject these simulators to performance deteriorations including the ones you have mentioned and see the effect on engine performance.