11 replies to this topic

[imanshu]

I have to re-rate Centrifugal Compressor for the revised operating condition (Sction Pressure & Temperature....which are lower than the current op.condition) and revised gas composition.

I have performance curves of the original compressors...These are Polytropic head (ft-lbf/lb) Vs ACFM. I have read in books that Polytropic head Vs Vol Flow curves are independent of gas composition and inlet condition. However, I am confused as the polytropic head is expressed on a unit mass basis which means it doesvary with change in MW and temperature or pressure as density will alter.

Is there a calculation methodology to estimate revised head based on the exiting curves OR I have to approach existing equipment manufacturer to obtain the revised curves ?

Please share your experience.

[Zauberberg]

The polytropic head is independent from the gas molecular weight - it depends only on the suction volume flow (for a given speed of centrifugal machine). But, the discharge pressure will strongely depend on the gas composition.

Let's assume that you have to compress lighter gas, and at a lower suction pressure, up to the same discharge pressure. Reduced molecular weight of the gas, coupled with lower suction pressure, will call for more polytropic head - meaning that the compressor capacity is robbed. If the gas becomes too light or if the suction pressure is too low, you could end up in the situation where the compressor will not be able to develop required differential pressure at all.

Here is one interesting article you can go through, and also look for chapters from Norm Lieberman's or Heinz Bloch's books.

http://www.revamps.c...cuments/166.pdf

[Zauberberg]

And yes, for any serious re-rating case (= significant deviation from design conditions) I would definitely contact the original equipment manufacturer.

[ankur2061]

I have to re-rate Centrifugal Compressor for the revised operating condition (Sction Pressure & Temperature....which are lower than the current op.condition) and revised gas composition.

I have performance curves of the original compressors...These are Polytropic head (ft-lbf/lb) Vs ACFM. I have read in books that Polytropic head Vs Vol Flow curves are independent of gas composition and inlet condition. However, I am confused as the polytropic head is expressed on a unit mass basis which means it doesvary with change in MW and temperature or pressure as density will alter.

Is there a calculation methodology to estimate revised head based on the exiting curves OR I have to approach existing equipment manufacturer to obtain the revised curves ?

Please share your experience.

imanshu,

I don't know where you have read that the polytropic head is independent of the MW of the gas. This is totally incorrect & in fact the MW has a very large influence on the polytropic head & change in MW by a very small number results in an immediate change in the polytropic head. I am quoting from a very reputed book on compressor:

To determine the number of stages (using the impeller and diffuser defined as the stage), assume 10,000 ft-lb/lb of head per stage. This value can be used if the molecular weight is in the range of 28 to 30. For other molecular weights, this initial value must be modified. As a rule of thumb, lower the head per stage by 100 ft-lb/lb for each unit increase in molecular weight. Conversely, raise the allowable head per stage 200 ft-lb/lb for a unit decrease in molecular weight. The rule of thumb gives the best results for a molecular weight range of 2 through 70. Because this sizing procedure is being used only to establish the rough size of the compressor, the upper range may be extended with some loss in accuracy.

The source of the above quote is from: Compresors: Selection & Sizing, 3rd Edition by Royce N. Brown

Why don't you send me your new set of conditions (suction P & T, discharge P, gas composition, mass flow) & I will send you the calculated new polytropic head, new shaft power, discharge temperature, rotational speed & no. of stages per section.

Regards,
Ankur.

[Zauberberg]

Quoting from N. Lieberman's Working guide to process equipment: "We cannot change the feet of head developed by the compressor by altering the physical properties of the gas compressed."

The following chapter gives a good example.

Attached Files

•  Centrifugal Compressors.pdf   254.26KB   402 downloads

[ankur2061]

Quoting from N. Lieberman's Working guide to process equipment: "We cannot change the feet of head developed by the compressor by altering the physical properties of the gas compressed."

The following chapter gives a good example.

Zauberberg,

Thanks for the attachment.

The differential head (delta P) (discharge pressure - suction pressure) in feet does not change with the change in gas characteristics is what has been stated by Lieberman - and I stand in agreement with it. It is a function of both vapor density and polytropic head. Liebeman's statement essentially means that if the vapor density decreases, the polytropic head increases & vice-versa; i.e., if the vapor density increases the polytropic head decreases - thus keeping the delta P as constant.

Differential head = vapor density*polytropic head

This is exactly what I have said in my post: that the molecular weight (vapor density) strongly effects the polytropic head.

I hope you are not confusing polytropic head with the differential head - as obviously both are not the same.

Regards,
Ankur.

[Zauberberg]

Thanks for getting back on the subject, Ankur. This is really an interesting topic and perhaps we should start a separate discussion on the polytropic head issue.

I see centrifugal compressors in the same way I see centrifugal pumps - they are dynamic machines, meaning that they develop certain head totally independent of the fluid that is being compressed. What I know from reality is that, if you want to compress lighter gas from the same suction pressure to the same discharge pressure, that would either result in flaring or increase in the suction pressure (for a fixed-speed device), or you have to run it at a higher speed (= higher polytropic head for the same suction volume flow) in order to push the gas up to the fixed discharge pressure. And I see it as simple as that - not so much interested in theory and equations.

I'm not saying that your statement is not correct - my intention was to give a practical meaning of the effect of gas MW on compressor operation.

[imanshu]

Based on the discussion, I understand that existing Polytropic Curves can be used for predicting the revised performance of the machine. One thing that I forgot to mention was the driver is Gas turbine hence speed variation can be done to increase or decrease the comp speed to acheve required pressure and flow demand. As regards to power, will key in the same performance curve (i.e. Polytropic head Vs ACFM & Polytropic efficiency Vs ACFM) in HYSYS to confirm driver power suitability. Hope with this approach I will be able to predict the performance with a reasonable level of accuracy.

I assume Polytropic head (workdone/unit mass) unit of ft-lbf/lb or N-m/kg is the same as ft or m of head with required correction for g value.

[ankur2061]

imanshu,

Many filed/operations engineers (even exeperiened ones) confuse between differential head and polytropic head when dealing with centrifugal compressors by considering them to be the same. Polytropic head is nothing but the change in enthalpy of the gas from state 1 (suction) to state 2 (discharge) whereas the differential head is the change in pressure of the gas from state 1 (suction) to state 2 (discharge). Gas characteristics do not effect the differential head if the machine configuration remains the same but the change in MW of the gas does have an impact on the polytropic head.

As regards to your question:

delta P (N/m2 or Pascal) = vap density (kg/m3)*polytropic head (N-m/kg)
delta P (lbf/ft2) = vap density (lb/ft3)*polytropi head (ft-lbf/lb)

Regards,
Ankur.

Edited by ankur2061, 16 April 2010 - 03:39 AM.

[Zauberberg]

Based on thediscussion I understand that existing Polytropic Curves can be used for predicting the revised performance of the machine. One thing that I forgot to mention was the driver is Gas turbine hence speed variation can be done to increase or decrease the comp speed to acheve required pressure and flow demand. As regards to power, will key in the same performance curve (i.e. Polytropic head Vs ACFM & Polytropic efficiency Vs ACFM) in HYSYS to confirm driver power suitability. Hope with this approac I will be able to predict the performance with reasonable level of accuracy.

You could try to plug the new figures into Hysys (I assume you have a model developed based on original performance/curves) and see what's the new operating window will be (i.e. the maximum capacity of the compressor @ lower suction pressure and the lowest gas MW). That should be good enough for a conceptual approach. Still, for changing the conditions in the real plant I would definitely contact the vendor and confirm the basis of my own calculations/modelling.

Best regards,

[imanshu]

Based on thediscussion I understand that existing Polytropic Curves can be used for predicting the revised performance of the machine. One thing that I forgot to mention was the driver is Gas turbine hence speed variation can be done to increase or decrease the comp speed to acheve required pressure and flow demand. As regards to power, will key in the same performance curve (i.e. Polytropic head Vs ACFM & Polytropic efficiency Vs ACFM) in HYSYS to confirm driver power suitability. Hope with this approac I will be able to predict the performance with reasonable level of accuracy.

You could try to plug the new figures into Hysys (I assume you have a model developed based on original performance/curves) and see what's the new operating window will be (i.e. the maximum capacity of the compressor @ lower suction pressure and the lowest gas MW). That should be good enough for a conceptual approach. Still, for changing the conditions in the real plant I would definitely contact the vendor and confirm the basis of my own calculations/modelling.

Best regards,

Many Thanks for your valued feedback.

Regards,

Himanshu

[fallah]

The differential head (delta P) (discharge pressure - suction pressure)

Would you please explain:

Why you apply "Differential Head" for "Differential Pressure"?