41 replies to this topic

[narendrasony]

I identified a mistake in my previous analysis that "for temperature reduction and at fixed pressure ratio power may increase or reduce depending on the operating point on compressor curve".
Based on the following affinity laws, I would like to correct this statement : " For fixed speed compressor, with suction temperature reduction and at fixed pressure ratio power will always increase and it is nearly proportional to the actual volumetric flow rate (ACFM) ". I apologize for that.

Subscript i, f and f ' are for parameters at
1) Initial suction temperature Ti
2) final suction temperature Tf - As per compressor curve and
3) final suction temperature - assuming constant mass flow


H_{f ’} = H_f = Polytropic head at final temperature T_f = (T_f/T_i) X H_i

Q_{f ’} = Actual volumetric Flow rate at T_f (ACFM) - At constant mass flow = (T_f/T_i ) X Q_{i < Qi}

Q_f = Actual volumetric Flow rate at T_f (ACFM) - As per the compressor curve > Q_i

d_{f ’} = d_f = Gas density at final temperature T_f = (T_i/T_f) X d_i



W_f / W_i = (Q_f X H_f X d_f) / (Q_i X H_i X d_i)
= Q_f/Q_i X H_f/H_i X d_f/d_i = Q_f/Q_i X (T_f/T_i) X (T_i/T_f)
= Q_f/Q_i > 1

W_{f ’} / W_i = (Q_{f ’} X H_{f ’} X d_{f ’}) / (Q_i X H_i X d_i )
= Q_{f ’}/Q_i X H_{f ’}/H_i X d_{f ’}/d_i = Q_{f ’}/Q_i X (T_f/T_i) X (T_i/T_f)
= Q_{f ’}/Q_i = T_f / T_i <1

<sub>(Variations in compressibility factor, specific heat are ignored)

As seen here, Wf / Wi = Qf/Qi > 1 for actual Qf on compressor curve since at reduced polytropic head, actual volumetric flow Qf increase. Mass flow increases even more (as pointed by Sheiko earlier). This is as per compressor curve.
Wf' / Wi = Qf'/Qi = Tf/Ti < 1 for the assumed case with constant mass flow. This is not possible for fixed speed compressor since operating point has to follow compressor curve.

So, power will increase in all the cases with suction temperature reduction. Apologies again for the mistake in previous posts. More insight from learned forum members is awaited.

Regards
Narendra</sub>

Edited by narendrasony, 13 May 2012 - 12:20 PM.

[ankur2061]

Narendra,

I disagree with you. Use the example posted by OP and do the calculations using a process simulator such as HYSYS. The inlet volume flow decreases with the decrease in suction temperaruture due to increase in density at a lower suction temperature. Also the poytropic head is higher with the higher suction temperature and lower with the lower suction temperature.

As mentioned earlier since absorbed power is mass flow rate times the polytropic head, the absorbed power is lower with the lower suction temperature.

Regards,
Ankur.

[ankur2061]

All,

The attached figure shows constant-speed performance curves indicating the effect of suction throttling.

For a capacity requirement of 80% it can be seen that the horsepower reduces to 86% when using suction throttling as a method to reduce capacity to 80%

It is important to note that a variable speed drive will give a higher horsepower reduction compared to suction throttling viz. 81% compared to 86% for the suction throttling method for capacity reduction.

This establishes that a decrease in inlet volume flow due to reduction in suction temeprature will cause the horsepower to decrease and not to increase.

Regards,
Ankur.

Attached Files

•  Centrifugal_Compressor_Constant_Speed_Curves.JPG   238.18KB   57 downloads

Edited by ankur2061, 14 May 2012 - 12:49 AM.

[narendrasony]

Ankur,
Thanks for the attachment.
As far as the example by OP is concerned, he has got it answered precisely by your calculations and it can be achieved by turbine driven compressor.
Our discussion is now centered around fixed rpm compressor only.

So far, I’ve come across conventional ‘polytropic head vs. volumetric flow (ACFM at suction)’ compressor curves only. I’m uncomfortable with pressure rise (Pd-Ps) vs. flow rate curves since they are of limited use in the sense that they assume a constant density , please correct me if I’m wrong. Can you please help me understand them better?
Product of abscissa (% flow rate) and ordinate (% pressure rise) in the top curve of your attachment is exactly equal to % power. That is not the case with conventional head vs. ACFM curves, power is calculated as the product of 1) head (ordinate) , 2) ACFM (abscissa) and 3) the third term gas density at suction.

I’ve used Suction throttle for compressor capacity control. This is scenario-2 in Post-21. Here mass flow is constant and volumetric flow (ACFM at suction) is reduced – we all agree, now as per conventional compressor curve polytropic head will increase, typically suction pressure will reduce with discharge pressure fixed – you will agree. And then things are straightforward, power will increase in proportion to polytrpoic head as per the equation for power.

Thanks again for valuable discussion

Regards
Narendra

[ht73]

Dear All,

I fully agree with Mr. Ankur theory that by reducing inlet suction temp, the power consumption of compressor reduces. By reducing temp, actually you are removing out saturated moisture from the gas, whereby total mass flow reduces to compressor suction. So, by doing so, either of following two objectives can be achieved:
1. Keep constant mass flow ( after moisture removal), Overall power consumption reduces.
OR
2. increase compressor throughput keeping same compressor power consumption.

I have personally experienced this benefits by implementing in our projects covering syngas, Co2, air application.


HT

[kjubo]

Dear all,

i would like to only say, that HYSYS simulation in my opinion CAN NOT be used as reference in power consumption change in this situation.
As far as i know, HYSYS computes required power consumption from differences in enthaply and dont takes into acount compressor geometry. Because if you change only suction temperature, for HYSYS mass flowrate remains constant, but in real life compressor will vary, so in my opinion much more rigorous computation is needed than just HYSYS simulation.

Regards

JK

[jitendraprocess]

Dear Ankur Sir,

As inlet temp decreases power required for compressor decreases. For that purpose we keep intercooler between two stage.
Am i right?

pls. reply.

Thanks & regards.

[narendrasony]

Dear Jitendraprocess,
Power will indeed increase with increase in temperature if you can in reality achieve constant mass flow at same suction and discharge pressure. If you are dealing with same fixed RPM compressor, this is not possible and the power actually reduces.

To make it more clear I’ve attached the performance curves of two fixed RPM centrifugal compressors “A” and “B”. Originally compressor “A” is operating at a lower suction temperature say TL (Point -1, Curve "A") and compressor “B” is operating at a higher suction temperature say TH (Point-3, Curve "B")

Now, Power in any compressor is proportional to : Head X mass flow
Mass flow = Suction volumetric flow (Q ) X Density at suction temperature
Head isproportional to suction temperature (assuming discharge and suction pressure are same)

Compressor “B” (operating at TH, Point-3) will take more power than compressor “A” (operating at TL, Point-1) at constant mass flow, it will be simply proportional to head or suction temperature. But mind you, they are different compressors at their fixed RPMs.

Now suppose compressor “A” has to operate at higher temperature TH (Point-2). It may develop the required higher head for new suction temperature (if it doesn’t surge), so it appears that power should increase again due to higher head. But the same compressor “A” can either develop higher head or higher suction volumetric flow but not both. While increasing the head, compressor suction volumetric flow (Q, ACFM) reduces. Mass flow will reduce even further. What about power ? See now for the same compressor “A”,

Power is proportional to : Head X Suction volumetric flow X Density
Head isproportional to suction temperature
Density is inversely proportional to suction temperature

Effect of head and density is nullified here, so power will ultimately reduce in proportion to the suction volumetric flow for a given machine at fixed RPM.

Suction boost : There is another possibility, suction boost (opposite of suction throttle as mentioned by Ankur) or discharge pressure lowering. Here with the same compressor “A”, we are trying to achieve the same mass flow at higher suction temperature , but it can be done either by increasing suction pressure or reducing discharge pressure if process conditions permit. In this case, head is reduced and power is reduced just proportionately since mass flow is constant. (Operatin point 4 in attached sheet)

Variable RPM compressor: One can achieve the given mass flow at given suction and discharge pressures by increasing the RPM. Power will increase in this case since head has increased at constant mass flow. (Operating point 5 for another variable RPM compressor "C").


To summarize, at higher suction temperature:

• For fixed RPM machines operating at same suction and discharge pressure

• Power will increase in proportion to the head or temperature for getting constant mass flow. This is what simulation or calculations will indicate. But it will be altogether different machines. You cannot get constant mass flow with same compressor (while maintaining same suction & discharge pressures)
• For same compressor, power will reduce in proportion to suction volumetric flow.

• Constant mass flow can be achieved with same (fixed RPM) compressor, but by suction pressure boost or discharge pressure reduction. This will reduce the head and Power will reduce in proportion to the head.
• With variable RPM machine, you can achieve both: required mass flow and required suction & discharge pressures. Power will increase in proportion to head or temperature.

So from 1.b and 2 above, power will reduce for the same compressorwith fixed RPM when suction temperature increases. Constant mass flow , suction & discharge pressure - all cannot be achieved at same time with same fixed RPM compressor. They can be achieved with variable RPM machine or another bigger fixed RPM machine, in both cases power will increase.

I hope I'm able to make it clear. Comments are awaited from learned forum mebers.

 compress_system.pdf   174.74KB   219 downloads

Regards
Narendra

Edited by Chris Haslego, 19 November 2012 - 01:33 PM.
Fixed corrupted attachment

[Lenard Lutino]

<sup>If I may add to the discussion, for academic sake, here are few items that I would like to add which might solve the confusion in Ankurs analogy.
In Ankur's basis he lowered the inlet temperature but maintained inlet pressure and mass flow using HYSYS. This should not be done. Lowering the inlet temperature will affect either the density or pressure of the inlet gas. Since we are modeling it in a closed system, where the gas flows through a rigid pipe section then we can say that the volume is constant and by law of conservation of mass the density must be constant. This implies that the reduction in inlet temperature should result to lowering of inlet pressure instead. This can easily be modelled in HYSYS, use adjust tool to adjust pressure with constant density across the cooler as reference.

@jitendraprocess
intercooler in between stages are to lower the temperature of gas to acceptable value before it is sent to next stage. Imagine if you dont cool the gas inbetween the stages, you can have very high temperature which poses problems with safety and material selection.</sup>

Edited by Lenard Lutino, 10 September 2012 - 03:28 AM.

[Shuraim]

Dear All I would like to calculate the centrifugal compressor power for Natural Gas which is consist of:

 

95% Methane

 

2.5% Ethane

 

1.6% Nitrogen

 

with the following parameters:

 

Suction Temperature :40 C

 

Suction Pressure:18 BAR

 

DischargeTemperature:137 C

 

Discharge  Pressure:42 BAR

 

Flow rate: 342 MNM/hr

 

I need to know  which else parameters I need to have, in order to do the calculation?

[S.AHMAD]

You need the polytropic and mechanical efficiency

[ankur2061]

Ahmad,

 

Polytropic efficiency need not necessarily be an input for centrifugal compressor head and power calculations. A fair approximation of polytropic efficiency can be made from the inlet volume flow also commonly abbreviated as ACFM or ACMH. Refer the following link:

 

http://www.cheresour...et-volume-flow/

 

Mechanical losses are required to be added to the gas power to arrive at the shaft power required for the compressor. Certain guidelines are available for mechanical losses as a percentage of gas power requirement which can be applied for arriving at the shaft power for compressors. They are as follows:

 

Gas Power Requirement, kW                Mechanical Losses,%

0-2500                                                  3

2500-5000                                            2.5

5000-7500                                            2

7500+                                                   1.5

 

These guidelines are available in the book "Pipeline Rules of Thumb Handbook" by E.W. McAllister

 

Regards,

Ankur

[sivaprasadbcpl]

Dear Friends,

 

By decreasing the suction temperature the power required by the compressor will be less without varying all other parameters. when you decrease the temperature the density of the fluid goes up and for maintaining the same mass flow rate you will be requiring less volume.

 

The implication will be your compressor suction Inlet Guide vane opening will be reduced to maintaining less flow rate keeping mass flow rate constant. hence your power will be less.

[dhaliwals]

Ankur,

I need your help!

 

As you are aware our ambient temperatures are much higher than the ISO conditions for rating the performance of air comprssors.

 

I am invoilve in low energy cooling for many years. but have no exposure to air compressors.I have designed a "Hybrid Air Density Correction Unit"  for Air compressor inlet cooling.

This is how it works:

Keeping in view high opex panelty of refrigerative cooling, the inlet air is cooled by an indirect evaporative cooling unit to remove nearly 85% of the sensible heat and  refrigerative cooling is used to reach 15 deg saturated air temperature on continous basis. (nearly ISO conditions)

 

It is commonly understood that reducing the inlet temperature ,reduces the input power for a compressor.We took a simple thumb rule of Lowering the temperature by 3 deg C saves 1% energy and calculated the annual energy saving based on 8760 hours of operation for Pune city.We did not include the effect of lowered moisture load.

 

Our calculation based on hourly ambient data showed saving  of only 3.3% on annulaised basis. Are we going the right way.Can you please suggest an appropriate concept/metodology to accurately asses the savings for reci/centri/screw  compressors?

For instance,one of the "lubicated screw compressor"  manufactuer claimes that increase in density make no difference to energy used.Is that right?

Edited by dhaliwals, 03 May 2014 - 05:21 AM.

[ankur2061]

Dhaliwals,

 

My assertion about compressor power reduction due to inlet gas temperature reduction is based on the theoretical equations of compressor power. I have never been involved in efficiency and power savings auditing of operating machines and hence regretfully cannot provide you any concrete guidance on the methodology for estimation power savings.

 

Based on the little understanding I have on these matters, a test bench would be required to be rigged up under a well defined set of constant external conditions and the machine run with varying inlet temperatures at a constant mass flow rate for defined periods of time to conclude about any power savings.

 

Regards,

Ankur.

[planck12]

Dear Ankur, 

I tried this simulation again in HYSYS and the values I get are different from your so I am not sure why this is the case: 

 

Power for first stage is get is: 16.6 MW (this is coming off the energy stream to the compressor) 

Polytropic Head is 215.1 KJ/kg (72100 lbf-ft/lbm) 

Polytropic Efficiency is 77.3

 

I basically used Peng-Robinson EOS and all the conditions as described in this dicussion. So I am not sure why my poly head is different from yours. 

 

Would appreciate your help. 

 

Thank You 

Ums 

[gunnarhole]

I believe the confusion here is the difference between rating and re-rating.  If you are looking at a new application and trying to assess the effect of lowering the inlet temperature on the required HP then the basic textbook formulas presented here by Mr. Ankur and others will properly show that lowering the inlet temperature will reduce the power requirement. 

 

However the problem as presented here is more properly considered as a rerate question.  In the case of lowering the suction temperature  to an existing machine, running at a constant speed, Mr. Montemyor and Mr. Sheiko are correct, the inlet density, massflow and required power will increase. However since we have a turbine driven unit we have an option.  A centrifugal compressor driven by a variable speed drive will be able to slow down a bit and run it a little further back on its curve to hit the desired mass flow rate and discharge pressure. In these cases it is likely that a lower power requirement will result.

 

ASME PTC 10 is where Mr Anup should look if he wants to study the basics of re-rating centrifugal compressors.  His is an interesting question and I encourage him to study this document closely.  It will lead him on to a better understanding of his work.  However if he does not have the luxury of time I suggest he go back to the Compressor OEM and ask them to investigate the rerate he is proposing for his machine.  Using the software tools they have in-house they should be able to quickly answer his question.

 

Regards  Gunnar Hole