Propane Compressor Efficiency

[L Adams]

Hello, I have a question about calculating the efficiency of a three stage propane compressor. The compressor has three inputs of propane with different temperatures and pressures, coming from 1st,2nd and 3rd stage knock out drums, and I know the flow, pressure and temperature of the input streams and I know the same for the discharge out of the compressor. First of all is it normal that a multi stage compressor does not have intercoolers? Second of all how would I go about calculating the efficiency? Lastly, I'm was not able to successfully use Hysys to simulate this compressor. I would like to verify my calculations using Hysys. I would really appreciate the support in this regard.

 

[pmover]

is your schematic similar to the attachment?

for each stage Td, the Td was guessed based on OEM curves and using actual rpm and inlet flow rate data.

good luck.

 

[PaoloPemi]

I have a different software (Prode Properties) and I solve these problems in Excel by iteration to find compressor's efficiency (from OEM curves),
not too different from procedure described by pmover,
in this case Excel offers some additional flexibility (while the library solves the polytropic stages) but you should be able to solve with many simulators.

 

[georgeverghese]

Intercoolers are only required when the gas temp rises beyond 150degC for oil/gas/petrochem applications. And for each stage, a max developed polytropic head of about 10000 to 12000 ft.lbf/lb is the limit. In most simulators, the simulator would backcalculate the compressor eff with all the info you've stated.

 

[L Adams]

Thank you pmover, it is very similar. the green value in the attachment are given and the blue is calculated/estimated?

 

[L Adams]

thank you PaoloPemi, what is the efficiency equation you are utilizing?

 

[L Adams]

Thank you georgeverghese, I've used Aspen HYSYS, however, the eff is calculated to be more than one. now in this simulation I've used a multistage compressor and I've stated there is 3 stages. I believe my Aspen Hysys setup is not right. or there is something causing the eff to exceed 100%.

 

[PaoloPemi]

for design case, if you have a table / graph of polytropic efficiency versus volume / mass flow rates enter those values,
or define some default values (for example 0.75-0.8) in a table,
Excel can estimate polytropic efficiency at operating point (interpolation within existing data set),
then you can solve with a library (as Prode) or a simulator

differently, if you wish to calculate polytropic efficiency, preferably solve separately the different stages, that is possible if you know intermediate T, P conditions,
or, for solving multiple variables, you may start with some default values (for example 0.8) and run the minimizer in Excel to find the correct values,
I have a Excel page with a few lines of VBA code for that, simulators normally include predefined procedures,
in case of erratic behavior you can force the direction of search in Excel assuming some relation between flow and efficiency in the three stages,
this converts the multivariable initial case into a single variable problem, easy to solve,
finally, if you get uncommon values (for example efficiency < 0.6 or > 0.95) check your input values, maybe some wrong number or unit,

Paolo

 

[georgeverghese]

Check that you have single phase gas only going to each stage, else install a phase separator before the compressor and drop out all the liquid.

 

[pmover]

L_Adams,

correct - method/technique i use to readily distinguish input/calc data.

um, after looking at this image, i see improvements that can be made. too bad i do not have the P&IDs.

 

[georgeverghese]

You may most likely require intercoolers for all 3 stages to allow for full recycle flow with no fresh feed, given the 140-150degC limit on Td for each stage - this will occur during startup and when the plant is idling hot. This full recycle case will also influence the design of the compressor.

 

[rotw]

I think the procedure mentioned by pmover is the approach to follow.

If you would like a gain in precision (not sure if it is justified or not, a word on this later) you could determine the mixing temperature with enhanced procedures depending on your working assumption:

1/- if gas properties are assumed constant:
T_after_mix = (G_sidestream x T_sidestream + G_incoming x T_incoming) / (G_after_mix)

2/- if gas properties are assumed to change linearly:
T_after_mix = (G_sidestream x Cp_sidestream x T_sidestream + G_incoming x Cp_incoming x T_incoming) / (G_sidestream x Cp_sidestream + G_incoming x Cp_incoming)

3/- if gas properties change significantly, I suggest to do enthalpy balance and proceed via EOS to calculate mixing temperature.

But like I said I do not know if there enhancements have justified costs. I am confident the simplified approach per 1/ is proven and could work effectively in this propane case refrigeration machine.

So consider the rest of the methods for a useful future reference.

If you plan an escape, you must succeed as if you fail, you will be punished for trying. Never say or write down your plan. Heart is the only place where secrecy is granted.

 

[AnhsirkT]

1.IN this propane compressor no need of inter cooler because there is facility of quenching if temp goes hign then normal
and second flow coming from 2 KO drum will cool the first stage discharge gas same for next stage.
2. Effeciency of every stage can calculate by

EFF=== (Log(P2/P1)power(k-1/k))/Log(T2/T1)