Power Consumption of a Compressor

[amiruw]

Dear Friends,

There are comprehensive mathematical relations in the literature for calculating the power consumption of compressors (in terms of [kW]). But just as an approximate solution (without considering the parameters such as specific heat ratio, compressibility of gas, and ...), which one of the following is correct?

Power consumption of a compressor is proportional to:

1. (flow rate).(discharge pressure-suction pressure)

2. (flow rate).{ln(discharge presure/suction pressure)}

(ln: natural logarithm)

Any comments or sharing any reference in this regard is highly appreciated. Thanks in advance.

 

[UmeshMathur]

There was a recent very spirited but informative exchange, which I recommend to anyone interested in this topic, in the following ENG-TIPS thread: thread391-151370.

In answer to your specific question:

(1) Power consumption is proportional to mass flow rate multiplied by compressor head.

(2) Head is proportional to [(P2/P1)^{(n-1)/n} - 1], where n is the polytropic exponent for the gas, P1 is the suction pressure, and P2 is the discharge pressure (both pressures in absolute, not gauge, units).

Therefore, neither of the relations you suggest is valid.

 

[dcasto]

Here is a simple formula for you HP = 22 * Ratio,
Ratio = Pd/Ps (in absolute pressures) for:
2 < Ratio < 4:
If Ratio > 2, HP = Square root (Ratio) * 22 * 2 or
HP = Ratio^.333 * 22 * 3, ect

 

[UmeshMathur]

dcasto:

I'm not sure that the formulas you have provided are not approximations. I ran a few numerical examples and found that they do not match the more rigorous formula for compressors within what I consider engineering accuracy (1-2%).

 

[dcasto]

UmeshMathur, you are correct. I use the above formulas in talking. Then I run similations to get actual physical conditions with my own estimated effiencies, then I run the manfactures programs to get thier effiencies, which I then rerun the similations with the efficencies from the manufacturer. In my opinion, only the manufacturer can give you a 1% to 2% value, but most post a disclaimer that they are +/- 3%.