Reciprocating & Centrifugal Compressors in parallel

[stk123]

Have anyone come across a Reciprocating & a centrifugal compressor operating in parallel ?
how will be the behaviour of such a system from discharge pressure point of view ?

with regards

 

[Montemayor]

This is a query posed as if it is a unique situation and a challenge. It is neither.

I have seen and been around compressed air systems where the initial recip-based installation was augmented by a centrifugal machine for the purpose of adding quick, up-dated equipment. This is not usual because most installations would rather keep a common type of machine in order to keep common spares. However, a main gas header can be filled/flowed with gas coming from any type of machine or source. The header doesn't care where the gas originates. However, the CONTROL of the sources is what may be of concern.

If you really are thinking of doing this, I'll spend some time explaining what I stated:

1. You have nothing to worry about concerning the operation and control of the recip. This machine is nothing more than a metering device and it will dole out the rated volume it is supposed to deliver with every stroke. The recip will control its capacity regardless what the discharge pressure may be - and this can be done in a variety of ways which are a subject beyond this thread, since you were not specific in your query.

2. The centrifugal will also operate on its own independent control system - usually a return recycle in order to avoid a surge condition. But it will feed gas to the common discharge header at the rated discharge pressure.

You have not stated anything about varying the header capacity flow rate, so all I can state is that I have to assume you mean to supply a steady-state flow rate of gas through the common header (which would be unusual). But even if you had to vary the header flow rate, you can program a PLC to reduce the capacity of your compressors on a step-wise manner. I would prefer to start reducing the capacity of the recips through clearance pockets because this is the most energy-efficient way to reduce capacity. The centrifugal is the most inefficient of both types and it is the most sensitive to capacity variance - that's the major set back and handicap of a centrifugal. Since you've bought into the less-efficient centrifugal by installing it, then you should try to keep it on the line at its best efficient point in capacity and vary the recip's capacity as a preference.

The header pressure should not exhibit any variance if properly instrumented.

 

[tgmcg]

Discharge pressure control should not present any problem.

The centrifugal compressor capacity can be controlled via variable speed or suction throttling @ constant speed....both more desirable from an efficiency standpoint than recycle.

Modern centrifugal compressors are available with polytropic efficiencies greater than 80%, which starts to close the traditional efficiency gap between centrif and recip.

I can support Montemayor's recommendation to use the recip as the swing machine while operating the centrif at best efficiency. Though more costly than using clearance pockets, you might also consider using a stepless unloader control on the recips such as Hydrocom, manufactured by Hoerbiger. Very precise pressure control can be had this way.

 

[dcasto]

I've always run my Reciprocating Compressors first because the overall effiency of the engine and compressor package is higher than the turbine driven centifical.