3 replies to this topic

[sheiko]

Good evening,

As you may know:

- Polytropic compression process is generally considered for centrifugal (dynamic) compressors.

- For Positive Displacement (PD) compressors (Reciprocating/Rotary) the compression process is generally considered as adiabatic for most engineering calculations.

I am looking to understand why? Is there any theory behind these statements or are they only based on experience?

Edited by sheiko, 08 January 2011 - 04:33 PM.

[breizh]

Hi Sheiko,

You may find your answer reading this ebook :

http://books.google....epage&q&f=false

Breizh

[Art Montemayor]

Sheiko:

I mention this not only for your interest, but because many young engineers will be reading this thread and this subject is analogous to another subject that continues to come up frequently: the heating and cooling jackets used on reactors or process vessels. I will explain this further down.

Yes, ideally a reciprocating compressor is described as following a thermodynamic adiabatic and reversible process – which makes it an ISENTROPIC process. This, of course means that the entropy remains constant through the process and is a handy means to calculate the brake work required for the compression.

The really ideal compression process would be one where the compression step could be carried out in an isothermal fashion. But this is a thermodynamic daydream because this is not a feasible or practical application in real life. There simply is no manner - nor mechanical method - that can be devised (to date) that allows for continuous, differential cooling of the gas as it is being compressed. If you can invent or come up with a feasible method to do this, you will become a super billionaire by revolutionizing the application of thermodynamics. The ONLY method that has been applied is a vain and “cosmetic” attempt to use a built-in cooling jacket on reciprocating cylinders. The compression cooling obtained in this manner is only a “best effort” type – you get what is available due to the mechanical size and constraints. I have operated reciprocating compressors since 50 years ago – some were cooled, others were not. For example I operated reciprocating compressors handling air, CO₂, Oxygen, Nitrogen, Argon, Acetylene, Nitrous Oxide, Hydrogen, natural gas, etc., etc. and most of these were jacket-cooled with cooling water. The acetylene application is an exceptional one because it is a special case where the cylinder and intercoolers are all immersed in a cooling water box. Acetylene is VERY special and has its own rules to follow. I read, in Hydrocarbon Processing, an article in the 1960s that opposed the use of cooling jackets – for various practical reasons. I tried operating my compressors without cooling water in the jackets and found – to my surprise – that the article was correct: there was NO NEGATIVE EFFECT on the compressor’s performance and operation when the jackets were run dry. It was then that I realized that I had been naïve in thinking that such a small cooling area could have a marked effect on a gaseous chamber that involved an inherently poor gas film heat transfer coefficient. You will recall that the worse film coefficients are those of gases. In fact, a static gas film is used as a natural insulator. Nature does this in giving us body hair – much as animals, such as Polar bears, are able to be insulated with only a fur coat. Our engineering insulation materials all depend on the effect of a static gas film.

In summary, it is my experience that a cooling jacket on a reciprocating compressor cylinder cannot be relied upon to lend any marked, positive effect on the efficiency or operability of the compressor. It exists only as a token attempt to lend some heat removal – but it is so inefficient as to be practically negligible. The same effect takes place in a reactor or process vessel when one considers a cooling jacket and tries to “optimize” the cooling/heating effect with external coils, dimples, etc., etc. The net effect is negligible because you are limited to the vessel size. You get what you get – and no more. A process reactor or vessel is designed for the reaction and the process capacity – not for the heat transfer. The heat transfer is an after-thought and is a best effort process.

Further, and direct to your query is the fact that since the actual, field compression process does not transfer any significant compression heat, it should then be considered adiabatic. And here, my design and process experience in the field has also confirmed the advice of many authors – specifically the NGPSA – in recommending the use of an isentropic compression with enthalpy data when calculating the thermodynamic results of a gas compression in a reciprocating machine. My experience has been that when I assume an isentropic process and use accurate enthalpy data – especially on pure gases – I get excellent design results on the work requirement and the discharge temperatures of each stage in a reciprocating machine. All this confirms for me that the performance of a reciprocating compressor is very close to the presumed isentropic process. I have never employed polytropic calculations with a reciprocating compressor simply because I had good results with the simpler and more direct isentropic assumptions. The centrifugal compressor is another animal altogether different and more complex in its calculations – something a practical engineer would expect taken that this machine is so mechanically simple. The expected trade off in a more complex process calculation given a simpler mechanical design is experienced.

A centrifugal machine is vastly more inefficient and limited in control than the reciprocating model. This is yet another trade off. Reciprocating machines are not only more thermodynamic efficient, but their process calculations are much simpler and the machine has much more positive capacity control than the centrifugal. Where the reciprocating machine cannot compete is in size, capacity, and price.

Note that when you increase the number of stages (and associated intercoolers) in a given reciprocating compressor, you are practically applying one version of isothermal compression. You could, theoretically, achieve isothermal compression if you had infinite stages and intercoolers. This is the positive effect a user profits from when multiple stages are employed in a given compression operation. The multiple stages result in a lower total horsepower compression requirement. Of course, more stages mean more capital monies; but an optimization exists.

I hope this experience helps to settle your query on the thermodynamic behavior of a reciprocating compressor.

[Erwin APRIANDI]

Good topic, especially when Art comes into pictures

Thanks Art for sharing the experience