Recip Compressor VE 2

[HookedOnHalo]

Can anyone explain why recip compressor performance calculations often have a limit on the VE to not be less than 10% or 15%? If the VE is calculated to be less, the operating condition is said to be in error.

 

[Montemayor]

Hooked:

Could you please identify what performance calculations you are referring to and who is saying that if the Volumetric Efficiency (VE) is less than 10 or 15%, the operating condition is said to be in error. Also explain what is the context under which this was said or written.

In a reciprocating compressor the piston does not travel completely to the end of the cylinder at the end of the discharge stroke. If it did, it would destroy itself and you would soon run up a very expensive maintenance cost replacing compressor cylinders and/or pistons, piston rods, crankshafts, and cranks.

Some clearance volume is required by common sense and it includes the space between the end of the piston and the cylinder head when the piston is at the end of its stroke. It also includes the volume in the valve ports, the volume in the suction valve guards and the volume around the discharge valve seats. This “Clearance Volume” is usually expressed as a percent of piston displacement and referred to as percent clearance, or cylinder clearance. From this, the term “Volumetric Efficiency” is derived and it refers to the actual pumping capacity of a cylinder as compared to the theoretical piston displacement and multiplied by 100.

One of the prime, effective ways of controlling the capacity of a reciprocating compressor is to vary the Clearance Volume. This is conventionally done with external clearance pockets. It is quite possible to build and employ clearance pockets that can yield a 100% clearance – i.e., the piston displacement is available as clearance and there is no compression work taking place – merely compression and expansion within the cylinder. Under this condition the compressor has 0.0% VE and zero flowrate. I can assure you that this operating condition is not “in error”. This operating condition is a pre-requisite for starting and stopping a reciprocating compressor to ensure that there is no load on the machine and its driver when it is undergoes startup and shutdown.

Are you sure you are talking about the Volumetric Efficiency (VE)? That’s why I hate the use of acronyms without the proper definitions stated before hand.

 

[Yorkman]

Montemayor,
I value you expierence a lot, but just a question and clarification. It is quite possible to build and employ clearance pockets that can yield a 100% clearance – i.e., the piston displacement is available as clearance and there is no compression work taking place – merely compression and expansion within the cylinder. Under this condition the compressor has 0.0% VE and zero flowrate.
Are you referencing a hydrualic cylinder unloader?
I have seen Copelands Discus compressors in the field but never really looked up clearance volumes on these.
The best I seen published when it comes to clearence volumes is 1-1.5% of the displacement value of the cylinder. The volumetric efficiency of the compressor is going to be a function of two variables and one constant, as I see it. The variables being the specific volumes of the suction vapor and discharge vapor, the constant would be the clearance volume.
I've never really run any numbers on re-expansion of the discharge vapor versus suction vapor but it would be interesting to calculate a best case senairio. That 10-15% might be pretty close if Hooked is talking about re-expansion and not efficiency.
That would almost be the compliment
100% cylinder volume - 10% re-expansion = 90% volumetric efficiency. Probably not very acurate but it might explain what the published reference Hooked mentioned is eluding to.

I'm not a real engineer, but I play one on T.V.
A.J. Gest, York Int./JCI

 

[insult2injury]

For a polytropic process, the volumetric efficiency is defined as:

eff = 1 + C - C*(Pdischarge/Psuction)^(1/n)

where n is the polytropic coefficient and C is the clearance volume fraction defined as:

C = Vdischarge/(Vdischarge-Vsuction)

 

[Yorkman]

insult,

Is p going to be in psia?

Also is V volume cubic inches?

And finally how/where do I find the polytropic coefficeint, does it change for different gases/refrigerants?


I teach a class that explores pressure enthalpy diagrams and the vapor compression cycle, volumetric efficeincy always seems to come up, as it should. The best I could do although not correct was use Boyles Law to describe the action of re-expansion. But I knew there was alot more going on there because of heat being added to the gas ect. Thanks for the formula if you could fill in some of the blanks it would be much appreiciated.

I'm not a real engineer, but I play one on T.V.
A.J. Gest, York Int./JCI

 

[insult2injury]

Since both the pressure and volume terms are ratios....the resulting values are non-dimensional so it doesn't matter what the units are as long as they are consistent.

As far as the polytropic coefficients, they are determined experimentally. For a compression process, 3 simple models are generally employed:

1. Isentropic - implies that no cooling occurs during process
2. Isothermal - implies that maximum cooling occurs, therefore, may be accurate for slow processes
3. Polytropic - implies that some cooling applies, therefore, somewhere in between isentropic and isothermal, real compression procession processes are generally polytropic in nature

All three of these models relate a function of pressure and volume (or specific volume) to a constant. In fact, the first two are just special cases of the polytropic equation. The governing equation is:

Pv^n = constant

This is the equation as it appears for a polytropic process. For an isothermal process, n=1; therefore:

Pv = constant

For isentropic compression, n = k which the ratio of specific heats or Cp/Cv (which is about 1.4 for air), therefore:

Pv^k = constant

The isothermal and isentropic cases result in the two extreme cases and therefore for the polytropic coefficient must be between 1 and k. All of the previously mentioned models descend from the ideal gas equation of state and can be derived readily with only minor manipulation.

Since the models are all equal to a constant, two states (suction and discharge) can be equated as follows:

P_1*v_1^n = P_2*v_2^n

or rewritten,

P_1/P_2 = (v_2/v_1)^n

These equations form the basis for equation to calculate the required compressor work.

http://img60.imageshack.us/my.php?image=voleffks8.jpg

Hope that helps!

 

[Yorkman]

It will certainly give me something to gnaw on as it were, could keep me occupied too. Thank you very much

I'm not a real engineer, but I play one on T.V.
A.J. Gest, York Int./JCI

 

[HookedOnHalo]

My question was referring to recip compressor performance calculations which many of the manufacturer's supply, such as Ariel and DR.

VE is Volumetric Efficiency. VE is defined as the % of the stroke that the suction or discharge valve is open compared to the total swept volume. You can refer to the Ariel Application Manual for more details if you wish. Go to

http://www.arielcorp.com

and download and install their software.

I found same answers to my question. Seems a value of 10% to 15% is typically used as a minimum value to ensure there is enough flow in and out of the cylinder to prevent over-heating of the cylinder. Furthermore, prediction of loads and flows becomes difficult since at low volumetric efficiencies, the effects of pulsations are more dominant.

API 618 specifies a minimum of 15% for performance guarantee points.

One small correction to a comment from Montemayor regarding 100% clearance. It is possible to have 100% clearance or greater than 100% clearance and still get flow in and out of the cylinder. I conducted studies of many compressors with >100% clearance.

Also, unloading or starting a recip compressor of the type typically used for natural gas compression or refinery service is typically done by means of a recycle or bypass that circulates gas from the final discharge back to the suction. This minimizes the compression ratio on the compressor resulting in little load. To load the compressor, the recycle is closed and the discharge valve is openned resulting in increasing pressure on the final discharge.

 

[25362]

When defining volumetric efficiency as the ratio of the actual intake volume to the piston's swept volume, I'd like to clarify the theoretical formula for volumetric efficiency brought by insult2injury, referring in particular to the value of C.

I say theoretical because apparently it doesn't consider other factors such as leakages, and pulsation or "fluttering" of valves as can be seen on indicator card plots.

V_discharge would be the so called starting volumetric clearance at the end of the discharge step just before starting expansion.

(V_suction - V_discharge) would be the piston's swept volume.

Besides, when measuring the volumetric efficiency it would be advisable to quote the P,T reference conditions.

Insult2injury, do you agree ?

 

[insult2injury]

25362,

Yes, that's correct. If you refer to the diagram that was attached for the compression process, the cylinder is discharging from 2-3 and the actual equation on that figure is:

C = V_3/(V_1-V_3)

It's also true that there are several minor losses that are not taken into account in the ideallized equation. If the compressor is well-designed, those losses should be minimized and the equation's efficiency prediction should be very close to the measured volumetric efficiency.

 

[BigInch]

Just because there is little point making a compressor that can fit all flow ranges. If you could adjust the volume of your cylinders from 0 to 200% of some mid range, there would be a lot of wasted iron around if you were running it at 15% of its true capacity. You really should have bought the small machine. Its simply not good economics to build machines or shoes in a one size fits all. A 10-15% variation allows some flexibility, yet doesn't waste too much iron.

Going the Big Inch!

http://virtualpipeline.spaces.msn.com

 

[zdas04]

Let me inject a couple of points that I would like future readers of this thread to make sure that they understand:
1. Above Insult2injury said (in answer to the question "is pressure in psia?") that since it was a ratio the units didn't matter. For temperature and pressure ratios the units may not matter, but you always have to do those ratios in absolute terms. If the question was psia vs. psig then it matters a lot. Say you are going from 1 psig to 50 psig. Using relative terms gives you 50 ratios while this is actually (at sea level) 64.7/15.7 or less than 5 ratios.
2. The original question was about recips. The assumption that best matches actual field performance is isentropic (PV^k=constant). The "k" values are very well known and often published. You need to use experimental polytropic exponents for dynamic machines like centrifugal compressors, but for positive displacement equipment isentropic ia a reasonable assumption.

Now back to the VE discussion

David

 

[insult2injury]

I concur with zdas04. I assumed that absolute units was understood...maybe I shouldn't have. The equation for volumetric efficiency is a general equation for all compressors not just recips which is why I presented it in terms of n instead of only for k specifically.

 

[25362]

Insult2injury, some experts would say that volumetric efficiency is a concept that doesn't apply to all compressors...

 

[insult2injury]

for what compressors does a volumetric efficiency not exist?

 

[Yorkman]

I think it does apply to reciprocating compressors, Not as much to scroll and screw type because of the nature of their design. Although the proportional stop or
(slide stop) on some screws can improve efficeincy.

I'm not a real engineer, but I play one on T.V.
A.J. Gest, York Int./JCI

 

[insult2injury]

I was under the impression that volumetric efficiency was indicative of performance for all compressors with the exception of turbos, but for turbos, i still believe that it's defined just with no significance. Many times, the volumetric and mechanical efficiencies are lumped into 1 parameter but it's still there. At least that was my understanding.

 

[Montemayor]

An explanation was asked by Hooked about Volumetric Efficiency on a Reciprocating compressor and all of a sudden, as zdas04 points out, the subject has reciprocated to other things – and some of which are erroneous – mostly because of the tendency to deal in generalities. The comments and explanations I’ve offered are literally found in such classic texts as the GPSA Engineering Databook, Lyman Scheel’s classic textbook on gas compressors, Ingersoll-Rand’s “Compressed Air and Gas Data”, personal notes and seminar information from Clark Brothers and Cooper-Bessemer. I haven’t quoted the information, but I’ve certainly proven it to be true in the field after many years. For example, the most accurate manner of calculating compressor horsepower is by an enthalpy change and not with analytical equations. This is straight out of the GPSA (& other experienced authors) and it works. And yes, David is 100% correct. The process taking place inside a reciprocating cylinder is so close to the isentropic cycle that it always converts a neophyte into a believer that the cooling jackets do absolutely nil or no cooling. I’ve run cylinder dry just to prove this point and it worked. The process is not polytropic. The only people who think so are those who don’t understand recips and haven’t operated them for years on a variety of gases (such as college profs). I have, with oxygen, nitrogen, CO₂, acetylene, nitrous oxide, air, carbon monoxide, Hydrogen, natural gas, ammonia, most of all the Freons, propylene, propane, ethane, off gas, syn gas, and many more. And in all of these cases the compression was identical to Isentropic – never “polytropic”. I’ve also operated centrifugal compressors (which are the subject at hand here) and these are strictly polytropic in nature – and what’s more, no one in his right mind will try to calculate the “Volumetric Efficiency” on a centrifugal. Whatever for? And just exactly, how would one go about calculating the VE for a centrifugal (which would be of no practical use anyway)? The answer is: no one calculates this useless information for a centrifugal. It is strictly a reciprocating machine characteristic and of practical use for this type of machine.

Every reciprocating compressor I’ve operated had some sort of capacity control. And each one employed some variation of cylinder clearance control. As an experienced compressor operator will tell you, there are several ways to control the capacity of a constant-speed recip using cylinder clearance:

1. Finger unloaders depress the suction valve plates and allow the valves to remain open, allowing the displaced gas to return back to the suction inlet chamber without any compression taking place. This, by simple logic, constitutes the equivalent of 100% clearance.
2. The suction valves can be lifted entirely off their seat and, thereby also create 100 % unloading;
3. The suction valves can be designed and built with “plug unloaders” – a method whereby the suction valve has a center hole that is kept closed by a pneumatic-actuated plunger when the valve is functioning normally; the plunger is raised (allowing gas to return to suction without compression) when 100% unloading is desired;
4. Clearance pockets can be built into a compressor cylinder both externally and internally. These are kept closed similar to the mechanism explained in the plug unloader, above. The advantage of clearance pockets is that they can be designed for various “steps” of clearance unloading and, consequently can yield capacity turndowns between 25-90% of rated cylinder capacity.

Relatively fast refrigeration recips like those manufactured in the USA by York, Dunham-Busch, Vilter, and others normally employ finger unloaders to unload their cylinders. This is so because refrigeration compressors are made by the thousands and are designed to be more “profit effective” and are not expected to yield the same service life as industrial gas compressors. Any machine running with a piston speed of 750 – 1,000 ft/min is, in my opinion, a “fast” recip and one in which wear and tear will develop much earlier than slower machines. Refrigeration machines are designed with an entirely different scope of work and they can never conform to standards such as API 618. As Yorkman infers, refrigerating machines always have a very small clearance volume (high VE). The reason for this is that they are trying to save metal. The bigger and more robust industrial models have lower VEs because of several reasons. One outstanding reason is that some margin is included to safeguard against possible liquid entrainment. Another is that all industrial compressor cylinders have already been designed and the closest model design is always picked that will handle the job. No one in his right mind would expect to have a manufacturer design and cast a “custom-sized” compressor cylinder for his machine – unless he wanted to pay the huge costs of such a project. As an experienced compressor engineer, I wouldn’t try to compare an Ingersoll – Rand (Dresser) compressor with an Ariel machine. Each has been designed and built with a different set of scopes of work. They are both good machines for what they are designed to do and what they are intended to accomplish over a certain time period. But one will almost always result higher-priced when applied to “gas patch” natural gas compression – usually the I-R machine. Valve life and wear are the differences when one looks at the piston speeds employed. A compressor doesn’t have to last 50 years to be a good one. If it can make a 25% return on investment in a 5 year period, it can justifiably be trash-canned and labeled as successful.

Hooked is wrong when he states: “Seems a value of 10% to 15% is typically used as a minimum value to ensure there is enough flow in and out of the cylinder to prevent over-heating of the cylinder. Furthermore, prediction of loads and flows becomes difficult since at low volumetric efficiencies, the effects of pulsations are more dominant.” There is no way that the effect of cylinder clearance affects the expected discharge temperature out of a compression cylinder. This is not my opinion; this is what is factually stated by the equation for determining the discharge temperature:

T_D = T_S (r^k-1/k)

where,

T_D = Discharge temperature, ^oR
T_S = Suction temperature, ^oR
r = cylinder compression ratio, psia/psia
k = ratio of gas specific heats, Cp/Cv

You will note that the clearance nor the Volumetric Efficiency are a factor.

Additionally, there are no pulsations felt in a well-designed compressor that are due to cylinder clearance. I’ve never experienced this and this is the first I’ve heard someone state it. I have always balanced a US nickel on the cylinder while it is operating and my machines never vibrated it; it remained stable. That’s my test of a well-balanced and operated machine.

Hooked also states: “One small correction to a comment from Montemayor regarding 100% clearance. It is possible to have 100% clearance or greater than 100% clearance and still get flow in and out of the cylinder. I conducted studies of many compressors with >100% clearance.” I don’t believe that I stated anything contrary to the fact that gas is flowing into the cylinder and out of it while it is unloaded. That, by the way, is the definition of 100% unloaded! As I stated above, there are many ways to get the clearance increased for the purpose of unloading the cylinder(s). Finger valve unloaders as well as plug unloaders will yield total unloading and the gas will, as a result, be pushed out of the cylinder and sucked into the cylinder every revolution of the crank.

Natural gas reciprocating compressors (and I’m presently handling a project with two large ones) sometimes employ a simple by-pass around the machine for capacity control not because it’s better (it’s not; it’s very costly from fuel consumption point of view) – but simply because it’s simple and CHEAP. It’s also traditional in the oil patch where instrumentation is kept to a bare minimum. Hooked states “This minimizes the compression ratio on the compressor resulting in little load.” This is also not true! The mere fact that gas is being recycled around the compressor means that the suction pressure is being kept constant by the recycle. This is the way this type of control is installed and operated. That being the case, then the compression ratio across the compressor is kept constant and so are the individual stage ratios (in the case of a multi-stage unit). There is no minimization of the compression ratio by this method. As gas producers are becoming more and more cost conscience and educated in field gas compression requirements, they are employing more and more clearance unloading techniques – and not finger unloaders either – because of the simplicity, the positive control, and the fuel saved.

Insult2 also states: “It's also true that there are several minor losses that are not taken into account in the idealized equation. If the compressor is well-designed, those losses should be minimized and the equation's efficiency prediction should be very close to the measured volumetric efficiency.” This statement is totally false according to industry standards and to what is published in the GPSA. It is also not true due to practicality: as I stated before, there are inherent configuration clearances due to the valve ports, the valve guards, and other mechanical clearances. The effect of the gas contained in the clearance volume on the pumping capacity of a cylinder can be represented by:

VE = 100 – r – C {(Z_S/Z_D) (r^1/k – 1}

where C = cylinder clearance, %

However, this strictly theoretical stuff and akin to what Insult2 alludes to. The real, practical, and empirical relationship used by manufacturers is more like:

VE = 96 – r – C {(Z_S/Z_D) (r^1/k – 1}

And when a non-lubricated compressor is used, the volumetric efficiency should be corrected by subtracting an additional 5% for slippage of gas. This clearly shows that there is a significant variation with the theoretical equation presented by Insult2.

My intention in responding in such a lengthy manner is to clearly leave no misconception of what Volumetric Efficiency is and what it signifies in a reciprocating compressor. It is not to show where others are wrong. If something is wrongly stated, I feel it my duty to clear it up in order to give worth to the discussion. A relatively low VE, in my opinion, simply means that you are getting a machine that has more flexibility – but not necessarily! A low VE doesn’t necessarily mean a less “efficient” machine.

 

[HookedOnHalo]

I think we are "on the same page", just some misunderstanding because of the limitations in communicating by this newsgroup format.

10% to 15% Volumetric Efficiency is generally recognized as the minimum normal Volumetric Efficiency for recip compressors. ACI has a great web site with some information on this topic.

A head end or crank end clearance of greater than 100% is possible and the compressor cylinder will still be moving gas. The HE could have 120% clearance and still have compression on the HE and move gas. A finger unloader or other unloading device results in no compression or no net flow. Flow simply moves in/out of the cylinder. 100% clearance does not equal "unloaded"

The comment about by-pass control was regarding an earlier comment about using clearance as a method of unloading the compressor during startup. Flow is often by-passed to minimize the discharge pressure which minimizes the load on the compressor allowing the driver and compressor to stabilize (temperature, lubrication, etc). The by-pass is then closed and discharge pressure increased thereby loading the compressor. There's a difference between by-pass and recycle, or at least some people distinguish between the two terms. As my boss says "The English language is a precise language when used precisely"

 

[nashc0]

Regarding Montemayor's comment:

<< Hooked is wrong when he states: “Seems a value of 10% to 15% is typically used as a minimum value to ensure there is enough flow in and out of the cylinder to prevent over-heating of the cylinder. Furthermore, prediction of loads and flows becomes difficult since at low volumetric efficiencies, the effects of pulsations are more dominant.” There is no way that the effect of cylinder clearance affects the expected discharge temperature out of a compression cylinder. >>

I think Hooked is actually correct in this instance because a very low Ev calculation can mean that gas is being compressed in the cylinder and never discharged. When this happens continually, the cylinder can overheat due to accumulation of heat due to friction. What Hooked is talking about is not discharge temperature of the gas, but the temperature of the actual cylinder itself. The equations Montemayor refers to is simply the thermodynamics of the work done on the gas.

Clayton