Recip Low Discharge Temp 1

[zdas04]

I'm looking at an Ariel JGT/4 3-stage compressor. I've been trying to match measured data with the Ariel Performance program and one cylinder is looking weird. Cylinder #1 has a suction pressure of 44 psig at 47F and a discharge pressure of 165 psig at 195F. Ariel performance predicts a discharge temperature 191F so that cylinder is fine.

The second first-stage cylinder has a measured discharge temp of 179F and I'm having a devil of a time explaining why Cyl #2 is 16F cooler than Cyl #1 (which has about the right temp for the compression ratios). I don't have any problem explaining why a cylinder would have temperatures higher than predicted, but I can't think of what is broken to give me a lower temp.

Could a discharge valve be leaking so much that the cylinder is doing 2 ratios instead of 3.056 (which would give a discharge temp of around 135F), and then add a chunk of 195F gas to get the blended volume up to 179F?

Could cylinder rings be leaking enough to cause it?

Thanks for any help.



David Simpson, PE
MuleShoe Engineering

http://www.muleshoe-eng.com

Please see FAQ731-376 for tips on how to make the best use of Eng-Tips Fora.

"Everyone is entitled to their own opinion, but not their own facts" Patrick Moynihan

 

[pmover]

David,

you probably have already investigate these, but . . .

do both 1st-stage cylinders have the same valves, springs, etc.?
-one cylinder could have less "pressure losses" than the other,
-their could be a pressure pulsation occurring such that the one cylinder has a slightly higher compression ratio,
-any possibility of the temperature sensor/transmitter/indicator is bad?

any leakage, valve or rings, would most likely result in higher temps.

any possibility of measuring temps & pressures at cylinder inlet/outlet and using an equation of state to determine change in enthalpy, thus efficiencies. a little work involved with this.

any possibility of using a listening device (I called them mickey mouse ears & have long since forgotten the make/model/device), but this device had about a 12-inch long rubber cone shaped nozzle connected to a sound sensor of specific frequencies-ultrasonic?) to listen to cylinder valve operation? I kid you not, for poppet valves, one could easily hear the poppet valves open/close on every stroke, easily hear the leaks (high pitch hiss). For rider bands or rings, more of a "whooshing" sound. One could also hear/detect piston slap on ICE. A great tool for diagnosing potential problems.

hope this helps.
-pmover

 

[zdas04]

Yeah, the two cylinders have the same valves and (plate) springs. Same clearance.

Of course the temperature indicator calibration could be bad, but I put my hand (gently) on the two cylinders and the cool one feels cooler to the touch.

Funny thing, the two first stage cylinders are both seeing the same external suction pressure and external discharge pressure. Those ratios call for a discharge temp around 195F. One cylinder has a discharge temp of 179F. Now if the cool cylinder valves were too stiff then the actual ratios within the cylinder would be higher (partial fillage) so the temperature would be higher than expected.

I'm thinking that with leaking rings, the suction stroke gets gas both from the suction valves and from the other compression chamber--resulting in fewer compression ratios than the external pressures would expect.

Whatever the problem is, the cool cylinder is moving a substantial portion of its share of the load (or interstage pressure would be a lot lower) so the problem is smallish.

I don't have any Micky Mouse Ears (or a stethoscope), and my recommendation is to replace the machine with a 2-stage anyway so I don't have to solve this one, I am just looking for an explanation that satisfies my curiosity.

David

 

[pmover]

ok, thanks!

Is the cooler cylinder exposed to cooler ambient air? Any radiant heat being directed at the temperature indicator? Yes, when process temps differ significantly from ambient temps, (hot cylinder interior with cooler cylinder exterior), one cannot "guess" or "assume" cylinder temps are the same. Firsthand experience taught me that situation.

I was also thinking that if there is cylinder cooling, that more heat of compression is being removed by the jacket cooling system on one cylinder than the other. Likely not the case though.

I will tell you that I have tested many a cylinders (mostly pipeline compression), with varying pressures & temps on the cylinders, but when determining the efficiencies, the results were nearly all the same. Yet, when further data was collected (cylinder pressure vs. cranks angle & PV data), the pressure pulsations had an effect on other cylinders. One must remember that the pulsation designs back in the 1950s or earlier was likely more guesswork (i.e. a rule of thumb) than scientific. We were not about to change the suction/discharge bottles as the cylinder efficiencies were nearly the same.

Regarding leaking rider band/rings, the cylinder capacity would likely decrease. The gas within the cylinder would continue to compress & expand, and eventually over time, increase in temperature (depending upon the degree of leakage).

Have a good weekend David.

-pmover

 

[dcasto]

i've seen 5 or 6 degrees difference between cylinder where on has the valves removed on the head end of one.

I believe the low temp is correct and the the high temp is because of bad valves in that cylinder.

 

[zdas04]

The high temp matches Ariel Performance and my hand calculations. The low temp doesn't.

David

 

[dcasto]

there is no thermodynamics that get you to a lower temperature. Time to get out the yellowbacks and test gauges

 

[zdas04]

That was kind of the reason I started this thread. The only scenario I could come up with was a source of lower temperature gas that reduced the number of compression ratios in the cool cylinder. If the gas on the compressing end leaked into the suction end during the entire stroke, then I can see where the blended stream could be something like 1/3 leaked gas at an average temp of 100F and 2/3 suction gas at 40F (blended average temp 60F) for a total compression ratio of 2.56 and a discharge temp of 179F. Comparing the suction header to the discharge header was 3.056 ratios so I think this is credible.

David

 

[dcasto]

nope, that is not correct. What you are describing is rings leaking.

What happens if you compress a gas, it gets hot. If you let the pressure off it cools, but not as cool as before. If you let the gas pressure off through a expander, the temperture gets colder than a straight valve, it still won't get cooler than the original gas, that pesky second law and entropy.

So if gas leaks back by the rings, the cylinder will get hotter. also, it will pump less new gas. This means that you'll have less gas going to the second stage, which means the second stage cylinder will be at a lower pressure, yet you state that the ariel model matches on both temperature and pressure?

I'll call you and review the data if you wish, I just live up the hill from you.

 

[quark]

Any updates on this issue?

 

[zdas04]

No, I finished my report by saying that there was no earthly reason for taking a 300 psig pressure drop across a suction controller (using 650 hp of the 1,200 being used) so that a 3-stage compressor could be in its comfort zone. The application is clearly a 2-stage and the compressor in question needs to go away. The packager refused to even talk to me because I'm "just a contractor" so I can't get any detailed data beyond what I started this thread with.

My interest was more acedemic than urgent. I still can't come up with a scenario that results in a low temp in a cylinder (and even if the instruments were off, one of the cylinders felt noticably less too hot to touch). I can explain high temperatures easily, but if the machine is doing at least the number of compression ratios that the external sensors indicate, and one cylinder matches the temp rise calculations then why is the other one cooler?

David

 

[quark]

May be a silly suggestion as you might already have looked into it. Are these cylinders water jacketed and how is the integrity?

Some of the air compressors I operated have open cooling water pathways between the cylinder jacket and cylinder head jacket, with gasket sealing. If the gaskets are damaged internally, leaks are not visible from outside.

 

[zdas04]

The machine is an Ariel JGT (no water jacket on Ariel compressors).

David

 

[dcasto]

take the valves out of the first stage cylinders.

 

[zdas04]

I looked at that scenario and if I took the valves out of the first stage cylinder I could raise the suction controller from 50 psig to 175 psig, but it only increased the throughput marginally and I'm still wasting a bunch of hp across the suction controller. Opening the suction controller all the way (with the 1st stage valves out) exceeded the MAWP of the second stage cylinder by quite a bit.

My recomendation was for it to go away.


David