liquid ammonia pump calvitation

Hello

At one of my customer system we have two hermetic ammonia recirculation pumps install on a horizontal expansion receiver. One is the backup of the other. The system is operating since three months, the first pump is always calvitating compare to the other is never calvitating. I am wondering why.

The pump could calvitate more than 6 times in a hour at least once every two weeks. During a day we occasionnaly see the pump calvitating. My refrigeration capacity is mainly stable, because it is a holding freezer at -12dF.

When we operate with the other pump it is never calvitating, same system, same receiver, only opposite connection.

The pump were design at a 3ft NSPH and we always have a 4ft NSPH, the amps of the motor 3Hp is 1amps below the motor maximun design. My piping design are the same. But one maintain a better pressure rise than the other one. They are both new pumps. A pump guardian is install on each pump.

We opened the pump which is always calvitating and iis seems ok.

My operating system is at 2inHg succion pressure
and at 160psig high pressure.

kathleen, P. eng

[URL=http://www.refri-ozone.com]

A pump cavitates when the pressure at the eye of the impellor drops below the saturation pressure for the temperature of the liquid. Evaporation takes place, forming a vapor bubble. As the bubble moves outward in the impellor, the pressure rapidly increases, condensing the vapor, violently collapsing the bubble. This can and will destroy the impellor.

The pump that is cavitating either has more restriction at it's inlet (valve not fully open?), or less restriction at it's outlet. If the inlet is fully open, restrict (partially close) the discharge to eliminate the cavitation.

Quote:

---

The pump were design at a 3ft NSPH and we always have a 4ft NSPH

---

Not that the Prof is a particular expert in this area, but he recalls there is some significant variances in manufacturer's stated NSPH.

It would be interesting to swap the pumps to see what happens.

You've pretty much answered the obvious questions. What are the inlet and outlet pressures on each of the pumps while operating? You stated system pressures and I am not going to assume them to be pump pressures. Do you valve out either pump while operating the other? If not are there any check valves present? My thoughts are a leak on the inlet side if the pump suction is below atmospheric or restriction in the inlet piping to the troubling pump. Or maybe you have 2 pumps that are supposed to be the same but yet are not, wrong impeller, shroud, etc. Check all part #'s, measurements , etc.
Mike Hopkins

I trust Gary's advice. I don't have anything to add. Well, maybe I do.

Your question proposes that there is a differnce between two pumps doing equal duty. Gary most likely is right. My sense is that you are missing an important clue that Gary's advice will help you to discover.

Dan

Hi Kathleen

Are both pumps fitted with pump delivery vent lines? And if so, the pump that is giving you the hassle - is the vent valve open?

Have you checked for oil lying in the pump? I know its pretty trivial, but it will cause cavitation!

Frosty

Thank you to all of you

Frosty I do have a vent line with a flow min control and at the outlet after the vent a flow regulator and a check valve. Do my flow regulator not working well.

One pump give me a pressure rise at the outlet of around 30psig and the other 40psig. The design was 42psig.

Tomorrow i am returning to the project and I will verify Gary advise.

Prof and Mike I you ever encounter a difference in the pump manufacturer that cause intermittent calvitation. The fisrt pump was the one who did the job of lower the freezer at the beginning. Which during this period we often encounter calvitation and the dirt of the system.

Have a nice to all of you

Kathleen, P. eng

Hi, Kathleen.
Just as a matter of interest what way is the inlet connections conected to the drop leg? They should be connected at a tangent, not a straight stab in. Think of water going down a plug hole, by tapping in at a tangent you catch this affect more effectivly. My thinking is that one pump is tapped in with greater effect than the other. Maybe you just haven a good enough liquid column above the pumps to service both and the one with the more direct flow/better tapping is robbing. Try increasing the level in you vessel and the evaporation to increase pump priming.
Regards. Andy.

Generally speaking, there are three ways to eliminate cavitation.

1. Lower the temperature of the liquid so that the pressure at the eye of the impellor is no longer below the saturation pressure of the liquid.

2. Increase the pressure at the inlet and therefore at the eye of the impellor, e.g. increase liquid head.

3. Reduce the flow at the outlet of the pump, which increases the pressure at the eye.

There is in fact a fourth way. There are centrifugal pumps with a positive pressure device at the inlet called an "inducer". These are generally used for special purposes. For example, the DeBakey artificial heart is a centrifugal pump with an inducer at it's inlet and an elongated impellor.

Gary,
Those are all keen ways to treat the symptom but not cure the disease. You would agree that it's best to get the pump curve and the minimum NPSH for the pump and then maintain NPSH at a level above minimum? In the system Kathleen is describing this has been done, but with little room for error. It may be that one of these pumps has suffered some impeller erosion or wear at the eye seal to the point it cannot develop sufficient head to maintain pressure above saturation? Simply pulling the pump apart and looking is not enough, these clearances need to be measured with feeler gauges and compared to manufacturers data. It is possible that some piping anomaly, valve appears open but is not, debris in the inlet line of one pump could be causing these problems. An air leak would probably afffect the other pump that runs ok if not valved out. But if piping is identical and pumps are identical, no air leaks, restrictions, etc. then I'd go after the impeller and volute housing. The intermittent problem is suggestive of an air leak or wear problem. The liquid recirc. pumps I have dealt with on R22 systems used orings and had leaks on them every year until we went with viton orings. These are small pumps and the clearances were critical on them, .002
from eye seal to volute was reccomended, max. was about .005.
Similar to the centrifugal chiller clearances on impellers, very tight.
Pumping water is one thing, pumping halocarbons is another, pumping ammonia is a whole different ballgame.

Mike Hopkins

Actually, these are general principles, which hold true for ANY liquid. The specific problem may be a little more elusive. :)

Any impellor (excessive) clearance problem increases the pressure at the eye of the impellor, making cavitation less likely, not more.

I agree with your statement about piping anomalies, valve problems, inlet debri, etc. Anything that drops the pressure going into the eye can cause cavitation.

I would be looking for differences between the two pumps and their associated piping, valves, regulators, inlet screens, etc.

Also, Kathleen hasn't told us which of the pumps provides the higher outlet pressure, and exactly where this pressure is measured. These would be important clues.

"One pump give me a pressure rise at the outlet of around 30psig and the other 40psig. The design was 42psig."

Most likely, the pump giving 30 psig is also making the worst cavitation. The flow regulator should be checked for dirt or sticking.

"The fisrt pump was the one who did the job of lower the freezer at the beginning. Which during this period we often encounter calvitation and the dirt of the system."
If there is a pump suction strainer, clean it. This can be a major cause of cavitation problems.
First, eliminate system problems such as dirty or stuck control valves, dirty strainers etc. After you know there isn't a simple malfunction you can begin to check the engineering (liquid head, pipe design etc.) I once flew from Canada to France for a client to "check the engineering" of a new system with poor liquid pump performance. I cleaned the pump suction strainer and went home!

Thinking more into this problem, we can only speculate that cavitation is indeed the culprit. What if the pump is not cavitating, but surging. There is another name for it I cannot recall. Possibly due to wear incurred during start up of the system. Alternately picking up flow and then when the pump cannot overcome the head proposed to it reversing flow. The pump experiencing this phenomena may at times be able to run ok with reduced head, but have trouble when flow increases or greater head is required. Sounds and acts like cavitation. The pump "giving 30 psig". Can we assume it to be the one giving trouble? The importance of having accurate gauges at the inlet and outlet of the pump or at least ports that are readily used shows here. If we knew the exact pressure and temperature entering and leaving the pump, and the fluid pumped, while the problem is occuring we could then say for sure what is going on. Just as in troubleshooting a refrigeration system. This post is almost like one of Marc's challenges!:) No?
Mike Hopkins

I would assume it to be the one with 40psig.

Marc eventually tells us the answer.

Kathleen may or may not.

Don't you just hate it when people don't come back and tell you what they found? It's like missing the end of a movie.

Anyone considered oil ingress ?? (I maybe missed it, sorry if I did)

Quote:

---

This post is
almost like one of Marc's challenges! No?

---

LOL. I learned something about the behavior of centrifugal pumps despite myself during Marc's challenge.

The surging puzzled me and I attributed it to NPSH differences as I attempted to justify what I was observing with my wife's waterfall pump.

The more efficient pump would be the more likely to surge because it has the greater reduction of NPSH during its efficient cycle. Thus I would guess the one with the 40 psig gain to be the culprit and that it experiences cyclic reductions in NPSH caused by some problem with the inlet piping relative to the design flow.

Quote:

---

Don't you just hate it when people don't come back and tell you
what they found? It's like missing the end of a movie.

---

Ain't that the truth! Kathleeeeeeeeeen?

Hi Kathleen
I have had the same trouble with a Witt pump.Cheap and most logical way was check incoming strainer found welding slag and other insolubles cousing a partcial blockage at the first pump started. Also got pump supplier to site to check empellor as pump was still under warrenty.
This is a large ammonia system.
robinf@ij.co.za

1. As the plant is new, were strainers installed to catch any welding slag? This stuff is pretty good at opening up pump clearances...
2. Is the cavitation a constant feature, or intermittent?
3. Do you get large variations in pressure \ level in the reciever?
4. It is not unknown for isolation valve seats to fall off partially and block off or restrict a pipe
5. Although it is a new plant, is the refrigerant clean and not been contaminated by water or by oil?
6. Out of curiosity, what make of pump are they?

We have a couple of Witt pumps suffering the same sort of problems, and we have experienced burnt out seals behind the impellor after only a short time, blown bellows on the pressure differential switches and blown pressure gauges....but at all times, the problem is occuring at low pressures, typically 0.70Bar(a) and light loads. Plus we seem to have a water content of around 200kg in a 4500kg system.

We are looking at a software fix to shut off one pump at low demand and pressure and to prevent operators pumping a plant down whilst a liquid pump is running. This is far easier that climbing into the surge drum to see if someone dropped a spanner into the suction pipe, (yes, it has happened....)

Regards

Steve (suffering with similar cavitation problem on Witt pumps a the moment)

I have ran into this problem in the past with a recirc. system i found the problem to be that we had a blockage at the inlet end of the pump. believe it or not their was a pipe cap left in the piping when installed, lost a lot of sleep on that one.

Thank for the reply Ammonia

We solved the problem this september, because the customer had to make a big shutdown, therefore I was having the time to empty the system and had a look. The both pump were fine, no sign of calvitation inside of it. but I change my connection to the expansion receiver, I do not have any problem anymore. Why one was more picky than the other one I do not know, but the one who was calvitating is not doing it any more. Oups we also change a pressure control on the pump, which was probably the cause. Because the pump was not showing any sign of calvitation.



Thanks to all of you

Was the possibility of a broken checkvalve eliminated?
In that case, if the other pump is working you got a short cycle over the pump with the broken checkvalve, hence too high flow = more pressure loss everywhere = gas = .........!
That is, if no maximum bleed was installed.
Nico.

All ways check pump rotation!

Hi, Valentino :)

Quote:

---

Originally Posted by Valentino

All ways check pump rotation!

---

welcome to RE forums .... OP is from 2002 ;)

Best regards, Josip :)

Hi Josip.
good call, this must be a record longevity thing. There seems to be more of them since the new site start up.