Pump Cavitation Indicator?

[dtse86]

We are currently in the process of starting up an open system with large Armstrong Inline Pumps (~100hp). The pumps are currently controlled by the BAS through VFDs. The balancer has now come in and claims that he is hearing cavitation as he ramps up the pumps and that we are not meeting the NPSHR (27'H2O). Through our calculations we have calculated a NPSHA of ~40'H2O. With the pumps off, he is claiming that the pressure on the suction side of the pump is only ~18'H2O. Something here doesn't seem right and I had personally ramped up the pumps to full capacity and only heard a little noise from the volute. The majority of the noise at that time was coming from the motor. Is there a rule of thumb to determine at what decibel level we can assume cavitation?

 

[tys90]

I wouldn't go by sound level necessarily. Even though it might not matter much in the end, I would reconcile the differences between the reading of only 18' on the suction side versus the 27' required and 40' calculated before making any decisions.

 

[HVAC68]

I suggest that you should go back to basics. Look at where the pump is operating. i.e. measure the flow and head and plot it on the pump curve, to see where it is operating.

HVAC68

 

[KLH]

With 18 ft. water pressure on the suction side of the pump, plus 33.4 ft. water for atmospheric pressure, you have a NPSHA of 51.4 ft. water. You are well above the NPSHR.

Unless you have high temperature water and are on the moon, I doubt you have cavitation.

I think you're fine.... maybe bad bearings in the motor???

Good luck!

 

[dtse86]

tys90,
After looking into it more, I believe that KLH is correct in that the discrepancy between our calculated and the 18' is in the atmospheric pressure. I think that the gauge on the suction side of the pump is only reading gauge pressure and not absolute, and so we have to add additional pressure to get the NPSHA. We had shown our NPSHA calcs to the balancer and he was convinced that NPSH does not include atmosphere, but I am sure he is incorrect.

HVAC68,
Yes, we did go back to the curves to make sure we were getting the proper performance that everything was matching up. It all seemed to align, except that a few individuals thought the pumps were running too noisy.

KLH,
Thanks, I forgot to take into account that the gauge measures GAGE pressure and so we have to add for atmosphere. I think we are at the point where most have sided with me and have thrown out cavitation. However, the lead engineer is saying that he has never heard pumps so loud and that he thinks there may be a problem. You may be right that there could be noise from the motor. That I believe is the loudest portion during operation. Currently these are 100hp split coupled inline pumps that are supported purely from the piping. So I believe that it may be an issue with the motor vibrating and it transferring the vibrations into the piping.

 

[trashcanman]

Ask the Armstrong pump rep to check it out.

 

[dtse86]

Yea, we have come to that decision and the rep's coming out next week now. In my opinion, I think that everything is running just fine and that for some reason everyone thinks that 100hp pumps should be extremely quiet and that you should barely notice that they're on. I have a feeling that the rep is going to come out and say that nothing is wrong, but if we need them to be quieter, we can get the motors rebalanced to a higher spec. But these pumps are located on a mechanical floor and the noise shouldn't be an issue for anyone.

It's just a little frustrating at this point that no one seems to be listening to me because they feel that I don't have enough experience to know what I'm doing although I have been on working on site now for longer than anyone else involved.

Thanks all for the suggestions, I'll let you know how it goes...

 

[KLH]

dtse89,

I was looking for an old post and came across this one in the process.

How did you ever make out with your pumps?

 

[dtse86]

Well, it turned out that the flotrex valve on the discharge of the pumps were only 20% open and so when the balancer started ramping up the pumps on the VFD we were getting cavitation at around 30 Hz, but that was only because we were pretty much dead heading the pumps. I ended up telling the balancer to open up all of the discharge valves to 100% open and then ramp up the pumps to full speed. They still made a little bit of noise but nowhere close to when we were dead heading them. I finally convinced the balancers, contractors, and lead engineer that it was ok for them to make a little noise, especially because these were inline pumps and were directly supported by the piping and pipe supports. The pumps were on VFD's and so we decided that the pumps did not need to have a flow balance on the discharge valves because the VFD's could be used to balance the flow. Really the only valves that needed adjustment for the balancing was the balancing valves on the entering water side of the condensers.

Really the most frustrating thing of this entire debacle was that everyone jumped to conclusions that the system wasn't properly designed and that we had screwed something up when in reality, they just needed to properly balance and run the system.

 

[Artisi]

Nothing like back to basics and checking everything before jumping onto the "pump doesn't work", "the system hasn't been designed properly" etc bandwagon.

 

[pipesnpumps]

It is unbelievable how misunderstood the concept of cavitation is. Everybody is an expert on cavitation, like your TAB firm, but none of them bother to even read up on the concept enough to calculate NPSHA/R correctly.

About the only time you will ever find cavitation in the HVAC realm is on condensate receivers or boiler feedwater, where the pumped liquid has a low NPSHA because it is very close to its boiling point. In the case where it is exactly at the boiling point the NPSHA is gravity head on the suction less the suction pipe losses.

You can definitely get flashing across a valve though.. If you are pumping hot water across a valve that is severely pinched closed, and there is alot of pressure (dead-headed), then the velocity at the throat of the valve can lower the normal pressure enough to cause flashing (Bernoulli's theorem) which is very loud.

 

[quark]

27' NPSHr seems to be too demanding for HVAC inline pumps in any case.