Suction pluggage cavitation 4

[mjpetrag]

I have a cooling tower with a 2' water level sitting 20' above 3 pumps that routinely swing from 190' to 130' about every 30 seconds. There are suction screens with a bunch of junk on them right in the pan. The suction screens sit on top of 3 respective 48" pipes and below that are vortex breakers for each suction. These 3 pipes feed the suction header to the pumps. The question I have is a loss issue. The difference in water height is about 6" across the screens. So that means a 1/2'(.22 psi) loss in the suction piping is causing the cavitation? It surprises me that the pumps are so close to the NPSHr. Unfortunately, I don't have NPSHr curves for these pumps because they were built before NPSHr curves were imposed on pump curves. I have a hard time believing there is another pluggage downstream of the vortex breaker causing an additional pressure drop, but maybe that is the case.

Thoughts?

-Mike

 

[1gibson]

Any way to get differential pressure readings across the screens? High flows could pull a slight vacuum.

 

[Artisi]

Any chance it is air entrainment?

 

[JJPellin]

I am having difficulty picturing the configuration of your screens. Please provide more description or a diagram. Is it purely 6 inches of head or can there be a slight vacuum behind the screen as 1gibson asks?

Johnny Pellin

 

[mjpetrag]

There are 3 suction screens. I'll simplify but describing one screen. There is a 48" diameter hole in the pan with a cylindrical bird cage screen 48" in diameter and 2' high sitting on top. So the top of the screen just reaches over the pan level. Below the pan level is a vortex breaker sitting in the 48" pipe. The pipe suddenly contracts from 48" to 20" after this and then dumps into the suction header.

-Mike

 

[JJPellin]

Based on that description, the top of the screen is above the liquid level and open to the air. Thus, there cannot be a vacuum on the downstream side of the screen. It is possible that you do have a situation where a change in suction head of only 0.5 feet can drive your pumps into hard cavitation. Does anything else change as the screens become restricted?

Artisi asked a very good question about air entrainment. Many of us may have been thinking only about vortexing. But, if the turbulence with the restricted screens is drastically worse, air could still be entrained and carried into the pumps even without a vortex forming. What is the flow rate expected down each of these inlets. If the velocity in the 20 inch line is high enough, air bubbles could be carried all the way down to the pumps.


Johnny Pellin

 

[QualityTime]

I had an experience with cooling towers made by Baltimore Air Coil a few years ago and I have talked about this in these forums a number of times. I had plenty of NPSHa but I was getting entrained air in the pumps. You could hear the air in the pump discharge and I could see the flow meters bouncing wildly because of the air. Do not mistake entrained air with cavitation. They sound the same, the damage to the pump is the same but the solution to fix the problem is different.

The local BAC rep had set the working water level to "what they have always set it to". I started looking into it closely and I called the BAC factory in the US to send me literature on what the working water level should be because their shop drawings did not show what it was supposed to be. I found a numerical discrepancy in their publications and drawings that they sent to me. In short one publication had a higher wroking water level than what was advised in another publication and my tower working water level was set to the lower level. After I discussed it with them they advised that there was an indeed an error and they had to talk to their local rep to raise the working water level in the cooling tower. Problem solved. It caused some havic on my pumps. You have to wonder how many installations were installed incorrectly in the past!!!

I had to get the BAC technical support lady in the US to talk to senior people in thier factory to get it straightend out and come back with an answer for me. You have to remember that as staff change over time, especially in a manufacturing environment, people forget the engineering behind what they manufacture

 

[QualityTime]

I will also add that the difference in the working water levels was a few inches and that made all the difference in the world

 

[QualityTime]

A double check would be to relate the top working water level to the cooling tower overflow and to the fan discharge beach plate.

The TWL should be just slightly below the invert of the cooling tower overflow pipe

The TWL should be just slightly below the fan discharge beach plate.

The low water level should only be about 1" below the TWL. Do you have a float operated water inlet valve or is the water level controlled by level probes?

 

[mjpetrag]

Ok, so I stuck a pressure gauge right before the pump suction . The gauge was reading around 6-7 psi while the pump wasn't cavitating/ingesting air. Then about 10 seconds after holding, the pump would start cavitating and the suction jumped around 1-4 psig. The discharge gauge would swing from 85-50 psi during this time. Then 20-30 seconds later, the discharge and suction pressure would hold and the cycle would begin again. Water level in the pan is controlled automatically and stays relatively constant at 2'

There is a lot of junk on the screens from what I can see, but the water is murky and there may be a lot of stuff plugging up the screens below the surface. Would these symptoms point to a suction blockage on the screens?

-Mike

 

[tr1ntx]

Yes, clean your screens.

 

[mjpetrag]

Cleaning them was my original thought. Wanted to make sure there wasn't something else I missed before shutting down the entire unit, like air ingestion.

-Mike

 

[Artisi]

Have the pumps always operated in this fashion or is it something new, if it's something new, what has changed?

 

[mjpetrag]

We had a lot of people retire that used to keep after this tower. I just found out this afternoon how often it was cleaned before they left. Now we start doing what they did again.

-Mike

 

[QualityTime]

Cleaning the screens is fundamental and is the first obvious thing to do. The fact that the suction pressure gage went from 6-7 psi to 1-4 psi and the discharge gage would drop from 85 to 50 psi still tells me the following:

The pump is still under a flooded suction. There will be no cavitation

You are still ingesting air

Do you know what the working water level in the tower is supposed to be.....check with the manufacturer

 

[QualityTime]

i am interested in what you have found so far

 

[mjpetrag]

Suction screens were cleaned and the problem went away. There was one of the 3 screens completely blocked and the level was extremely low. It actually went a few feet down from below the pan level and I couldn't actually see the level in the pipe

-Mike

 

[QualityTime]

With cooling towers, having the proper working water level is critical. If the water level is too low the entrained air from the cooling fan does not have enough time to float up to the surface before it gets sucked into the pump suction. When I say too low....I am talking even as low as an inch off makes a big difference

 

[QualityTime]

As I said in my earlier post....Do not mistake entrained air with cavitation. They sound the same, the damage to the pump is the same but the solution to fix the problem is different.....

The solution in this case to get rid of the entrained air was to clean the cooling tower screens in order to get the working water level up to thier proper level.

If it was cavitation you don't have enough npsha and you have to reduce the head loss in the suction piping.

Too many people confuse cavitation with entrained air and if you are not trained to know the difference you could chase a solution that will be costly and does not work. I am glad that you have solved your problem

 

[Artisi]

QualityTime,
Yes you are 100% correct regarding the confusion between cavitation and entrained air, and I must disagree with your statements that 1)they sound the same (in some cases it may be so) and 2) causes the same damage (no it does not).

I recommend the following links, 3 of the many references which refute the erroneous statement that entrained air causes the same damage as cavitation. Unfortunately, this statement has been repeated on a number of occasions by well meaning but ill-informed writers simply regurgitating the same incorrect information.

http://www.cheresources.com/invision/topic/10445-pump-impeller-damage-due-to-entrained-nitrogen/

http://news.lubipumps-na.com/blog/2009/05/effect-of-cavitation-on-centrifugal-pump-performance/

http://www.waterworld.com/index/dis...tion-air-on-centrifugal-pump-performance.html

If you have any information on detailed studies undertaken on the damage from entrained air, I would be pleased to see and read this information.