PUMP QUESTION

[habs]

I have a nursing home that has heat pumps in every room and a heat exchanger in the mechanical room tied to the heat pump loop and the cooling tower. The cooling tower pump keeps loosing its prime and the system shuts down. The strainers are clean and the pump will not pull the water from the tower.The tower is not piped properly, the suction comes out of the pump and goes up five feet and outside and down five feet to the tower.We thought the water was flowing back to the tower when the power went out and the pump would lose its prime but I found a check valve in the suction pipe yesterday so it was not flowing back and the power did not go out yet it still went down.I got it to run yesterday by removing the check valve and running the main pump (5 hp B&G) and the 7.5 hp back up at the same time. This got the system up and running but the pump is still cavitating(I can hear it).
The tower is an open BAC and the sump is full so I am pretty confident that I have a full column to the pump.
My question is will scaled plates on the heat exchanger cause my pump to cavitate. The water goes from the pump to the exchanger to the tower and back to the pump.The building is three years old, there is chemical in the system. The system has done this in the past maybe three times before this summer,(maintenance guy keeps great records)but this summer it has gone down maybe ten or fifteen times so something is wrong.

Sorry so long but want to giv e as much info as possible. Thanks for all replies

[chillrdude]

Originally posted by cardinal ac
The tower is not piped properly, the suction comes out of the pump and goes up five feet and outside and down five feet to the tower.

Sounds like you have an inverted trap there, is there any kind of vent in the horizontal line that goes thru the wall? I dont believe scaled plates will cause your problem, sounds like you have some kind of restriction in the suction piping leaving the tower(birds,racoon,bags,who knows). The inverted trap is going to need to be full of water to get the pump primed, can you get the air out of it.

[s tek]

Sounds like your suction is getting plugged. Check your suction pipe for a rust hole or leaking threads. Did you check your flow rate back to your tower? You might run a temperary hose and footvalve. This would let you know if the suction is a problem. If the discharge is the problem reverse the flow and collect the water so you get rid of the scale. The last time I did this we used a horse tank and some 2" hoses and a gas trashpump. (there pump if they let you)It worked good haven't been back.

[habs]

I closed off the pump discharge and bumped the water in the suction side to 30 psi with a hose, to check the piping and the check valve (this is how I found out there was a check valve buried in the insulation at the tower) the check held and the piping held.The only other place is in the tower but nothing is accessible. The screen is clean in the tower.
There are no vents on the suction line.
If I can't push enough through the exchanger then I can't be pulling enough through the pump, does that make sense at all or am I grasping at straws.
The sprayers in the tower do not seem to be spraying very hard but is that an indication of a suction or discharge problem.
The specs on the heat exchanger say it should have a ten psi drop and my inlet is 23psi and outlet is around five. The temp drop is around 25 degrees. There are no records of what the original pressure drop was or temp rise.

[s tek]

Are you sure the ck valve is opening. It can hold pressure but not open all the way, this would cause the problems your having.

[wolfdog]

Originally posted by cardinal ac
I The specs on the heat exchanger say it should have a ten psi drop and my inlet is 23psi and outlet is around five.

Specs say 10psi and you have 18psi accross the heat exchanger. Sounds like you should look at the heat exchanger.

[habs]

The flapper from the check is laying on the roof so it is not a problem anymore, but I still had a very hard time getting the pump going and it is still cavitating. The heat exchanger needs to be cleaned but I am not sure if it will solve the problem and to take it apart will take at least a day and the residents paying 2-4 grand a month get ugly when the air is not on.

[airworx]

make sure your piping doesnot have any place where it can trap air. then clean cooling tower, orfices, strainers,and condenser barrell. if that doesn,t work call for service.
also make sure pump isnt running backwards.

[sonc]

I just cleaned out a h/exchanger on a similiar system friday, it was plugged with all kinds of crap. (about 30 plates). You may want to clean it out with that much p/drop and td. Take an apprentice for sure.

[mdp]

The inverted trap on the suction side of the pump will kill you on an open tower. The centrifugal pump cannot lift the water up out of the tower through your inverted trap to the system. Centrifugal pumps have to have a specified amount of static suction pressure to operate. Sounds like if the tower had been piped straight through the wall into the suction side of the pump the system would be better off. Usually we like to see the tower a little higher than the pump to create some pressure on the inlet of the pump but I have seen pumps put only a foot below the tower in a properly designed system.

[mattm]

Couple of things to look at. You didn't mention if you've looked at the impellor in the pump. And another does this BAC have those perforated tubes on the suction line that act as strainers and assure that you only pull water from the bottom of the basin? I've seen those get plugged up pretty good. Lastly if those spray nozzles are those black plastic ones withe the slot in them pull a few and poke around those headers will get plugged up with scale really bad and reduce flow trhough the tower.

[acjourneyman]

MDP is right on the money.A pump like yours is not designed to lift water.I was involved in a similar job a few years ago and the engineers solution was to put a check in the suction side of the pump,well the pump did not have enough power to open the check and lift the water.It was a cluster and I was there weekly repriming the pump.I ended up having to change the spring in the check valve to a weaker one so you might want to do the same.it worked alot better after that but it still wasn't perfect.I don't know why engineers don't pull there heads out of there asses when they design condensor loops.The outlet of the tower always needs to be at a higher elevation than the pump or the pump will always cavitate.Also you mention you did pull the check so now you can't shut it down as all that water will flow back into the tower.You really should repipe the loop or reinstall the check with the weakest spring possible and see if that works, another thing you might do is set up the controls ,if there is any, to start both condensor pumps and then shut 1 down.

[Edited by acjourneyman on 07-18-2004 at 10:16 AM]

[weezer]

I agree with MDP.If your inverted trap is above the water line of tower sump, even by several inches you will have a problem. Unless that pump runs constant even a small pocket of air will stall a centrifugal pump. Especially since you now know that the check valve is out of the loop. Everytime the pump shuts down the level in the suction line will equalize with the level in the sump. A dirty HE will further exzasperate the problem.

[hw chiller man]

Hey Scott, it sounds like the same type of system as at Windsor Star, only the suction to the pumps there goes up about 10 feet and the pump runs at 7 inches vacuum. Every time there is a power failure the system looses its prime, but it does work OK normally...........eats out strainers though. These pumps continually cavitate and always will until repiping is done. I think you have the same deal.....good luck.......I'll guess who the engineer was HP

[habs]

We thought it was losing its prime too until I foud the check valve. Maybe the check valve was sticking but when I tried to get it going again it took two pumps and a hose hooked to the inlet to the pump to get it going again.
I did try it with the spring off the check but it would not pull. The system has run for two or three years this way without many problems until now. I am still not sure if the clogged plates will cause the pump to cavitate.
The sump does have a screen across the back end and it is clean and the pump runs all the time. It is also well below rla.
Wayne what did you guys do to fix the star tower or is it fixed? I will find out who engineered it Monday.

[hw chiller man]

I suspect your inverted trap is full of air. Check to see if there is a maid of mist or some other type of air eliminator in this line, which could be sucking in air if the internal float is sticking. Do you know for sure that the plates are plugged, maybe take a pressure across them to confirm. however I don't think a partially plugged HX will cause cavitation. Check to see if the suction side of the pump is in a vacuum. I'll bet it is. Feel free to call me at home if you want.

[habs]

The suction side is in a vacuum , it held thirty psi water pressure with the check valve in so I don't think there is a leak and the pipe goes out the wall three feet from the pump, it goes up about five feet and out the wall and down about five feet to the roof level then horizontal about 25 feet to the tower.Maybe the check was causing the problem and tommorrow all the air will be out of the suction line and it will run till the power goes out.
Gotta love these Fahrhall installs!

[hw chiller man]

I have seen at the star where the air bleeder will hold pressure,as soon as it gets into a vacuum situation it starts to suck in air. The air bleeder was isolated and the line primed and the problem disappears until the next power failure. The problem was never corrected because of the cost involved.

[habs]

I thought I could hear air at the vibration absorber but assumed it would leak if would suck air or maybe I am hearing something else, will find out tommorrow.
Thanks for the help Wayne.

[glgto]

if air is a problem and it sounds like you have a few definetly check the impeller cavitation will eat a impeller in no time