problems cooling centrifugal pumps

[bragamor]

We have several centrifugal pumps in our process pumping liquid at high temperatures back into the reactor. These pumps get really hot so we use cooling water to cool them down. The problem is that we cannot send this cooling water back into the cooling water return line because there is not enough pressure differential, so we have to dump the water on the floor...

Does anybody have any ideas on how to solve this problem?

Thanks

 

[Pumpsonly]

Are you cooling the pump by spraying water on it or just cooling the bearing housing and mechanical seal?
If the pumps are designed to operate at high temp, cooling the lube oil and mech.seal flushing line will be sufficient.

 

[ione]

http://www.ppcmechanicalseals.com/API/API-21.pdf

http://www.ppcmechanicalseals.com/API/API-23.pdf

Regards

 

[bragamor]

thanks both for your answers

We cool some of them by spraying...any ideas on how to fix this problem without having a cooler for each pump.

 

[Pumpsonly]

Install a big fan near the pumps then ....

What type of pump and operating temperature?

 

[maddocks]

You could blowcase the cooling water back int the system but this can get more complex than pumps....

 

[SNORGY]

What is the nature of the cooling water you are spraying them with? If you have added chemicals you probably can't just go to clean water sewer (CWS) or oily water sewer (OWS) with it; your regulator probably won't allow it. I don't think it's actually "cooling water" - or is it? If so, you will be paying for that in CT make-up makeup and in chemicals to control the increase in cycles of concentration if your spraying rate is high enough. So, I assume it is fresh water or utility water. Are you just opening a utility station wash water hose and spraying them down?

In any event, I doubt that the pumps were designed to be cooled in this fashion. I would address this to the extent possible with air cooling, or else look for a different type of pump.

The other thing you might try is give CSI (Controls Southeast Inc) a call and give them your pump vendor drawings. They might be able to design a jacket system for the pump casing, which would facilitate you towards then adding a closed CW loop. You could go, for example, from open top tank full of water, through a small pump to the casing and back to the tank and let the exposed surface area of the tank take care of the heat rejection; it would likely equilibriate at some acceptable point if the tank was large enough.

Regards,

SNORGY.

 

[SNORGY]

Apologies for my atrocious grammar.

Regards,

SNORGY.

 

[sshep]

Hey Bragamor,

There is something missing to you story. The pumps get hot and this is a problem because...(why). Reading into your story I think that the pumps were designed for the service they are in and have a cooling jacket, but the cooling water system cannot deliver the flow because of low differential pressure in the supply and return header. The spray is a work around.

On the assumption that the above assessment is correct and that the CW system was properly designed at one time; the cooling water system either has more load on it because of the addition of more users, or it is not balanced properly. These pumps are probably not the only thing struggling for cooling. Install some pressure gages as part of a CW survey, and try balancing the flows better- maybe more dP is possible by reducing CW flow elsewhere. Otherwise run a more adequate CW supply header by finding better tie-in points in the network.

best wishes,
sshep