pump leak mystery - looking for help

[USAeng]

Hey guys,

We have a thermal fluid pump at our plant that is leaking from the shaft seal. We have tried 2 different brand pumps now with the same result. It is a slow leak from the shaft seal. Both pump manufacturers have tried looking at the information for the system with no ideas. When disassembled - the seals look brand new. We would like to hire a 3rd party to help us with the issue. Can someone suggest an entity that has experience that might be able to assist? Google did not help us with finding someone so far. If this constitutes as "recruiting" - apologies. I am only hoping to get a couple good suggestions for company names

 

[1gibson]

Are you cutting the o-ring when the seal is installed? If it passes over threads, look for a little chunk cut out of the ID. If it passes over a keyway, look for small notches/flaps the width of the keyway, these can be much harder to see and the material won't necessarily be removed, just cut.

 

[Latexman]

Have you called your thermal fluid provider and asked for help? I have had good success with our supplier, Therminol, in the past. They have a group of highly experienced, practical Engineers to help their customers. I imagine the other big heat transfer fluid providers may do the same.

Good luck,
Latexman

To a ChE, the glass is always full - 1/2 air and 1/2 water.

 

[hydtools]

Look for a shaft finish that looks like a thread that carries the fluid passed the seal. A grind finish using a feed grind will create the kind of texture. Change the shaft finish method to a plunge grind to create a parallel texture finish instead of a helical texture finish.

Ted

 

[LittleInch]

John crane seals?

Have you tried your seal vendor direct?

What did you search on?

USA or somewhere else?

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[USAeng]

Thanks guys. Both pumps were brand new when installed and started leaking almost immediately. We have worked with Paratherm who is the thermal fluid supplier. We also worked with Dean pump, John Crane, and MP Pump. This is why we are at the point now of trying to hire someone to look at our system.

 

[USAeng]

When talking about the leak it is not pouring out - about 2 tablespoons per day leak if I had to guess. Not really trying to figure this out online - just hoping to get a couple names so we can consider someone to hire

 

[hydtools]

You have apparently taken the pumps apart to know the seals look like new. Inspect the shaft finish and seal gland finish yourself.

Contact the designer and builder of the system.

Ted

 

[USAeng]

we did not take the pumps apart - one was done at the pump supplier, and the other was done at the pump manufacturer with John Crane rep present

 

[Compositepro]

Almost all seals leak, it is only a question of how much. High temperature a temperature cycling are tough on seals. Many oil heaters that I have seen have a drip pan under the seal with a tiny gear pump used to return the drips to the system.

Otherwise, you probably need a different pump design, like sealess, or one where the seal is far from the hot parts.

 

[USAeng]

I have been told that too to some degree - but it happens that both pump manufacturers seem to think that the seals should not leak at all.

 

[georgeverghese]

If this Paratherm has a closed cup flash point lower than the max operating temp of this pump, then you should be looking at a pump with double seals; API seal flush plan 52 at the least, which should be made good for the max operating temp. You've got process safety issues if you have a flammable / hazardous area around this pump with this Paratherm leaking all over the place. There may even be health issues with operators inhaling the flash vapors.

 

[USAeng]

Closed cup flash point per Paratherm website >410F. MP Pump HTO 120 max operating temp = 650F per pump specs. Our actual fluid temp in the system is 350-375F

 

[georgeverghese]

Had a quick look at the Paratherm website - its clearly not Paratherm NF or Paratherm GLT. Still, if you want to eliminate leaks to atm altogether, double seals with plan 52 flush would be the way to go. On plan 52, the outboard seal leakage goes into a seal pot which may be vented locally at a safe location. Presume 375degF is the max op temp at the pump, which should include times when there is little or no heating demand. John Crane, Flowserve and Borg Warner are some double seal vendors - in severe operating conditions, opinions / advice from these vendors may differ.

If the (a) expansion drum is not inert gas blanketed, or (b) if the HTF fired heater is not designed to keep max tubeskin temp at less than max recommended tubeskin temp at max firing rate, then you may have thermal decomposition/ cracking of this HTF (or oxidation products) and this flashpoint value of 410degF may not be applicable. It is often next to impossible to confirm that (b) is actually seen in practice.

 

[USAeng]

Thanks. We had plan 11 on the first pump (Dean pump). The second MP HTO 120 pump - I am not sure - I just know it was recommended to us by one of our pump suppliers we trust and from the literature it seamed to be proper for the application. The plan 52 design looks expensive and seems a bit excessive for our application - however that is why we are hoping to hire a professional engineer to help us troubleshoot the system. The little secondary loops only are there to scavenge some exhaust heat- adding expensive pumps and seals will kill the savings idea. We do have much larger pumps in the primary loop that have clean fluid in a sump on the opposite side of the seal. Those pumps have never had an issue. I just have a hard time believing 2 manufactures of hot oil pumps both say the pumps we are trying are proven pumps for our application- yet they both leak. I just have a feeling something else is going on. Thanks for trying to help though. I’ll continue my google search.