Continue to Site

Eng-Tips is the largest engineering community on the Internet

Intelligent Work Forums for Engineering Professionals

  • Congratulations IDS on being selected by the Eng-Tips community for having the most helpful posts in the forums last week. Way to Go!

leak check of water coils in bearing oil reservoir 5

Status
Not open for further replies.

electricpete

Electrical
May 4, 2001
16,774
We have several large vertical motors with Kingsbury thrust bearings on top where the oil reservoir is cooled by water cooling coils submersed in the oil.

We recently got bit by a very slow water leak which appeared immediately after the motor was overhauled... even though our spec requires a leak check. The method used by repair shop for leak check was as follows:
fill cooling coil with water > 95% volume.
pressurize remaining space with 90 psig air.
Close air valve, remove air supply, wait for pressure drop over 10 minutes or so.

During troubleshooting of the coil we did our own air pressure drop test. This one completely full of air... for 30 minutes had zero pressure drop. BUT, the problem did show as a very small leak by snooping.... suggests our test (completely air filled) was not sensitive enough.

I suspect that even though the shop test was more sensitive (smaller volume of air), it may not have been sensitive enough for this small leak (they didn't snoop).

We are looking for other methods that are effective in finding small leaks, don't require unreasonable effort, and don't introduce any other problems.
Other methods under consideration:

- is there a way to do a water pressure drop with no air? Fill with water and then pressurize with tiny-capacity hand-driven positive displacement pump?

- air with snoop I don't think is good method. Relies too much on being able to get to all the possible leak sites. And wondering... how bad is it if snoop is not fully removed from the reservoir.

- how about air in the coils while cavity is filled with water or oil... look for bubbles.

I am interested to hear what other people find to be a good leak test to specify for this type of application (doesn't require unreasonable effort, and don't introduce any other problems, doesn't rely on perserverance/judgement of the testers)
 
Replies continue below

Recommended for you

by the way, cooling water system pressure in this case was approx 80 psig
 
Two things come to mind- Compressed air and an ultrasonic leak tester, clean & leaves no residue but has a minimum leak-rate for reliable detection, or R314A refrigerant and detector, extremely sensitive but requires appropriate safety precautions.
 
When you say you are filling the coil with 95% water and remaining 5% with air, I couldn't understand how you are pushing air through water to the entire length of the coil. Or am I missing something? What is the need to fill the coil with water?

You can do hydraulic test with a hand operated pump but this requires higher pressure to force the water through the leak and you have to check the design coil pressure than what you require for compressed air (because of high density).

Dipping the coil in water and checking for air bubbles seems to be ok, but it is a qualitative test.

Why can't you test only with compressed air as suggested by RAMConsult? You can also do a vacuum drop test if you have a vacuum system near by. These two tests can quantify the leak rate (and hole diameter) if you know the coil volume.

Regards,


Believe it or not : A Neutrino is so mass less(and electrically neutral) that, a conservative estimate states, it can pass through a 50 light year thick lead block.
 
RAMConsult's tip is the way the industry is going today.
There are several makes of ultrasonic leak detectors (ULD). I used one, but am afraid of giving the name so the posting is not red-flagged... ;-)
The ULD is just a hyper-sensitive microphone so sensitive and directional that i could hear a the noise made by the rubbing of the thumb and index finger of a guy 60ft away!!!
We used the ULD to:
Find leaks in piping systems, detect sparking in motors and switchgear with the doors closed, detect problems inside transformer (bubbling - i.e. gases being generated). Problems inside Generators - etc.
It can also be used during the greasing of motor bearings the change in noise pattern indicates when the bearing has got just enough grease.

In my opinion the ULD is the way to go... and matches your specs.
HTH



saludos.
a.
 
The method your shop used ( 95% water and 5% air ) may give does not account for some of the air disolving in the water.
It's usual to completly vent air out of systems before a hydraulic test. If there is air in the system it will expand and maintain almost the same pressure with little pressure drop.
Hydrotesting with air and partial air is potentially dangerous ( depending on the system volume ) because of the stored energy.
I recomend picking a few brains over at the piping forum.
 
Suggestion: It appears that it would be better to use one medium, either water or air, not both mixed, since an air pocked may prevent the water leak to be detected within 10 minutes.
 
Questions: Is any industry standard followed?
If not who created the procedures?
 
pete,

Since liquids are not as readily compressible as gases, we use hydraulic test at about 1.5 to 2.0 times rated pressure to check generator and motor air-to-water coolers. This hydraulic pressure test usually lasts about two to three hours during which we see large pressure drops in the presence of leaks. It is ofcourse essential to drive out all air inside before closing the water outlet valve and pressurizing the cooler. A positive displacement hand pump is used for this test.
 
Status
Not open for further replies.

Part and Inventory Search

Sponsor