any NDT method to find heat exchanger tube leakage? 2

[cloonypan]

hello,
Fin-tube Heat exchanger internal tube leakage problem was encountered on a large electrical motor. This problem is not easily found exact leakage position when lots of half inch tubes inside the exchanger especially tiny leakage. We use a water pump to check every tube to see if water pressure can be hold or not. However this method cost much time and patient. Therefore I wonder if there any NDT instrument (ex: Ultrasonic) can be used to find out exact tube damage location. Thank you.

 

[mjjmecheng]

Unfortunately, I am not aware of any diagnostic tools that are specifically made for identifying the precise location of a leak within a fin-tube heat exchanger (particularly when there is a large array of tubes). One thought that comes to mind, aside from the mentioned pressure test, would be to use a thermocouple and measure the inlet/outlet of each pipe. It would seem that the leaking tube would have a different delta-T than the others in the array. This might help to at least isolate the search to a specific tube.

 

[racookpe1978]

Doesn't seem even the thermocouple method would work, since every different area of the motor would have very slight (or very major!) differences in heat generation. It would be almost impossible to get a baseline, so you would know when a change from the baseline occurred due to a plugged line, a leak, a un-plugged line, a change in motor load.

And you'd need to run the test at load, shut down, move the thermocouple, re-fill the coolant, re-run the test, shutdown , drain the coolant, move the thermocouple ....

Consider going the other way. Freon detectors are very sensitive to low levels of gas. Can you get to a area where all of the coolant tubes are accessible? Plug all of the outlets with the rubber stoppers you used for the hydrotest.

Put a slight Freon pressure on the other side of the tube bank. Sniff for very small traces of the leaking Freon gas, rather than for the very infrequent and necessarily large drips of leaking liquid water.

 

[macmet]

Have you tried acoutic testing? I have used this. Efore on boilers that were slightky pressurized with compressed air( < 50 psi). Be careful not to use too much air pressure as it can be dangerous.

 

[macmet]

This is the sort of thing I meant....

http://www.uesystems.eu/our-instruments/ultraprobe-9000/

 

[crshears]

Put phosphorescent dye in the cooling water and pressurize it slightly, then use an ultraviolet light to search for the leak.

If the tubes are horizontal and the gas side is vertical, even though you may not immediately pinpoint the leak you should at least be able to localize it to one vertically oriented cone or frustrum, at which point endoscopic inspection should precisely locate the leak.

If the tubes are vertically oriented, the coolant should track down only the surface of the leaking tube due to surface tension [spray of coolant onto an adjacent tube should be unlikely if the leak is in fact 'tiny'] and form a nice glow-in-the-dark ring around the defective tube.

CR

 

[davefitz]

It is my understanding that, after draining the HX, one can insert an ultrasonic noise generator into a header and then use a ultrasonic micrphone/detector to listen to where the leak is. This method has been used on large surface condensers and will probably work on smaller HX's.

"Whom the gods would destroy, they first make mad "

 

[StoneCold]

We have an ultrasonic leak detector. So you pressure up the coils with gas and listen for leaks. It works, but it is easier to pressure up with air and pour soapy water over the thing to find leaks. It is crude but just about a guarantee that you can find the leak if it is big enough. The cooling water with dye in it works better than the bubble method if you have really small leaks or the system is hot and you can't keep the soapy water on it.
My two cents.

Regards
StoneCold