I got this "Maintenance-Tips from Reliabilityweb.com" by email:
quote:
Motor Testing Tip When evaluating test equipment, those that incorporate a multiple technique, or “Swiss army knife” approach, tend to be able to diagnose a wider variety of motor and winding faults. Just like a toolbox.
For example:
Many motor repair facilities have Hi-pot testers. In some cases these Hi-pot testers are older than the technician operating them. Let’s face it, using a modern GPS to navigate is accepted as easier than using an antique sextant and the North Star.
The same situation applies with test equipment. There have been many test techniques implemented into modern equipment that allow quick, easy, and non-destructive diagnosis of a wide variety of motor faults.
One of these new, non-destructive [emphasis added] motor testing techniques is the DC Step Voltage test.
This technique uses modern, microprocessor control to automatically plot the leakage current at standard intervals. This automatic control allows reliable, repeatable test results to be obtained. 15 or more years ago, to attempt this test required a technician to have great patience, concentration, (and a stop-watch) to time 15 or more 1-3 minute steps!
Test equipment that integrates this technique into one box with other proven methods, such as the Meg-Ohm, Kelvin Bridge, Continuous Ramp and Surge test, will help you keep your motors running!
I would agree that dc step voltage test reduces the risk compared to hi-pot, and microprocessor test reduces the risk compared to manual dc step voltage test.... because the operator or microprocessor MAY see a non-linear trend indicating impeding failure and remove voltage before breakdown voltage is reached.
But I have never before heard anyone suggest that this is a fail-safe approach which renders the technique non-destructive. We performed a dc step voltage test on a 13.8kv machine which was linear from 5kv to 23kvdc, and then failed at 24kvdc. No abnormal currents were noticed... the first thing noticed was that the test set breaker tripped. Motor had to be rewound.
Granted, dc step voltage test is better than simple hi-pot (without plotting current vs voltage) and microprocessor step/ramp test may be better than manual test. But I think the underlying failures are too varied and unpredictable to claim this test is not destructive. There will be always be risk when you apply a higher than nameplate voltage to your machine and you'd better recognize that before you do the test and have contingency plans in place for possible failure (spare motor or time to rewind the tested motor).
Good catch Pete, I guess the sales teams are writing the maintenance tips here. I read this and didn’t give it a second glance, but I’m not in the market for another blade for my Swiss Army knife.
Posts: 5 | Location: USA | Registered: 16 November 2006
EP, The reasoning behind the test is that it needs to stress the insulation system slightly higher than nameplate, due to the fact that the windings will be subjected to 8 -10 times the nameplate rating in current on startup. EPRI has also performed studies that show starting across the line can also (with slightly out-of-phase voltage) stress the insulation system to 4 to 5 times the nameplate rating. So, as a user, would you want to know of a marginal insulation system during a testing outage or just after you put the 'pedal to the metal' on a plant startup?
Posts: 204 | Location: Philadelphia,PA | Registered: 18 July 2006
I can draw from this that maybe PdM inspections of motors should actually NOT be done during production or brief interruptions BUT rather during a lull in activity when a motor damaged during testing can then subsequently be removed for overhaul.
Posts: 30 | Location: Canada | Registered: 19 January 2008
If the pdm inspection includes a dc step voltage test or a surge test then, I agree 100%. We do those tests, but only when a spare motor is available and we have enough time available to cope with a failure during the test.
