Rotor Earth Fault Protection Alternative

I am sorry to say, that in many years of correspondence, this is the toughest question I encountered. With slip rings it is straightforward. Without a set of one, one has difficulty maintaining magnetization in a resistive control fashion. Measuring it relatively straightforward,be it wired or wireless. As a matter of fact, I prefer wireless for simplicity.

Logically, I modified the question to magnetization as a relevant question. An earth fault, so to speak, can result from any imbalance. Be it conductive or capacitive. A modern measuring technique has the "fault" of a higher frequency limit never contemplated of the original designer.

Looking to further conversation.

I have found the below product that would seeming allow for wireless rotor earth fault protection, however this is a comprenehsive device that does much more than I want to (I imagine that it would be quite expensive). Also, not sure what standard for protection this unit meets (ANSI etc).

http://accumetrix.com/demo/apps/Rotor%20Health%20Monitor%20data%20sheet%20Rev%205.pdf

Interesting.

And compared to the cost of a repair it also may be a bargain.

I guess you need to weigh in the likelyhood that this actually happens over the life of the generator package as well. PM programs could be put into pace to regularly check the health of the windings on the rotor.

Have you seen any statistics on when this sort of fault would typically occur on a machine if the load is pretty benign (DOL motors and lighting)?

Interestingly, mostly rudimentary.

Everything else is, an occasional Appendix.

I keep in mind, that not anybody has the latest version of a geegaw, causing all sorts of grief, if violating that "rule".

Can you plz let me know type of excitation system in your machine...??

The alternator in question is a Stamford P80 series HV alternator with class F windings, see link below:

https://www.cumminsgeneratortechnologies.com/en/products/stamford/P80/P80-High-Voltage/

A PMG is fitted to the machine and the MA330 AVR is used.

Here's another device "...The PRISMIC R10 is an electronic unit designed to provide a brushless generator with an alarm in the event of a rotor earth fault...", there's probably others out there that can be Googled. The real trick is to be sure that you don't violate the machine's warranty, as well as finding room on the rotor shaft.

If you find the cost prohibitive then you'll have to revert to checking the insulation resistance on a fixed interval to "catch" the first fault, but that needs to be balanced against the cost of an extended outage and the consequent loss of production/revenue.

Checking the excitation level into the brushless exciter won't tell you when the first fault occurs, by the time someone notices that the excitation current increase unexpectedly without a corresponding increase in terminal voltage/VARs will not give much warning about the second (damaging) fault unless it is very close to the first one.

Pay now or pay later.

Thanks for the link to the alternative product, I will make an enquiry with the supplier as to what standards the protection functions in the device are designed to. I have hope that it is a proven design though as it has DnV stamps on the brochure.

The space on the rotor shaft was my primary concern as well as the possible warranty issue on the machine.

While not directly measuring for the rotor earth fault, I would have thought that vibration monitoring on the bearings could have detected the effects (i.e. rotor imbalance and the bearing casing shaking more because of this). Have you or anyone else here ever seen the effect a second rotor earth fault, I am curious as to whether vibration monitoring would pick up the effects as well?

If yes, then with the vibration monitoring, a PM program would be easy for the machine as it is a standby unit and the winding health can be easily checked when offline.

I will also ask the manufacturer whether they have any mean time between failure (MTBF) data for the stator and rotor windings of this machine, this will help my management make a decision as to whether the rotor earth fault protection is worth the expense if it is quite high. Have you, or anybody else here, come across any typical MTBF data for machine rotor windings and can to give me a rough idea until the manufacturer responds?

From this technical paper: "... Shorted turns in the generator field winding can cause distorting the field across the air gap. This is due to the unsymmetrical turn current of magnetomotive force in different parts of the field winding. The accrued much distorted forces can effect rotor deformation or severe vibrations which can destroy machine bearings. The machine can be protected before serious damage by vibration detector which can alarm or trip unit. Shorted winding is often due to rotor ground faults at two different places..."

The trick is to set the alarm and trip levels correctly. You may get some false alarms and/or trips from other causes, just make checking the field integrity an absolute requirement before any restart.

Regarding failure rates, such data is very hard to find, you might try to get a copy of this book and see if it helps in your analyses, it does have some interesting ways of determining the "health" of the insulation system:

Electrical Insulation for Rotating Machines: Design, Evaluation, Aging ...

By Greg Stone, Edward A. Boulter, Ian Culbert, Hussein Dhirani

Thanks for the paper and the name of the reference book, I like to make the most of an opportunity to add to my ever growing library. I have read some of theory on insulation elsewhere (possibly MG Say's Performance and design of alternating current machine but I cannot remember right at this moment), but I do not have a reference text purely on machine insulation which will be nice.

Vibration levels would be easy to set, I can get live data with a transient load step and rejection test and a full load run. From there I can set warning and shutdown alarms above the worst case level. On other similar sized diesel generator packages, I have used 7.2 mm/s as a warning level and 10.8 mm/s as a shutdown level for accelerometers on the x and y axis of each bearing on the alternator and while they were close enough to the worst case vibration level, there was no intermittant warning alarms.

That paper also suggested loss of excitation which is part of the protection scheme already.

If the machine was continuously operating, then I would be pushing for the rotor earth fault protection, but as the machine would be lucky to run 1 or so hour per month during maintenance runs, the cost of retrofitting the protection would need to be justified further to the people with the money if current protections and PM programs could almost do the same job. Maybe if I can find some paricularly good photos or reports showing the effects of a second earth fault on the rotor windings...

Operates 12 hours a year??? Anything you do is overkill, unless you just happen to have a spare vibration monitor laying around.

Yeah, I know its over the top, but everything else for this site is too, so why not...

It would be surprising if the original equipment manufacturer of the <...2MVA generator...> were not able to answer these sorts of questions over the telephone.