Alternator rotor earth fault 1

[SA07]

Hi
Our alternator 11kV 60 MVA rotor has a rotor earth fault. The alternator exports energy on the national grid through a step-up transformer. There was lightning and a 3 phase fault on the grid. The alternator islanded but tripped after a few seconds due to high vibrations. Vibrations has rose to 349 micron. The relay rotor earth fault was activated. When we megger the rotor the insulation is low.
The stator, rectifying diodes, excitation winding have been checked and are good. We have dismantled the rotor. There is no sign of burnt winding.

Recording of grid fault in relay SEL 311 is attached.

What might have caused this rotor earth fault? Thks

Regards

 

[ScottyUK]

So the steel retaining rings have been removed prior to the photograph being taken? I can't imagine anyone building a rotor of that size and not needing retaining rings in the design, even on a four-pole machine where the rotational forces are much lower than on a two-pole design.

 

[edison123]

Scotty - I am guessing the diameter of rotor around 1.3 m. At that dia and speed and power, it has to have steel retaining rings. Insulation bindings not gonna cut it. OP has to post more info to help.

Muthu

http://www.edison.co.in

 

[ScottyUK]

Well if you're right then that makes it nearly twice the diameter I'm guessing. I reckon about 800mm. Still way too big to avoid the use of steel rings, at least in the machine designs I'm familiar with.

If there's a shorted turn caused by winding movement due to centrifugal forces or thermal expansion then it may not be apparent until the rotor is at operating speed and temperature, with the fault going open-circuit at standstill and/or as the metals contract back to their rest positions. That will make fault location more awkward to say the least.

 

[waross]

An anecdote to illustrate a problem evaluating an intermittent earth fault.
This was the exciter on a very small machine but the principles of differential expansion and/or centrifugal forces are the same.
We had an exciter on a small machine that would fail once the machine came up to operating temperature.
We took the rotor into the shop.
It tested good.
The shop heated the rotor in an oven.
When hot, it tested good.
The shop was hesitant to rewind a rotor that tested good and suggested that the problem may be with the field or exciter.
We installed the rotor and ran it.
It failed again.
I did a brush null test.
This is applying AC to the stator winding and measuring the AC voltage appearing at the brushes.
The brushes are then shifted until the null point (zero voltage) is found. That is the best starting setting for the brushes.
We found that the voltage varied as we rotated the rotor. This is a clear indication that there is a fault causing the windings to be asymmetrical.
We took the rotor back to the shop and described the test and the findings.
The shop agreed that the test did show a fault in the winding.
The oven heated the rotor evenly.
When the rotor was in service the heat was generated in the windings and conducted to the core.
This caused heat expansion of the windings relative to the core.
The differential expansion happened only in service, not in the oven.

I don't know if this test will be feasible in your situation, but I hope that the information will be of use.

Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[ScottyUK]

I can well imagine it Bill. The difference in expansion coefficients for copper and steel is about 30% or so, plus the copper is normally the hotter of the two.

It will require a fairly big DC source to bring that winding up to operating temperature, and even if such a test was set up there would be second-order effects such as a thermal bend caused by heat rising to the top of the rotor forging causing it to expand more than the bottom, in turn causing the rotor to develop a slightly arched shape which will introduce additional stresses not seen in normal operation.

 

[waross]

I realize that spinning a machine that big for a test may be unrealistic.
How about installing the rotor and putting a suitable voltage on the stator field?
The rotor could then be shorted to bring the wire temp up quickly. Remove the short and measure the induced voltage over one revolution.
Do you think that we could get the windings hot enough, quick enough to avoid too much temperature rise in the iron?
Turning gear would be an asset.


Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[edison123]

Bill

Your test seems to be for DC machines (which we normally use to check interpole polarity without dismantling the machine). Don't see how it will work in this case.

Muthu

http://www.edison.co.in

 

[waross]

Yes. I agree. I forgot that you need a commutator. It may work to develop the differential heating.
The voltage applied to the stator would be much less than rated voltage.

Bill
--------------------
"Why not the best?"
Jimmy Carter