250 KVA Hydro Generator - Pick apart the design 4

[mhulbert]

I have a client that has a small hydro generating setup. It's about 30 years old and experiencing some problems. I think there are a few areas of the design that may need fixing. Some of this is outside what I deal with typically, and I would like you guys and gals to give me your opinion.

The system runs parallel to the grid. All of the water side items seem to work correctly, so I will focus on the power side. The generator is a 250kw 3 phase induction motor, labeled for generator use (by GE). It is connected for 480V. This goes to a contactor to connect it to the line, fuses, current monitoring, in-phase/in-voltage monitor, revenue metering, and then out to a transformer. There is also a capacitor that is brought on or off line for PFC (I think). Electrically speaking, the generator is hooked up just as you would a DOL motor starter, except with fuses instead of OL's.

The 480V heads to a pad-mount transformer(300 KVA, 12470Y/7200-480 Delta, 1.9% impedance), and here is where things get a little uncertain for me. 480V goes in on the delta side of the transformer (X1, X2, X3). 12kv comes out of the Wye side (H1, H2, H3). H0 is tied to the building/site ground through a resistor and 12KV/120V single phase transformer arrangement. The 120V winding feeds a voltage relay which I believe opens the contactor when voltage goes above a threshold. The 12kv lines go underground 300', and are connected through fused cutouts (with lightning Arrestors) to the 12kV grid.

So, what problems have we had-
-Hydro plant is running, owner is present fine tuning controls
- A 480V lightning arrestor shorted out. It was unfused so a lot of current and sparks flowing. The owner was present and disconnected the system at a 480V disconnect just before the transformer. The lightning arrestor was connected on the line side, so it kept burning up! The rest of the 480V equipment is disconnected at this point
-Transformer starts thumping, lightning arrestor still burning up
-After about 5 minutes, one of the fused cutouts blows out, everything stops.
-Upon removing the lightning arrestor, and refusing the cutout...fuse blows immediately
-Not sure on a plan of action at this point. I have meggered all 480V wiring and equipment, tested fine

After reading all this, and thanks to anybody who got this far- what problems do you see with this setup? I am concerned that there is no grounding of the 480V side of this system so L-G voltages could become elevated. That may be why the 480V lightning arrestor shorted out. Could this have triggered a fault in the transformer? Would a transformer fault only take out one fuse on the 12kV side? Is this how these are usually grounded?

We are thinking of having an oil analysis done on the transformer, but are hesitant to bring out a firm to test the underground cabling as it is $3500+ My gut says the transformer is faulting to ground, but I find it odd that it did it exactly when the 480V lightning arrestor faulted out too.

Any thoughts on this are incredibly appreciated. I have pictures and can answer questions if you have any.
Thank you Mike

 

[davidbeach]

Transformer probably damaged by the arrester failure. Arrester on the 480V system should be rated 500V or more to deal with the issues of an ungrounded system. One conducting arrester wouldn't carry any current on an ungrounded system, so something else took one of the other 480V phases to ground and you had a phase-phase fault through ground and the arrester. Could have been a phase-ground fault in the transformer, so maybe both transformer and arrester failed together.

 

[rhatcher]

I agree with davidbeach.

 

[waross]

I suspect a transformer fault in a primary winding that raised the common mode vboltage of the secondary to ground. As the fault progressed, it caused the thumping and eventually took out the primary fuse.

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

 

[mhulbert]

Thank you for the insight guys, I had not thought of the lightning arrestor creating a phase to phase fault. I have not worked with many ungrounded systems, and so I take ground for granted.

Why would you build an ungrounded system like this, and what would be the preferred way to establish a ground on the 480V side?

Waross, I'm not sure I fully understand your description. Are you saying that the 12KV winding faulted, and this raised the 480V side in relation to ground? Can you elaborate on common mode voltage?

It sounds like we will have to replace or rebuild the transformer. The unit is 30+ years old and I don't believe it has ever had any oil analysis or testing done to it.

Keep the comments coming, I am curious to hear more.
Thanks,

Mike

 

[wolf39]

Mike:

Sometimes problems are being placed in this forum with little or no relevant information, making it difficult if not impossible to deal with. Your thread is a shining example how a particular problem should be described. This star is well deserved.

Wolf

http://www.hydropower-consult.com

 

[rasevskii]

for mhulbert:

Was the lightning arrestor connected phase to ground in the conventional way or phase to phase? If it was phase to ground then there had to be another ground somewhere on the 480V side to have had it burn up. It can be that the transformer had a ground fault on the 480 side, inasmuch as you said that upon opening the contactor and disconnecting the generator, it continued to burn.

As davidbeach said, the transformer is likely faulted and has to be replaced. The primary fuses on the pole were likely oversized, catering only for a dead short in the cable or transformer and not as any sort of conventional overcurrent protection.

Now you should megger the transformer primary and secondary after disconnection to verify that it is faulkted or ground faulted.

On the generator: is there a neutral nad if so is it connected to anything?

rasevskii

 

[mhulbert]

Thanks for the additional help everybody.

I'm not sure how the 480V lightning arrestor was connected, as all that remains of it is a pile of debris and a scorch mark on the concrete wall. I will ask the owner to see if he remembers. I'm not quite sure why it was even installed, and putting it on the line side of the main disconnect, without fuses, was certainly the wrong way to do it.

I'm sure the fuses are for cable protection only, as they are provided by and coordinated with the utility. Ideally this setup should have fuses at the high side of the transformer as an underground cable fault would just burn up if the generator was online. There is some sort of ground fault protection on the 12KV side of the transformer but I'm not sure if it is appropriate for this setup.

The generator has no ground, as it's essentially a 3 phase induction motor being run by the hydro turbine via gearbox. So we have a delta connected motor feeding into the delta side of a transformer. This seems wrong to me but I don't have a good reason!

Wolf- I enjoyed looking at your website! You are dealing with some huge facilities, makes my project look like a toy.

Thank you for all of the help, I will update this thread if we find anything else

 

[rasevskii]

Well, if the generator was online, and a HV cable fault occurred, taking out the pole-top fuses, there would be no voltage from the induction generator, as it can only excite itself from the grid.

In this case an undervoltage and/or overspeed trip has to open the 480V contactor, and shut down the turbine (close the wicket gates)
allowing the unit to come to a safe stop.

There should also be a loss of phase protection included in case only one pole-top fuse were to fail.

Let us hope this is implemented in the control scheme.

rasevskii

 

[wolf39]

Mike:

I don't know whether this important for your problem, but large synchronous hydro generators usually are star connected, feeding into delta winding transformer windings.

Wolf

http://www.hydropower-consult.com

 

[mhulbert]

Wolf,
that is what I thought. I am very much comfortable with a star arrangement. This generator has only 3 leads so we are stuck with what we have. It seems like there should be some sort of grounding scheme on the 480V side, perhaps a zig-zag transformer.

Rasevski, I believe there is phase loss and undervoltage protection. The control cabinet has seen a lot of modifications over the years, but it looks like the required relays are all still working. I had not thought about losing excitation voltage from the grid. Luckily the whole thing fails closed. There are a pair of NO solenoid valves that open on power loss, moving a hydraulic cylinder which diverts the water to the outlet.

Thank you for all the insight into this setup, this site is a great resource.
Mike

 

[waross]

What I am trying to say is that rather than a failure to ground, I suspect that a primary winding faulted to a secondary winding. That would explain the arrester arcing until a primary fuse cleared. It may also explain the noises coming from the transformer.
If you want to ground the 480 volt system consider a wye:delta transformer bank rather than a zig-zag.
Decide on the value of your NGR and the resulting ground current.
For a ground current of 5 amps, select a transformer that will carry 5 amps in the primary winding. That is 5 Amps times 480 volts or 2400 VA. Use three 480 Volt dry type transformers rated at over 2.4 KVA.
The secondary voltage is not important.
Connect the 480 Volt windings in wye and connect the wye point to ground through your neutral grounding resistor. Connect the secondaries in delta. The delta winding will stabilize the wye point of the primary. This is a conservative design as the transformers see only 277 Volts on the 480 Volt windings. The current in the delta is proportional to the current in the NGR.
This is an off the shelf solution rather than a special design and build zig-zag transformer.

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

 

[mhulbert]

Bill,
I will check for connection between primary/secondary with a megger. We meggered everything at 1000V (obviously not that helpful for the 12 KV side, but a good sanity check on the 480 side)., but I did not check from X1--H1, etc.

That is an interesting solution to grounding the system, I like it. So the secondary winding are connected in Delta only, with no other connections to them? Interesting

Will have to float this by the owner once we get things up again. Have to find an affordable 250kva transformer first...

 

[waross]

Yes no connection to the delta. There is a variation where a resistor inserted in one corner of the delta will limit the current to ground through the primary wye point. You may also use the wye:delta scheme without the NGR. The wye point is solidly grounded to provide an artificial neutral. The transformers must be large enough to carry the maximum unbalanced single phase current. Solidly grounded wye:delta banks should be evaluated to ensure that they have sufficient capacity to withstand fault related currents until the protection clears a fault.
I sugest you stay with the delta on the generator side of the GSU transformer. The delta winding often does a better job of distributing single phase loads and unbalanced loads on the generator but this may be moot if you have a delta generator winding.


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

 

[edison123]

I would recommend the following tests before you trash the trafo.

1. HV to Ground IR

2. LV to Ground IR

3. HV to LV IR

4. Surge comparison test of all 3 HV phases

5. Surge comparison test of all 3 LV phases

Muthu

http://www.edison.co.in

 

[electricpete]

I don't work with transformers much any more, but I haven't heard of "surge comparison test" for transformers. One problem you would face is that the phases will probably not be symmetric to begin with because the center leg phase has a much different magnetic environment. This could in theory be overcome by comparing a phase against itself with varying voltage... change in waveform with increase in voltage indicates breakdown.

I think more standard tests for transformer would include Transformer turns ratio test which can be fairly sensitive indicator of shorted turn.



=====================================
(2B)+(2B)' ?

 

[mhulbert]

Any good write-ups on a turns ratio test? We are trying to avoid paying a testing company $3000++ to come out tell us we have a bad transformer. If we can ascertain that it is bad, we'd rather put the money towards a rebuild or replacement. Given that it has never had an oil or any other test done, I can imagine the internal condition as it is.

We are trying to get the local utility to close the cutouts with the transformer disconnected from the 12KV line to make sure we don't have a bad cable on top of this. With a 1000V DC megger, one phase measured 3.3 G-Ohm to ground, while the other 2 were over 11 G-Ohm (the range of my meter).

Wish I had access to more test instruments on this one.

 

[DTR2011]

Turns Ratio Test-

#1 I would recommend renting a small single phase unit. It can be performed without a dedicated test set, however if you are considering replace/repair of the xfmr, better to have dedicated equipment.

#2 - Completely isolate transformer from system. Apply voltage (10,40 or 100V) from H1-H0 to corresponding phase on LV side. (look for the parallel lines (H1-H0 / X1-X2) for secondary connection. The connections can be easily determined from the vector diagram on the transformer nameplate. The calculated ratio will be (12470/1.732)/480, assuming you are in the middle tap position. Repeat for the other 2 phases. Your measured ratio vs. your calculated ratio should match within 0.5%. A decent TTR should also measure exciting current. Compare phases. You should see a pattern of 2 highs and one low.

Another test to consider is a low voltage backfeed of the unit, using a variac (0-150V). Again follow all safety precautions. Compare currents (connect ammeter on the neutral (grounded) side of the variac output). You will be backfeeding the transformer, at 1/4 normal voltage (120v output). I have used this technique when there is a suspected transformer is involved and the budget is tight. If the variac trips, or shows excessively high current this is a very good indication of a failed transformer.

Good luck.

 

[edison123]

epete

I routinely do surge comparison tests on test trafos and have not seen magnetic flux playing any role in the results. I have attached two reports One for HV tested at 12 KV & one for LV tested at 1 KV) on a 11 KV/415 V, 1.6 MVA recently rewound by us. You can see no difference in the waveforms in all 3 phases in both HV & LV. The magnetic flux plays no role in surge testing since the current taken is so low that the flux is nowhere (since amp-turns is so low) near the rated levels.

mhulbert

Your IR readings are confusing

"With a 1000V DC megger, one phase measured 3.3 G-Ohm to ground, while the other 2 were over 11 G-Ohm (the range of my meter"

If these readings are about the same winding (HV), then you can't have different readings for each phase of the same windings (unless you separated the starts and ends of each phase or the each of these windings got cut inside, in which case you will read an insulation resistance value between the phases)

Muthu
www.edison.co.in

 

[edison123]

Apparently, only one attachment sticks to your post.

Here is HV winding surge test photos.

Muthu
www.edison.co.in