Busbar Failure Between Excitation Transformer And Excitatation Board

pretty odd size for zucchini, and if the previous one was squash, I would not expect it to hold up to UHV either.

I see you are now using copper. That is a good start. You are having a voltage spike sometime, either through lightening or through a start-up transient where the field and the local generator is not synchronized properly, and thus you are sinking more power through the local bus bar than its rating.

Can you explain t to me how cqn i explain how can the voltage increase and how can i make sure of that

And mitsubishi in the first fault they told us that this fault due to bad design

You sound like you are out of your depth.

You need to get an Electrical Engineer that has experience with high power systems to look into it for you.

Already mitsubishi send two qualified engineer to collect data and analysis the problem but they didn't know the reason for this fault when i ask abut voltage spike i know what spike mean but i don't know the reason for it at the busbar

It sounds as though Mitsubishi has criticized its own design. Did they give any details, such as solid versus laminated bus bars, number and thickness of laminations, spacing of braces, etc? Large overcurrents could cause adjacent bus bars to attract/repel with enough force to crush the insulating material in the overall sandwich.

The mitsubishi bus bar each phase contain two busbar each of them 240 *8 so each phase cross section area 2*240*8 and each of them is solid and type compact and sandwich

Am sorry i mean what is the source of voltage spicks at ths location between the excitation transformer and excitation panels and how is the effect of voltage spicks on the bus bar

This is more likely to involve current spikes rather than voltage spikes. (Please note the spelling; your version is an ethnic slur, albeit unintentional.)

I thought this was on the field bus bar leading to the field transformer (grid). Now you are saying excitation, which is quite a different animal. What is the power feed to the exciter circuit? Is it 480V, 3 Φ, or higher voltage?

Maybe you have a dirt problem? Maybe the insulation is getting wet, who knows. Do you have electrical storms there on a frequent basis?

It's very difficult for any of us to troubleshoot your problem, given that none of us were there or can see what you can see.

What was happening or what was done just before the events happened? Was there anything in common between the two events? Maybe that's where to start.

Thanks eixter

The common between two accident the two fault was the load for the unit in each case the generation was 500mega watt

And for the first one the operators try to reduce the generation but the fault occured when he try to go down

The protection device is connected to the primary of excitation transformer which is connected delta delta and no protectiin device at thr bus bar which connected excitation transformer to excitation panels and the protectin divice trip at the forst over curent trip and when we see the trend we find that the beginng of protection signal is earth fault and after that three phace to ground fault

And the busbar was destoryed at the point of fixed pins totaly distroed and make three holes at copper conductor and in th second fault one phase to ground and the new busbar one phase destriyed and we found effect of sc in many points

So this is not the excitation circuit, but really is the field circuit. Thanks for clearing that up.

If you were simply decreasing the load on your generating unit from peak load, this usually takes places rather smoothly....however things to watch out for are the VARS, power factor, etc. while transitioning the load. You do not want to over current the generator by reducing real load on the system while holding VARS constant.

Without physically seeing the damage, all any of us on this end can do is suggest this or that as a while guess. We might as well be feeling of the elephant blindfolded, and having you expect us to say it is a leopard.

Send some pictures into your next response if you had, or have any.

Your description correlates with an insulation breakdown at the mounting point.

You didn't mention the operating voltage. What is it?

Is the excitation AC or DC?

If the transformer wasn't damaged, it is not likely a current over load unless it feeds many of these un-protected busbars. What type of protection are on the transformer and what is its voltage ratio?

Is it possible that the generator went out of sync and produced an over-voltage to ground in the excitation circuit with the busbars insulation being the weakest point? Any other collateral damages?

As others already mentioned, get an independent expert to look into this. This cannot be solved over the internet.

Operation voltage 840 voltage Ac voltage

Transformer turns ratio 24kv to 840v

Excitation panel input Ac 840 volt and output DC 600 volt

We do all electrical test on transformer and transformer in healthy condition

The protection is over current short circuit protection and earth fault protection

We have previous problem on generator synchoronization at grid and we change the rotor of transformer to change the inerthia because it absorb very big phase sequence in synchronization and mitsubishi change the rotor to solve the problem

The generator was in service but i don't know this phenomena of generator went out of synchronization

So plz marcot i need more illustration about phase to ground over voltage due to fault at synchronization

Really i know it is very difficult but internet and people like you will learn me more about the phenomena of this acction

Is the transformer secondary hard grounded WYE? or is it the field that is grounded?

Is the excitation panel thyristor or IGBT based? Does it have line fuses? What value?

If nothing is grounded, you have the potential for large voltage swing with respect to ground.

A theory is that the busbars's insulation may be the weakest link. Once an arc to ground develops, the transformer's primary protections are not likely to detect it and let large current flow resulting into the destruction of the busbars. The ~30:1 ratio doesn't help.

The cause for the over voltage can be multiple. Generator transient, excitation panel glitch...

If the xfo sec. is grounded, this theory is unlikely. Look for errors in voltage rating or bad installation procedures (mounting screws through the insulation, hardware dropped between busbars or case, mechanical stress that breaks the insulation material, mices chewing on the insulators, large cockroaches, water dripping on the assemblies...)

Mices or meeses? I thing this guy is seriously jerking our chains.

You have a good working theory about a short through the insulation, and due to the large excitation currents even at 600 VDC for a generator of this size, there is the likelihood of an unnoticed ferrous metal burr from drilling, etc. being attracted to the bus bar, then all bets are off the table.

I feel that the work is being done in a slack-jawed, and hurried way, without careful inspection of final assembly.

Pretty strong chance of getting movement of any metallic particles present when currents are so high.

I'm sorry for rate in replay

1-the transformer connection is delta delta so it not grounded

2-the excitation is thyristor based

3-yes contain line fuses high rapture fuse 650 volt

I get two pictures send to me your mail and i will send pictures to you

It is then very likely that you had some over-voltage event with respect to ground that destroyed the insulation. Fast ground leakage detection could be used to protect the next installation against fire but it will not prevent the insulation degradation. You need to figure out if a ground id necessary in the circuit.

Get a local expert to make sure that your installation and the operation procedures you are using are appropriate.

Sorry I'm late to this thread, but it is very strange that a piece of high current, 1kVdc buss bar from two different manufacturers would fail twice in the same circuit, especially after such a short period of operation. It is also highly suspect that the OEM wouldn't know why their own equipment failed to perform properly, so we must look at the terminations and the circuitry at both ends.

The rotor of a 650MW alternator is a very significant inductance capable of storing quite a bit of energy for a relatively long time, with a time constant measured in seconds. Under normal operation the circuitry on the output of the excitation system is built to withstand the sudden collapse of the field current and the reultant inductive "kick"; therefore, I suspect that there is an unexpected combination of events that is causing the problem.

1. How is your 4m bus bar braced, by what means, and at what spacing?

2. Exactly at what location is the insulation failing; at a termination, at a bracing point, on a particular polarity, polarity to ground, or polarity to polarity?

3. You said the rotor was previously rewound/repaired; exactly what caused the need for the rewinding, what was the nature of the repair, and most importantly how was it tested?

Without seeing any pictures of the damage (and I've seen many other failed excitation system/rotor field failures in person), my first guess is that you have an intermittent shorted turn or two in the rotor. As the field current rises and falls randomly, the buss bars have to deal with very large transient current spikes which, just like the ignition coil in a car, in turn cause very large transient voltage spikes far in excess of what the insulation can safely withstand. The current transitions may also be setting up unexpected vibrations in the buss bars, which are causing the insulation to crack and/or flay off, which would allow the transient voltages to puncture through.

Your answer to the earlier questions plus any TDR (Transient Data Recorder) graphs and pictures of the failure points, will go a long way to determining what is going on here.

Ram consult

I apprechiate very much your answe

The old busbar was destroyed at distance 4m the copper area beside fixition pins was destroied and this area was converted to holes in the basbar not in certain distance all 4meter

the The location of the fault in the second time zucchini busbar after the termination by 50cm beside the support which carry the busbar

The previous problem in the rotor was that when we connected the generator at the grid the rotor has big inertia so when we get out from service it will not stop and absorb very large amount of mvar from the grid and the seond problem occured it's end ring was destroyed due to thermal and mechanica stress and cause damage inside the generatord so they rewind some windings

Mitsubishi solve the problem by change the rotor and give us one additional new rotor and take the second one to japan and reduce the rotor weight

How it was tested i will check with my staff and i will replay fast possible as could

I hope to ask you some questions

First the new busbar worl around 10days before destroying so how turn to turn or turn shory not affect on it in this period

Second we have generator circuit breaker closed after the fault and the only the excitation transformer was connected to the generator

How can i make of sure that turn to turn short circuit or turn short is the reason and what type of trend will helpme in that

Thanks for your help

There are many things going on here, and they are not necessarily dependent upon each other. Machine inertia should have little to do with the damage to the rotor and the end rings. Machine inertia is the source of stabilizing real power output during transient fault conditions, and should play no part during the normal unloading prior to disconnecting of a machine from the system. The end ring damage that you describe is consistent with that which is found in severely underexcited machines and/or those machines that are operated with severe phase imbalance/high negative sequence currents.

A possible sequence of events is 1. the DC bus shorted, 2. this reduced/removed field current 3. which resulted in a huge inflow of reactive current into the machine as it tried to remained synchronized, 4. and none of the protective relaying operated, 5. forcing the operators to improperly shutdown the machine by reducing the steam flow to the turbine while the alternator was still connected to the system, 6. which in turn caused the machine to draw real power into the machine until 7. the reverse power, undervoltage, overcurrent, and possibly the underfrequency relays operated to finally disconnect the turbine-generator from the system.

A thorough review of the mounting means of the laminated buss bar, protective relay settings, and proper operation of the turbine controls is in order. I suspect that you will find that the laminated buss system requires very specialized hardware to secure it, and that no form of drilling is permitted to be done to it. You will also find that the T-G was improperly operated and that safe operating guidelines were not followed.

I realy thankful for your replay

First i call oem and know much information about rotor ring damage they told me that the mean reason is that the frequency of rotor equal to grid frequency so it cause rotor squeeze

And also tell him internal short in rotor as acause of bubar damage and they told me that they put fluxprob in generator and will detect this problem

Second we will use new bubar which can withstandmore than 35 kv

To put the unit in service and will change it with new oem busbar after two months

Untill now we don't know the reason for the bubar damage

What is your opinion about flux prob detection the fault

I really need your email and your time at forum because i need your opinion in this problem

Having read the "initial and following anonymous posts", I have some observations.

1. It seems a fault occurred where the generator was not automatically disconnected from Grid when a fault occurred but motored on - despite the turbine having been stopped. Unfortunately, I feel that the correct commissioning settings/tests and operational procedures cannot be assumed for equipment.
2. The protection mentioned for the excitation transformer 24kV/840V appears to be on the primary HV side. It will give only backup protection for the LV side and be set for the limits of the transformer (which has not been damaged, despite 2 LV busbar shorts). There are LV "660 volt" fuses, which protect the busbar/transformer against a fault in the excitation controller, estimated to blow in about 0.1 second for a short, but this is no protection against the effect of the 50,000 amp "arc welder" on busbars - let loose by a busbar short.
3. There is thyristor control - which would have "quick blow" semiconductor fuses.
4. Were any of these semiconductor fuses or the LV fuses mentioned in 2 above found blown after the busbar shorts??
5. Thyristors have a lower overvoltage withstand than the transformer or busbar and usually have overvoltage-limiting protection provided for them. This would also limit line-line voltages on the busbar. However, this would not control busbar to earth voltages if earthing mentioned in 6 below not OK.
6. The protection/earthing on the LV rotor/automatic voltage controller/busbar has not been described. It would be usual for the rotor to be earthed by a resistor from one pole to earth. A single earth fault in the rotor would then cause a limited DC current through that resistor and the voltage across resistor can operate protection. This earthing also limits rotor/busbar to earth voltages (the thyristors are connected from busbar to the rotor).
7. Has the earthing and protection of 6 above been checked and found OK??
8. Is the 4 metre long busbar on this generator identical length/type to that on other well-proven generators or is it just on this generator?? If different, in what way??
9. Busbars in this service should ideally be a "never fail" item designed to have adequate clearance without insulation and then be fully insulated to guard against the unexpected. They are usually enclosed by earthed metal but this very enclosure can be a problem. If totally enclosed and well sealed, damp air can be held-in and condense into a cloud when the ambient temperature drops - it is usually essential to have anti-condensation heaters working while the unit is stopped . Some busbars are virtually sealed and fed with dried air to prevent condensation.
10. What are the enclosure/ventilation arrangements of generator/transformer/busbars/excitation cubicle?? Is it possible poor sealing (even on access panels/doors) is allowing undesirable flows bringing dust or damp into busbar??
11. Busbars should obviously be safe against a leak in the building roof etc. Is it possible the busbar was not assembled properly at first construction [oil rig disaster due to blow-out prevention valve upside down comes to mind], due to time pressures or error - and rebuilt exactly the same after the first fault, perhaps with bolts too long - projecting sharp threads cause high volt/metre [aircraft window blew out at 30,000 feet after replacement - technician had carefully replaced all bolts with new (same type as removed) - but window had been replaced a long time before using short bolts]??
12. Did faults occur when weather abnormally hot/cold/humid/heavy rain??
13. It is probable the busbar assembly & insulation must be done on site. Was this done by the same contractor both times?? Was it Mitsubishi or a subcontractor??

Ram consultant

Whe have flux prob device to check the in balance in field and to dect turns short circuit and we don't find any alarm for that

I measure the thermal lmage for the tempreture and i find the maximum tempretuw is 70