Generator Protection

As the technician replaced the original AVR with <...a homemade one...> and the thing has been rewound any original equipment manufacturer's [OEM] warranty on the equipment is now null and void. The OEM can only be persuaded to help now by paying it for service and parts. Recommendations:

1. Open a dialogue with the OEM with a view to getting OEM's parts fitted, instead of <...a homemade one...>.
2. Have a representative of the OEM's organisation arrive to recommission the equipment using OEM parts.
3. Review the spare parts stock and assess whether spares of the OEM brand need to be stocked for this equipment.
4. Review the commissioning records for the equipment. Ensure that recommissioning records are added to the archive.
5. Review archiving procedures and ensure that equipment archives across the facility are being correctly managed.
6. Review the incident, the repair and the outcome of the reviews with the technicians that carry out repairs as a training exercise as, clearly, the facility's maintenance procedures are wanting attention. There is something that the facility's personnel can learn from this incident that will be of value going forwards. Review the maintenance procedures, particularly planned-preventative-maintenance, and ensure these procedures are up to date, carrying out further re-training as appropriate and at appropriate intervals.

Regardless of how mean and difficult to understand PWSlack's answers are, he is usually right about the solution of the problem.

In the first world you need to ensure reliability if you want to have enough revenue to pay for your spendings. Doing a home-made design and putting it into production is very likely to lead to unexpected failures.

Even if you had some type of protection, you would still get power cuts, and I doubt your customers will be happy about it.

However, if you still want to go your way, there are circuit breakers for you application, but you might have to spend more money on the circuit breaker than on the winding repair, so it would not be worth it.

You didn't really say why the winding burnt out, initially.

Was the generator being run at full speed for an extended period, without adequate ventilation perhaps ?

If you don't want the same thing to happen again, regardless of the reparation, then it would be prudent not to operate the unit under the same conditions.

More information on the design of the homemade voltage regulator might be helpful. I wonder if excess ripple in the excitation might cause extra direct heating in the generator stator laminations, so that the generator overcurrent protection (current sensors on the stator leads) might not be able to sense the extra heat.

Simple overcurrent protection should normally be adequate for a machine that size, which it should already have. Perhaps the protection is not operational, it can be tested before putting back in operation.

If the stator windings are getting abnormally hot, your exciter should require a slightly higher level of excitation than normal to deliver a given power. (I2R losses higher than expected due to higher copper temperature.)

Under what conditions do you run this generator? Parallel with other generation or as the sole supply of electricity? Grounded or ungrounded system

Balanced/unbalanced load?

I really have nothing further to offer, and it is just a guess at this point, but you should compare the rectification scheme of the original exciter to the home-made one, and perhaps that might shed some light?

i.e. full wave bridge,half wave bridge, perhaps more subtle differences,smoothing filters?

The exciter should not be able to heat the stator except indirectly, and the design current for the exciter, if exceeded, should cause other troubles. Over-excitation can cause excessive current flow in the stator, but your stator overcurrent protection should protect the windings, even the most basic protection will protect for that.

I would look first to confirm that your overcurrent protection is set properly according to the manufacturer's instruction, and then confirm that it works. Primary injection test method is warranted considering your history.