DER exceeding transformer load

[davidbeach]

Our distribution system is a pretty normal North American system where the transmission/sub-transmission system feeds the delta winding of the delta-wye transformer(s) in the substations and distribution feeders extend outward as radial load serving, multi-grounded neutral, feeders. Just like they've been for 125+ years.

Now we're facing the situation at some of our rural substations of having enough PV installed that the distribution system could become a source to the transmission system. Obviously any fuse protected transformers will need to get relays and circuit switchers. What's got me a bit puzzled is what to do about lack of a ground reference for the transmission. If I could start from scratch, I'd put wye-delta-wye transformers in the substations but that's impossible now as we have to maintain the existing phasing of that load can be moved seamlessly between substations.

Has anybody else faced this? I can't imagine we're the first. What do you do? Add a grounding transformer? Just change out the arrestors? Ignore it? Something else? Thanks.

 

[waross]

Do you require the PV installations to disconnect on the loss of the grid?

Obviously any fuse protected transformers will need to get relays and circuit switchers.

Will adding wye:broken delta PTs give you the input that you need for the relays?
Obviously this wouldn't work with the original gravity drop re-closers of a century ago and probably not with the second generation electro-mechanical re-closers, but:
Can you send a trip signal from the relay to an electronically controlled re-closer and use the re-closer as a "Poor Man's" breaker?

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

 

[davidbeach]

Perhaps I was not as clear as I might have been - I'm looking at the impacts to the transmission system.

For instance, consider a fuse protected 57/13kV 7.5MVA transformer, delta on the 57kV, grounded wye on the 13kV and feeding a couple of 13kV distribution feeders. The station is tapped from the transmission line and does not have line breakers or line protection. During the spring and fall the load on that transformer might be 2.5MVA. Somebody comes along and submits an interconnection request for a 5MW PV installation one of the feeders. What we need to do on the distribution system seems well enough understood. It is expected that the anti-islanding protection built into the inverters will be slower than normal when the amount of generation is significant in relation to the amount of load.

Where it gets complicated is on the high-side. Obviously the 57kV fuses have to go, replaced with a circuit switcher and relays. But there's the issue of that delta winding feeding a ground fault on the 57kV system for the time between the normal transmission sources tripping and the inverters shutting down. Now we have an ungrounded source connected to a system designed around all sources being effectively grounded.

The lightning arrestors on the transformer are rated for line-ground voltages but would be exposed to line-line voltages. The line insulators are also sized for line-ground voltage.

 

[Marmite]

It sounds like you need neutral voltage displacement protection on the 57kV side of the transformer. How would you currently detect a wire down on the 57kV line which is backfed from the transformer via the other two phases?

Regards
Marmite

 

[jghrist]

I don't know why the fuses would have to be replaced with relays. There will be phase and ground fault sources on the 57 kV system that will not change with the addition of the PV on the 13 kV system. Fault current through the fuses for a transformer fault will not change.

If there is a ground fault on the 57 kV system, there will be no contribution from the PV inverters and the normal 57 kV protection will clear the fault. Three phase faults on the 57 kV system will have some contribution from the PV inverters, but it will be inconsequential. See

http://quanta-technology.com/sites/default/files/doc-files/Solar PV Inverter formatted.pdf

for a study of fault contributions from distributed PV installations. There could be a fault contribution up to 1.2 times the PV inverter ratings. The inverters will trip if nothing else clears the fault.

There will be some inverter contribution to a transformer fault, but I don't think it would increase damage to the faulted transformer significantly if it is not cleared before the inverters trip.

 

[waross]

I believe that the concern is what happens AFTER the 57 kV clears and the iverters have not yet cleared. When the grid clears, the ground reference may be lost. The system will become a corner grounded delta and the voltage to ground on two thirds of the aresters and insulators will rise above their rating.
Is this correct, David.


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

 

[davidbeach]

Yep. Getting rid of the fuses will, among other things, provide differential and tripping of both sides of the transformer for any transformer faults.

 

[krisys]

Simultaneous tripping of primary and the secondary side of the breaker is one of the feasible option.
In spite of that, the capability of surge arrestors need to be ascertained.

In the event of such faults, it is not a voltage surge. It will be a over voltage. Sometimes, the surge arrestors can explode for the sustained over voltage.

 

[stevenal]

Davidbeach. How are you dealing with voltage regulation when the power flow is reversed?

The arresters should have a TOV curve you can compare to the expected clearing time of the inverter. Polymer housed arresters will not fragment upon failure the way porcelain will.
Have you considered transfer tripping the PV? Or using a PMU system to detect slipping between the systems and trip the PV if it occurs?

 

[davidbeach]

Thanks Steve. Regulation is being looked at by the Distribution Planners. PV will be transfer tripped; it's easy to know when to transfer trip it for feeder faults, not so easy to know when we need to transfer trip it for transmission faults. We've been batting around the idea of sending transfer trip to the tapped transformer(s) involved so that they trip every time the line trips and then close back in once the line holds. Not what the system was designed for...