Fault level limiting in 11kV distribution

[Bung]

I'm interested in hearing about how people out there deal with the problem of controllong fault levels in 11kV distribution substations. Our nominal maximum (design) fault level is 13kA (250MVA). Recent plans to augment some substations by adding another transformer means that the fault level will go above this (and beyond the CB capacity) unless we take action to limit it. I'm sure our frineds in the petro-chem and other heavy industries would know all about this problem!

Methods I have come across include:
1] reactors in bus-sections between transformers
2] high Z trnasformers ( I can think of lots of downsides to this one, and few upsides).
3] electronically controlled explosive powered "fault current limiters
4] saturable reactors, traditional magnetic amps or superconducting magnetic amps.
5] running split bus and putting up with short duration interruptions for transformer faults (I like this one -KISS)

Any thoughts out there?

Bung

 

[bigamp]

Is changing out switchgear an option for you?

I was on a job recently where the source of supply for an existing installation was changed. The new source could provide significantly greater fault current than the downstream switchgear could handle so reactors were added. These were installed outdors, and took up some space. They were also a bit of a challange to cable up to but ended up OK.

High Z transformers, I agree, are a marginal solution.

I have had no experience with the other methods you describe.

Regards

 

[Bung]

The between - bus reactors are a real challenge on indoor metalclad switchgear when done as a retro-fit!

The costs of change-out of switchgear are probably so close to the costs of putting a new sub in the right place it wouldn't be worth it.

I think that putting a new sub in the right place is the real answer (in a distribution system - not necessarily so in industry).

Bung

 

[abspark]

Site growth took our fault level beyond the existing equipment's 250MVA rating with a predicted level of 350MVA. Our short term solution involved two tapped reactors (operating in parallel) to segregate the older 250MVA network from the new 350MVA rated equipment. As more site load is added and extra embedded generation installed, the taps are changed to limit the site fault level.

I too would be very interested in the progress of superconducting limiters. I am considering using fault limiters as bus ties rather than as serial elements in the network to avoid the effects of spurious outages/trips.

Long term plans involve connecting our 11kV generation at 33kV in parallel with the site incoming supply. This will involve an additional 11/33kV transformer and associated switchgear. Network modelling indicates that site fault levels will drop by 40-45% with this connection mode.

We also operate with one of our three 24MVA 33/11kV supply transformers held open, ready to close if either of the duty transformers trip. As site generation is increased, we may run on one transformer holding two on standby to put off the 33kV connection expense as long as possible.

Do not forget the fault level rating of the installed 11kV cables.

Best Regards

 

[peterb]

Bung -
I am presently working on a project where we are expanding an existing plant 500 MVA rated system. The system was expanded a few years ago, using fault limiters (your item 3) to retain the old 500 MVA gear in service. The limiters do work (we have seen two 3-phase faults where the limiters operated), but at some expense to replace/rebuild the elements afterwards.
The approach that we are taking for the present expansion project will use 1000 MVA gear for new distribution and future generation, with the ties to the existing system being via current limiting reactors. Any further expansion will take the distribution system to the 24 kV level.
I definitely agree with your item 5 (split bus operation), if that is a feasible option. Perhaps a judicious automatic transfer scheme can be installed to cover supply restoration after transformer faults. Also consider an automatic scheme to minimize parallel time during manual bus transfers.

 

[dpc]

Avoiding parallel operation of the transformers is probably the least costly approach. You can open the feeder loop somewhere and close it in (automatically or manually) if you lose one source.

I would consider reactors as an absolute last resort.