Continue to Site

Eng-Tips is the largest engineering community on the Internet

Intelligent Work Forums for Engineering Professionals

  • Congratulations MintJulep on being selected by the Eng-Tips community for having the most helpful posts in the forums last week. Way to Go!

GIS ratings and standards 2

Status
Not open for further replies.

Mbrooke

Electrical
Nov 12, 2012
2,546
Basically I am evaluating a system that requires 6 40/50/60MVA transformers with the 33kv secondaries operated in parallel. Considering the sheer amount of fault current, I plan on using standard 72.5kv (69kv) GIS gear and adapting it as needed, however I am wondering if another option exists like standard 35kv air insulted gear- and if so what the advantages would be.


Also on the same subject... the distribution load will be around 250MVA. I take it that I do not need 4000amp rated bus bars if load is even between each feeder and bus partition?


Open to all and any answer/opinion.


 
Replies continue below

Recommended for you

Do you have to have all 6 paralleled? Could you normally run two groups of three and group up when one or more are out of service? What impedance are the transformers?

One thing to consider would be adding enough impedance in the neutral to reduce the current of a single-phase to ground fault down to the level of a 3-phase fault.

How many load circuits? Hopefully a multiple of 6.

One option, but it takes a bit of space, would be to insert current limiting reactors between transformer/load groups with bypass switches around the reactors. With all transformers in service there's only limited current across reactors; enough to keep the voltages uniform. When you have to take out one transformer you close some or all of the bypass switches. You might be able to keep the fault current down to 5 transformer's worth.
 
Preferably all 6, but 5 is ok. The goal N-2 coverage, so at least two transformers can out of of service during peak load. However because the load consists of spot and secondary networks it is desired to keep them paralleled. There will be 16 to 20 feeders, 18 can be achieved if needed. All the secondary transformers are delta-wye, so an earthing impedance has no effect- possibly even preferred to keep fault currents down in vaults and man holes.

I am thinking a transformer impedance around 12% at "0" tap position.


Though this is not to say I dislike your idea, if this was an over head or non networked installation the above would be an ideal solution.
 
I was looking at two out. You say that the load is 250 and two out is 240. There's a lot of strategies for managing two out that work if 240 is enough that don't work if 250 is required. Personally I'd try to stay at less than 90% of necessary capacity at 4 of 6. If you can stay there, normally running two segments with three transformers and half the load on a three transformer segment allows you to drop a single transformer on the loss of one and close everything up. Run two three transformer segments or run everything on four transformers, normally with a fifth as an energized standby. You never need to deal with more than four transformer's worth of fault current, but in an N-1 condition always have one more transformer to offer up. The 10 MW between 240 and 250 is significant.
 
I can do two out and keep the network at 240MVA, perhaps 90% of 240, but I am uncertain at how POCOs normally bring a hot reserve online automatically. Or set the tap changers on the hot reserve to tap in step with the normal operating units.
 
Why not 3 numbers 120 MVA at 33 kV? Once I supplied 8 numbers 174 MVA 220/33 KV transformers to a refinery to meet their load.
 
Resistors and reactors on the neutral terminals are feasible, but would you ground each one through its own impedance or parallel all 6 neutral terminals to a common bus and then ground that bus via desired impedance?
 
prc said:
Why not 3 numbers 120 MVA at 33 kV? Once I supplied 8 numbers 174 MVA 220/33 KV transformers to a refinery to meet their load.

What advantage would 3 larger units give over 6? My understanding is that the X/R ratio also goes up but could be wrong about that.

Biggest desire for me is that 40/50/60 is a very common standard power transformer in addition to its smaller size and experienced ability of personnel to transport/install.
 
Less cost, less transformer losses, less fault level on paralleling. Unit cost and unit losses of transformer will not increase linear to MVA rise,but exponentially raised to 0.75.
 
Regarding fault current: Assuming all units are paralleled, won't fault levels increase in the sense that you will end up having 120MVA (technically 75MVA, on air natural oil natural) of iron contributing fault current instead of 40MVA of iron with 5 paralleled 60s?
 
I considered 5 numbers 60 MVA units each with 12.5 % impedance gives a net 2.5 % impedance while 2 numbers 120 MVA with 15 % impedance results in 7.5% impedance,
 
Thanks prc.

Ok, so I looked at transformer sizes and what would "roll" down the street. I'd need to get the dimensions to a 120MVA, but it looks like I can comfortably work with a 95MVA unit.

 
But, going back to the original post:

1. What type of gear would be best to consider and why? GIS, metal clad, ect. This will be housed an indoor, roofed, dry and climate controlled substation.

2. What bus configuration, ie single breaker double bus or multiple bus sections with ties?
 
My preference for something like that would be breaker and a half. One transformer and one load connection per bay. If the placement of things allows I'd also alternate transformers and loads on each bus.
 
I've never seen BAAH done in metal clad switch gear. Is this possible if I go that route?
 
I was focusing on the GIS bit and sort of ignoring the 33kV part. GIS baah can be done at 115kV.

There's a lot that would go into the decision making process. Local familiarity with the selected option is important. Make it a complex as necessary to meet the operational objectives but any more complex than that.

Make it as maintenance friendly as you can. If uptime is king you'll have something much more complex than if outages can be readily scheduled.
 
Ahhh, my mistake then in assuming. I was thinking single breaker double bus or ring bus for the 115kv portion, both are simple while simultaneously being maintenance friendly allowing for any bus partition to be repaired.
 
Status
Not open for further replies.

Part and Inventory Search

Sponsor