HV Substation Protection Philosophy

[rockman7892]

I was recently reviewing a concept for a 161kv - 34.5kV Substation that served as a substation for connecting a BESS to the grid. I had some general questions about some of the protection philosophys shown on the one-line sketch that I was hoping to learn from some of the utility folks here what is usually considered the "norm" or "standard".

Attached is one-line sketch I developed showing general protection philosophy with some questions below referencing questions in red on one-line

1) Are PT's with (2) secondry windings that are used for metering and protection typically located on line side of incoming disconnect switch as shown or are there reasons to locate it on the secondary side of switch?

2) Is it typical to see (2) sets of transformer differential relays as shown with one diff zone looking all the way down to secondary of low side feeder breakers and the other looking at ct's on xfmr secondary. I'm not accustomed to seeing (2) relays and typically are used to seeing on xfmr diff that only looks to CT's on secondary of transformer. Is it typically to have (2) for redundancy or a specific needs for the (2) zones?

3) Currently I only shows xfmr neutral CT's connected to one of xfmr differential relays? Is there a need to connection another set of CT's on each neutral to connect to the second diff relay?

4) I shows a feeder protection relay (SEL-751) on high side of breaker. I usually see this relay in this application but was never quite sure of the reasoning considering both the xfmr diff relays are capable of providing many of the same functions with high side CT's (50/51, etc...). Is there usually a specific need to including a feeder protection relay here (perhaps use for BF,etc..?)

5) One-Line shows (2) different line differential relays which will have fiber connection to upstream substation. Is it typically to see (2) different line differential relays in this application? Is one simply a redundant backup?

6) Is there typically a need to have a feeder relay on the secondary of the transformer? I would think not again considering that the xfmr relay with secondary CT's provides many of the same functions as this secondary feeder relay would. Is there a need if certain number of radial feeders is exceeded?

7) The feeders from the 34.5kV collector bus will feed down to large BESS units. I know that some inverters associated with BESS systems have ability to act in "grid forming" mode with capability to operate in islanded mode. If this is the case is there a need to perform sync check functions across feeder breakers when re-syncing to grid after a loss of utility?

Appreciate the help or any other comments anyone has.

 

[davidbeach]

1) Depends on who's switch and who's meter. Also depends on local practice.

2) Utility protection is almost always redundant. Industrial protection is less likely to be so. In my experience, that some of this/some of that approach is older; today we'd have all matched sets. But I've done transformer with two feeders that don't reclose all with a pair of 487Es. Wrap the whole thing with a diff and use the winding overcurrents for the feeders. No need for the 587Z or a bunch of other things. Wouldn't do that in most applications though.

3) I'd take the neutral currents to both transformer relays.

4) Preferences

5) Dual is nearly universal for the utility. Different or identical can border on religious war territory. I'm firmly in the identical camp.

6) Preference and other things. The 487E has 5 current inputs, it could wrap a transformer with four feeders. On the other hand, a low-side main can be beneficial, particularly if there's another transformer close by that could pick up the load.

7) Synch check is useful most any time there are sources on both sides of a breaker. That could be a good reason not to rely on the 487E for everything, the 751s can do the synch check. Close supervision is actually a better term, that includes addressing all of the hot/dead combinations. You may, for instance, not want to allow close dead-dead.

Lots of preferences in the design, both from the BESS developer and the serving utility. Things that could be routine in one area could be entirely unacceptable elsewhere. Nice to see that wye-delta-wye transformer from the beginning, too often developers want to put a delta on the high-side (our side) and we have to send them back to the drawing board.

Drawing CTs and connecting to the middle isn't how somebody doing protection would show it, the connection would always be to one end of the CT or the other. The unconnected end being the wye point of the CT secondary circuit. Connected to the middle it's impossible to tell how the CTs are installed. For non-directional overcurrent it doesn't matter. But for anything with directionality, and that includes differentials, it matters.

I’ll see your silver lining and raise you two black clouds. - Protection Operations

 

[bacon4life]

David's comments are great.

2)Having overlapping zones allows the use of fewer relays, at the cost of increases complexity of settings and increase risk during testing. My preference is to avoid installing relays that utilizes physically paralleled CT inputs such as shown for the SEL-387.2 and the SEL487E. CT testing on any of the 6 feeders will require disabling both relays. I have already experienced technicians incorrectly shorting paralleled CT's, and I assume folks will become much less familiar with paralleled CTs as utilities standardize on using dual input SEL-411L relays.

6)I am not sure that an additional 751 relay on the low side the transformer adds much value unless you also add a main breaker. One possible advantage of a main breaker would be having the station service connection upstream of the main.

 

[rockman7892]

David Thank you the great comments. Very helpful for me to learn from experienced perspective. And yes good catch on CT representation, I simply drew that that way for quick illustration.

With your first comments being so helpful I figured I'd see what kind of experience of opinions you had on others that came to mind:

8) Where would you typically see surge arrestors for an application like this. From what I'm used to seeing I'd expect to see on on incoming 138kV line (ahead of first disconnect), transformer primary, and transformer secondary. Would their be a need to have an 138kV arrestor between the 138kV switch and breaker for a scenario where the switch was open?

9) Is there an advantage to using a 138kV CCVT vs a PT in this application? I've seen design that use either-or, or use a PT for metering and CCVT for relay protection. Is it always common for both of these to have a 2-winding secondary?

10) You mentioned the wye-delta-wye transformer configuration as being a good idea. I'm assuming that delta tertiary winding blocks zero sequence current from passing through transformer? I have seen some other BESS applications with similar voltage class use wye-delta transformers (wye side - utility & delta side - BESS site) with a grounding transformer connected off of the collector bus to source ground fault current on the delta system. Are their pro's/con's to one vs other? My thought is that the grounding transformer requires more relaying and complexity. Is there an issue with using a wye-wye transformer?

11) With the 34.5kV bus diff on the collector bus is it more preferable to use the Hi impedance bus diff or low impedance bus diff for this application, or is this usually a utility preference?

12) There are several overlapping zones on this application but from what I see no real benefit in terms of tripping selectivity. A fault on 34.5kV collector bus or transformer secondary will have the same effect (primary and all feeder breakers tripping) whether detected by bus diff or transformer diff. Similarly a fault on secondary of 138kV breaker will have same effect (primary and all feeder breakers tripping) whether detected by line diff relays or transformer diff. Is the overlapping zones in this application primarily for redundancy?

 

[rockman7892]

bacon4life - Thanks for the comments as well.

In regards to your comment for #6 i'm not quite sure I follow what you are saying with main breaker providing and advantage for upstream station service connection?

My thought with having a relay only on the secondary with no physical breaker would be to have the CT's act as a virtual breaker with tripping of primary breaker. You could still provide all the secoondary functions as you would with a breaker but physically trip the primary breaker. But I guess that is somewhat redundant considering that the transformer diff secondary CT's can provide this same functionality with the transformer diff relay tripping the primary breaker?

 

[davidbeach]

8) I might not, but others might see lots of them. I live in an area that gets away with unshielded 500kV transmission. I believe you live or work in Florida. Two very, very, different places in regards to what gets done. I'd expect arrestors on the high-side of the transformer at that's it.

9) Preferences, costs. Dual secondary is very common in my experience, but might not be common in industrial applications.

10) As the transmission provider I want the BESS to present as an effectively grounded source. The grounded wye on my side and the delta makes that possible. The BESS owner probably also wants and effectively grounded source on that side as well; that box gets checked as well. There is transference of zero sequence currents from one side to the other but the protection system can sort that out.

11) Preference, equipment costs.

12) Overlapping zones is a design philosophy. It goes along with the use different relays to do the same function philosophy. Both might have made good sense 20-30 years ago when the numeric relays were new and unfamiliar. Both are an approach to minimize the chances of a fault getting missed. But at this point that's not the problem the industry has as a whole. Looking at the misop statistics for facilities subject to PRC-004 reporting one sees that essentially all (over 95%) misoperations are unnecessary trips, not failures to trip. The do the same thing twice approach, with fully independent sets, reduces the opportunities for things to go wrong, both operationally and during testing. Testing one of redundant pair while the other remains in service is a huge time saver and avoids a lot of system shutdowns. Knowing what you're into really helps. But in many ways the most important aspect is that there's a limited amount of time to develop the design and the settings. With two of the same, 100% of the available time can be spent on that single approach. With the two different approach, you get to spend 50% of the time on one and 50% of the time on the other. Lack of time for the design and settings is ultimately where most of the misoperations come from. But that's just my opinion. ;-)

I’ll see your silver lining and raise you two black clouds. - Protection Operations

 

[bacon4life]

6)Regarding main breakers, I was thinking about some issues I have seen in distribution stations and remote generation stations. Having station service tapped of upstream of the main breaker can allow use of stations service even if most of the switchgear is out for maintenance. If your site has a reasonable alternative station service, this may not be an issue for your site.

If the individual units have grid forming, another possible use for a 34 kV main breaker would be as the synchronization point for the whole plant. Putting sync checks on all six feeders would leave you with 6 separate islands. Putting sync check at the 161 kV level would require that the BEES be capable of handing the inrush of energizing the GSU from the low side.

12) Even when the same breakers trip, restoration can be much faster if relays contain targeting information to identify the difference between a transformer fault, a bus fault, or a line fault. I would expect a line reclose after simple visual inspection, whereas I would expect a transformer trip to require a full set of oil/insulation testing.

Other) I have been curious how DER resources like this are coping with resilience risks. On the utility side, most utilities have a collection of spare equipment that can be cobbled together for emergency use. It seems like a failed transformer would have the whole plant out of service for many months.