HV GIS Substation Considerations

[rockman7892]

I’m in the preliminary stages of concept design for new 345KV GIS substation and was hoping to hear any experiences or best practices others have had with GIS layout.

Substation topology will be incoming 345kV lines to GIS which will then feed (4) distribution transformers.
Proposed method currently is to have 345kV lines coming into substation to transition from cable to Gas Insulated Lines (GIL) for connection to GIS breakers/bus. The transformer feeders will also have GIL lines within substation between GIS and transformer bushings.

Curious to hear any experiences from others related to HV GIS in substation related to the following:
1) Is GIS typically located in an enclosed building or is it also standard to have located outdoors?

2) Is the standard approach for connections between GIS and incoming lines and transformers in yard gas insulated lines or is
it usually more practical to transition to AIS OH bus or even underground cable?

3) Is 345kV underground cabling common or practical for routing within substation? Any pros or cons?

 

[thermionic1]

I've worked in several GIS stations in our footprint, as well as other utilities.

1- Most GIS I've seen has been indoors. Some buildings have the transformers indoors with suitable fire suppression. I was at a 345 GIS station last fall that was about 40 years old. Built in the middle of nowhere. Most others are in downtown environments. I've seen a 138kV installation, more like 2 lines, two breakers that was installed outside. If you build indoors, you will also have to install suitable gantry cranes for initial installation and future work.

2- Our last new installation was 138kV, 6 breaker ring with 4 lines and two transformers. The existing oil filled lines were cut into and out of the station. For the new circuits, it was XLPE cable. Our newer 138 station had XLPE cable to the transformers. The 345 had gas bus to the transformers outside. I believe EHV XLPE cable has come a long way in the pas decades and it's acceptance has become more common. From an environmental perspective, the cable can't leak gas and cause environmental problems to deal with.

3- If in a downtown area and OH lines aren't there, there will be some kind of existing UG tunnel / duct system. If you are dealing with any oil cables, the splicing is a very specialized task. We call in specialty contractors from the East coast. Last year there was a 138kV Pothead that was damaged and during the replacement, the device used to freeze the oil failed during installation, which resulted in a discharge of about 50 gallons of oil near a river. That was an expensive mess that also got the attention of the local news. For XLPE systems, I'm seeing online Partial Discharge systems being installed with the new cables and FO cables that are also installed with the new system.

Any GIS will have quite a few motor operated switches for breaker disconnect, grounding, line disconnect and grounding. The 138kV installation I checked out in 2020 had several motors fail. When inspecting the burnt motors, I was very surprised to see plastic gears driving the switches. None of this stuff is local, so be sure to have adequate spares on hand.

In AIS stations it's very clear when a switch is open or closed or if grounds have been hung. In GIS there are position indicators, and also small viewing windows. IME, I haven't met a switchman that liked this second arrangement, so plan on having training for the switchmen.

All the motor operated switches, ground switches, line disconnect switches and line grounding switches need to be electrically interlocked, as well as their mechanical interlocks. You might be tempted to use high tech relays/bay controllers/61850 verses standard hardwired mechanical relays. Everybody knows how to troubleshoot the mechanical aux relays. Not everybody is going to understand a vender supplied bay control unit, the software, the HMI, etc. Each gas zone will have it's own density monitor. Our latest station has an HMI with a mimic screen that shows physically where each detector is located.

We are retiring the first GIS the was installed downtown in the late 70's. The gear has held up well over time, largely due to being indoors. The OEM has changed hands multiple times and spare parts are no longer available and some components are showing their age.

Accessing PT's, CT's and Breakers for testing can be a challenge. There are removeable ground links, but sometimes a breaker will be grounded on both sides and not all breaker timing test sets can accommodate this arrangement. If the end user has older equipment, it might be wise to include new test equipment in the project.

 

[crshears]

Accessing PT's, CT's and Breakers for testing can be a challenge. There are removable ground links, but sometimes a breaker will be grounded on both sides and not all breaker timing test sets can accommodate this arrangement.

Too true; the utility I worked for had a standing rule that for work on SF6 equipment some manner of grounding had to be applied at all possible points of infeed. Since ground links had to be opened on one side of the breaker for timing tests, the isolated zone had to be expanded on one side of the breaker to the next groundable component prior to the opening of the breaker ground links on that side.

Indoor siting is far gentler on equipment than outdoor. Also, count on insulating and heating the building, as there will be far, far less equipment issues due to fluctuating temperatures, reducing operating cost. Crews are also more productive working in a temerate environment.

And whatever you do, DO NOT, I repeat, DO NOT use an uninsulated building to house SF6 equipment, unless the locale is in desert conditions; micro-climates can develop within uninsulated buildings, leading to condensation all over the equipment, which can be especially problematic where maintenance crews have been lazy and not re-installed all equipment covers.

CR

"As iron sharpens iron, so one person sharpens another." [Proverbs 27:17, NIV]

 

[prc]

Answer to your question 2&3 Avoid a cable connection between the GIS and the transformer. Either directly connect the transformer to GIS by oil to SF-6 bushing or connect the transformer to GIS through a short overhead line with a surge arrester in between them.

 

[Kiribanda]

From my GIS experience my conceptual design will be, from the $$$ point of view,
1)Incoming 345kV line should be overhead and not cable.
2)The 345kV GIS (better with double bus bar) should be inside a pre-fabricated building
3)Outgoing feeders should be also O/H lines
4)That means the pre-fabricated building should have three wall mounted 345kV bushings
for the incomer and four sets of wall mounted bushings for the four outgoing feeders
5)Your outdoor sub will have
5.1 One set of 345kV surge arresters and a 345 kV disconnect before the wall mounted bushings at the line side
5.2 Four sets of 345kV surge arresters and a 345 kV disconnects after the wall mounted bushings at the load side

 

[bacon4life]

For the layout, ensure there is sufficient physical space between components to removal/reinstallation of every component. Ensure there are sufficient electric isolation points and gas isolation points to allow removal/reinstallation of every module without require a full station outage.

 

[rockman7892]

Thanks for all the great responses here.

I've attached a very basic layout sketch which shows the scale of GIS lineup (from vendor) as well as the (4) feeder transformers. The 240MVA transformers are quite large so layout pretty wide compared to the GIS footprint.

Ideally I was trying to position to where GIS had SF6-Air bushing right at main GIS unit with overhead cable out to each transformer. Was hoping that transformers would align directly with GIS breaker module for short direct connect run (with GIS in building or outside) but due to layout will have to have some creative OH cable runs or put in lengths of GIS bus to get closer to transformers.

1) I was considering putting blast walls to pull transformers close together. Is there a recommended distance on how far
transformers should be from GIS as shown? If there is a blast wall between transformers and GIS can these be pulled in
tight?

2) Is it typical in most applications to have some offsetting lengths of GIS bus run between GIS and transformer before a SF6-
Air bushing as shown with red lines in sketch? It seems like these additional GIS bus runs would add cost including added
structural support, etc..

3) Any ideas for better optimized transformer or GIS layout that allows for direct OH cable connection or less GIS bus?

4) On the incoming 345kV GIS lines is there any recommended distance between dead-end structure and GIS? Ideally I'm thinking that we reduce GIS bus as much as possible and bring OH lines as close as possible to GIS to connect to SF6-Air bushing?

Thanks for all the comments.

 

[rockman7892]

@kiribanda

Are the surge arrestors you recommend the ones that would be on the transformer itself?

Are the 345kV switches you are suggesting part of GIS or separate 345kV air switches outside of GIS on feeders? Is purpose for these switches to be able to isolate GIS without having to rely on disconnect switches integral to the GIS itself?

 

[Kiribanda]

1) Yes, pl. refer to the attached.
2) Yes, pl. refer to the attached.

https://files.engineering.com/getfi...8ab46221b&file=345kV_GIS_CONCEPT_Comments.pdf