Underground LV network

[Mbrooke]

Is there any special way- or rather how do you- distribute the MV circuits in to the LV system in an underground network? So far classical engineering text books typically show feeders A, B, C, D, E, F, G, H each supplying an individual network transformer sequentially one after another. All 8 feeders loop and siz zag around the LV mesh from start to finish with each segment or quadrant of that LV mesh receiving an equal input from all 8 feeders, (ie upper west zone 2 network transformers feeder A, 2 network transformers feeder B, 2 from C... middle north zone 2 network transformers feeder A, two B, two C, ect.


Now- another scenario.

Each quadrant is fed by only a few feeders rather than all of them. Ie North Quad 1 is fed via A, B, C, North Quad 2 B,C,D quad 3 C,D,E, North Quad 4 D,E,F, ; middle Quad 1 E, F, G middle quad 2 F, G, H, middle quad 3 G, H, A middle quad 4 H, A, B ; south quad 1 A,B,C and so forth...


The idea is that should the network ever exceeded N+2 or N+3 (such as a bus fault or serve vault/manhole fire) a certain LV area will always be deprived of an MV source. The intent here is that this will produce depressed secondary voltages reducing KW consumption for each building in turn protecting the remaining MV feeders from over load and prevent LV network cables from overheating.


Is this possible? Is it done in the real world?


I know people will bring up critical customers and that this will back fire on them due to being predominately motor and none linear based, however these will be fed from either dedicated spot network feeders (I, J, K, L) or be able to transfer from feeders A,B,C,D to E,F,G,H automatically being spared this operating procedure.

 

[krisys]

Mbrooke (Electrical),
Your question is quite complicated to understand. Make it little simple.

 

[Mbrooke]

I'll start with this...


Picture a secondary 208 or 480 volt LV mesh. What are the different ways to distribute the MV circuits into this LV mesh?

 

[magoo2]

A low voltage grid has been incorporated in most of the major cities since the early 1900s. They are called secondary networks. These are all 208 V wye/120 V grids. I would not consider doing something similar for 480 V because of arc considerations. The 208 V grid works quite well because 120 V faults tend to burn clear. With higher voltages like 480 V, this is not the case. They (the 480 V) may be used on spot networks and not grid networks.

There's a ton of information on what you're asking about in the Electric Utility Reference Book: Distribution Systems published by Westinghouse back in 1959. This covers the physical arrangements and interleaving of feeders necessary to may it work satisfactorily. This might be what you referred to as one of the classical engineering texts on the subject.

If you want to get some better information, I'd suggest you contact someone from Con Ed in New York City or Com Ed (Exelon) in Chicago since both of these cities have extensive secondary networks.

With a 208 V grid, you can't supply huge amounts of power over any significant distances. For this reason, today most secondary network utilities have gone to spot networks. Spot networks are available today at both 208 V and 480 V. Spot networks are just 2 or more feeders and 2 or more network transformers and network protectors.

Finally, you might also reach out to Dave Smith on this. He had worked for Westinghouse on secondary networks and retired from Siemens/PTI and is still going in Pittsburgh. He a good friend of mine and is probably the best expert on this topic in the world.

 

[Mbrooke]

"With a 208 V grid, you can't supply huge amounts of power over any significant distances."



So I am guessing that if the 5 boroughs of NY were wired today grid networks would not be used?


Thank you for all the information though! Does Dave Smith post here by chance or how I can reach him? If you would like you are welcome to forward this to him.

 

[davidbeach]

I've not seen an eight feeder network system before, all of ours use four feeders. There are two types of networks, the area network and the spot network.

Area networks, also know as street networks and a handful of other names, are implements at 208Y/120V and will have network transformers in vaults along the street with the 208V conductors running from vault to vault and loads tapped off the 208V conductors as needed between the vaults. A single network can span many blocks and have many transformers connected to it. This is where the 200kA type T fuses (even upward of 300kA) find their use, large networks can have hugemongous fault currents. Faults in the 208V conductors will generally burn themselves out.

Spot networks are much more likely to be 408Y/277V and feed a single building. Large buildings may even have multiple spot networks within the building at different levels, perhaps in the basement and in the mechanical penthouse. Tall enough and there may even be intermediate networks. The transformers of a spot network will be grouped together but there will be no connection between the secondaries of different groups.

 

[Mbrooke]

Thanks, I understand. FWIW Con ed has 22 feeder networks Page 3:

http://www.dps.ny.gov/06E0894/06E0894_CE_Initial_Power_Outage_Report.pdf

The question when dealing with an area network is on how to distribute those feeders. Its seems that confining a set of feeders to one area might help in some cases- at least theoretically from my thinking.

 

[davidbeach]

A set of feeders would be confined to a defined geographic area; then the trick is trying to keep the load balanced as evenly as possible between feeders. Some vaults will have more or fewer transformers than other vaults and not every transformer will be as heavily loaded as other transformers. Probably balanced by the seat of the pants as much as by another means.

 

[bacon4life]

Our network uses 3 feeders from each of two substation. History and conduit availability play a large part in how the present set of transformers are distributed. One important item to consider is the secondary side cable loading and the thermal impact this has on shared duct banks.

Are you considering building a new grid network? Arcflash mitigation is a huge challenge.

 

[Mbrooke]

Wait, 3 feeders from two substations? I am not saying this is wrong or not right, but how do you keep phase angles and load on each sub balanced?

And yup, this is a new grid network, or at least a consideration for an urban area. From everything I read I somehow gather that is ConEd, PEPCO, First Energy, Seattle ect would not be using secondary grid networks today if they had to start all over?

 

[bacon4life]

It has been in service for decades, so things must stay balanced enough. The load definitely shifts some between the stations, but the network group is still looking for a piece of software that can model both a multisource networks and single phase distribution loads.