Underground cable system- how many?

[Mbrooke]

I am designing an underground transmission system consisting of a 330kv loop. This loop will supply several bulk substations which in turn each will supply 110kv area substations. The load on each bulk substation is about 1000MVA, and each area substation feeds about 180MVA of network load. Parts of the loop will carry approximately 3000MVA when all is said and done.


The question is, how many and what size cables should be used? Fewer cables of higher ampacity or many cables of lower ampacity. I am thinking 6 362kv solid dielectric cables from bulk substation to bulk substation and 4 121kv solid dielectric cables between the bulk and each area substations.

Second- what contingency level should I design this for? I am currently thinking N-2 on both 330 and 110kv, as failures of underground cables take much more time to repair.

Third, though not really the scope of my question but still partly relevant- cable charging compensation shunt reactors should be placed on each cable bay or lumped and placed on the bus?

I have no reference to go by... Or where to begin. The substations will be GIS btw.

 

[aussiejohn2]

G'day MBrooke,

That's an awfully big system to learn/practice on...

John.

 

[Mbrooke]

Just an urban load pocket, thats all.

 

[Power0020]

Seems interesting to study

I would think about that in this way:

1- Carry out a system modelling on ETAP or PowerFactory to identify load flow aspect, especially reactive power for cables and the sizes and locations of compensating reactors.(I think this should have been done by the owner, at the very early plans for the project).

2- More cables of lower ampacity doesn't seem tempting to me, as with the additional cost incurred and greater reactive power with charging, cable easement will also be bigger. Again, joints will cost a lot at these voltage levels. I don't know how such cables are routed but I think it may be running through an urbanised area, easements may make life difficult and lots of consents to go.

3- The other option with a smaller number of cables with higher ampacity, I think it will be more economical, since insulation of 362 kV cables isn't cheap. If these cables are to run in an underground services tunnel, it will be much easier.

4- Contingency level should be dedicated by grid code. I think N-2 is more than sufficient, but some places will go with N-1 at these voltage levels depending on other costs. A reliability analysis should be made based on outage severity, probability and risk.

5- Lumped or distributed reactors is also governed by the amount of reactive power (depends on length), you may need them at different substations as the reactive power doesn't travel much. Note the other costs involved, 330 kV GIS bay has a substantial cost and must be considered for connecting reactors. Also, the intended operation sequence of the cables and possible maneuvers will give a good picture to think about.

Good luck with the big job!

 

[Mbrooke]

Assuming easements aren't that much of a concern, are 3000 to 4000amp rated cables thinking to much?


I hear you on N-1, but imagine the scenario where 2 cables failed and load shedding had to take place.

 

[itsmoked]

I would consider any cables that run in the same service tunnel as not redundant. Remember the recent airport debacle. This would point to fewer cables -> fewer tunnels or cable-ways.

Keith Cress
kcress -

http://www.flaminsystems.com

 

[Power0020]

The scenario of two cables out will cause an outage, but how likely is that? the cables should run in different trenches to minimize the risk. Tunnels can simply compromise that as itsmoked mentioned.

 

[Mbrooke]

I was going to ask that about the common tunnel, good point and thank you. What do you guys think about duct banks? Or separate steal tubes?

 

[bacon4life]

Seems like three tunnels with 2 cables each would be a bit more feasible than 6 separate tunnels for the 330 kV loop plus 4 duct banks for 110 kV cables. I doubt you can route 10 cables into any substation without having some common mode infrastructure. You should certainly consider N-2, but the performance criteria is likely different than for a single cable failure. It might be acceptable to rely on some amount of demand side management load reduction or local generation for an outage of multiple cables.

It would also be prudent to do some evaluation for the consequences of having 1 cable plus 1 multi-cable tunnel out of service. This is an extreme scenario where load shedding is reasonable, but this scenario shouldn't collapse the whole system.

Also, be sure to include provisions for isolating, maintaining and replacing every piece of GIS switchgear. Some GIS designs require the entire substation to be de-energized to replace a bay in the middle of the lineup.

 

[Mbrooke]

Also, be sure to include provisions for isolating, maintaining and replacing every piece of GIS switchgear. Some GIS designs require the entire substation to be de-energized to replace a bay in the middle of the lineup.

Well, we are on the same thought train it seems. I was planning on starting a thread on just that particular issue.

In regards to common mode failure I think that will be inevitable. I am not sure how to separate what is essentially up to 12 330kv cables. (6 in 6 out). It will be "worsened" by the fact each in/out pair will alternate to prevent bus sections from overloading.

The other option is having each area substation rely on more than one bulk substation meaning that you could loose an entire bulk station and have the remaining bulks seamlessly pick up the area load. I am not sure how economically feasible that is, however its not entirely off in theory as each bulk substation will have spare capacity and in theory the number of feeder cables from anyone bulk could be reduced (ie 2 from bulk A and 2 from bulk B instead of 4 from bulk A)- so its more of a redistribution of assets so to speak.


Also- anyone see anything prohibitive in say 500kv vs 330kv for the underground system? Just entertaining the idea as I know the trend is to increase voltages over time. If the system is designed for this now, 60 years down the road it might come in handy.

 

[wroggent]

Are you talking about 6 per phase or 2 per phase when you say 6 cables?

Regarding the redundancy, are you talking about having an extra cable or two per phase for a single circuit or how many redundant circuits you need? I wouldn't install any extra cables within a single circuit; although you should definitely buy spare cable and leave it on reels. As far as how many redundant circuits you need, I think this would be a regulatory requirement.

With 6 cable (2 per phase) 69kV underground circuits (in a duct bank) it is a struggle to get 800A per cable with 2500MCM aluminum cable and a 0.75 load factor. This of course is based on local conditions. I'm not sure how large of a cable you can get but I thought most manufacturers top out at 4000 or 5000MCM. You're definitely not going to get anywhere near double of 800A out of a 5000MCM cable. What is your load factor?

Are you considering installing a climate controlled tunnel? I suppose that might work if you can find the space, but I don't see how you could realistically get a duct bank to work in an existing built out urban area with other utilities already present.

I think looking at the economics of 500kV vs 330kV is worth your time.

 

[Mbrooke]

Are you talking about 6 per phase or 2 per phase when you say 6 cables?

One per phase. If any factors say otherwise I may go two per phase- which will probably be the case if installed in ducts as appose to racks in a climate controlled tunnels. Certainly the case if I go above 1000-1500amps per circuit.

Regarding the redundancy, are you talking about having an extra cable or two per phase for a single circuit or how many redundant circuits you need? I wouldn't install any extra cables within a single circuit; although you should definitely buy spare cable and leave it on reels. As far as how many redundant circuits you need, I think this would be a regulatory requirement.

How many redundant circuits. In my mind a circuit is 3 phases of one or more conductors per phase protected by one breaker at each end (two at each end if ring or BAAH)

With 6 cable (2 per phase) 69kV underground circuits (in a duct bank) it is a struggle to get 800A per cable with 2500MCM aluminum cable and a 0.75 load factor. This of course is based on local conditions. I'm not sure how large of a cable you can get but I thought most manufacturers top out at 4000 or 5000MCM. You're definitely not going to get anywhere near double of 800A out of a 5000MCM cable. What is your load factor?

Starting on page 12 are the sizes I have in mind:

https://www.generalcable.com/assets...igh-and-Extra-High-Voltage-SILEC.pdf?ext=.pdf

I think looking at the economics of 500kV vs 330kV is worth your time.

I'm all 5 senses!

 

[bacon4life]

You may not want all 6 cables terminating at every substation in order to guard against a catastrophic complete substation outage. If you have a couple of the cables go to every other station in the loop, you will still have transfer capacity even if you loose a whole substation. Not to mention your routing problems get a lot easier if you 8 instead of 12 terminals at each substation.

 

[Mbrooke]

That is a good point @Bacon4life- a very good point definitely worth considering- I will weigh the pros and cons in doing so.