Ungrounded 11KV system

[electrofelon]

Because by NEC 240.21(C)



Therefore, we always use a delta configuration on the secondary side to avoid adding an 11 KV circuit breaker at the secondary.

I don't see any way around having some sort of OCPD in between the transformers. See 450.3, your second transformer needs primary protection and I don't see any allowance to protect it by the first transformer OCPD. Pad mounts typically have bayonet fuses built in that can take care of that.

 

[julianov]

Keep in mind through this discussion that the POCO-owned transmission facilities are governed by the NESC, not the NEC. The NEC only comes into play when the circuit(s) meet the consumer (or anything outside the system of the POCOs involved.)

In this case is for a datacenter system. So, is not part of POCO.

There is 110.10 which is a catch all that generally requires you can clear a fault without excessively damaging the equipment. Systems over 600V typically have to be designed and stamped by a PE around here, I don't think the regular electrical inspectors even inspect them. One of the few over 1000V I know of the customer pays the utility to have their linemen maintain it.

I know, but I want to find where I can locate the safety measures for ungrounded systems over 1000 volts, also for academic purposes. I don't believe protections or considerations for ungrounded systems over 1000 volts are left to chance, whether in the NEC or NESC. I understand that a phase-to-ground fault must be detected, not taking the system out of service because it's not a clear fault, but it should be clearly indicated to prevent a second fault. Furthermore, a phase-to-ground fault, while not taking the system out of service, generates overvoltages.

Let's start there: What size is the conduit, and while I'm asking, what type, any interruptions, etc?

We need to sort and define the constants before juggling the variables.

To meet the 2% voltage drop requirement over 1400 feet with a 350A current at 480V, two parallel runs of 1000 kcmil copper conductors would be necessary. However, there isn't enough space for that. Additionally, the transformer is configured as delta-delta to avoid using an Overcurrent Protection Device (OCPD) at the secondary. But the issue I see is that among the transformers, we have an isolated system. I'm looking in the NEC or NESC for rules that clearly indicate we should add a ground fault detector to the isolated system.

Since no one has come out and said it, No you do not need to ground MV systems, and you do not even need ground detectors.

But how would you detect the first ground fault? The MCB won't be able to detect it

 

[tortuga]

Center pivot irrigation machines are all 480 volt, never seen one that was any other voltage. Well pump motors are only stocked in 480 volt. For whatever reason that is what they started out with long ago and it likely not going to change now with so many out there that are 480 volts. Under 75 HP you do find them as 230/460 dual volts but 75 and up is almost always single voltage 460,

Interesting to see what people do in different areas, Franklin Electric makes '575 volt' pumps.
They have a 8" 75HP 2396136241 submersible. It would run off a typical 600V 3ph pump service.

 

[electrofelon]

To meet the 2% voltage drop requirement over 1400 feet with a 350A current at 480V, two parallel runs of 1000 kcmil copper conductors would be necessary. However, there isn't enough space for that.

First, not sure who on earth would use copper for this. Also, it seems you are ignoring voltage drop in the transformers. Off the top of my head, I would guess you should figure about 8% VD for a non transformer system to be comparable to a transformer system. I am not sure why everyone ignores this. A step up step down system has lots of losses and horrible voltage regulation, trust me. It's just simply absurd to compare say you need 2% VD and compare it cost wise to a transformer system.

But how would you detect the first ground fault? The MCB won't be able to detect it

I'm not sure, I don't do much MV design, other than small residential step up step downs now and then, and those are grounded. Probably the thinking is a fault is quite rare, and even if there is a first fault it's really not a hazard, you just have a grounded system now. You can certainly add fault detection, it's just not required by the NEC.

 

[electrofelon]

I I'll bet you a milkshake that if you use 2 sets of 350 AL with no transformers, you will have more voltage stability and less losses than using a step up step down, and dramatically less cost.

 

[tortuga]

View attachment 2572223

All voltage drop issues and 11kv aside how are you calculating the required ampacity of the motor feeder and branch circuit if the 300HP motor was right next to the panel?

 

[julianov]

But Why Section 250.21(B) requiered Ground Detectors for AC systems operating at not less than 120 V and 1 kV or less. Why is there a requirement for ground detectors in this specific voltage range, but not for systems exceeding 1000 volts?

Ground detectors shall be installed in accordance with the following:

1. Ungrounded ac systems as permitted in 250.21(A)(1) through (A)(4) operating at not less than 120 volts and at 1000 volts or less shall have ground detectors installed on the system.
2. The ground detection sensing equipment shall be connected as close as practicable to where the system receives its supply.

 

[LarryFine]

Have you built in the poor voltage regulation/voltage drop of having two transformers into your evaluation? Most people simply leave that out and of course that creates a very unfair comparison.

Not to mention continuous energizing power consumption.

 

[tortuga]

Its not really an issue of ground detectors its an issue of long feeders. Feeders over 1000V are pretty rare and the NEC defers to a PE see [2023 NEC] 225.356 and 495.48. The code covers most things well but outdoor feeders so long that they in effect become transmission lines are not well addressed.
A under 1000V feeder could also be so long that a phase to ground fault would not trip a standard breaker so a PE might require ground fault relays on those as is done by the utility, again see 110.10.

 

[electrofelon]

Not to mention continuous energizing power consumption.

Yup, there's a zillion step up step downs out there that have worse performance and higher cost than a no transformer system. I have a step up step down at my house, voltage swings a lot under load even though the VD in the cable is negligible.

 

[LarryFine]

To meet the 2% voltage drop requirement over 1400 feet with a 350A current at 480V, two parallel runs of 1000 kcmil copper conductors would be necessary. However, there isn't enough space for that.

What size is the conduit?

 

[electrofelon]

What size is the conduit?

Enough space for two transformers though?

 

[julianov]

Conduit size is 1''C

 

[tortuga]

Conduit size is 1''C

I think the smallist MV cable your going to find in a 15kv is a #2 and that has a OD of just over 1".
You could do a 2000V feeder and PV wire

 

[LarryFine]

Conduit size is 1''C

I'm getting #2 XHHW-2 as the largest wire size you can fit three of in 1".

 

[julianov]

The conduit size is 3''. It was a typo, and there is a MV tripolar cable available for that size from Prysmian. Beyond that, I don't understand why the NEC sets rules for ground fault detectors for ungrounded systems below 700V but not for those above 1000V. Although we have a Delta configuration on the transformer secondary and a phase-to-ground fault does not cause overvoltages on the healthy phases, the capacitive currents of the other healthy phases do change. These currents can potentially generate electric arcs or damage the insulation of the conductors.

 

[LarryFine]

3"?! Well, that's a course of a different holler (overheard in cheer-leading class).

I get three 900kcm, six 400kcm, or nine 250kcm in 3" EMT.

Or, three 700kcm, six 300kcm, or nine 4/0 in 3" Sch 40 PVC.

Would any of those combinations suit your expected load?

 

[julianov]

Thank you.

Yes, I believe that 6x300 kcmil is an excellent choice.

 

[LarryFine]

Thank you.

Yes, I believe that 6x300 kcmil is an excellent choice.

You're welcome.

The next question is: how high do you need to go in voltage to deliver the power you need, at the current that pairs of 300cm will carry, with 6-conductor derating?

 

[julianov]

Six sets of 300 kcmil conductors can carry 350 A over a distance of 1400 feet. The voltage drop would be approximately 26.25 V, which represents a 5.47% voltage drop. So we are not good. It should be six of 400 kcmil. I thinks its a lot.