Issues going from delta-delta to delta-wye transformer? 1

[wthall]

I've got a client who is getting a new 750KVA transformer from his utility. The new transformer is delta on the primary and wye on the secondary. His existing transformer is a 500KVA bank that is delta on both the primary and secondary.

What effects, if any, will this service change have upon his equipment...motors and lights specifically?

Thanks In Advance

 

[cuky2000]

The following information could be helpful to answer your question:

a) Secondary voltage (new & existing).
b) Existing grounding method (Ungrounded, corner or center grounded).
c) Proposed new secondary grounding system.
d) Actual single phase load for lighting and others
e) There is any ground fault detection system ?

SUGGESTION: Double check the interrupting rating of the protective device(s) and compare with the increase prospective short circuit provided by the 750 kVA transformer and any future motor load.

 

[electricpete]

You should give the info cuky asked for. I'm sure he will give you a careful and considered response.

I'll throw out my thoughts...

I'm assuming if the original is delta that you'd want to hook up the replacement in ungrounded wye. (Introducing a ground would require very careful separate consideration depending on your system).

You should check with the manufacturer before leaving the neutral of a transformer ungrounded. Some (ge) will recommend that you attach a surge arrestor to the neutral bushing. I have seen very bizarre damage at neutral bushing of an ungrounded wye transformer. Don't ask me why.

 

[wthall]

I hope this helps you help me...

a)Secondary Voltage, New and Existing is 480VAC, 60HZ

b)Existing Ground = Ungrounded

c)Center of wye, connected to ground rod adjacent to new pad, which is then connected to ground bus of new/existing switchgear. A neutral will also be pulled in for future 277V needs.

d)Lighting load is not known at this time. Their single-line diagrams show a feeder to a 200A box (still 480VAC). This panel contains approximately 20 three-phase circuit breakers. 15 of these breakers feed lights. The others feed transformers and heat strips. We are a bit confused since most (if not all) lighting circuits are single phase regardless of voltage...so what's with the three-phase circuit breakers? Maybe their lead electrician will know...I'll have to get back to you on this item (D).

e)I'm aware of no ground fault detection system.

f)Transformer impedance (new) is 5.99% ~ 20KAIC
New Main Breaker rating = 30KAIC
Existing Transformer and Main Breaker Ratings are not on the single-line drawings.

 

[peebee]

If the 3-pole breakers you describe directly feed lighting branch circuits, per NEC the lighting circuit arrangement you describe is illegal as you can't have more than 10% of your panelboard poles serve lighting if there's more than 42 poles, See NEC 408.14 & 408.15. I'm a bit confused by your description, though -- are those breakers actually serving transformers for 120v lighting? It would seem that you couldn't have any existing 277v lighting, and I'd be surprised if you had 480v lighting. If the 3-pole breakers are serving downstream distribution panels then you're fine.

An ungrounded system can be more reliable than a grounded system as it requires 2 faults to drop the system. You'd need ground-fault indication to make use of that, though, which you indicate was never installed.

A grounded system such as the proposed new one will limit overvoltages on the system, which will make life a little easier on your motor and cable insulation.

 

[electricpete]

Don't hold me to it but I still think ungrounded wye is your best bet. (and add a lightning arrester at the neutral).

If you add a ground there are many implications. One is that you will dramatically increase fault current on downstream devices. IF memory serves me right (and you'd better double check), grounded wye would give substantial increase in fault current downstream (due to zero sequence path), but ungrounded wye would not. You don't want to push fault current beyond capability of downstream interrupting devices.

There will also likely be protective relaying considerations to adding a ground to your system. Protection strategy is different for grounded system than ungrounded. Sorry to speak in generalities. Please dont take my word for it. Otehr comments?

 

[wthall]

I just got back from the client's plant.

They use 480 for their lights...no transformation to 120 (for lights).

The way it was explained to me by the electrician is this: Phase A&B would be taken to light 1, skip light 2, then proceed to light 3, skip 4, proceed to light 5, etc. Phase B&C would be taken to light 2, then skip 3 and go to 4, skip 5, go to 6, etc. Phase A&C would go to even more banks of lights...all of this off of one 3 phase breaker.

The voltage from phase A to B is 480
The voltage from phase B to C is 480
The voltage from phase A to GND is 480
The voltage from phase C to GND is 480
The voltage from phase B to GND is 0

I can't find article 408 in my NEC1999 book.

They would not only get a new main but also new distribution breakers...8 at the moment...ranging from 400 to 800 amps and rated at 30KAIC like the new Main 1600 amp breaker.

The 1600A main will have a selctable/interchangeable trip so when they need more power, all they'll have to do is get a bigger transformer. The current pad is suitable for up to a 1500KVA transformer and the wires are rated for 1600A.

 

[brupp]

If you're really planning on having 277V loads in the future, then a solidly grounded system would provide that directly. However, changing to a solidly grounded neutral is a big change as you will need to carry a neutral to quite a few places. If you don't need the 277V or expect to only need it for a few places your best bet is probably to use a high resistance grounded system. You can always add 480 to 277V transformers for your 277V loads.

In a high resistance grounded system you put a resistor between the neutral and ground limiting the ground fault current to 5 to 10 amps (27.7 to 55 ohm resistor). You connect an voltmeter/voltage relay across a portion of the resistor to detect a ground and send an alarm to an operator. To locate the ground fault you activate a pulsing system that is easily accomplished using a motorized pulsing relay and vary the resistance, say 30 to 50 times per minute. Then a clamp-on ammeter can be used to find the fault as only the faulted circuit and upstream circuits will show the pulse.

I think it's safe to say that the vast majority of North American power plants built in the past 15 or 20 years have this type of system.

One warning, though. NEC article 250-5 includes the requirement that "The conditions of maintenance and supervision assure that only qualified persons will service the installation." Might this disqualify your installation?

There are a lot of web pages that discuss this further. One good one is

http://www.geocities.com/SiliconValley/Pines/3406/high_r.html

 

[peebee]

I somewhat disagree with you, epete.

If they are planning on running any 277v or 4-wire loads directly from the service (that is, with no intervening 480:480/277 transformation), the service is required to be solidly grounded or at least high impedance grounded per NEC 250.20(B) or 250.36.

Also, seems to me that a 3-phase bolted fault would be of the same magnitude regardless of system grounding, and would require the same fault bracing on the downstream equipment. True, a single-phase ground fault would be of almost negligable magnitude, probably less than 20 amps, but unfortunately this also makes detection/relaying of that fault current almost impossible. And if you get a second phase ground-fault, your system is still coming down, and at a higher magnitude of fault current than the single-phase ground fault on the grounded system.

The overvoltage imposed on the system insulation during a single-phase ground fault means that such systems require a higher insulation rating. Overvoltages can range from 173% on a steady-state fault to 500% on an arcing fault (due to the capacitance of the system). That is the difference between breakers rated 277/480 and just plain 480 -- you should never apply a 277/480 volt breaker on an ungrounded system.

You are right that ground-fault protection per 215.10 would be required only on the solidly grounded system (assuming this is a 1000 amp service), but if it is ungrounded, then it would be strongly recommended to provide zero-sequence voltage relaying to help detect the fault.

If you ask me, the best of both worlds is to provide high-impedance grounding. This provides for easier detection of ground fault while still maintaining operations during a single-phase fault. This configuration is not widely used due to the increased cost and size, except where very high reliability power is required. If you're not going to do high impedance grounding, then I'd recommend the grounded neutral.

The IEEE Red Book, Chapter 7, backs most of this up.

 

[electricpete]

I have no disagreement with what you have said peebee.

Although I work at industrial facility with a very reliable ungrounded 480 volt system. I'm no code expert but I don't think we are violating any codes.

You are correct that I was wrong is my characterization of impact on fault magnitude. Most unbalanced faults will be more severe in presence of a ground, including simple single line to ground fault. But 3-phase symmetric fault won't depend upon the grounding.

As discussed, the protective relaying will have to be looked at closely before adding a ground to a system designed as ungrounded.

 

[Codemaster]

Here are my comments for what they are worth:

Per ELECTRICPETE on 10/23, I have never heard of that phenomena. Most of the ungrounded systems that I have seen out here in California are ungrounded wyes but PG&E never used GE padmount transformers, usually Cooper.

Per WTHALL on Oct. 24, since voltage of phase B to ground is 0, it sounds as if they have a corner grounded delta system to me. First check the ratings of the existing breakers, many breakers are not rated for corner grounded delta (since voltage to ground on two legs are 480, which is much higher than 277). For a good discussion of this see the NEMA web site and look for this article, "Molded-Case Circuit Breaker Individual Pole Interrupting Capability". It is in their "white paper" section. Also I don't agree on feeding two phase loads with a three pole breaker, especially where a fault on one light will take out all the lights. It would be better to feed them with 2 pole breakers.

I agree with everyone that high resistance ground with fault detection and tracing is the best way to go for an industrial plant. Utilities here in California won't let you do that unless you are primary metered, i.e. you own your own transformer and secondary. Ungrounded delta is fine as long as you have fault detection and correct faults ASAP. Many VFD's are not recommended on ungrounded or corner grounded delta systems.

If phase B is presently grounded, they may find to their chagrin many areas in the plant where phase B is grounded unintentionally when the conversion to the new transformer is complete.

 

[electricpete]

Codemaster - If I am lucky the following link will provide some details of my misadventure with an ungrounded wye transformer.

http://groups.google.com/groups?hl=...e2&seekm=85jade$2k$1@nnrp1.deja.com&frame=off

 

[resqcapt19]

peebee,
You can't feed 277 volt loads from a high impedance grounded system. See 250.36(4). Also if the lighitng is 480 volts, the 42 pole limit does not apply. A 480 volt lighting circuit on a 480 volt wye system is not a "lighting and appliance branch circuit". See 408.14. The 42 pole limit only applies to panels that have more than 10% of it OCPDs protecting "lighting and applinace branch circuits".

wthall,
Article 408 in the 2002 NEC was Article 384 in previous codes.

Don

 

[busbar]

Determine if the voltages are from intentional Bø grounding or from a temporary condition.

Aside from IEEE std 142 {Green Book} another good overview is: John P. Nelson and Pankaj K. Sen, High-Resistance Grounding of Low-Voltage Systems: A Standard for the Petroleum and Chemical Industry, IEEE TRANSACTIONS ON INDUSTRY APPLICATIONS, VOL. 35, NO. 4, JULY/AUGUST 1999

 

[jbartos]

Suggestions:
1. Check the local laws and codes. They may require the solid grounding of the transformer secondary wye. E.g. Electrical Code of the City of New York requires the solidly grounded system neutral.
2. The solidly grounded system has various advantages:
2a. Allows dual voltage system, e.g. 480/277V
2b. Stabilizes the voltage power supply in terms of the neutral and voltage stability (no circulating neutral)
2c. Trips faults fast.
3. Disadvantage:
3a. Relatively large fault that can create some local destruction.
3b. Good grounding path and its continuity is mandatory.
3c. Etc.
4. High resistance system grounding may be considered for the reduction of damages due to ground faults and continuity of the power supply in case of one fault only.

 

[peebee]

You are absolutely right, resqcapt, no 277 from a high-impedance ground.

Wthall, you need to pin down the existing installation details and the proposed new design criteria to figure out the best solution. You've indicated that the existing installation is both ungrounded and b-phase grounded, and that you're looking either 277 or 480-volt lighting. You gotta clear up those points of confusion.

 

[wthall]

The current system is supposed to be ungrounded. They have a fault somewhere in the system on B phase. I have their electricians double checking the voltage measurements even as we speak.

Their lights are 480...fed from combinations of the three phases like AB, BC, AC.

We just wanted to run a neutral from the new transformer to the new switchgear as an option. Much more cost effective to do the labor up front. There are no 277 V loads now...but if they want to add one in the future, then they may do so.

The client wants a grounded wye system because they believe it to be safer to personnel. I've been told of stories where the "one free fault" has been found by accident and the electrician has been shocked.

 

[busbar]

Don't forget that using 480Y/277V breakers or 300-volt fuses on other than a solidly-grounded wye system is a severe and sometimes deadly misapplication.

 

[peebee]

It is unlikely that the existing switchboard has a neutral bus. It is also unlikely that a neutral bus could be retrofitted without violating the UL listing (check with the manufacturer).

You may need to replace the switchboard to get neutral-wire distribution as well as a place to bond the neutral to ground. Depending on where and how the neutral is to be bonded to ground, you could possibly replace the switchboard in the future, and leave the new transformer ungrounded in the time being.