Generator bonding with multiple transfer switches

[swnelson]

We're having a little disagreement in our office and want your opinion.

The design includes a 480/277 volt, 1200 amp service with GFP. It also includes a diesel generator that feeds two automatic transfer switches from separate output circuit breakers on the generator. One ATS is four pole with a switched neutral and it feeds a 480/277 volt life safety lighting panelboard. The other ATS is three pole with no neutral and it feeds a motor control center with only three phase 480 volt loads.

Our disagreement is whether the generator is a separately derived system and does the generator neutral get bonded to ground. My opinion is the generator is a separately derived system and the neutral must be bonded. If the neutral isn't bonded and there is a phase to ground fault the generator breaker will not operate because there is no path for the fault current to get back to the neutral point in the generator.

Another of our engineer's opinion is that when we add the three pole ATS, the generator neutral bond must be lifted. If we don't, this will create two service ground points and will be a violation of NEC.

Who is correct?

 

[ron]

Re: Generator bonding with multiple transfer switches

I will begin by dancing around the answer.
What type of GFP is installed at the generator. zero sequenece or 3-wire (where are the CT's located to trip the breakers)?

 

[Nick]

Re: Generator bonding with multiple transfer switches

You are correct. If you lift the bond at the generator neither transfer switch will have an effective fault clearing path when fed by the generator. The fact that you are not carrying a neutral with the 3 pole transfer switch saves the day. If you were, you would be creating parallel current paths, de-sensitizing GFP etc. The lack of a grounded (neutral) conductor in the 3 pole switch is the key.

 

[ron]

Re: Generator bonding with multiple transfer switches

After thinking about it more, and reading Nick's post, I'll agree that since there is no neutral brought to the 3 pole xfer switch, and the four pole switch is by nature not a solid neutral connection, the n-g bond needs to remain at the gen too.

[ August 20, 2004, 07:54 PM: Message edited by: ron ]