Separately derived system with multiple gensets

Status
Not open for further replies.

RUWired

Senior Member
does a service have to be supplied by the utility company?
definitions;
Service. The conductors and equipment for delivering electric energy from the serving utility to the wiring system of the premises served.
handbook said:
The definition of service includes the statement that electric energy to a service can be supplied only by the serving utility. If electric energy is supplied by other than the serving utility, the supplied conductors and equipment are considered feeders, not a service.
 

Smart $

Esteemed Member
Smart, Rick,

I'd like your opinion of 250.20(D), in particular the new text added in 2008 NEC. Oddly

it's under (D) Seperately Derived Systems, but, IMO, it addresses this situation by stating

( in short ) that to be a non-SDS the grounded conductor of the on-site Gen.(s) are to be

solidly connected to the "service grounded conductor " or it is a " derived system " and

grounded per 250.30(A), which is Grounding SDa/cSystems. There is no service at this site

, so no grounded service conductor, therefore must be treated as a SDS. Opinions?
I've been up and down, over and around this situtation many times. The answer always comes back to intent... which is a grounded conductor shall only be grounded at one location in the system (with very few and stringent exceptions). In the case of two generators with a transfer switch, and no service on site, the grounded conductor can only be bonded at one location if the transfer switch does not switch the grounded conductor. If the transfer switch does switch the grounded conductor, then both grounded conductors on the line side of the transfer switch must be bonded individually to ground.

When the grounded conductor is switched, each generator is treated as an SDS. When the grounded conductor is not switched, both generators are treated as a non-SDS (which is to say the system grounded conductor gets bonded as if it were a service).

FWIW: Judging from the NEC 2011 ROP Draft, 250.20(D) is being omitted in its entirety.
 
Last edited:

benaround

Senior Member
The issue I have with the non-sds setup is the potentionally fatal circumstance it sets up.

Generators used every day will require maintenence much more than Back-up type.

A person goes to work on the gen. set, the one with the N/G bond, he opens the gen.

disconnect, the non-sds gen. is running. For whatever reason he lifts the EGC or the N as

part of his tasks, a ground-fault on the system puts this person in harms way as this is the

only path to the other gen.s' N.
 

Smart $

Esteemed Member
The issue I have with the non-sds setup is the potentionally fatal circumstance it sets up.

Generators used every day will require maintenence much more than Back-up type.

A person goes to work on the gen. set, the one with the N/G bond, he opens the gen.

disconnect, the non-sds gen. is running. For whatever reason he lifts the EGC or the N as

part of his tasks, a ground-fault on the system puts this person in harms way as this is the

only path to the other gen.s' N.
IMO, if the two genies are set up in non-SDS mode, a service-entrance-rated transfer switch should be used, and the N/G bond would be at the switch. Lifting any or all conductors at either genset while the other is operating (and the off one's MCB at transfer switch is opened) should pose no threat.
 

benaround

Senior Member
IMO, if the two genies are set up in non-SDS mode, a service-entrance-rated transfer switch should be used, and the N/G bond would be at the switch. Lifting any or all conductors at either genset while the other is operating (and the off one's MCB at transfer switch is opened) should pose no threat.
Smart,

Is there an exception to bonding n/g past the first disconnecting means ( gen c.b. ) ?
 

benaround

Senior Member
Not really an exception but rather a subsection: 250.6(B).
Where are you going with this ? Why would you bond N/G at the TS ?? it could be 500'

away from the source ( generator(s) ) Main Disconnect. You need 5 wires from 1st.

disconnecting means (at gen. ) to TS in any case.
 

Smart $

Esteemed Member
Where are you going with this ? Why would you bond N/G at the TS ?? it could be 500'

away from the source ( generator(s) ) Main Disconnect. You need 5 wires from 1st.

disconnecting means (at gen. ) to TS in any case.
I'm not going anywhere with "this". You asked a question... I answered. What more do you want?

Why would it make any difference if the TS and gennies are 500' apart?

The concept is to provide a ground fault path and the grounded conductors suffice. Also, the supply conductors are not service conductors, so 5 wires (ABCNG) are required regardless of the gennies having a disconnecting cb.
 

RUWired

Senior Member
Smart,

Is there an exception to bonding n/g past the first disconnecting means ( gen c.b. ) ?
I know this question went out to Smart$ but i have a thought on this.
If the requirement to in stall the SBJ is at any point on the SDS from the source to the first OCPD, then the SBJ has to be installed at the generator because the "system" does not include any other pieces of equipment other than the derived conductors and the OCPD. The transfer switch in between is not part of the system. I'm not sure if i can support the terminology of whats included in the "system", but i seems to make sense.
 

Smart $

Esteemed Member
I know this question went out to Smart$ but i have a thought on this.
If the requirement to in stall the SBJ is at any point on the SDS from the source to the first OCPD, then the SBJ has to be installed at the generator because the "system" does not include any other pieces of equipment other than the derived conductors and the OCPD. The transfer switch in between is not part of the system. I'm not sure if i can support the terminology of whats included in the "system", but i seems to make sense.
In the case of a solidly connected neutral at the transfer switch (i.e. 3-pole TS), both gennies and the TS are of the same system.
 

RUWired

Senior Member
In the case of a solidly connected neutral at the transfer switch (i.e. 3-pole TS), both gennies and the TS are of the same system.
So it wouldn't really matter that the SBJ was down stream of the gen set. LIke in a typical non SDS and a s/n transfer switch the MBJ in the service equipment is used. The only difference is the OP's down stream SBJ is not part of service equipment.

Rick
 

Smart $

Esteemed Member
So it wouldn't really matter that the SBJ was down stream of the gen set. LIke in a typical non SDS and a s/n transfer switch the MBJ in the service equipment is used. The only difference is the OP's down stream SBJ is not part of service equipment.

Rick
That's the way I see it (if I interpret your comment correctly).
 

benaround

Senior Member
So it wouldn't really matter that the SBJ was down stream of the gen set. LIke in a typical non SDS and a s/n transfer switch the MBJ in the service equipment is used. The only difference is the OP's down stream SBJ is not part of service equipment.

Rick
Rick, It would matter, NEC says SBJ is installed anywhere from source to 1st disco. On a

typical set-up like you described, the MBJ moves from the 'old main panel' to the TS, and

the neutrals and grounds get seperated at the 'old main panel'.

From the gen to the TS to the "sub-panels " are all feeders (5 wire ), you never bond the

N/G on feeders. Where is the, Effective G.F. Current Path in the above example ? I can't

believe this conversation!!
 

Smart $

Esteemed Member
... From the gen to the TS to the "sub-panels " are all feeders (5 wire ), you never bond the

N/G on feeders
. Where is the, Effective G.F. Current Path in the above example ? I can't

believe this conversation!!
So tell me how you would bond N to G and avoid objectionable current using a 3-pole TS??? For the sake of discussion a 4-pole TS is the last-resort option.
 

RUWired

Senior Member
I have to side with Frank on this because there is no prohibition on where to put the SBJ. I would be able to install 20 panels downstream of the transferswitch; install the jumper and put in anothet 20 panels never knowing where the jumper is later. At least with a non SDS you know the jumper will be back at the service equipment. It seems that the combo of gen sets or single gen set are SDS's untill tied into the utility as 250.20D suggests. Rick
 

Smart $

Esteemed Member
I have to side with Frank on this because there is no prohibition on where to put the SBJ. I would be able to install 20 panels downstream of the transferswitch; install the jumper and put in anothet 20 panels never knowing where the jumper is later. At least with a non SDS you know the jumper will be back at the service equipment. It seems that the combo of gen sets or single gen set are SDS's untill tied into the utility as 250.20D suggests. Rick
So common sense doesn't tell you the bond would likely be where the supply enters the structure or before? Granted, you will not have "service" equipment in the sense there is no service... but there will likely be service rated equipment used. I suggested earlier to use a service-entrance-rated TS. Wouldn't that give you a clue as to where to bond? If you didn't know the site was running strictly on gennie power, would you not assume one supply of the TS was POCO?

If you used a service-entrance-rated, 3-pole TS and had POCO power on one side and gennie power on the other, how would you bond? Now replace the POCO power with another gennie. What do you have?
 

benaround

Senior Member
You have a gennie with a main OCPD, using your example, if the POCO supply was a

meter/main combo, would you still use the TS for N/G bond ?
 

RUWired

Senior Member
So common sense doesn't tell you the bond would likely be where the supply enters the structure or before?
To me common sense or logic would say a generator or transformer are seperately derived untill the grounded conductor of the SDS is tied into a system with a grounded conductor and that systems grounded conductor is always made at the source or first OCPD. If I had one gen set it would be SD. If I add another gen set to the system and add a switch with a solid neutral the second set would be non SDS. So it would make sense to install the jumper in one or the other like when having a gen set that feeds two separate buildings with utility to each.
 

Smart $

Esteemed Member
To me common sense or logic would say a generator or transformer are seperately derived untill the grounded conductor of the SDS is tied into a system with a grounded conductor and that systems grounded conductor is always made at the source or first OCPD. If I had one gen set it would be SD. If I add another gen set to the system and add a switch with a solid neutral the second set would be non SDS. So it would make sense to install the jumper in one or the other like when having a gen set that feeds two separate buildings with utility to each.
You can do it that way, and be code compliant. But consider the following: a ground fault happens while the second gennie is operating, the fault current must go to the first gennie on the grounding conductor, then return to the TS on the grounded conductor, then on to the second gennie on its grounded condutor.

What if the gennie's are 500' away as Frank asked? Your ground fault current path is 1500' from structure to second gennie.

What if this event occurs during maintenance or replacement of the first gennie, as Frank brought up? The only way to eliminate the hazard is to have the second gennnie shut down while performing those tasks... and that defeats the main the purpose of having the second gennie... right?
 
Last edited:
Status
Not open for further replies.
Top