The engineering side of standby or portable generator power sources

Status
Not open for further replies.
F

fastline

Senior Member
Location
midwest usa
Occupation
Engineer
It dawned on me through some inet chats and my experiences with other pro installs that I might need to run down this rabbit hole for a minute to reaffirm my basis. In my profession, I am taught not to simply rely on "codes" as a basis for doing things, but use the "intent" and my own brain and education to ensure safety. Not that I think NEC is in any way wrong, but may not cover everything, in which people seem to find a shoe that fits and just run with it.

Many people install portable gensets on their homes for an emergency. I am not talking proper gensets, I am mostly talking about he Lowes variety. Everyone clings tight to this "bonding only in the load center past the meter" thing as an excuse to either disbond the generator, or run that ground wire right in. They use say 10/3 to a receptacle, maybe an interlock if you want to be proper, and code man signs off. However, if you connect a bonded generator this way, you will have current on the ground conductor between the generator and panel. That is not cool. Further, there seems to be some notion that if the generator is "grounded", it will create some sort of elusive ground loop.

I don't think people realize utility 1P power is 2 hots and a neutral. The utility establishes ground many times between power generation and distribution to avoid errant ground currents. The point being I could never agree to disbond a generator chassis from the neutral for obvious safety reasons! But this is somehow recommended!

Obviously disbonding in a load center is an option for a genset, but not always practical. What I have done many times is NOT even run a ground conductor between the genset and load center! There is a certain level of common sense here that anything conductive needs to be grounded but when you think about how people run these, they set a generator that is grounded to itself, never drive a ground rod, run current through a ground conductor, and call it "code".... It is a guarantee that if you carry a ground conductor between genset and panel, and do not disbond, there will be current on the ground conductor.

As well, it is ultimately about keeping the neutral at ground potential through proper earthing thus a ground rod for a genset!

I am curious what you guys run into and accepted practices? I just cannot agree with disbonding a metal chassis genset and let that chassis float! Can you imagine someone playing with that in a storm? But it also stands to reason that using an insulated ground conductor on a genset and letting it carry to a panel AND drive another ground rod may be the safer alternative. People talk about all these wild ground currents, but I think missing the main reasons we don't bond in other places. We do not want current on a ground!
 
If you remove the neutral to frame bonding jumper in the generator the metal frame and all of the metal parts get connected to the EGC in the 4-wire cord feeding the inlet making it a non-SDS. I don't see an issue with that.
 
But what are you really doing there? you disconnect a nice solid ground path so you can make it take a tour that could become unreliable? the second you unplug, you have no ground at all! Now, this is totally common in small inverters, but they are all plastic and designed this way. I also see PRO installs not using an insulated ground. Is there an NEC or NFPA reference on this?
 
Let's say that the generator has the neutral bonded to the case and you don't have an EGC connected from the generator to the panel, as was mentioned in post #1. What happens if the neutral conductor between the generator and panel has an open circuit? Since both the neutral and EGC are disconnected in this case, let Np be the panel neutral and Ng be the generator neutral, Gp be the panel EGC, and Gg be the generator EGC. As always with an open neutral, the L1-Np and L2-Np voltages in the panel can get significantly unbalanced depending on how well the loads on the two busses are balanced. Under the worst case scenario one L-Np could be 0V with the other L-Np at 240V if it doesn't have any loading on that phase. And so, for example, if L2-Np = 240V then L2-Gp also will be 240V because of the panel's bond. In other words there will be 240V between L2 and the grounding electrode conductor (GEC) of the home. But with the broken neutral conductor, the L1-Ng voltage and the L2-Ng voltage will still remain balanced with each at 120V on the generator. That means that Ng will be at 240V - 120V = 120V relative to the GEC. And because of the bond from Ng to the case of the generator, the case will be at 120V relative to the earth potential. That possibility seems worse than having currents flow in parallel paths on the neutral conductor and ECG.

Now consider the case where a portable generator without a bond to neutral is not connected to an inlet on a home. If someone uses the 20A receptacles that have GFCI's, the GFCI's will not trip with an external fault unless there's also a fault from the input side of the GFCI device to the case of generator (which is what a N-G bond would do). It has been argued in other threads whether this is a problem or not, and whether you might provide a switch that would select whether neutral is bonded to the generator case or not.

I think another option would be to have a resistor from N to EGC in the generator. For example, with a 1K resistor the current through the EGC "parallel path" when connected to an inlet would be negligible (one milliamp for each volt from N to EGC).
However, a bolted ground fault on a circuit powered by a 30A generator receptacle without GFCI protection would not cause the internal breaker to trip without a solid N-EGC bond.
 
Last edited:
Well there is little difference between talking about a neutral loss from utility power and any genset. As well, I beg to differ on a GFCI failing to trip. It actually would not even matter if a genset was bonded or not. The only thing that GFCI will know in a home's main panel is current ended up taking the "ground route", which it should not do. But ultimately that panel is bonded so all those grounds and neutrals end up returning to the generator on the neutral conductor, just like utility power. We can talk about neutral losses but you can't give it any more regard than utility. It happens!
 
There is another similar thread on this subject which I can't find.
Over the last few code cycles the NEC, now has very clear information in Section 250.30, 250.34, and 702.7.
Basically you match the transfer switch type to the bonding of the neutral in the generator. Certainly most homeowners don't understand the signs required in 702.7 and some electricians are likely confused.
Mike Holt has information on generators here:
www.mikeholt.com

Free Resource - Generators and Standby Power Systems

CLICK HERE to read the newsletter; please share it with your colleagues.
www.mikeholt.com
 
Status
Not open for further replies.
Top