Re: NFPA petition decision on 2014 NEC 445.20, for portable generators <15kW
Guy, et al.,
On pp. 10 of the Canadian report you cited in your post and on your website, the wording indicates that a 0.6mA fault current to ground was observed. This, of course, would not trip a GFCI that was thresholded at 5mA. The exact wording that follows in the Canadian document is:
Briefly playing the devils advocate here, the wording above not conclusive that the GFCI failed to do its job (i.e. trip at 5mA fault current). Now, if the wording said that the fault current increased to above 5mA and the GFCI did not trip, then I this would be clear cut. The reason that the GFCI did not trip in the first case is that there was not a sufficient fault current return path to exceed the 5mA threshold. In the case of this primary fault, insufficient current flows to trip the GFCI as there is no current path. This is not an unsafe condition for the generator user, presuming we are comfortable with the 0.6mA leakage current level.
The Canadian report acknowledges as much on pp. 14:
But the report then goes on to discuss the dual fault condition on pp. 15:
In my extensive email discussions with PGMA, I made essentially the same double fault case argument as above, though the Canadians crafted it more succinctly.
I think you can make a further argument that the as-chosen wording for 445.20 going forwards is in violation of NEC 250.4(B)(4) "General Requirements of Grounding and Bonding - Ungrounded Systems." From the 2014 NEC code draft:
I also think an argument could be made for the TIA 1117 wording for NEC 445.20 as violating 2014 NEC 250.21(B)(1). Again from the 2014 NEC code draft:
Said detectors are to indicate the first ground fault condition of a normally ungrounded system, so that the operator can address this before a second fault condition causes a problem.
