time2roll wrote:
OK and does this mean a N-G bond? Or simply the chassis of the generator needs to be bonded to the RV frame.
I think it could have been worded more clearly. Sometimes NFPA standards and codes (of which NEC is one of many) has wording assuming the user of that standard or code is well versed in the application of that standard or code. I know, as I served on two committees of one of their other codes and we often had to rewrite some statements for more clarity.
A ground fault is typically, but not exclusively, a short between neutral and ground or chassis frame. Such a short would divert a portion of the NEU return current from the neutral circuit to the ground circuit. That would create the possible imbalance to trip a GFCI. So an N-G bond at the source (generator in this case) would provide the required path for a ground-fault detection.
A short between hot and ground would typically trip the respective circuit breaker. But, if that short was high enough resistance to restrict the errant current flow to a level below the circuit breaker rating, then a portion of the HOT current would be diverted through ground while the rest would continue through neutral. That would also cause the possible imbalance to trip a GFCI. Again an N-G bond would provide the required path for a ground-fault detection.
An additional note on a hot to ground short. When the ground is not bonded to neutral (open ground) the related circuit breaker would not trip on a hot to ground short as there would be no path for current flow from hot to ground and back to the source (the generator). That would mean the ground (including RV chassis) would be hot. If a person somehow touched chassis frame and neutral at the same time then current would flow through that person. If that short was downstream of a GFCI circuit then the GFCI would trip. If the short was on an unprotected circuit then there would be no protection. If there was an N-G bond there would be no danger as neutral and ground would be at the same potential.