Was doing a new construction inspection today and it just so happened that the city electrical inspector and I wound up at the panels at the same time. We were comparing notes on deficiencies and I mentioned that I had not yet seen a bonding connection to the gas piping. He had not either. I told him I’d take a look for it when I was in the mechanical room in case it was there and we each proceeded with our individual inspections.
He came and tracked me down later to tell me that according to his office, they have been permitting the electrical contractors to rely on the egc at the furnace as the bonding means. I pulled the make-up cover off the furnace so we could make sure that the unit was grounded and discussed it a bit further. The egc inside the furnace was only 14AWG and it was on the appliance side of the gas appliance connector. It just didn’t seem right to rely on this as the bonding means. He said he would follow up as did I.
Well the AHJ tracked me down a little later and handed me a sheet of paper and said the reference was in NEC article 250-104 (this was a good city inspector to work with).
So back at the office, I peruse the article and find this under subsection B
Interested in thoughts on this alternative bonding means from our resident master electricians. There was no separate jumper from the egc in the furnace cabinet for the metal gas piping so the control valve and the flexible appliance connector with be part of the bonding path. It just doesn’t “feel right” to me. Is a separate jumper required from the metal gas pipe to the egc when doing it this way?
I don’t understand Chuck. Isn’t the NEC saying that the equipment would most likely energize the piping and so the grounding system for the equipment is what would take the majority of the current? So in theory, the egc in the equipment and the grounding conductor in that circuit would be in parallel with the gas piping if a fault were to occur, but the grounding conductor would provide a much better path and so the majority would go through it. Even though the control valve and the flexible line are in the path, the amount passing through them would be minor.
The NEC allows the EGC in the branch circuit of a gas appliance to bond the metallic gas pipe. No other bonding jumper is required. If there were no electrically connected appliances then no gas pipe bonding would be required. All of this changes with CSST since the manufacturer may require additional bonding.
Rob,
I have been in a discussion with some folks about the requirements with CSST gas piping. Being that the bonding requirements are found in the manufacturers installation instructions and the mechanical contractor is the one who is doing the installation how can an electrical inspector hold the electrician responsible for the installation of the bonding?
It is not the electrician that is making the installation of the CSST and might not even know that CSST is being used. Personally I have never seen the installation instructions of CSST pipe while on a job. The only time I have ever saw the installation instructions is on web sites such as this one or in articles written on the subject.
Should I be doing a job where CSST is being installed all I have to do any bonding by is the NEC unless someone approaches me with the instructions concerning the bonding. I think this needs addressing somehow in the NEC or the CSST pipes needs to be outlawed, what are your thoughts?
Just to be clear for myself. Does the egc need to be connected directly to the gas piping in this scenario or is the bond simply relying on the chassis of the appliance to have a conductive path the the gas piping within the appliance?
Regarding there not being a need for bonding if there are no electrically connected appliances: I always thought, perhaps incorrectly, that bonding was also important in the case of a lightning strike, hence the connection to metal water piping, metal framing, etc. Is this not the case?
Not concerned about the special requirements for CSST - I’m familiar with those.
Thanks Jim. That makes sense. I think the city inspector was going by “if the furnace had an egc, it was OK” The furnace egc was not connected to the gas pipe. I still wrote it up as deficient even though the city inspector let it go.
This was helpful - I still learn something new every day.
I agree and I’ve been echoing your sentiment since the entire CSST problem surfaced, that it’s not the EC’s responsibility. Here in NJ, the State has made it the electricians responsibility.
Manufacturers and most other research studies state a direct bond is required and that indirect is not adequate. That is I go by when CSST is installed.
I think so, appreciate you following up. This stuff is interesting and confusing at the same time. You said you had written it up as deficient, do you think that still after the discussion?
Yes. Because there was no direct connection from the egc to the metal gas piping. As configured, it would be dependent on their being adequate continuity from the egc to the furnace cabinet through the gas control valve assembly and the flexible appliance connector before we actually got to the metal gas piping system.
Although, in the future, if I see one with a connection from the furnace egc to the metal gas piping, I will let it go. Prior to this discussion, I would not have. I have never seen one configured this way however.
IMO it’s not required. If you look at the NEC section that Chuck posted the requirement is for gas piping that will likely become energized which would be by the circuit used to power the gas appliance. One may even argue that the word likely is so strong that even bonding though the appliance EGC is not required in some cases.