Hartwell on the Code: Is a meter disconnect switch a service disconnect (and so what)?

This first part of a two-part series looks closely at a utility initiative to allow certain "meter disconnect switches" ahead of service disconnecting means. Their proposal (4-129) has been accepted for the 2002 NEC, and now awaits public comments before it takes final form. Many local utilities are already requiring use of these disconnects in service conductors, which means you need to be thinking about the Code issues they present even now. This is because these disconnects can be installed even without a code change, provided they are installed as service disconnects with overcurrent protection and not considered to be some other equipment that happens to be ahead of a conventional service disconnect.

When would such a service disconnecting means not be a service disconnecting means? If it were, as mentioned, a "meter disconnect switch" ahead of the supposed actual service. The change, as now accepted (with new wording underlined), makes Section 230-82(2) read: "(2) Meters, meter sockets, or meter disconnect switches, nominally rated not in excess of 600 volts, provided all metal housings and service enclosures are grounded in accordance with Article 250." This is the section that details what equipment can be connected ahead of a service disconnect. The meter socket reference simply corrects an obvious oversight and isn't controversial, but the other wording is a major change.

Meter disconnects have been around for a very long time, normally consisting of a multipole circuit breaker mounted within a multifunction meter enclosure or in a self-contained metering pedestal. Theoretically a manufacturer could make any of them as convertible to either "hot sequence" (meter ahead of switch) or "cold sequence" (switch ahead of meter) in the field, to suit local utility requirements. At present, most of this market consists of hot sequence units that aren't field-convertible (see drawing). If these breakers are on the load side of the service point (the usual case), and if they provide overcurrent protection for the conductors they supply (also the usual case), then what they supply is a conventional feeder, and not a continuation of service conductors.

Why worry about classifying them as service disconnects? One major reason is — if they are so classified — a grounding electrode would have to be provided at the metering point. If the meter is on the outside of the building, it isn't a big problem. If the meter is hundreds of feet away, it would involve an additional electrode that would meet code but accomplish very little in terms of safety, since there would be no electrical loads at the remote metering point. It would be like requiring a grounding electrode conductor to be brought to every conventional meter socket.

There is another major impetus for this proposal, as cited in the substantiation. Utilities are showing increasing interest in cold sequence metering, especially on self-contained 480Y/277 Volt metering systems, because of the greater safety it affords their service personnel. Pulling a meter under load at 277 volts to ground can result in a severe arc, which is why the NEC has required GFPE on 480Y/277 volt services for the last thirty years. In fact a number of major utilities are now mandating simple switches ahead of such metering without overcurrent protection. (See drawing.) Remember that bypass switches in meter sockets are to maintain load continuity, not load interruption, and opening a meter bypass switch under load may destroy the meter socket.

Some utilities are also interested in this concept where the metering is to be at a roadside, with the service running to the building served typically using an underground wiring method. This is true even on ordinary 120/240 Volt single phase services to single family dwellings. Utility representatives point out, correctly, that here as well a remote disconnect adds an additional level of safety. Often electricians have been in the position of needing to pull a meter in order to deenergize service equipment in a flooded basement; a remote disconnect is much safer.

Unfortunately, barring clarification in the comment period, CMP 4's pending rationalization of this common (in some areas) utility practice for the 2002 NEC threatens enforcement chaos. Given that countless NEC rules depend on a common understanding of exactly where the service is, allowing two devices, often widely separated, on the same property that each potentially qualify as service disconnecting means will be extremely troublesome. In addition, the switch without overcurrent protection effectively creates, per long established code definitions, a feeder without overcurrent protection, which requires correlation with Section 230-91. This section requires service overcurrent protection to be located as an integral part of the service disconnecting means or immediately adjacent thereto. This doesn't invalidate the utility safety argument, but it does mean that the utility proposal as presented should be modified to accommodate these issues.

It was only in the last cycle that the following similar allowance was deleted from Section 230-82: "Fuses and disconnecting means or circuit breakers suitable for use as service equipment, in meter pedestals or otherwise provided and connected in series with the ungrounded service conductors and located away from the building supplied." The reason this provision was deleted (Proposal 4-159 in the 1999 NEC cycle) was that such disconnecting means are in fact service disconnects and the normal requirements in Part B of Article 225 should generally apply because the conductors they supply are feeders. Further, the existence of this provision (which originated in the 1971 NEC, long before building disconnects moved from old Section 230-84 to Article 225) was leading to confusion and inconsistent application of the rules because of conflicts with Article 225. That action was essentially correct.

The concluding part of this article covers another change pending in Article 230 that makes this even more complicated. It concludes with a detailed discussion of everything that would need to be changed (in at least two NEC articles) to really clear the way for this initiative to succeed.

Acknowledgements
The designations "National Electrical Code" and "NEC" refer to the National Electrical Code, which is a registered trademark of the National Fire Protection Association.

About the author…
EC Online columnist and nationally recognized NEC expert Fred Hartwell is widely recognized as one of the most prolific contributors to the NEC, with nearly one thousand proposals and comments over the years making it into print. He has three code cycles experience on one of the code making panels. He has many years of contracting experience in Illinois and Massachusetts. He also served as an electrical inspector for fifteen years and he was head electrician on a college campus. (Back to top)