Sign In  |  Register Now  Why Register?

The Technical Correlating Committee (TCC) handed CMP-9 a hot potato for the 2002 code cycle in the form of a series of comments on a 1999 NEC proposal to CMP-2 [in Section 210-70(a)]. The proposal would require that an occupancy sensor set in its "off" position either completely disconnect the ungrounded conductors or contain an engraved notation indicating the likelihood of downstream voltage when in that position. The original proposal responded to the use of occupancy sensors as a replacement for wall switches in dwelling unit applications, as covered now in Section 210-70(a)(1) Exception No. 2. However, the issues clearly have a broader application, and the TCC responded accordingly.

Why do electronic switches leave a voltage on its downstream side? Not all do, actually. Those designed for use with a grounded circuit conductor present usually won’t do this, because the representation in the switch location of both sides of the circuit means full power is always available for the electronics in the switch, even in its off position. Nevertheless today, even if the grounded conductor is available, manufacturers reported that conventional designs shy away from the traditional air-break design due to size constraints.

In addition, an electronic switching device essentially designed to replace a single-pole switch with no grounded circuit conductor present has a much tougher problem. With the switch off, obviously you want it to be able to turn back on without walking over to it (or why use an automatic one at all). If the conductor that runs to the light is completely open, then there isn’t any potential across the electronics of the switch. No potential means the switch loses its brains.

The only solution in all these cases is to allow a very small amount of current to flow through the load when the switch is nominally "off." The product standards allow these switches to pass small amounts of current (typically 0.5 mA) through the load, for just this reason. Other designs pass the same current into the equipment grounding system (also allowable), but the issue here involves the ones that use the load. This amount of current isn’t enough to pose a shock hazard; however, it can be a startle hazard. Suppose you need to change the ballast, for example. If you don’t shut off the circuit breaker (preferably under a lockout/tagout protocol) and treat the electronic device the same as a dumb switch instead, you’ll get a nasty surprise once you disconnect the ballast. The ungrounded supply lead will be hot.

Remember also that this only happens when the ballast is disconnected. When it is connected, the current flow is so low that there isn’t any meaningful voltage drop across the ballast; almost all the voltage drop occurs across the occupancy sensor. The residual voltage to ground at the ballast supply lead isn’t enough to indicate as a problem on most testers. CMP-9 received documentation of instances where workers were startled to the point of falling off ladders when this happened, after turning a sensor marked "on-off-auto" to its "off" position and then conscientiously testing for voltage at the ballast.

Often economic pressures make disconnecting the branch circuit problematic, particularly on 277V installations where one circuit might supply literally dozens of small individual offices. What to do? The original proposal suggested these switching devices come with an engraved warning indicating the presence of full voltage in the "off" position.

In the action on proposals for the 2002 NEC, CMP-9 agreed that "off" should mean off, that is, fully disconnected. However, the panel wasn’t willing to require the engraved warning on devices often too small to carry it legibly. The solution lies in precedents located elsewhere in the NEC, particularly Section 424-20(a).

Not all room thermostats for electric heat open all ungrounded conductors. The Code allows for this, simply prohibiting such devices from having a marked "off" position. We’re used to this provision, and we know not to work on room heaters as if these devices qualified as disconnecting means. The panel applied the same concept in this case. Section 380-15 will contain (subject to further action in the comment period) a new subsection assuring that any switching device that has a marked "off" position must be one that completely disconnects all ungrounded conductors to its load.

The rest, of course, will be up to field personnel who will need to learn that a switching device that shuts off the lights might not be actually disconnecting its load, unless it actually says "off." Furthermore, older occupancy sensors (this concept is now in the product standards, but there are many older designs still in use) may say "off" and not actually disconnect their load.

The best idea is always to find some way to shut off and tagout the circuit. Recall the admonitions on most dimmer switches suggesting their circuits always be installed and worked cold due to the likely dimmer element failure if it were subjected to fault current; occupancy sensor manufacturers reported the same preference to CMP-9. If you can’t do this, then be aware of the limitations of these devices. They are being installed in enormous quantities as energy saving initiatives take hold. At least "off" will begin to mean off once again.

Anthony Montouri, employed by New York Board of Fire Underwriters, has served on Electrical License Board and the Advisory Board of the Department of Buildings of New York City since 1984. He is also a member of the Steering Committee and Panels 5 and 21 of NYC Electrical Code Revision. Montouri represents IAEI on NEC Code Making Panel 9, and is presently International 2nd Vice President of IAEI.

Home  | Join | Advertising | Seminars | Contact Us | Privacy Statement | Legal Notices
Copyright © 1997-2008 IAEI. All Rights Reserved.