Revisions to the UL standard governing ground-fault circuit interrupters have raised the bar for safety but brought into question who has the power to bring about change.


When ground-fault circuit interrupters were first introduced in 1972, more than 700 people were dying per year as a result of household electrocutions. In the 30 years since then, that number has dropped to 200, according to figures compiled by the U.S. Consumer Product Safety Commission (CPSC). GFCIs aren't solely responsible for that drop, but it still suggests a powerful correlation. Yet in an industry where consumer safety is one of the chief tenets of design and production, even a 71% reduction in electrocutions wasn't good enough. Representatives from CPSC and NEMA, manufacturers, and other members of the electrical industry began work late in 1994 to improve GFCI receptacles to reduce that number even further by strengthening the requirements put forth by Underwriters Laboratory (UL) in standard UL 943 (Sidebar at the bottom of the page). Spurred on by the results of a joint study conducted by NEMA and UL that revealed nearly 10% of GFCI receptacles in the United States were nonoperational, the various groups involved came to the table in the middle of 2001 to hammer out changes to the standard that would make the product even safer.

The changes at hand, which take effect Jan. 1, 2003, call for improvements to a GFCI's ability to handle higher surge current, a greater immunity to damp, corrosive environments, and the inclusion of a new feature that would cause the device to trip and cut off power indefinitely in the event of miswiring (i.e. line-load reversal) — changes that will most certainly make GFCIs safer, but their adoption, and more importantly, the denial of one proposal in particular have raised questions as to whether there were other motivations for making the changes.

The answers offered by most suggest a genuine desire on the part of manufacturers to better protect the public. But there are those in the industry who believe the situation highlights an increasingly prevalent issue: When industry standards are updated, groups independent of the manufacturing community routinely lobby for the inclusion of proprietary features, requiring manufacturers to incorporate them into their products and pay licensing fees along the way.

Simplifying the equation.

Eight years ago, Bill King, chief engineer for electrical and fire safety at the CPSC, approached then commissioner Ann Brown with a concern he and other engineers at the government agency shared: GFCIs weren't as safe as they could be. The devices could degrade and lose their ability to protect against lethal ground faults yet still provide power. If homeowners tested the devices once a month as UL suggests, they could detect any problems and have the receptacle repaired or replaced. However, King and Brown both knew consumers could neither be trusted to conduct those tests as often as recommended nor take the necessary steps to fix the situation if they found a receptacle to be faulty.

The two agreed the consumer should be given as little responsibility as possible in ensuring the safety of the product. If homeowners weren't going to conduct the tests as often as they should, there needed to be less of a reason to test at all. The devices should be strengthened against things like surges and corrosion so failure was less likely. In fact, why not remove the consumer from the equation altogether? Design GFCIs that could test themselves. The device the CPSC had in mind would effectively shut down and cease providing power if it detected a hazard. The technology was there, or at least close — GFCI manufacturer Leviton had already begun work on a receptacle that would lock and refuse to reset if installed incorrectly, and Pass & Seymour/Legrand has since then developed a similar power-denial technology — so the next step was getting the industry to support a standards change.

Brown, who now heads SAFE — Safe America For Everyone — informally presented the CPSC's case to members of NEMA at one of her first speeches as chairman in 1994. Looking back, she acknowledges that those in attendance didn't exactly jump out of their seats to applaud her assertions that GFCIs needed to be improved and equipped with indicator lights to show whether they were working properly. In fact, the response fell somewhere at the other end of the spectrum. “I must say, my talk wasn't met with as much happiness as I would have liked,” she says. “But I do think they have made great improvements to them [since then].”

In fact, NEMA, in general, and GFCI manufactures, in particular, could have chosen to ignore Brown's request — the CPSC has no regulatory power to effect changes to UL test requirements. Yet contrary to the initial hesitance Brown witnessed when she first suggested the change, NEMA collaborated with UL in January 2001 to collect and test a random sample of GFCIs across the country for failures attributed to surges and corrosion, finding that roughly 10% of installed receptacles were continuing to supply power without providing ground-fault protection (Table above).

It's important to note, though, there was no evidence included in the study to suggest any of those failed receptacles were responsible for an electrocution. However, the results of the survey were clear enough: One out of every 10 GFCI receptacles was failing to protect users from ground faults — a statistic unacceptable by anyone's standards. Even if there was debatable evidence as to whether one of those devices had electrocuted someone up to that point, there wasn't anything to say it wouldn't happen in the future.

The situation was exacerbated by the frequency of miswired GFCIs, in which the line and load side connections are reversed at installation. Receptacles wired backward will continue to provide power without ground-fault protection, and the consumer may not discover the problem until it's too late. For several years, UL has required manufacturers to place yellow tape on a GFCI's load side terminal to ensure those installing the devices knew which side was which. To miswire the device, the installer — whether an experienced electrician or a weekend do-it-yourselfer — would have to ignore the warning and deliberately remove the tape. Despite this added level of security, John Drengenberg, UL's consumer affairs manager, says the receptacles are still often miswired.

“We still have found that there seems to be a tendency to miswire,” Drengenberg says. “Now whether it's the homeowner who installs one in a lifetime and really isn't clear on the concept of line and load, or the electrical contractor who goes into a brand new apartment building and puts the GFCI in the receptacle and doesn't know which is the line and which is the load wires coming out of the conduit in the wall, it's still a problem.”

Together, the results of the NEMA study and the anecdotal evidence of miswiring was enough to bring the interested parties together to address Brown's concerns.

Coming to the table.

Changes to UL standards aren't simply unilateral directives handed down by the organization that manufacturers must follow or else. The process begins with a standards technical panel (STP) meeting open to everyone from manufacturers to government agency representatives to end-users at which ideas and proposals for changes to UL standards are offered up and debated. Once a list of proposed changes is agreed upon, UL sends a bulletin to all voting members, who may comment and offer suggestions. Drengenberg says all comments and concerns — regardless of size or scope — are addressed by the organization. After the comments are considered, the changes are put to a vote. Manufacturers are then notified of the new test requirements their products must meet. In this case, UL sent out notice of the changes in November 2001, giving manufacturers just over a year to make the necessary design changes.

In accordance with revised standard UL 943, all GFCI receptacles manufactured and sold after Jan. 1, 2003, must pass UL's new tests to receive the organization's mark. Contractors, take note: These changes will not change the appearance of the receptacle, nor will the UL mark offer any indication the receptacle you purchase after that date complies with the standard revisions. If in doubt, ask your distributor. However, Drengenberg points out that even those GFCI receptacles manufactured prior to 2003 and still bearing the UL mark are acceptable for installation after January 1st. Similarly, receptacles approved by UL and installed prior to 2003 need not be replaced with the newer models.

By and large, the proposed changes were met with approval from GFCI manufacturers. Representatives from Pass & Seymour/Legrand, Hubbell Wiring, and Leviton agree the updates were positive additions.

Bill Grande, product manager of safety products, Leviton, is chief among them. “All of these improvements — the resistance to surges and overvoltages and corrosion — should result in a much more reliable product with a lower instance of failure over time,” he says.

The changes to UL 943 put less of a strain on Leviton's product development teams than others — the company's SmartLock GFCI already met some of the new standards, including a feature that “locks out” the user if the receptacle is wired backward, only supplying power once the receptacle has been wired correctly. The device has been available since 1999, and after updating its surge resistance capabilities, Leviton is now offering the receptacle in full compliance of the 2003 standards. It even goes a step further and shuts down completely if a test finds it no longer capable of providing ground-fault protection; after such a failed test the reset button can't be depressed.

It's this very feature that is at the center of a debate between other manufacturers, UL, and the CPSC. King applauds the new requirements in the UL standard, but he says the commission believes they don't go far enough. “We're telling the industry, every manufacturer, and their trade association that we have higher expectations than where we're at right now,” he says. “We're pushing the application of current technologies through Code proposals, and then we're looking at the product itself. We're trying to raise the bar on the product.”

And in order to raise that bar the CPSC hoped to include a requirement in the updates to UL 943 that GFCIs employ a lock-out technology similar to what Leviton's receptacle offers. In a position paper drafted by the CPSC engineering staff and sent on April 20, 2001 to the STP overseeing the updates, the commission spelled out this desire to mandate a more “intelligent” GFCI.

“GFCIs can now be designed and manufactured to be much more reliable and to reduce reliance on consumers to determine if the shock protection capability is functioning properly,” the letter states. “GFCIs should be required to deny power when the reset mechanism has been actuated and the GFCI is inoperable. This requirement should have the earliest effective date for receptacle-type GFCIs since such ‘lock-out’ technology has already been demonstrated.”

John McFarland, product manager, Hubbell Wiring, is satisfied that the required changes offer two ways to address the issue of surge currents — either surviving the surge or locking out power to the receptacle if it can't. He says Hubbell has completed testing on its updated GFCI line and will be ready to send it to market by the end of the year, but is concerned by the differentiation regulatory and testing bodies make between safety requirements and competitive features. “If you have a patented feature on your product, you'd love to have UL make that a requirement because you're the only one who has it,” he says. “It's good business from their standpoint, but whether it's good for the customer is another thing altogether.”

James Ruggierri, P.E., a forensic analyst for General Machine Corp., who served on the STP for the revisions to UL 943, has other problems with the proposed lock-out requirement. “It seems like a good idea on the face, until you start looking at it more closely,” he says. “What will people do [if the receptacle locks out and shuts down]? They won't repair the GFCI in most cases. Most people will delay it, and they'll start running extension cords from the adjacent space, and that's a very sloppy fix.”

Regardless of the concerns people like McFarland and Ruggierri may have, Drengenberg dismisses the idea that groups lobbying on behalf of certain manufacturers are capable of influencing UL's final decisions with regard to standards changes. “Some manufacturers may have designs they favor and have even patented that they think are better, and for them, of course, it's a profit motive,” he says. “But ultimately, when we put our UL mark on a product, it's what we think had to be done to those products, and [it represents] the fact that it has to pass everything that we think it should pass before it carries our mark.”

Talk of lobbying on behalf of proprietary technologies, however, serves only to cloud the one issue that can be proven: The changes have made GFCI receptacles safer. Even more encouraging is the fact that manufacturers took a more proactive approach to solving the problem. Manufacturers and regulatory bodies often conduct cost-effectiveness studies to determine whether the threat of wrongful deaths and the resultant civil suits is justification for changing a product or the standards that govern it. If the estimated cost of such a suit is greater than the cost of changing the product, it's in the manufacturer's best interest to make the changes and avoid litigation. However, as was the case with GFCIs, if a product hasn't been proven to have caused a consumer death, it doesn't make financial sense to pay to make it safer. But that's just what GFCI manufacturers did.

Although he took issue with the CPSC's attempt to push for the lock-out feature, Ruggierri believes the industry's willingness to bring about the other changes is laudable. “This is a good example of industry putting their heads together and making changes to improve safety that are not casualty-based,” he says. “Generally, when federal regulatory agencies go to rule making, you won't see changes in the case of safety unless we're losing people everyday. In this particular case, there were no casualties.”

How long these new changes last, however, is another matter. King insists that CPSC's suggestion for a smarter GFCI should be taken more seriously and eventually adopted. According to Drengenberg, the standard can be revisited at any time — even before the latest revisions take effect — but he makes no promises about whether the issue will be addressed. “Our documents are dynamic,” he says. “If we find as time goes on that even higher levels of safety would be beneficial, we will revisit that for sure.”

But if nothing else, the fact that the revisions came about in such a positive way reveals the potential for more proactive approaches to safety in the future.




Sidebar: Changes to UL 943

After nearly eight years of discussion, planning, and testing, Underwriters Laboratory has released the following changes to its standard covering ground-fault circuit interrupters, UL 943:

  • A more stringent voltage surge test to ensure the GFCI can handle a higher surge current.

  • A corrosion test to demonstrate greater immunity to moist conditions.

  • An operating test to verify proper operation of the GFCI can't be prevented by manipulation of the GFCI controls.

  • A line-load miswire test that requires the device to trip when miswired.

  • An abnormal overvoltage test that requires the GFCI not to become a fire or shock hazard during extreme overvoltage conditions.

  • Increased requirement for the GFCI to operate properly after exposure to conducted radio frequencies.