Standard Interpretations - (Archived) Table of Contents|
| Standard Number:||1926.505|
February 16, 1994
|MEMORANDUM FOR:||ALL REGIONAL ADMINISTRATORS|
|THROUGH:||H. BERRIEN ZETTLER, DEPUTY DIRECTOR|
DIRECTORATE OF COMPLIANCE PROGRAMS
|FROM:||ROY F. GURNHAM, DIRECTOR|
OFFICE OF CONSTRUCTION AND MARITIME
|SUBJECT:||Testing Ground Fault Circuit Interrupters|
The purpose of this memorandum is to reiterate the policy for testing ground fault circuit interrupters (GFCIs) used at construction sites.
GFCI testers which are not an integral part of the GFCI shall not be used as a means of determining compliance with 1926.404(b)(1)(ii) because such testers often do not produce accurate results. Only the testing circuit which is manufactured as part of the GFCI shall be used to determine whether or not the device will function as intended.
A copy of the Electrical/Electronic Technical Note 91-1, authored by the Office of Electrical, Electronic, and Mechanical Engineering Safety Standards, is attached for additional information and guidance.
If you have any questions, please contact me or Dale Cavanaugh of my staff at (202) 219-8136.
December 18, 1991
How to Test the Operation of a GFCI
The Office of Electrical, Electronic, and Mechanical Engineering Safety Standards does not recommend the use of GFCI testers as a means of determining compliance with 1926.404(b)(1)(ii), as such testers may not produce accurate results. Ground-fault circuit interrupters incorporate a testing circuit that can be used to determine whether or not the device itself will function as intended. No further tests are necessary.
Any GFCI tester that puts a resistive load between the ungrounded circuit conductor and the equipment grounding conductor(1) to measure the current at which the device trips is subject to errors due to voltage fluctuations. If the circuit voltage is 100 volts, the tester could indicate that a GFCI tripped at 7.2 mA when it would actually have tripped at 6.0 mA. (See attachment.)
Testers like the Greenlee model 5708,(2) described in Mr. Loebach's memorandum, cannot produce a reliable indication of the trip level of a GFCI. This device sends a 200-millisecond pulse through the grounding conductor at various current levels. A GFCI may not trip at minimum current levels (that is, 6-20mA) in such a short period of time. (For example, UL Standard 943 allows trip times of up to 1.5 seconds at 15mA.) This tester provides a 4-second interval between pulses and cannot be adjusted to provide a longer pulse or a shorter interval.
Additionally, an employer cannot reasonably be expected to know at what level his or her GFCIs trip. A reasonable person would only expect the employer to check them periodically using their built-in test mechanisms. Assuming that test equipment could accurately detect a GFCI that trips at too high a current, the employer should not be penalized for conditions beyond his or her normal control. A citation in such cases may be warranted--penalties are not.
The supervisory circuit built into a ground-fault circuit interrupter is designed to cause tripping even when the circuit voltage is 85 percent of rated voltage (102V for a 120-volt device). At rated voltage, the current employed by the supervisory circuit may not exceed 9mA. Thus, it gives an indication of the operability of the GFCI at currents approaching the trip level. It may not give an indication of whether the GFCI actually opens the circuit; however, this can easily be determined by plugging utilization equipment into the circuit in question.
For these reasons, the Office of Electrical, Electronic, and Mechanical Engineering Safety Standards recommends that the compliance staff use the test button on a GFCI in combination with an attached load plugged into the circuits to be tested rather than a GFCI tester. A plug-in ground continuity tester would suffice as an attached load. If the lights on the continuitytester go out when the test button is pressed, the GFCI can be assumed to be operating correctly. If the lights stay on or if the test mechanism fails to operate, the GFCI is faulty, and a citation would be warranted.
December 21, 1993
Mr. Roger Clark
Director of Compliance Programs
US Department of Labor
Occupational Safety and Health Administration
Washington, DC 20210
SUBJECT: GFCI Testers Your Letter of December 9, 1993
Dear Mr. Clark:
Thank you for your response to my November 10, 1993 letter to Assistant Secretary Zeigler Concerning proper testing of GFCI's.
We appreciate your investigation of this matter. By copy of this letter to Misters Gurnham and Cavanaugh, we are offering to provide any information which would be of assistance.
Please advise us of the actions you take upon completion of you investigation.
Very truly yours,
PASS & SEYMOUR/LEGRAND
Footnote (1) Deliberately putting current through the equipment grounding conductor is undesirable and, under certain conditions (for example, in a hazardous location), is unsafe. [Back to Text]
Footnote (2) This device is not currently approved by a nationally recognized testing laboratory and does not meet OSHA's electrical safety standards. Therefore, it should not be used by OSHA compliance staff. [Back to Text]
|Standard Interpretations - (Archived) Table of Contents|