RESIDENTIAL BY DAVID SHAPIRO
Does this mean, at some places
behind an LED, your nonmetallic wiring is grilled by 100°C or 105°C, while
it would have been safe behind incandescent lamps? Not unless the designer
goofed in choosing the LED’s heat sink
and the testing lab slipped a cam, too.
“For any given combination of incan-
descent lamp, application, LED wattage
and fixture construction, one or the
other could be hotter,” Shaner said. “It
really makes no difference as long as
the parts of the fixture that touch com-
bustible material do not exceed UL’s
The critical question is how much the
canopy and outlet box can heat the wire
installed to power the lighting outlet. Is
this likely to reach the temperature of an
LED’s electronic circuit board itself? In
a luminaire designed for incandescent
lamps, is it likely to reach the tempera-
ture of the lamp socket and screw shell
of the lamp?
It shouldn’t, but this hinges on the
layout. The modern ANSI standard for
luminaires, UL 1598, lists types of fixtures
that are exempt from the tests that others
have to pass. If the heat from lamps can
escape by convection (heat-driven movement) of the air around them, rather than
being focused back on the canopy by an
outer enclosure such as a globe, the wiring is at much less risk.
Being able to radiate heat away rather
than trap it close to the wiring matters a lot.
Take porcelain lampholders, which hold
lamps straight out from them and don’t
enclose them, letting the heat dissipate.
The expert who pointed this out
is Michael S. O’Boyle, a member of
Code-Making Panel (CMP) 18. He has
officially participated in all work on the
lighting-related National Electrical Code
(NEC) sections for more than 20 years,
since he was the alternate to the CMP
18 American Lighting Association rep-
resentative for the 1996 NEC. He also
sits on the technical advisory panel for
UL’s lighting standard and has been part
of other panels involved in important
O’Boyle, who works for Philips Light-
ing, started out with UL and was kind
enough to explain the changes in the
lighting standard that took place over
the years. Was there a record of horrific
fires started by luminaires manufactured
to the old standard? Not exactly. It was
more of what I call a “whoops.”
In the 1960s, UL 57, the standard
for luminaires at the time, trusted that
those luminaires that passed its tests
would not expose the building wir-
ing associated with them to more than
60°C, which the conductors’ insulation
could withstand cheerfully. Unfortu-
nately, it took UL a while to realize there
was a mismatch between the test design
and how electricians wire luminaires.
Here is O’Boyle’s explanation: To
test the temperature that the building’s
wiring needed to withstand over long
periods of time, UL mounted thermocouples against the ears of the outlet
box, essentially flush with the ceiling.
The testers figured the building wire
would be some ways back from these or,
as UL 57 indicated, “within the outlet
box.” The test setup, then and now, used
1½-inch-deep octagon boxes, not 5/8-
Yes, the pigtail or pigtails coming
from the lamp socket or sockets had to
start out by the canopy, good and hot, but
they would then meander back, radiating
their heat into the air for an inch or so
until they reached the points where the
splices to the supply wire nestled. That’s
what UL figured.
But the UL panel was wrong. Sure, it
may work this way in some luminaires.
In many others, though, whether due to
box crowding or just chance, the building wire bustles forward from clamp or
connector or locknut-and-bushing into
the canopy space. Maybe it even pushes
against the back of the canopy.
You’ve seen this, the same as I have.
So, as hot as the canopy gets with incandescent lampholders lying parallel to
it and perhaps a glass or plastic dome
holding the heat in, that’s how some of
the building wire’s insulation is being
toasted. Or, nowadays, as hot as the much
flatter canopy gets with the LED printed
circuit board right there.
I hope you don’t think, “Ain’t it
awful?” or “How dare they?” Like the
NEC and other product standards, this
one gets revised both in response to
product inventions and improvements
and to discoveries of human error. It’s
not just politics and manufacturers’
maneuvering that lead to changes.
In next month’s column, I will discuss how the problem was fixed and
how it affects us.
How Hot Is It Back There?
Materials, construction and design make a difference
LAST MONTH, JEFF SHANER, director of design engineering with Acuity Brands,
Atlanta, reported that, if you replace a 40–60-watt incandescent lamp with a listed
light-emitting diode (LED) fixture that’s just about as bright, it will use a heat sink to
balance out temperatures that can exceed 100°C (212°F) at locations on the printed
SHAP IRO is the author of “Old Electrical Wiring,” “Your Old Wiring,” and co-author, with W.
Creighton Schwan, of “Behind the Code.” He is a contractor, consultant, writer and local IAEI
officer in Greenbelt, Md. He can be reached at
email@example.com. L E