LED Lights and Dimmer Switches

Status
Not open for further replies.
I posted the following information in an engineering inspection report, pertaining to a lighting retrofit project at a large hospital that was built in the mid-1960's:


"A majority of the interior areas on campus are primarily lit by T-8 fluorescent light fixtures using 32-Watt (W) fluorescent lamps of various color, temperatures, and standard efficiency electronic ballasts. Since much of the lighting is on for long periods of hours, the fluorescent lighting represents a significant portion of the entire campus lighting electrical consumption. Long burn hours produce quick lamp failure, making these fixtures disproportionately burdensome to maintenance staff. A retrofit to LED will provide the highest possible efficiency -- reducing lighting electrical load by 40-50%.

Since LED rated life greatly exceeds that of the existing fluorescents, the maintenance burden will also be significantly reduced. Patients and staff can expect less frequent disturbances to maintain these fixtures. Replacement of existing metal halide fixtures in the parking garages with new LED luminaries will likewise provide similar benefits, reducing that electrical load by up to 80%, along with extending maintenance intervals by a factor of ten.

But what we have here is a mid-20th century building – with mid-20th century electrical wiring – that's now had 21st century lighting installed.

While the lighting electrical contractor may be able to make some accommodations in some areas to adjust the lighting levels to meet the standards of ASHE (American Society of Healthcare Engineers) criteria, the hard fact is that much, if not most, of the lighting in this 50+ year old hospital may not be able to be properly adjusted to the proper luminescence level, simply because the facility and office spaces are not properly wired to do so.

Some areas clearly need double-pole light switches, so that only half, or all of the lights could be turned on and off (when the sunlight through the windows is bright), as it suits the occupant.

And as far as dimmers, some of the 1960's electrical wiring in these buildings could not accommodate the quickly rising and lowering amperage spreading across the distance of the wire, without possibly leading to a potential electrical fire hazard."


The hospital electrical engineer became livid, and replied via e-mail with: "This paragraph is not true. Installed LED lighting takes less wattage and amperage to operate. For dimmers installation, the dimmer requires two smaller size wires (14ga.) and uses Direct Current (DC) Voltage to operate in addition with the existing AC wiring (10 or 12ga.) to power the LED rectifier. The 1960’s electrical wiring has nothing to do with what you stated in the above paragraph and is not a potential electrical fire hazard."


My argument is that it is not only just the reduction in magnitude of current that we should be concerned with, but given the age of the wire itself, plus the rate of change in the current (dA/dt), still presents electrical/electronic friction, which could cause a rise in temperature, and thus subsequently, an electrical fire hazard.

I need to come back at the hospital engineer with cited NEC code that backs up my premise:

1) Age of wire is a factor. Even if all copper wire was used during hospital construction in ~1965, there could be aluminum components in the switches and junction boxes. Or unless the spacing of wires in the junction boxes has been inspected by the electrical contractor before reconfiguring with dimmer switches, all of these factors still present a hazard.

2) Rate of change is current (dA/dt), not just magnitude of current, could present a rise in temperature in electrical wiring. Unless current flow tests are conducted in the existing 50+ year old wiring at the hospital, or unless all of the lighting electrical wiring is replaced to meet 2017 NEC code, the electrical contractor should not reconfigure the old wiring with LED lights plus dimmer switches.


Thank you for your insight and feedback.
 

retirede

Senior Member
Location
Illinois
There may be reasons to object to LED retrofits and there are members of this forum who will likely chime in on this. But overtaxing existing wiring is not one of them.

I second MAC702's position.
 

480sparky

Senior Member
Location
Iowegia
.......Some areas clearly need double-pole light switches, so that only half, or all of the lights could be turned on and off (when the sunlight through the windows is bright), as it suits the occupant. ..................

How does a DP switch allow only half the lights to be turned on?
 

gar

Senior Member
Location
Ann Arbor, Michigan
Occupation
EE
171105-1521 EDT

CEO@TRCEnergy.com:

I have been accused of scaring away first time posters, but I have to negatively response to your post.

You list your self as an EE and a PE. For someone with this background a statement that dA/dt relates to heating of the wires just makes no sense. Temperature rise of something is time dependent and a function of energy in and energy loss.

Under steady state conditions the RMS current into an invariant resistance will determine the resistor's average temperature rise. What is RMS? Go back and look at your textbooks.

From a practical point of view you need to do some experiments to answer your concerns.

A Kill-A-Watt EZ is a low cost voltage, current, volt-ampere, power, power factor, and kWH meter. About $30 at Home Depot. Reasonably accurate, but product quality places it in the junk category. You need one of these to do some experiments. I keep having to buy new ones because of their failure rate.

Here are some measurements:

Kill-A-Watt EZ was plugged into AC line, input voltage about 123 V, sine wave waveform reasonably good, minor flat topping, some erraticness in readings, probably from the load and needing longer time averaging.

The dimmer I used is what I call a three wire phase shift dimmer. This gives me fairly good control of the turn on time with different loads. The dimmer requires some small amount of power. The dimmer had to be placed after the Kill-A-Watt because the dimmer interfered with the Kill-A-Watt when it was on the output side.

Minimum time delay to turn on of the Triac was 2 mS. Maximum 6.5 mS. One half cycle is 8.3 mS.

75 W incandescent

Turn on at 2 mS.
0.61 A, 72.4 W, 75.2 VA, 0.96 PF

Turn on at 4 mS.
0.52 A, 49.0 W, 67.2 VA, 0.75 PF

Turn on at 6.5 mS.
0.28 A, 8.8 W, 34.1 VA, 0.25 PF


CFL, non dimmable.

Turn on at 2 mS.
0.38 A, 19.2 W, 44.6 VA, 0.42 PF


CREE 9 W dimmable with either sine wave voltage or phase shift. Phase shift here.

Turn on at 2 mS.
0.09 A, 10.1 W, 11.2 VA, 0.93 PF

Turn on at 4 mS.
0.08 A, 5.9 W, 10.8 VA, 0.50 PF

Turn on at 6.5 mS.
0.03 A, 1.0 W, 4.4 VA, 0.30 PF


The PF values are calculated by Kill-A-Watt. These are fairly close to calculated values from W and VA. But since the meter readings were jumping around this could be part of the reason for he difference between my calculations and the meter's.

What these measurements show is that in the real world there is no increase in the RMS current as the level of dimming is increased as a result of using the dimmer at different phase shifts.

There are other considerations with respect to using LEDs. See all of Electric-Light's many comments in many different threads.

I will add another and that is RFI.

.
 

MAC702

Senior Member
Location
Clark County, NV
How does a DP switch allow only half the lights to be turned on?

As mentioned, he probably was picturing two separate switches, and used the wrong term, which meant an entirely different kind of switch wholly unsuited to the purpose. Either way requires rewiring the switch legs with another conductor, and would be a major remodeling project, which seems an odd recommendation from a guy trying to prevent the disturbance to occupants by maintenance personnel changing a light bulb.

I'm curious what was the motivation for this inspection, and the benefit for finding something "wrong."
 
Status
Not open for further replies.
Top