VD on Lighting circuits

[Isaiah]

I'm looking at a Voltage Drop calculation. In this case circuits 1, 3 and 5 (i.e. phases A, B and C) plus neutral (shared) and ground are ran together in 3/4" conduit to the LED's, which operate @ single phase, 120VAC and draw 48VA or .4A. I am treating each circuit as a single phase load even though all three phases are ganged together at the lighting panel board, three phase 4W, 120/208VAC. Is this correct?

 

[ptonsparky]

You have a total of .4 Amp per phase and you're worried about voltage drop?
I would use 208 V and 3ph for calculation if they all come on at the same time. Switched or controlled separately I would use 1ph, 120 V for VD. IDK how harmonics, if any would affect the calculations.

 

[Isaiah]

You have a total of .4 Amp per phase and you're worried about voltage drop?
I would use 208 V and 3ph for calculation if they all come on at the same time. Switched or controlled separately I would use 1ph, 120 V for VD. IDK how harmonics, if any would affect the calculations.

There are actually 20 lights at .4A per circuit - three circuits, A, B and C total of 60 lights - all driven at 120V. The CB is three-pole, 20A (NEC requirement) since the neutral is shared. My understanding is harmonics are usually not an issue with LED's. I dont think you can treat the lights like you would a three phase motor, i.e. calculate at 208VAC 3 phase.

 

[ptonsparky]

Use three phase calcs if you want to know the VD at the contactor or at a common junction point before the circuits are split. Use single phase if you want to know the VD at each 120v fixture from there on.

Most LED fixtures are rated 120-277 anymore. Use 208 distribution and forget the neutral.

 

[ptonsparky]

There are actually 20 lights at .4A per circuit - three circuits, A, B and C total of 60 lights - all driven at 120V. The CB is three-pole, 20A (NEC requirement) since the neutral is shared. My understanding is harmonics are usually not an issue with LED's. I dont think you can treat the lights like you would a three phase motor, i.e. calculate at 208VAC 3 phase.

All things being equal the VD at each location of three grouped fixtures using a single common neutral would be the same as if it were a 3 phase motor. The neutral would be a non current carrying conductor, essentially a 3 phase load.

 

[Isaiah]

Use three phase calcs if you want to know the VD at the contactor or at a common junction point before the circuits are split. Use single phase if you want to know the VD at each 120v fixture from there on.

Most LED fixtures are rated 120-277 anymore. Use 208 distribution and forget the neutral.

"Use 208 distribution and forget the neutral" - I wish I could do that, but I am only performing a check - its not my design. I would definitely agree with using three phase calc at the contactor since its switching the primary side of the Xfmr. But for each branch I think it should be based on single phase loading.

 

[synchro]

There are actually 20 lights at .4A per circuit - three circuits, A, B and C total of 60 lights - all driven at 120V. The CB is three-pole, 20A (NEC requirement) since the neutral is shared. My understanding is harmonics are usually not an issue with LED's. I dont think you can treat the lights like you would a three phase motor, i.e. calculate at 208VAC 3 phase.

Do you know if the lights have a spec on total harmonic distortion (THD) of their current waveform of 10% or less? If so, then to a good approximation you can treat them like three phase L-N loads where the currents cancel at the common neutral connection, and you can ignore the comments below.

If LED lights have a high THD this is typically because they draw current mainly at the peaks of the voltage waveform. So with such lights on each 120V L-N of a 208V 3-phase system, the current pulses on the neutral from each of the three 120V circuits will not overlap as they do with pure sinusoidal waveforms. That means that the currents will not cancel at the neutral. Instead, the total RMS current may be as much as sqrt (3) times that of each phase (for equal phase currents). The non-overlap of current pulses also makes the peak voltage drop on each 120V L-N circuit relatively independent of eachother. So you can calculate the peak voltage drop on each 120V circuit as if the LED loads on the other two phases were not there.

 

[Isaiah]

Do you know if the lights have a spec on total harmonic distortion (THD) of their current waveform of 10% or less? If so, then to a good approximation you can treat them like three phase L-N loads where the currents cancel at the common neutral connection, and you can ignore the comments below.

If LED lights have a high THD this is typically because they draw current mainly at the peaks of the voltage waveform. So with such lights on each 120V L-N of a 208V 3-phase system, the current pulses on the neutral from each of the three 120V circuits will not overlap as they do with pure sinusoidal waveforms. That means that the currents will not cancel at the neutral. Instead, the total RMS current may be as much as sqrt (3) times that of each phase (for equal phase currents). The non-overlap of current pulses also makes the peak voltage drop on each 120V L-N circuit relatively independent of eachother. So you can calculate the peak voltage drop on each 120V circuit as if the LED loads on the other two phases were not there.

.

The only common neutral connection would be either at the panelboard or the Xfmrs' 'Wye' secondary since typically, lighting circuits are spliced at dozens of points along their path - this of course includes the neutral (and EGC). So I think the second part of your write up would mostly apply for purposes of branch VD.
Very nice response. Thanks.

 

[charlie b]

You have 60 LEDs rated at 48W each, for a total of 2880W. At 208V three phase, that results in an 8 amp current. With #12 wire, your run could be 200 feet without exceeding a 3% VD. If you had separate single phase circuits, each feeding 20 LEDs, the current would still be 8 amps. With #12 wire and operating at 120V, you would have to stay under 125 feet to keep VD below 3%. The reason it is different is that for a single phase circuit you have current travelling the same distance twice (out on the black wire and back on the white wire). For a three phase circuit current that leaves on one phase returns split between the other two phases.

You didn't tell us the distance from the source to the last LED in the circuit. If it is under 100 feet, you won't have to be concerned with VD.

 

[Isaiah]

You have 60 LEDs rated at 48W each, for a total of 2880W. At 208V three phase, that results in an 8 amp current. With #12 wire, your run could be 200 feet without exceeding a 3% VD. If you had separate single phase circuits, each feeding 20 LEDs, the current would still be 8 amps. With #12 wire and operating at 120V, you would have to stay under 125 feet to keep VD below 3%. The reason it is different is that for a single phase circuit you have current travelling the same distance twice (out on the black wire and back on the white wire). For a three phase circuit current that leaves on one phase returns split between the other two phases.

You didn't tell us the distance from the source to the last LED in the circuit. If it is under 100 feet, you won't have to be concerned with VD.

Thanks Charlie. At this point, I don't actually know the distance to any of the LED's since I was only trying to get the basis down before looking into the details. But I am guessing the last LED would be much further than 100 feet since this is a very large Refinery and circuits go 'all over the place'; not just piperacks but up and down towers,vessels, platforms etc.

 

[charlie b]

I suggest a "worst case followed by refinements" approach. To start off, assume you have the 60 LEDs all located at a distance corresponding to the farthest distance any single LED is from the panel. Put all the load at that distance, and calculate it as a balanced three-phase load. For example. if the farthest LED is 600 feet from the panel, a #8 wire will get you to a 3% VD. That might be a bigger wire than you would want to use.

You can refine this by dividing the run into smaller chunks. For example, place 30 LEDs at a distance of 300 feet and the rest at 600 feet. The first section will include the load from all LEDs, but the second section will only have half the LEDs. With #10 wire, the VD from the panel to the first point (with 8 amps of load) is 2%, and the VD from the first point to the second (with only 4 amps of load) is 1%. Calculated this way, you can use #10 wire for the entire run.

You can further refine this by putting a third of the load spaced out at distances 200 feet. I will leave that math to you.

 

[Isaiah]

I suggest a "worst case followed by refinements" approach. To start off, assume you have the 60 LEDs all located at a distance corresponding to the farthest distance any single LED is from the panel. Put all the load at that distance, and calculate it as a balanced three-phase load. For example. if the farthest LED is 600 feet from the panel, a #8 wire will get you to a 3% VD. That might be a bigger wire than you would want to use.

You can refine this by dividing the run into smaller chunks. For example, place 30 LEDs at a distance of 300 feet and the rest at 600 feet. The first section will include the load from all LEDs, but the second section will only have half the LEDs. With #10 wire, the VD from the panel to the first point (with 8 amps of load) is 2%, and the VD from the first point to the second (with only 4 amps of load) is 1%. Calculated this way, you can use #10 wire for the entire run.

You can further refine this by putting a third of the load spaced out at distances 200 feet. I will leave that math to you.

Great advice Charlie. I have a program created specifically for lighting voltage drops, called a 'segmented VD calculator' - it looks at the VD to each individual LED from its source (panelboard). You have to estimate the distance to each light - this can be done of course using scaled plan dwgs or a 3D model, if available.

 

[gar]

191203-0921 EST

Isaiah:

Very many LEDs don't care what the input voltage is so long as it is not too high. In other words they are dimmable. Some work equally well on phase shift dimming, and variable sine wave.

So long as you can provide the needed light level why would voltage drop be a problem?

.

 

[Isaiah]

191203-0921 EST

Isaiah:

Very many LEDs don't care what the input voltage is so long as it is not too high. In other words they are dimmable. Some work equally well on phase shift dimming, and variable sine wave.

So long as you can provide the needed light level why would voltage drop be a problem?

.

Interesting commentary. Would footcandle levels be reduced if enough volts are 'dropped' due to load/circuit distance? I will investigate with Crouse-Hinds - we're using their CHAMP VMV series LED's.

 

[gar]

191203-1121 EST

isaiah:

Fundamentally the light output of an LED chip (the basic light emitter) is proportional to the current thru the LED.

Something has to control the current thru the LED chip. In the simplest form this is a resistor with a voltage drop of at least as much as the drop across the LED, and fed from a moderately constant voltage source. At a given current level thru an LED the voltage drop across the LED is quite sensitive to temperature. Thus, to work over a wide temperature range you want a constant current source to feed the LED. A high source voltage and a high resistance is a moderately good constant current source for a small, but varying, load (LED) resistance.

All LED light sources will have some sort of driver to control current to the LED. This driver is as simple as a series resistor, or as complex as some control circuit that monitors light output, and maintains that light constant independent of input voltage.

You have to know your LED product, and what its characteristics are.

.

 

[Isaiah]

191203-1121 EST

isaiah:

Fundamentally the light output of an LED chip (the basic light emitter) is proportional to the current thru the LED.

Something has to control the current thru the LED chip. In the simplest form this is a resistor with a voltage drop of at least as much as the drop across the LED, and fed from a moderately constant voltage source. At a given current level thru an LED the voltage drop across the LED is quite sensitive to temperature. Thus, to work over a wide temperature range you want a constant current source to feed the LED. A high source voltage and a high resistance is a moderately good constant current source for a small, but varying, load (LED) resistance.

All LED light sources will have some sort of driver to control current to the LED. This driver is as simple as a series resistor, or as complex as some control circuit that monitors light output, and maintains that light constant independent of input voltage.

You have to know your LED product, and what its characteristics are.

.

Gar, this is great stuff and deserves investigation. However, I am still bound to a client specification that limits the branch VD to 3% - or 5% if the power source (transformer) is very close to the panelboard. I will still have to crunch the numbers but its nice to know LED's will (probably) function just fine if the VD exceeds the client specs.