Install heavy load breaker first in panel to avoid lights dimming

[kwired]

190706-1424 EDT

My post #35 had nothing to do with code, or any likely condition, or the original post.

My post #35 was in response to romex jockey's question in post #34 as to why. I used simple values that you could juggle in your head to try to illustrate the problem.

.

Well you didn't make that clear ahead of time, one could take that to mean 200 amp supply to each end of a bus is something commonly seen.

 

[gar]

190706-1530 EDT

kwired:

My mistake in not titling the comment directly to romex jockey. But my post immediately followed his post in sequence.

.

 

[jaggedben]

190706-1424 EDT

My post #35 had nothing to do with code, or any likely condition, or the original post.

My post #35 was in response to romex jockey's question in post #34 as to why. I used simple values that you could juggle in your head to try to illustrate the problem.

.

My response was intended to respond to the whole discussion about PV. Apologies for not making that clear.

A general point, as a response to romex jockey's original question, is that location matters when adding a source to a panel or circuit in a way that it does not matter with respect to loads (assuming that the loads are otherwise compliant with code and listings).

 

[gar]

90706-1709 EDT

jaggedben:

Assume that you uniformly load the bus bars in a panel to 200 A, and feed all this current from one end. Then the source end is hottest defined by 200^2 at that end. Suppose that produces a temperature rise of 1 unit at that end.

Feed this panel from both ends, 100 A at each end. As a very rough approximation the temperature rise at each end is about 1/4 unit. This is not a very accurate approximation because there is thermal conductivity in the copper bus bar, but it points in the direction of what happens.

.

 

[jaggedben]

90706-1709 EDT

jaggedben:

Assume that you uniformly load the bus bars in a panel to 200 A, and feed all this current from one end. Then the source end is hottest defined by 200^2 at that end. Suppose that produces a temperature rise of 1 unit at that end.

Feed this panel from both ends, 100 A at each end. As a very rough approximation the temperature rise at each end is about 1/4 unit. This is not a very accurate approximation because there is thermal conductivity in the copper bus bar, but it points in the direction of what happens.

.

I imagine the CMP knows little of that. My understanding is that UL testing of panelboards involves a load breaker equal to the panel rating installed right next to the main feed. (i.e loads are not uniformly spread out, because that is not worst case scenario) This setup is then tested at max current and temperature rating. Now imagine the same thing going on simultaneously at the other end. UL and the CMP do not have confidence in this. So we are limited to 20% at the other end, even though Kirchoff's law otherwise says we should be able to more.

If someone like you could lay out a robust theoretical basis for a new kind of UL standard, that might shift where the code is on this.

 

[gar]

190707-1140 EDT

This post will get back to a discussion of the first post.

Neither UL or NEC is a direct connection to how a circuit works. But both are generally, and sometimes very closely related to actual circuit operation. UL and NEC are rule based systems to provide some simple defined ways to achieve a safe system without the user of the rules having to really understand too much about the system. Most of the rules are very good at achieving the goal of a safe system.

Neither UL or NEC directly relates to this post.

Now to the original post.

The title is "Install heavy load breaker first in panel to avoid lights dimming".

Then the post contents were

My instructor years ago told me to keep house lights from dimming when the AC turns on, to install the AC rwo pole breaker first or next to the main breaker, upstream from the lighting breakers. Is this what you also suggest ? Thank you.

Some measurements from my 200 A QO main panel with plugin main fuses. I don't know if fuse drop is greater or less than a breaker of the same rating. My guess is that a fuse is lower.

The tests were with a 10 A load. Voltage drop within the panel should be close to linear with respect to current. So one can extrapolate to higher currents from this small current.

Note: voltage drop (magnitude without any phase information) across some portion of a circuit, voltage between two points, subtracted from a source voltage may not give real good information on the voltage or change in voltage across some load. But if you assume everything is resistive, then the measurements may still be quite useful.

From top to bottom of my QO panel with a hot bus bar loaded at the far end the change in voltage over the length of the bus bar was a 0.3 millivolts change for a 10 A change. That is a bus bar resistance of 0.000,03 ohms. If starting current to an air conditioner was 200 A, and its breaker was at the far end of the hot bus bar, then this would produce a voltage drop along the total length of the bar of 20*0.3 = 6 millivolts. The lighting circuit breaker would also need to be at the far end for the light to see the 6 millivolt change. Note: 200 A at 240 V starting current would be a large air conditioner.

The average individual will not be aware of a 6 millivolt change in voltage to an incandescent bulb. 2 V is more likely to be noticed. That is a 333 times greater change.

Next we include the main fuse in my voltage drop change measurement. This change is 22 millivolts for the 10 A. Again at 200 A starting current this drop is only 440 millivolts. A trained individual with signal known exactly might detect the incandescent light change from this voltage change.

The next next includes everything to the point of origin of the power. Here the change in voltage to the load is measured rather than the voltage drop along some path to the load. The 10 A load change produced a voltage change of 1 V. Note: this includes all the loop impedance. So neutral is an added component. Assume this is 1/2 the loop impedance, but it is actually somewhat less. Doesn't have the main circuit protection, and meter in its path. Using 0.5 V at 10 A as the hot side drop, then 200 A across the 240 (air conditioner load) is a 10 V change to a 120 V light circuit. This change in voltage to a tungsten incandescent bulb probably will be noticed by a person not looking for the flicker.

Most of that measured 1 V drop is not from the 50 kVA pole transformer. This is easily seen by looking at the voltage change of the opposite phase when the 10 A load is applied. Mostly the service drop, meter, and main fuses are the important impedances.

So it really does not matter where you place that high load breaker on the QO bus bars from the perspective of light flicker. The voltage change difference along the bus bar is nil compared to other factors. Do some of your own experiments.

Also if my measurements and estimated values are correct 200 A thru the bus bar only dissipates 1.2 W (200^2*0.000,03 = 1.2) in the bar. This would not be a determining factor in where to place breakers.

I don't know if I have any mistakes here, check me.

.

 

[tortuga]

190707-1140 EDT

This post will get back to a discussion of the first post.

Neither UL or NEC is a direct connection to how a circuit works. But both are generally, and sometimes very closely related to actual circuit operation. UL and NEC are rule based systems to provide some simple defined ways to achieve a safe system without the user of the rules having to really understand too much about the system. Most of the rules are very good at achieving the goal of a safe system.

Neither UL or NEC directly relates to this post.

Now to the original post.

The title is "Install heavy load breaker first in panel to avoid lights dimming".

Then the post contents were


Some measurements from my 200 A QO main panel with plugin main fuses. I don't know if fuse drop is greater or less than a breaker of the same rating. My guess is that a fuse is lower.

The tests were with a 10 A load. Voltage drop within the panel should be close to linear with respect to current. So one can extrapolate to higher currents from this small current.

Note: voltage drop (magnitude without any phase information) across some portion of a circuit, voltage between two points, subtracted from a source voltage may not give real good information on the voltage or change in voltage across some load. But if you assume everything is resistive, then the measurements may still be quite useful.

From top to bottom of my QO panel with a hot bus bar loaded at the far end the change in voltage over the length of the bus bar was a 0.3 millivolts change for a 10 A change. That is a bus bar resistance of 0.000,03 ohms. If starting current to an air conditioner was 200 A, and its breaker was at the far end of the hot bus bar, then this would produce a voltage drop along the total length of the bar of 20*0.3 = 6 millivolts. The lighting circuit breaker would also need to be at the far end for the light to see the 6 millivolt change. Note: 200 A at 240 V starting current would be a large air conditioner.

The average individual will not be aware of a 6 millivolt change in voltage to an incandescent bulb. 2 V is more likely to be noticed. That is a 333 times greater change.

Next we include the main fuse in my voltage drop change measurement. This change is 22 millivolts for the 10 A. Again at 200 A starting current this drop is only 440 millivolts. A trained individual with signal known exactly might detect the incandescent light change from this voltage change.

The next next includes everything to the point of origin of the power. Here the change in voltage to the load is measured rather than the voltage drop along some path to the load. The 10 A load change produced a voltage change of 1 V. Note: this includes all the loop impedance. So neutral is an added component. Assume this is 1/2 the loop impedance, but it is actually somewhat less. Doesn't have the main circuit protection, and meter in its path. Using 0.5 V at 10 A as the hot side drop, then 200 A across the 240 (air conditioner load) is a 10 V change to a 120 V light circuit. This change in voltage to a tungsten incandescent bulb probably will be noticed by a person not looking for the flicker.

Most of that measured 1 V drop is not from the 50 kVA pole transformer. This is easily seen by looking at the voltage change of the opposite phase when the 10 A load is applied. Mostly the service drop, meter, and main fuses are the important impedances.

So it really does not matter where you place that high load breaker on the QO bus bars from the perspective of light flicker. The voltage change difference along the bus bar is nil compared to other factors. Do some of your own experiments.

Also if my measurements and estimated values are correct 200 A thru the bus bar only dissipates 1.2 W (200^2*0.000,03 = 1.2) in the bar. This would not be a determining factor in where to place breakers.

I don't know if I have any mistakes here, check me.

.

I happen to be doing some busbar calcs and read this.
If its 1/8" X 3/4" bussbar (119 kCMIL) the factory lists it as 87.7 micro ohms per ft.
Assuming a 200A load, one inch distance and a 75C ambient
Voltage Drop = 2 K I D / CIML
so would it be
2 X K @ 75C X 1/12 / 119kCMIL
or
2 X 12.9 X 200 X 0.083333333333333 / 119000
=
0.002913165266106 Volts per inch drop ?
Check my math.
A bolted short of 10000 amps would have a voltage drop of .14 volts per inch.
I'd say look for other weak spots.
A infra red camera might help look for hot spots.

 

[LarryFine]

I'll admit it: I don't arrange the breakers the way I do for electrical reasons; it just looks so damn good.

 

[jaggedben]

I'll admit it: I don't arrange the breakers the way I do for electrical reasons; it just looks so damn good.

Well if that's you're criteria then I'm gonna through the PV opposite end thing back into the discussion...

Please, if you've got subpanels in the panel, put them at the opposite end, so that if I decide to interconnect PV to the subpanel, I don't have to move your breakers.
That overrides any aesthetic considerations in my book. :lol:

 

[kwired]

Please, if you've got subpanels in the panel, put them at the opposite end, so that if I decide to interconnect PV to the subpanel, I don't have to move your breakers.

What if said subpanel has little load? That panel to the garage maybe that seldom if ever has much load on it?

 

[jaggedben]

What if said subpanel has little load? That panel to the garage maybe that seldom if ever has much load on it?

My point - make it easier for the PV guy - is not affected by load.

I mean, if it's a sub on a 30A breaker, no big deal really. It's when we get to 100A subs it starts to become a real pain.

 

[ggunn]

So then, why does everey PV guy that hires me insist on their 'panel interface' ocpd being the loowest and furthest away from the incoming mains/mlo?

~RJ~

Because 705.12(D or B)(2)(3)(b) says so. There are other reasons but that one should be sufficient.