Panelboard Sizing for Mechanical Loads

[Junior_EE]

I work for a consulting engineering firm as an electrical engineer. We have a difference of opinion in our office on sizing for a panelboard or switchboard serving mechanical loads.

Let's say the FLA of a dozen or more mechanical units (where the largest load is taken at 125% FLA) sums to 1250Amps, 3-phase, at 480Volts.

In the past, I typically have applied a reasonable demand factor to the load of, maybe 75-80%. 1250A * 80% = 1000A. Therefore a 1200A panel is acceptable.

My colleague believes the NEC does not allow us to take such demands because they are not code-driven and, so, one would need to provide at least a 1600A switchboard to serve the load.

Are either of us correct? Or is he simply being more conservative and I am being more aggressive in our design approaches? I feel like if we were to follow his approach for some jobs, equipment would be insanely oversized.

 

[ron]

I don't see where in Article 220 it allows the demand factor you are describing, unless some will naturally not operate at the same time, such as heating and cooling that are interlocked.

 

[Junior_EE]

I don't see where in Article 220 it allows the demand factor you are describing, unless some will naturally not operate at the same time, such as heating and cooling that are interlocked.

I guess the better rationale of using a 1200A panel instead of a 1600A switchboard would be applying a diversity factor to a slew of loads to account for fact that they’ll never all be running full blast simultaneously. There is nothing in the NEC that seems to allow it or prevent it.

 

[drktmplr12]

I work for a consulting engineering firm as an electrical engineer. We have a difference of opinion in our office on sizing for a panelboard or switchboard serving mechanical loads.

Let's say the FLA of a dozen or more mechanical units (where the largest load is taken at 125% FLA) sums to 1250Amps, 3-phase, at 480Volts.

In the past, I typically have applied a reasonable demand factor to the load of, maybe 75-80%. 1250A * 80% = 1000A. Therefore a 1200A panel is acceptable.

My colleague believes the NEC does not allow us to take such demands because they are not code-driven and, so, one would need to provide at least a 1600A switchboard to serve the load.

Are either of us correct? Or is he simply being more conservative and I am being more aggressive in our design approaches? I feel like if we were to follow his approach for some jobs, equipment would be insanely oversized.

I agree with your co-worker.

The code requires conductors serving motors or groups of motors to be rated for the connected NEC motor load + 25% largest motor +100% noncontinuous non-motor load + 125% of continuous non-motor load.

Find me somewhere it says you can do anything other than what is described above.

 

[petersonra]

I guess the better rationale of using a 1200A panel instead of a 1600A switchboard would be applying a diversity factor to a slew of loads to account for fact that they’ll never all be running full blast simultaneously. There is nothing in the NEC that seems to allow it or prevent it.

215.2 Minimum Rating and Size.
(A) Feeders Not More Than 600 Volts.
(1) General. Feeder conductors shall have an ampacity not
less than required to supply the load as calculated in Parts III,
IV, and V of Article 220. Conductors shall be sized to carry not
less than the larger of 215.2(A)(1)(a) or (b).

220.60 Noncoincident Loads. Where it is unlikely that two or
more noncoincident loads will be in use simultaneously, it shall
be permissible to use only the largest load(s) that will be used
at one time for calculating the total load of a feeder or service.

 

[drktmplr12]

I guess the better rationale of using a 1200A panel instead of a 1600A switchboard would be applying a diversity factor to a slew of loads to account for fact that they’ll never all be running full blast simultaneously. There is nothing in the NEC that seems to allow it or prevent it.

Mechanically speaking, some industrial processes are designed such that there is an installed standby unit. The is very common in water and wastewater. We use this to help size transformers and generators (DF = 0 for standby) but not for sizing conductors. That's strictly 430.24, which is clear and provides no demand or diversity factors. You must use the NEC table for motor amps unless it's a low rpm (1200 or less) motor, in which case you need to use the motor nameplate because it will be higher than NEC tables.

Sizing a transformer or generator, on the other hand, you can apply 80% DF to all motors if you want, depending on what type of generator (legally required, hospital, etc).

 

[Junior_EE]

I agree with your co-worker.

The code requires conductors serving motors or groups of motors to be rated for the connected NEC motor load + 25% largest motor +100% noncontinuous non-motor load + 125% of continuous non-motor load.

Find me somewhere it says you can do anything other than what is described above.

I guess the question is, is applying diversity factors allowed to electrical loads based on engineering judgment? I’ve seen this done broadly on many large projects, but never thought it was an issue.

 

[don_resqcapt19]

I don't see where in Article 220 it allows the demand factor you are describing, unless some will naturally not operate at the same time, such as heating and cooling that are interlocked.

I agree that is what the code says, but I have never installed an MCC where the rating even equaled the sum of all of the connected loads. These were industrial installations where there was no actual interlocking to prevent some loads from running at the same time as other loads....it has always been based on the design engineer making a demand factor guess based on some knowledge of the process operations.

 

[petersonra]

I guess the question is, is applying diversity factors allowed to electrical loads based on engineering judgment? I’ve seen this done broadly on many large projects, but never thought it was an issue.

I think the answer is that many times it is done even though the code prohibits it.

 

[Junior_EE]

I agree that is what the code says, but I have never installed an MCC where the rating even equaled the sum of all of the connected loads. These were industrial installations where there was no actual interlocking to prevent some loads from running at the same time as other loads....it has always been based on the design engineer making a demand factor guess based on some knowledge of the process operations.

This is similar to what I’ve seen on a service switchboard. The service switchboard capacity is often less than the sum of the receptacle demand load + mechanical connected load.

 

[petersonra]

This is similar to what I’ve seen on a service switchboard. The service switchboard capacity is often less than the sum of the receptacle demand load + mechanical connected load.

There are a lot of times when engineers design stuff that violates the NEC. You sometimes have to choose between complying with the letter of the code, or supplying a cost effective design.

 

[Junior_EE]

I think the answer is that many times it is done even though the code prohibits it.

I’ll say from firsthand experience that it is often seems nearly impossible to provide service equipment equal to or greater than the NEC calculated load, so diversity is required. Ditto for sizing a generator. Especially on large projects where the total NEC load can be over 25,000A.

 

[Junior_EE]

There are a lot of times when engineers design stuff that violates the NEC. You sometimes have to choose between complying with the letter of the code, or supplying a cost effective design.

I don’t see in the NEC where it says one can’t apply a diversity factor.

 

[petersonra]

I’ll say from firsthand experience that it is often seems nearly impossible to provide service equipment equal to or greater than the NEC calculated load, so diversity is required. Ditto for sizing a generator. Especially on large projects where the total NEC load can be over 25,000A.

perhaps you should submit a proposal to modify the code to allow for an engineered solution when the code solution costs more than you want to spend.

 

[petersonra]

I don’t see in the NEC where it says one can’t apply a diversity factor.

The code has procedures for determining the load. It is a code requirement that the procedure be followed. The only so called diversity factors allowed are those found in the load calculation procedures.

The code is very clear about this.

220.1 Scope. This article provides requirements for calculating
branch-circuit, feeder, and service loads. Part I provides
general requirements for calculation methods. Part II provides
calculation methods for branch-circuit loads. Parts III and IV
provide calculation methods for feeder and service loads.
Part V provides calculation methods for farm loads.

Your other option for code compliance is to be granted special permission from the AHJ. see 90.4.

 

[Charged]

Your point about engineering judgment and diversity i think has its place in a buildings design, I just don’t think service equipment and branch panelboards are the places to apply. I haven’t done a lot of large MCCs to know if that’s a place that it’s common practice. ….maybe some data about historical demand load of a simile facility even if it’s not the 12months utility bills like for existing loads would be a better way to present then engineering judgement demand factor. …..

 

[tom baker]

The code has procedures for determining the load. It is a code requirement that the procedure be followed. The only so called diversity factors allowed are those found in the load calculation procedures.

The code is very clear about this.


Your other option for code compliance is to be granted special permission from the AHJ. see 90.4.

And in WA state that requires written permission, via a form and fee. You are paying for a no.

 

[petersonra]

Have you considered putting in two 800 amp pieces of gear in lieu of 1,600 amp?

Has the advantage of not requiring GFPE.

 

[drktmplr12]

as a design engineer, you are taking a risk if you are knowingly not using the code-required procedure for sizing feeders and equipment. the risk is the building official challenges you. they are within their rights to reject the design for non-compliance or report to the board for inadequate design. extremely unlikely.

if they make you change it, hopefully you asked the architect for enough space in the electric room or the owner might be buying the contractor a boat or two.

 

[nhee2]

I agree with your co-worker.

The code requires conductors serving motors or groups of motors to be rated for the connected NEC motor load + 25% largest motor +100% noncontinuous non-motor load + 125% of continuous non-motor load.

Find me somewhere it says you can do anything other than what is described above.

430.26?