Multiple AC unit feeder diversity

[cppoly]

Can you take a feeder diversity factor for multiple AC units on the same feeder or are feeders taken at 100% even though not all compressors of the multiple AC units may be running at the same time and cycling on/off.

 

[Smart $]

No demand factor applies unless the equipment is interlocked to not run simultaneously.

 

[cppoly]

Thanks

 

[iwire]

No demand factor applies unless the equipment is interlocked to not run simultaneously.

I think that is an overstatement, I think as long as the loads are not likely to run at the same time you are good to go.

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.

 

[Smart $]

I think that is an overstatement, I think as long as the loads are not likely to run at the same time you are good to go.

Well, in theory, I agree. Say for instance an industrial facility where their documented switching program turns some equipment on or off manually by the breaker at different times. But how do you implement this for say a 100 room hotel with PTAC units where the units are never switched off other than thermostatically or by housekeeping but only when the room is vacant?

 

[iwire]

Well, in theory, I agree. Say for instance an industrial facility where their documented switching program turns some equipment on or off manually by the breaker at different times. But how do you implement this for say a 100 room hotel with PTAC units where the units are never switched off other than thermostatically or by housekeeping but only when the room is vacant?

Fair question but still does not change the fact that the code section only says unlikely, not impossible.

There is nothing requiring electrical interlocking.

 

[Smart $]

Fair question but still does not change the fact that the code section only says unlikely, not impossible.

There is nothing requiring electrical interlocking.

Turning one breaker or disconnect off then another on—manually—is a form of electrical interlock... just not the means that comes to mind when the word interlock is mentioned.

 

[iwire]

Turning one breaker or disconnect off then another on—manually—is a form of electrical interlock... just not the means that comes to mind when the word interlock is mentioned.

Sure it can be, but still no interlock required so I am missing the point.

 

[Smart $]

Sure it can be, but still no interlock required so I am missing the point.

The point is, how do you assure the plan checker that one or more loads are non-coincident?

 

[iwire]

The point is, how do you assure the plan checker that one or more loads are non-coincident?

You claimed an interlock is required however the NEC does not require an interlock for non-coincidental loads. We can go round and round with different scenarios but in the end an interlock is not NEC mandated. Might be a good idea, might be a good design, it might give a plan checker the warm fuzzies but it is still optional.

Now, am I saying that cppoly can claim 'non-coincidental loads' in this case? No, I think it is likely the AC units would run at the same time unless there is some underlying reason they would not.

 

[kwired]

Is or should be a design issue IMO. If you don't want to design it so they can all operate at same time, you risk shutdown because of overcurrent device operation.

If you don't want that unexpected overcurrent device operation because too many were running at one time - then design the supply to handle them at same time or use some sort of positive interlocking means or load management scheme so they can't all run at same time.

Most installations the worst case scenario is on the hottest day of the year - they all could be running at same time at least part of the time.

Next worst case scenario is a building that is only occupied at limited times - they maybe turn them all on at same time to cool the place down for the upcoming event, and maybe it is filled with people for only a few hours and gets high demand during that time as well, then it sits nearly idle for maybe a week, two weeks, a month.... but you still need to handle that peak when it happens.

 

[petersonra]

You claimed an interlock is required however the NEC does not require an interlock for non-coincidental loads. We can go round and round with different scenarios but in the end an interlock is not NEC mandated. Might be a good idea, might be a good design, it might give a plan checker the warm fuzzies but it is still optional.

Now, am I saying that cppoly can claim 'non-coincidental loads' in this case? No, I think it is likely the AC units would run at the same time unless there is some underlying reason they would not.

I think the underlying reason they won't all run simultaneously is because they won't, given as they are on thermostats, and some rooms are likely unoccupied and the A/C won't run there at all. However, I don't see any way you can take some kind of "credit" for that because you can't quantify it, unlike other similar situations where the code calculations have quantified it for you.

 

[Smart $]

You claimed an interlock is required however the NEC does not require an interlock for non-coincidental loads. We can go round and round with different scenarios but in the end an interlock is not NEC mandated. Might be a good idea, might be a good design, it might give a plan checker the warm fuzzies but it is still optional.

Now, am I saying that cppoly can claim 'non-coincidental loads' in this case? No, I think it is likely the AC units would run at the same time unless there is some underlying reason they would not.

For the record, I did not say an interlock was required. I implied it. But it was with respect to the type of circuit I assumed it to be... and I'd wager that I'm correct.

I agree interlocking is not required. But you best be able to substantiate to the AHJ a load reduction where not interlocked.

 

[Smart $]

I think the underlying reason they won't all run simultaneously is because they won't, given as they are on thermostats, and some rooms are likely unoccupied and the A/C won't run there at all. However, I don't see any way you can take some kind of "credit" for that because you can't quantify it, unlike other similar situations where the code calculations have quantified it for you.

Yes, odds are all will not run concurrently most of the time. But there's no way to assure the A/C units will never all run simultaneously if they are just thermostatically controlled.

 

[petersonra]

For the record, I did not say an interlock was required. I implied it. But it was with respect to the type of circuit I assumed it to be... and I'd wager that I'm correct.

I agree interlocking is not required. But you best be able to substantiate to the AHJ a load reduction where not interlocked.

Actually you did state it was required, although you did not use the word "required".

No demand factor applies unless the equipment is interlocked to not run simultaneously.

 

[Smart $]

Actually you did state it was required, although you did not use the word "required".

You two can twist it any way you want but as I said, I did not say an interlock was required.

 

[kwired]

I think the underlying reason they won't all run simultaneously is because they won't, given as they are on thermostats, and some rooms are likely unoccupied and the A/C won't run there at all. However, I don't see any way you can take some kind of "credit" for that because you can't quantify it, unlike other similar situations where the code calculations have quantified it for you.

My understanding of sizing heating/cooling units for best overall efficiency and comfort means they will be sized to run quite a bit of the time during the maximum extremes they were designed for. You may not reach those extremes in some areas but a half dozen or less times a year but when you do, or if you exceed the maximum extreme it was designed for the duty cycles run near or at 100%.

If you oversize a unit by too much you get less run time but comfort levels do suffer.

 

[ike5547]

Yes, odds are all will not run concurrently most of the time. But there's no way to assure the A/C units will never all run simultaneously if they are just thermostatically controlled.

I agree with this. I live in an area that gets very hot in the summer (100 deg+ for weeks during a bad heat wave). AC units are often not spec'd out sufficiently to handle this. If you have four units, for example, on a single feeder they are all going to be chugging together for certain periods. If you have 16 units on a feeder it's not so likely, but it can happen.

 

[iwire]

For the record, I did not say an interlock was required. I implied it. .

Yes you did imply that and that was incorrect.

But it was with respect to the type of circuit I assumed it to be... and I'd wager that I'm correct

Correct about what? That you can't predict if a group of thermostatically controlled equipment will never run together? Well congratulations on knowing what we all did.

 

[iwire]

Yes, odds are all will not run concurrently most of the time. But there's no way to assure the A/C units will never all run simultaneously if they are just thermostatically controlled.

We can say the same thing about demand factors for dryers and ranges.

There is no way to assure that all the ranges or dryers will not be used at the same time.