Multifamily Dwelling - Demand Factors

[Jim_SWFL]

We apply demand factors (table 220.84) to feeders/services for units in a multifamily dwellings.


The recent impact of hurricane Irma in FL, has caused developers of high and mid-rise condos to ask whether the planned onsite generator can be upsized to supply a limited number of circuits to each residence unit for refrigeration, one or two receptacles, and a few lights? (the buildings in question are still in various phases of design and development)


It was thought that we might install a small dedicated panel for these emergency circuits which would be fed via an ATS.


My first swag at this was to extract the load from the current dwelling calcs, and apply it separately as a new load on the generator. This caused a discussion/debate about whether I could use a demand factor on these loads?


Can a demand factor be applied to these loads in one instance, and not in another, slightly modified instance?


I do not have a dogmatic opinion about this, if a demand factor cannot be applied, I would like to understand why so I can avoid a similar mistake in the future, and explain why the generator/feeder/service size has increased proportional to the newly added loads.


As a secondary question - I just noticed that 220.84(A)(1) permits the use of the demand factors in table 220.84 as long as the unit has a single feeder.


Does the use of a second feeder for the "emergency" panel in the unit, as described above, remove our ability to apply a demand factor on the normal feeder/service?

Thanks for the input.

 

[Dennis Alwon]

IMO, The demand factor in article 220 can only be used on the entire dwelling unit not specific loads unless specifically stated. Such a case may be the 75% demand for 4 or more aplliances

 

[Jim_SWFL]

Thanks Dennis

 

[joeytaft]

Not to poke a hole in a good idea, but i sounds like you are using one ATS for all the condo units? Is there only one meter on the building?

 

[ron]

I don't think you can use demand factors unless it is the whole dwelling or whole building,

Section 403.4.8 of the Building Code likely ended up with a relatively large generator already, so some 180VA receptacles shouldn't be a problem.

 

[kwired]

We apply demand factors (table 220.84) to feeders/services for units in a multifamily dwellings.


The recent impact of hurricane Irma in FL, has caused developers of high and mid-rise condos to ask whether the planned onsite generator can be upsized to supply a limited number of circuits to each residence unit for refrigeration, one or two receptacles, and a few lights? (the buildings in question are still in various phases of design and development)


It was thought that we might install a small dedicated panel for these emergency circuits which would be fed via an ATS.


My first swag at this was to extract the load from the current dwelling calcs, and apply it separately as a new load on the generator. This caused a discussion/debate about whether I could use a demand factor on these loads?


Can a demand factor be applied to these loads in one instance, and not in another, slightly modified instance?


I do not have a dogmatic opinion about this, if a demand factor cannot be applied, I would like to understand why so I can avoid a similar mistake in the future, and explain why the generator/feeder/service size has increased proportional to the newly added loads.


As a secondary question - I just noticed that 220.84(A)(1) permits the use of the demand factors in table 220.84 as long as the unit has a single feeder.


Does the use of a second feeder for the "emergency" panel in the unit, as described above, remove our ability to apply a demand factor on the normal feeder/service?

Thanks for the input.

Code aside, you are supplying loads that will likely have a high demand during situations when they will be energized by standby power. Normal load diversity used to determine allowable demand factors mean nothing in this situation. With limited number of outlets being supplied you can almost plan that most occupants will get whatever they can out of them. And if they have at least some power and their friends or relatives don't, there will be guests using some power as well.

 

[gadfly56]

Code aside, you are supplying loads that will likely have a high demand during situations when they will be energized by standby power. Normal load diversity used to determine allowable demand factors mean nothing in this situation. With limited number of outlets being supplied you can almost plan that most occupants will get whatever they can out of them. And if they have at least some power and their friends or relatives don't, there will be guests using some power as well.

If you have dedicated lighting circuits you can calculate actual loads, which will most certainly be less than 15 amps worth, or 12 if you want to derate. If you have a dedicated circuit for the refrigerator you could calculate a diversity for that load among all the units, also going to be much less than 15 amps. If you allow one 20 amp SABC to be supplied by the generator, that's enough for folks to plug in a microwave. What do you have now, maybe 30 amps tops per unit?

 

[kwired]

If you have dedicated lighting circuits you can calculate actual loads, which will most certainly be less than 15 amps worth, or 12 if you want to derate. If you have a dedicated circuit for the refrigerator you could calculate a diversity for that load among all the units, also going to be much less than 15 amps. If you allow one 20 amp SABC to be supplied by the generator, that's enough for folks to plug in a microwave. What do you have now, maybe 30 amps tops per unit?

For short term standby that is probably fine, OP mentioned recent impact from hurricanes, which got me to thinking if power is out for long term - any working outlet will get anything imaginable plugged into it. My answer to that situation is it is too hard to predict what the load will be. In Florida maybe you won't have the space heaters you might find being used where I live if this were to happen between October and April.

 

[Jim_SWFL]

Not to poke a hole in a good idea, but i sounds like you are using one ATS for all the condo units? Is there only one meter on the building?

Yes, there one meter for the entire building including house and residence loads.

We are using a single ATS for the loads I describe above. There is a separate, larger ATS, for the house loads including life safety equipment.

 

[Jim_SWFL]

I don't think you can use demand factors unless it is the whole dwelling or whole building,

Section 403.4.8 of the Building Code likely ended up with a relatively large generator already, so some 180VA receptacles shouldn't be a problem.

You would think, but this will also tie into other projects with many more units.

Even a few receptacles, lights, and a refrigerator multiplied by 58 units (as one example) becomes a large load without a demand factor.

I agree that most of these kinds of buildings will have relatively large generators to begin with, but it would be easy to come up with another 200kva of loads by just summing up all of the units.

 

[Jim_SWFL]

Code aside, you are supplying loads that will likely have a high demand during situations when they will be energized by standby power. Normal load diversity used to determine allowable demand factors mean nothing in this situation. With limited number of outlets being supplied you can almost plan that most occupants will get whatever they can out of them. And if they have at least some power and their friends or relatives don't, there will be guests using some power as well.

I understand your point about how some residences might utilize a receptacle during an extended outage, there could also be a lot of unnecessary generator capacity and fuel storage that will go unused.

 

[Jim_SWFL]

If you have dedicated lighting circuits you can calculate actual loads, which will most certainly be less than 15 amps worth, or 12 if you want to derate. If you have a dedicated circuit for the refrigerator you could calculate a diversity for that load among all the units, also going to be much less than 15 amps. If you allow one 20 amp SABC to be supplied by the generator, that's enough for folks to plug in a microwave. What do you have now, maybe 30 amps tops per unit?

Yep - just over 20A calculated per unit.

However, the initial plan was a 60A feeder shared by 3 units - using demand factors. Without a demand factor these feeders get bumped to 80A. Not a big deal. The bigger problem is the load multiplied by the number of units - it really starts to add up.

This is one project. We have other high rises where the developers are asking about similar solutions, and as I've mentioned in a couple of other replies, this will add hundreds of kVA to the overall generator size.

 

[Jim_SWFL]

For short term standby that is probably fine, OP mentioned recent impact from hurricanes, which got me to thinking if power is out for long term - any working outlet will get anything imaginable plugged into it. My answer to that situation is it is too hard to predict what the load will be. In Florida maybe you won't have the space heaters you might find being used where I live if this were to happen between October and April.

The intent was that residents could use phone chargers and maybe a fan or two, and the receptacles weren't being placed in areas that we convenient, but nothing stops them from using extension cords...

 

[kwired]

The intent was that residents could use phone chargers and maybe a fan or two, and the receptacles weren't being placed in areas that we convenient, but nothing stops them from using extension cords...

My reply based on my own experience of being without power for about two weeks - and in cold weather on top of it. First couple days aren't so bad, after four or five days you long for anything normal and find ways to power things up to get as close as possible to normal as you can. Once the thermal mass temp of the house drops it is miserably cold no matter what you try to do to keep it warm in just the main area of the house.

We don't typically have extended outages during summer months unless you were directly hit by a tornado - then you probably don't have a home to go to anyhow, but distribution damages in such cases are concentrated to small areas and there is no widespread outages for any extended time periods.

 

[Jim_SWFL]

I appreciate the input from everyone.

What I'm wrestling with is the idea that these specific loads, when lumped in with the rest and used to calculate the overall service to the building, are a tiny fraction.

Then when they are broken out separately for the purposes of providing backup power, that isn't required, we have to provide generator and wire capacity to carry 100% of the calculated load.

Granted, it is a small load when viewed unit by unit, but in summation they grow very quickly.

As I've mentioned in other replies, when these discussions get going for buildings with 50 or 150 units (these are the typical size projects we deal with), these loads will easily grow to 300, 500, .. kVA.

I guess the more I think about it, I can see where it is worth sizing the wire at 100%, but not the generator, since these are optional loads and can be disconnected if needed.

 

[Jim_SWFL]

My reply based on my own experience of being without power for about two weeks - and in cold weather on top of it. First couple days aren't so bad, after four or five days you long for anything normal and find ways to power things up to get as close as possible to normal as you can. Once the thermal mass temp of the house drops it is miserably cold no matter what you try to do to keep it warm in just the main area of the house.

We don't typically have extended outages during summer months unless you were directly hit by a tornado - then you probably don't have a home to go to anyhow, but distribution damages in such cases are concentrated to small areas and there is no widespread outages for any extended time periods.

Same here - we were without power for a few days after Irma, and we were seeking any way of getting comfort from the heat and humidity. I can't imagine going through that in the winter. We moved here from Cleveland, so we understand what winter is like, but not without the ability to heat our home - yikes.

 

[kwired]

Same here - we were without power for a few days after Irma, and we were seeking any way of getting comfort from the heat and humidity. I can't imagine going through that in the winter. We moved here from Cleveland, so we understand what winter is like, but not without the ability to heat our home - yikes.

I have electric heat, geothermal heat pump with resistance back up heat. I though a wood burning fireplace would be sufficient for heat during power outages - not really, though there are some designed to put more heat into the room then what I ended up using. Otherwise a majority of the heat goes right up the flue, most of what you get in the room is radiant heat off the back wall of the fireplace, and it is very time consuming keeping that fire stoked up to get maximum effect.

 

[gadfly56]

I have electric heat, geothermal heat pump with resistance back up heat. I though a wood burning fireplace would be sufficient for heat during power outages - not really, though there are some designed to put more heat into the room then what I ended up using. Otherwise a majority of the heat goes right up the flue, most of what you get in the room is radiant heat off the back wall of the fireplace, and it is very time consuming keeping that fire stoked up to get maximum effect.

Many homes, particularly tract housing constructed from the 50's through the 70's, draw the combustion air from the interior of the house. So not only does the fireplace not do a good job of heating the space it's in (Count Rumford's expertise nowhere in evidence), it pulls cold outside air through every available nook and cranny, chilling other spaces accordingly. I remember my parents' house where we would enjoy a fire in living room fireplace in the late evening and retire to freezing cold bedrooms afterwards.

 

[Smart $]

...
I guess the more I think about it, I can see where it is worth sizing the wire at 100%, but not the generator, since these are optional loads and can be disconnected if needed.

Load shedding has to be automatic if you are using an ATS. Or are you saying you'll just drop the 'convenience' load entirely via its ATS if the generator becomes overloaded?

In your original post you queried about "220.84(A)(1) permits the use of the demand factors in table 220.84 as long as the unit has a single feeder." You are correct. If units are supplied by more than one feeder, you cannot use the demand table. I think this will be the largest constraint of implementing the idea. Appears the only way to implement is to use a unit ATS set up.

 

[Jim_SWFL]

Load shedding has to be automatic if you are using an ATS. Or are you saying you'll just drop the 'convenience' load entirely via its ATS if the generator becomes overloaded?

In your original post you queried about "220.84(A)(1) permits the use of the demand factors in table 220.84 as long as the unit has a single feeder." You are correct. If units are supplied by more than one feeder, you cannot use the demand table. I think this will be the largest constraint of implementing the idea. Appears the only way to implement is to use a unit ATS set up.

We have used automatic load shedding in the past on other jobs, it may be necessary to use it here too.

Thanks for touching on 220.84(A)(1)... this is a concern of mine and how the AHJ will look at it. I considered implementing some sort of load transfer on a per-unit basis. Any recommendations on a manufacturer you'd consider for this?

My searches resulted in devices intended for lighting loads.