220.87 - Maximum Demand from Power Company?

Status
Not open for further replies.

BattleCat

Member
Location
NJ
I am curious how to satisfy 220.87 by data from the power company? If I were to call the power company about a particular house and ask for 12 months of data showing the maximum demand, what will they give me? This information can't be found by using normal KWh usage from typical monthly invoices, correct? Or is it? Are they just going to average out the usage over the last 12 months?

I've always done load calculations before, but I would like to try this method and learn how to do it, that are some circumstances that make it easier anyway. I need this data for a whole house generator installation. Thank you for your help.
 

G._S._Ohm

Senior Member
Location
DC area
U.S. houses pull 4A (yearly average) with short peaks possibly up to 200A and a generator may have a surge rating of 1.3x the steady state rating.


How long can a surge last for the generator to still handle it? If a gen follows an (I^2) T damage curve and can occasionally handle 20A for two seconds then it can maybe handle 28A for one second, if it doesn't stall. (20^2) x 2 = (28^2) x 1 = ~800.
Once you know this you can match the gen surge capability to the house demand.


Generally, the higher the peak demand the less often it will happen.


BTW, depending on the variation you see over a few days in the 15 minute period there are statistical methods to determine to some level of certainty the max demand you will ever see, provided your sample represents 'typical' usage for that house. But whether the AHJ accepts your reduced sampling period and its complex math is another question.

The NEC may care about 15 minutes but your gen is concerned about much shorter durations, depending on the hp and the size of its flywheel.
 
Last edited:

don_resqcapt19

Moderator
Staff member
Location
Illinois
Occupation
retired electrician
The typical residential meters are not demand meters. They have no way of providing the information that you are looking for. Maybe the new smart meters have that function, but I don't know.
 

kwired

Electron manager
Location
NE Nebraska
The typical residential meters are not demand meters. They have no way of providing the information that you are looking for. Maybe the new smart meters have that function, but I don't know.
I second that. Even light commercial often uses same type of meter as most residential services.

If meter is not capable of storing demand information, all you know is how much kWhrs were used, but you have no idea if they were evenly averaged during a billing period or if there were significant peak times of high demand.
 

BattleCat

Member
Location
NJ
I see. So there is no real way to do did. That makes sense, I didn't think the power company would have a way to find that out. I guess a typical load calc is our only option.

While we're on the topic, does anyone know of a good load calculator that doesn't require Microsoft Office? Everything I read always says to use the Mike Holt one, but it requires Office and that's a bit spendy for me.
 
Last edited:

kwired

Electron manager
Location
NE Nebraska
I see. So there is no real way to do did. That makes sense, I didn't think the power company would have a way to find that out. I guess a typical load calc is our only option.

While we're on the topic, does anyone know of a good load calculator that doesn't require Microsoft Office? Everything I read always says to use the Mike Holt one, but it requires Office and that's a bit spendy for me.

I use NEC art 210 and my desktop calculator:happyyes:

EDIT: Sorry, I meant art 220.
 

BattleCat

Member
Location
NJ
I downloaded OpenOffice, never knew about that! I was able to use the load calculator on Mike Holt's download page. I got a 2,900 sqft house that I ran through it, it's all natural gas so I knew the load wouldn't be terribly high, but with 3 AC units I figured it wouldn't be all that low. However the calculator came up with 51A. That seems way too low for a house with 3 central AC units (2.5, 2.5, and 1.5 ton).

I'll do it long hand in the mornings, but I was just wondering if that seemed wrong to anyone else?
 

Hv&Lv

Senior Member
Location
-
Occupation
Engineer/Technician
I am curious how to satisfy 220.87 by data from the power company? If I were to call the power company about a particular house and ask for 12 months of data showing the maximum demand, what will they give me? This information can't be found by using normal KWh usage from typical monthly invoices, correct? Or is it? Are they just going to average out the usage over the last 12 months?

I've always done load calculations before, but I would like to try this method and learn how to do it, that are some circumstances that make it easier anyway. I need this data for a whole house generator installation. Thank you for your help.

Go ahead, call and ask. Our new smart meters have demand displayed on the screen. We reset the demand every spring and fall. Many of the smart meters have the demand feature, it may or may not display.
 

BattleCat

Member
Location
NJ
Go ahead, call and ask. Our new smart meters have demand displayed on the screen. We reset the demand every spring and fall. Many of the smart meters have the demand feature, it may or may not display.
This house doesn't have a digital meter. That's why I found it odd that the PoCo would be able to accurately give this info.


I'm curious about the Load Calculator on Mike Holt's download page, does that include calculating the 25% of the largest motor to the load calculation? It doesn't display it, but it might be doing it behind the scenes.
 

Hv&Lv

Senior Member
Location
-
Occupation
Engineer/Technician
This house doesn't have a digital meter. That's why I found it odd that the PoCo would be able to accurately give this info.


I'm curious about the Load Calculator on Mike Holt's download page, does that include calculating the 25% of the largest motor to the load calculation? It doesn't display it, but it might be doing it behind the scenes.

Not sure. you can also use 30 days worth of measurements, I really don't like that method though. 30 days in September isn't the same as 30 days in January.
 

BattleCat

Member
Location
NJ
Not sure. you can also use 30 days worth of measurements, I really don't like that method though. 30 days in September isn't the same as 30 days in January.
Exactly, especially because all houses I service are natural gas. If I took a measurement now, it would be very low and wouldn't show the huge load that the house sees all summer.
 

G._S._Ohm

Senior Member
Location
DC area
To some degree of accuracy, and by sampling and calculation and past electric bills, IMO this power profile could be done reasonably within an hour or so.
Can an EC claim this 30 days is an undue burden?
 
Last edited:

G._S._Ohm

Senior Member
Location
DC area
I downloaded OpenOffice, never knew about that! I was able to use the load calculator on Mike Holt's download page. I got a 2,900 sqft house that I ran through it, it's all natural gas so I knew the load wouldn't be terribly high, but with 3 AC units I figured it wouldn't be all that low. However the calculator came up with 51A. That seems way too low for a house with 3 central AC units (2.5, 2.5, and 1.5 ton).

I'll do it long hand in the mornings, but I was just wondering if that seemed wrong to anyone else?
What is the chance that all three will be on at the same time within a random 15 minute interval? Assume a 50% duty cycle. My 2.5 pulls about 20A.

If there's a power failure they will be all be on for a while but that happens rarely.
 

G._S._Ohm

Senior Member
Location
DC area
Assuming they pull 20A, 20A and 12A, all combinations are listed below.
1/2 duty cycle, 3 compressorshoriz sum\/
0000
001212
020020
0201232
200020
2001232
2020040
20201252
26.00=avg pwr
0=1/8 of the time
12=1/8 of the time
20=2/8 of the time
32=2/8 of the time
40=1/8 of the time
52=1/8 of the time
So for avg.power 51A seems high to me.
 

mivey

Senior Member
The NEC may care about 15 minutes but your gen is concerned about much shorter durations, depending on the hp and the size of its flywheel.
A very good point.

The typical residential meters are not demand meters. They have no way of providing the information that you are looking for. Maybe the new smart meters have that function, but I don't know.
Most probably do not.

Go ahead, call and ask. Our new smart meters have demand displayed on the screen. We reset the demand every spring and fall. Many of the smart meters have the demand feature, it may or may not display.
I wish they all did, but many utilities are cheaping out and getting energy-only meters for residential and small commercial. I have seen some work-arounds that try to use frequent meter reads but the incremental data transmission is not reliable enough. Better to have a meter that stores the demand up front.

To some degree of accuracy, and by sampling and calculation and past electric bills, IMO this power profile could be done reasonably within an hour or so.
You would need a pretty reliable load profile for the particular home. You might as well make a load calculation as recommended by the gen manufacturer.
 

G._S._Ohm

Senior Member
Location
DC area
You would need a pretty reliable load profile for the particular home. You might as well make a load calculation as recommended by the gen manufacturer.
Since the code cite does not specify an accuracy goal for this method, what is a reasonable value? How accurate are the gen maker methods and how accurate do they need to be for a reasonable gen. lifetime and very few HO brownout complaints?

I can't imagine anyone renting a datalogger for a month to do this sampling test but I'm not an EC.
 

mivey

Senior Member
Since the code cite does not specify an accuracy goal for this method, what is a reasonable value? How accurate are the gen maker methods and how accurate do they need to be for a reasonable gen. lifetime and very few HO brownout complaints?
I don't know.

I can't imagine anyone renting a datalogger for a month to do this sampling test but I'm not an EC.
If you are going to use the sampling method for load calcs, that is kind of what the NEC requires. It does not have to be a full-blown PQ meter, just a demand recording meter. However, I would not necessarily use this method alone for genset sizing.
 

G._S._Ohm

Senior Member
Location
DC area
Here's a simplified version of this problem.


You have a sampling ammeter that records in 15 min. intervals.


Once/month a guy throws a 4 hr party during which time his house draws much more current than normal. Four hours is 16 ea. 15 minute chunks.


You don't know when the party will be and you have one month to look for this peak current draw.


If you ran the ammeter continuously, that is 2880 15 minute chunks, you have a 100% chance of detecting this peak current draw.
If you ran the ammeter once for 15 minutes you have 16 chances out of 2880 (0.00555, ~0.6%) of detecting this peak current draw so you have 1-0.00555= 0.994 (99.4% chance) of not detecting it.

How many times (x) would you have to sample within the month to have a 50% (0.5) chance of detecting the peak? How 'bout for a 95% (0.05) chance?


This is 0.994^x = 0.5 and 0.994^x = 0.05.
x = (log 0.5)/(log 0.994) = 115 times
and
x = (log 0.05)/(log 0.994) = 498 times
 
Last edited:
Status
Not open for further replies.
Top