220.87 Determining Existing Loads

[JohnHess]

I believe that the residential load calculations we perform are very bloated. Sometimes they can limit what we can install for the homeowner or cause them to spend unnecessary money on service upgrades or load management products. Instead of submitting load calculations with my electrical permit applications I would like to submit 1 year of the actual maximum demand data because I know that will be much lower and more accurate than the load calc.

My question is what the "maximum demand" is, exactly. Our area is served by a very large PoCo and is mostly smart meters now, so I assume that data will be available thru their website. But I am not sure exactly what form it needs to be in to tell me the average load in place of the VA number I get when doing a load calculation.

 

[brycenesbitt]

@JohnHess
Which large PoCo ?

There's a ton of data on this exact issue taken from thousands of homes. And in fact, that data was presented to the NEC committee at the technical meeting stage, resulting in new calc language in the 2026 code cycle.

TL;DR panels have been oversized by those calculations for a long time.

Most of the new formulas made it through the whole cycle, except for one, and that's the subject of an appeal August 18th 2025.
Send a private message I can get you a full summary.

---
For maximum demand it depends on the PoCo. Identify yours for more help. Generally, you can find it on a bill as a "demand charge" or "peak charge". In some cases, you can download 1- or 15-minute data and find the peaks yourself. However, in other cases you only get 1 hour data which is kind of chancy for making load assumptions.

There are now a TON of ways to "peak shave" these days. Or, to have a new device dance about and use power only when the panel has capacity.

------
Eaton's appeal on load related issues reads in part

NFPA Standards Council
1 Batterymarch Park
Quincy, MA 02169

RE: Appeal regarding CAM 70-87 18 July 2025

Dear Ms. Bellis,

I am writing to formally appeal the outcome of certified amending motion 70–87. This is a unique case of where the process created a situation that negatively impacts all parties involved and arguably negatively impacts electrical safety as well. ....
The issue stems from a correlation problem created by the recent process involving Certified Amending Motion 70–87. As per the motion made at the annual meeting, Public Comment 1854 was accepted, which proposed increasing the VA/ft2 for calculating the number of branch circuits in a dwelling unit from 3 VA/ft2 to 4 VA/ft2. ....
If we are not careful, the calculation for the number of branch circuits for a dwelling unit could shift to a 2 VA/ft2 basis....

This motion passed at the annual meeting, and according to NFPA's regulations, such a motion requires that the committee, in
this case CMP-2, be balloted. However, CMP- 2 did not support the ballot. As per the NFPA standards development procedures, when this occurs, the text must revert to the language of the previous edition, which in this case is the 2023 National Electrical Code (NEC). The
2023 NEC did not contain Section 120.13, and NFPA regulations dictate that if no corresponding text exists in the previous edition, the section must be deleted. Thus, Section 120.13 is being deleted.

 

[JohnHess]

@brycenesbitt Thank you for the reply!

The power company is PSE&G in northern NJ.

After doing some searching, I found that PSE&G's commercial bills list the maximum demand right in the sidebar of each bill. However, the residential bills do not. I was not able to find any information like this on my own bill.

PSE&G's website has a lot of data available in multiple forms. My concern is to find the exact type of data necessary and put it together into a way that municipal inspectors will have to accept in lieu of a typical load calculation.

These inspectors are accustomed to saying that a 100A dwelling service can't handle a 48A EVSE. And load calculations agree with the inspector since they are so bloated and usually come out to 80-85A. But I know that a 100A service can easily handle a 48A EVSE since our houses are all natural gas, and the maximum demand is under 10A.

Since many inspectors might not like this and there might be a disagreement, I would like to be sure that the data I submit is proper.

The changes you mentioned in the 2026 are interesting, I am going to read into that tomorrow.

 

[Birken Vogt]

What about the old way, install monitoring equipment for 30 days and get your own peak? Is that valid?

 

[JohnHess]

What about the old way, install monitoring equipment for 30 days and get your own peak? Is that valid?

That wouldn’t work so well for me, unfortunately. Most of the time I complete the installation within two weeks of the customer’s initial call. Doing my own monitoring would put the job off by 30 days.

There’s a few other reasons as well. In my area the only large load is air conditioning since everything else is natural gas. So if I did a 30 day monitoring during any of the cooler months, the inspectors would immediately say that it’s not going to work in the summer. Getting the information from the power company would be more authoritative as well as cover the whole year.

 

[brycenesbitt]

What about the old way, install monitoring equipment for 30 days and get your own peak? Is that valid?

Yeah, but.... seasons.

Note this is a well worn topic

Calculations for existing buildings - 30 Day Metering

Hi, I wanted to get a consensus on this topic because I have read conflicting arguments on how to calculate total load on a feeder using existing metering data and new load. Let's assume we have done a 30 day load study and have and multiplied the load times 1.25 as stated in the code, NEC...

forums.mikeholt.com

 

[Birken Vogt]

That wouldn’t work so well for me, unfortunately. Most of the time I complete the installation within two weeks of the customer’s initial call. Doing my own monitoring would put the job off by 30 days.

Well if you can't get a permit then there will be no job at all, so they can jolly well wait 30 days if they have a 100 amp service they do not want to upgrade. I'm sure you will already know well enough whether it will pass or fail anyway.

There’s a few other reasons as well. In my area the only large load is air conditioning since everything else is natural gas. So if I did a 30 day monitoring during any of the cooler months, the inspectors would immediately say that it’s not going to work in the summer. Getting the information from the power company would be more authoritative as well as cover the whole year.

Code addresses this in 220.87(1) Exception

Take the max demand and add in the AC if the 30 days you picked does not include AC.

I know my poco has 15 minute demand for commercial but not for resi so we are out of luck there.

 

[JohnHess]

I have been reading other threads on this topic. It seems like I can’t find any singular agreed upon dataset that could be used for 220.87.

 

[brycenesbitt]

From the substantiation for reduction in the formula in the next NEC edition:

 

[wwhitney]

However, in other cases you only get 1 hour data which is kind of chancy for making load assumptions.

One of the PIs I submitted for 2026 was to move the reference to interval data being every 15 minutes from the Exception (30 day monitoring) to the main text (one year of data). The CMP response was in effect that if using one year of data, the interval frequency does not need to be 15 minutes.

The upshot is I believe the CMP thinks it's fine to use one year of one hour data.

Cheers, Wayne

 

[JohnHess]

Through my power company’s website I’m able to go into their MyMeter portal and download some information.

I can easily download a CSV file listing the kWh every 15 minutes for an entire year. As you can imagine that’s a lot of lines and not something I could print out to submit with the permit application.

If I paste that into a spreadsheet and do a simple average of all those numbers, do you think it would be acceptable to just send in a printed out screenshot of that average kWh?

I need this to be something authentic because these inspectors are accustomed to seeing load calculations saying these houses use at least 80 A, while the actual usage is most likely less than 10 A.

 

[infinity]

I have been reading other threads on this topic. It seems like I can’t find any singular agreed upon dataset that could be used for 220.87.

You're probably correct and that's why your local NJ HHS electrical inspector isn't going to accept anything other than a load calculation whether that calc is bloated or not. If someone wants to add a 48 amp EV charger to their existing 100amp service they'll have to consider a service upgrade unless the standard or optional calc says that it's code compliant. They could also choose a lower level charging system which maybe more suitable based on the size of their existing service.

 

[JohnHess]

You're probably correct and that's why your local NJ HHS electrical inspector isn't going to accept anything other than a load calculation whether that calc is bloated or not. If someone wants to add a 48 amp EV charger to their existing 100amp service they'll have to consider a service upgrade unless the standard or optional calc says that it's code compliant. They could also choose a lower level charging system which maybe more suitable based on the size of their existing service.

Most people want either the ChargePoint Homeflex or the Tesla Wall Connector. Either because the ChargePoint is recommended by our power company since it’s eligible for all the incentives, or they got a free Wall Connector with their Tesla. So I typically just lower the maximum charge rate and lock it out, as well as put a sticker on it and the electrical panel stating this. That always works.

But it does bother me a little that I have to limit the customer based on a bloated calculation, when we all know that the house is using so much less power and could easily handle that charger.

Maybe I’ll contact the Office of Regulatory Affairs and see if they would be willing to back a way to present a dataset in accordance with 220.87. The DCA has been pretty good with various things dealing with EV chargers, such as allowing 6-2 NM for 48A chargers on 60A circuits.

 

[suemarkp]

Through my power company’s website I’m able to go into their MyMeter portal and download some information.

I can easily download a CSV file listing the kWh every 15 minutes for an entire year. As you can imagine that’s a lot of lines and not something I could print out to submit with the permit application.

If I paste that into a spreadsheet and do a simple average of all those numbers, do you think it would be acceptable to just send in a printed out screenshot of that average kWh?

I dont think you average it, you sort it and find the largest one. You just have to watch the units and interval period to adjust it and end up with kw.

 

[infinity]

Maybe I’ll contact the Office of Regulatory Affairs and see if they would be willing to back a way to present a dataset in accordance with 220.87. The DCA has been pretty good with various things dealing with EV chargers, such as allowing 6-2 NM for 48A chargers on 60A circuits.

Absent something of this nature I think that we're stuck with the current system of standard or optional load calculation according to article 220. HHS inspectors are always looking for guidance from the DCA because it makes their job easier.

 

[JohnHess]

I dont think you average it, you sort it and find the largest one. You just have to watch the units and interval period to adjust it and end up with kw.

Yes, so I just downloaded the data from a house, 1 year worth of 15 minute increments. I sorted it with the highest first, which was 1.7 kWh. That is 6.8 kW which comes to 28 Amps.

When doing a load calculation by hand I came to 79 Amps.

Biiiiiig difference. One says I can't install a typical 48 or 40 Amp EVSE. The other says there is plenty of ampacity for it.

Absent something of this nature I think that we're stuck with the current system of standard or optional load calculation according to article 220. HHS inspectors are always looking for guidance from the DCA because it makes their job easier.

I am going to try. I will let you know what I come up with. Or maybe you will see it yourself in the next CCC!

 

[curt swartz]

Yes, so I just downloaded the data from a house, 1 year worth of 15 minute increments. I sorted it with the highest first, which was 1.7 kWh. That is 6.8 kW which comes to 28 Amps.

When doing a load calculation by hand I came to 79 Amps.

Biiiiiig difference. One says I can't install a typical 48 or 40 Amp EVSE. The other says there is plenty of ampacity for it.


I am going to try. I will let you know what I come up with. Or maybe you will see it yourself in the next CCC!

KWh? That number is useless to determine loads.

Are you looking at usage or demand?

 

[JohnHess]

KWh? That number is useless to determine loads.

Are you looking at usage or demand?

Even when measured in 15 minute increments?

I don’t believe there’s any way to get the demand for residential customers.

 

[jim dungar]

I sorted it with the highest first, which was 1.7 kWh.

KW is KW. You do not need to do any manipulation of the 15 minute interval values, except find the largest.
Load size is measured in kW, energy used is kWHr.

 

[brycenesbitt]

Just be sure you're not off by a factor of four.
Kilowatts are a rate. A kWh is a kilowatt that ran for an hour.
One kW for 15 minutes is .25 of a kWh.

Saying "1 year worth of 15 minute increments. I sorted it with the highest first, which was 1.7 kWh" makes no sense.

What would make sense the highest 15 minute bucket averaged 1.7kW ( 7 amps ).

----
Anyone interested in this topic, there is a working group for it. PM me.