220.87 Determining Existing Loads

A lot of inspectors outright say that there’s no way a 100 amp service on a single-family dwelling can support a 48 amp EV charger because they’ve always seen load calculations of the dwelling being in the 80 amp range.
So when coming to them with data saying that the house is actually under 20 amp, many are just going to call BS. Having guidance from their governing body would prevent this from becoming an issue.
I've put EVSE on an apartment building with 14 units of housing served by a single 100amp breaker.
With no pushback from the AHJ. The cost is a $350 adder for active load management.

Your other approach is a cheap Empora load meter and get 1 second data for a week, including some tests of "turning everything on".
Your AHJ is going to have trouble calling that BS. I keep a loaner Empora just for that purpose, and tether it to a phone if WiFi is not available.

---
The POWER working group has been working for about three years to raise awareness of the magnitude of calculation based overestimation of actual load.

---
What won't work? Even in my area where power 4-6pm is 61 cents per kWh, I can't argue that the homeowner won't charge at peak times.
Because, they might.
 
P.S. I think we've discussed before the idea that for demand data, the utility might record, say, 1 minute average power usage data, and then use a 15 minute sliding window, and the peak demand would be the largest number you can get by summing up 15 consecutive 1 minute average power figures.

In that case, the underlying 1 minute data (not reported to you) is equivalent to kWh used each minute. If you only have 15 minute kWh usage data, you can still provide an upper bound for the 15 minute sliding window peak demand, based on the largest sum of two consecutive 15 minute kWh usage numbers. Because the sum for any 15 consecutive minutes within that 30 minute interval will certainly be less than the sum of all 30 minutes.

However, I don't see the text of 220.87 as requiring the usage of a sliding window, or even such fine grained measurement as 15 minute interval data (for the case of 12 months of data). The 125% factor applied to the maximum observed demand is apparently judged sufficient to cover any undercount attributable to a longer averaging interval.

Cheers, Wayne
 
I've put EVSE on an apartment building with 14 units of housing served by a single 100amp breaker.
With no pushback from the AHJ. The cost is a $350 adder for active load management.

Are you talking about the Wallbox? I am eagerly waiting for ChargePoint to include a reasonably priced load management system that could be added to their EVSE.

I just purchased a DCC-12 load shedder for $800 for a job. This thing is a pair of contactors, a circuit board, and a couple CT’s. It just shuts the charger down dead when the load approaches 80% of the service ampacity. I don’t like anything about this.
 
Multiplying by 4 is a waste of time. Just find the highest interval and divide it by 15. This is your maximum average demand per minute, but it is still up to the AHJ.
(kWh used during a 15 minute interval / 15 minutes) * 60 minutes/hr = average kW during the interval. And 60/15 = 4. So your procedure also multiplies by 4.

Cheers, Wayne
 
P.S. I think we've discussed before the idea that for demand data, the utility might record, say, 1 minute average power usage data, and then use a 15 minute sliding window, and the peak demand would be the largest number you can get by summing up 15 consecutive 1 minute average power figures.

In that case, the underlying 1 minute data (not reported to you) is equivalent to kWh used each minute. If you only have 15 minute kWh usage data, you can still provide an upper bound for the 15 minute sliding window peak demand, based on the largest sum of two consecutive 15 minute kWh usage numbers. Because the sum for any 15 consecutive minutes within that 30 minute interval will certainly be less than the sum of all 30 minutes.

However, I don't see the text of 220.87 as requiring the usage of a sliding window, or even such fine grained measurement as 15 minute interval data (for the case of 12 months of data). The 125% factor applied to the maximum observed demand is apparently judged sufficient to cover any undercount attributable to a longer averaging interval.

Cheers, Wayne
I recall that some utilize 5 minute intervals. Some may just use 15 min intervals, same as the residential smart meters. Some may measure it in kVA? I'm unaware of a standard. In any case I agree that the notion that there's some industry standard definition of demand that's more strict and rigorous than 15min interval energy consumption has not been supported here. AHJs may need some educating, but that can always be said.
 
A lot of inspectors outright say that there’s no way a 100 amp service on a single-family dwelling can support a 48 amp EV charger because they’ve always seen load calculations of the dwelling being in the 80 amp range.

So when coming to them with data saying that the house is actually under 20 amp, many are just going to call BS. Having guidance from their governing body would prevent this from becoming an issue.
Read through this thread and you'll understand why HHS inspectors won't touch these demand calculations.
 
The “demand” could be a very high amount of power drawn for only one second. We would have to go by that high amount of power, not what it averaged out to over 15 minutes.
@JohnHess I respectfully disagree.

A one second peak load does not automatically repsent anything of meaning. Load is a proxy for heat, which is a problem.
Look at a trip curve for a breaker.... you can draw 2x 3x or 4x the power for quite some seconds before anything bad happens.
This is the only reason motors can operate at startup.

Continuous loads (like EVSE) and peak loads (like compressor startup) are totally different beasts.
The 15 minute data averages all that out. I'd prefer something closer to 5 minute data, and actually collect 1 minute data,
but regardless "smoothing" that data to some extent is ok. Sliding windows are ok. Frankly THERMAL measurements should be OK if all breakers and wires could report them.

Standard_Trip_Characteristic_of_a_Thermomagnetic_Circuit_Breaker.svg.png

Peak amps are a nothingburger as long as they don't create a spark or build up heat.
View attachment 2579178

 
Here in NJ an HHS electrcial inspector is doing the plan review and it's probable that they won't approve a load calculation cobbled together with the historical data mentioned in this thread.
Eh, using 12 months of 15 minute kWh data clearly complies with 220.87, and if the electrical inspectors in your area don't accept that, they need some further education.

Cheers, Wayne
 
Eh, using 12 months of 15 minute kWh data clearly complies with 220.87, and if the electrical inspectors in your area don't accept that, they need some further education.

Cheers, Wayne
What is evident from this thread is that that calculation method is full of holes so it can easily be rejected by the plan reviewer which is why the question was asked in the first place. I'm curious to see if John can get an answer from the State.
 
I have found the optional calc pretty favorable to EVSE's.
P.S. I think we've discussed before the idea that for demand data, the utility might record, say, 1 minute average power usage data, and then use a 15 minute sliding window, and the peak demand would be the largest number you can get by summing up 15 consecutive 1 minute average power figures.
Can you clarify what you mean by 'sliding window' say that for 60 1 min values say every 5th minute reads at 10 and all the other values were 2.
 
Can you clarify what you mean by 'sliding window' say that for 60 1 min values say every 5th minute reads at 10 and all the other values were 2.
Your example is regular enough that the 15 minute fixed window method (looking at minutes 0-14, 15-29, 30-44, and 45-59) will give you the same answer as a sliding window method (which would consider minutes 0-14; then 1-15; then 2-16, etc up to 45-59).

An example illustrating the difference is where minutes 0-59 are all 0 except that minutes 14 and 15 both have 1 kWh of consumption (meaning 60 kW of average power over each of those two minutes). The fixed window method would report a maximum demand of 4 kW because the spike in consumption happens to occur at a boundary between the fixed windows. The sliding window method would report a maximum demand of 8 kW, which is captured by any of the windows 1-15 through 14-28.

But this discrepancy is not so likely to control the maximum demand, particularly over 12 months of data, absent active measures to achieve such discrepancies.

Cheers, Wayne
 
Top