220.87

[don_resqcapt19]

The utilities in this area can provide fifteen minute usage data for the previous year. Can you use the highest 15 minute kWH data for determining the maximum load on the service when you are adding additional loads such as an EV?

 

[tortuga]

The utilities in this area can provide fifteen minute usage data for the previous year. Can you use the highest 15 minute kWH data for determining the maximum load on the service when you are adding additional loads such as an EV?

Thats pretty cool I dont see why not as they also apparently can now monitor the load in real time.
last summer I had one utility remotely shut off a commercial service using the meter for exceeding 200 Amps for over two hours.
So once you add the car, they will make sure you dont exceed the rating of the service.

 

[wwhitney]

Is your question whether the 15 minute kWh data satisfies the 220.87(1) requirement for "maximum demand data"? That's not specified, so it's ambiguous, but as the exception refers to 15 minute kWh over a shorter time period, one could argue 15 minute data should suffice for 220.87(1).

On the other hand, the exception does specify that the "highest loaded phase" be monitored, based on the loading at the start of the 30 day period. So if the AHJ decides "maximum demand data" refers to a sampling period smaller than 15 minutes, and you don't have any per phase data available from the POCO, you couldn't apply 220.87(1).

Cheers, Wayne

 

[ron]

Depending on the unit of measure that they provide for the 15 minute kWh that you are referring to, if it were kWh, I select the highest 15 minute value and multiply it by 4. That satisfies 220.87 IMHO

 

[tortuga]

It would be better if you can access peak kW data as opposed to kWH

 

[don_resqcapt19]

The reason I ask is the meter that the utility uses also has a 15 minute demand register, and they do not have that one enabled. So I am thinking that a 15 minute peak kW is different from a 15 minute usage kWH.
This is the type of data that we can download from the utility.

Date/Time	Kind	Quantity	Units	Interval
1/16/2023 0:00​	Usage	0.34​	kwh	15min
1/16/2023 0:15​	Usage	0.31​	kwh	15min
1/16/2023 0:30​	Usage	0.27​	kwh	15min
1/16/2023 0:45​	Usage	0.29​	kwh	15min
1/16/2023 1:00​	Usage	0.22​	kwh	15min
1/16/2023 1:15​	Usage	0.33​	kwh	15min
1/16/2023 1:30​	Usage	0.22​	kwh	15min
1/16/2023 1:45​	Usage	0.28​	kwh	15min
1/16/2023 2:00​	Usage	0.24​	kwh	15min
1/16/2023 2:15​	Usage	0.3​	kwh	15min
1/16/2023 2:30​	Usage	0.36​	kwh	15min
1/16/2023 2:45​	Usage	0.28​	kwh	15min
1/16/2023 3:00​	Usage	0.34​	kwh	15min
1/16/2023 3:15​	Usage	0.26​	kwh	15min
1/16/2023 3:30​	Usage	0.33​	kwh	15min
1/16/2023 3:45​	Usage	0.33​	kwh	15min
1/16/2023 4:00​	Usage	0.3​	kwh	15min

 

[charlie b]

I would have to say no. But if you were to accuse me of splitting hairs, I would likely plead guilty.

220.87 is all about “power,” (i.e., KW, the rate at which energy is being used), not the total amount of energy that is used in a given amount of time (i.e., KWH).

Here is an extreme example: A panel has only one single load. It uses 36 KW of energy, runs constantly for 5 minutes, then turns off for 10 minutes. It repeats this cycle every 15 minutes. In each 15-minute interval, it would expend 3 KWH of energy. If you base your decision to add load to that panel on the 3 KWH value, then the system might be overloaded during the 5-minute run cycle of that one big load.

I like to use the following analogy: A driver made a 120-mile trip in 2 hours on a road that had a speed limit of 60 mph. He got a ticket for speeding. How is this possible?

 

[retirede]

I would have to say no. But if you were to accuse me of splitting hairs, I would likely plead guilty.

220.87 is all about “power,” (i.e., KW, the rate at which energy is being used), not the total amount of energy that is used in a given amount of time (i.e., KWH).

Here is an extreme example: A panel has only one single load. It uses 36 KW of energy, runs constantly for 5 minutes, then turns off for 10 minutes. It repeats this cycle every 15 minutes. In each 15-minute interval, it would expend 3 KWH of energy. If you base your decision to add load to that panel on the 3 KWH value, then the system might be overloaded during the 5-minute run cycle of that one big load.

I like to use the following analogy: A driver made a 120-mile trip in 2 hours on a road that had a speed limit of 60 mph. He got a ticket for speeding. How is this possible?

I agree. Demand is not (properly) measured in kWh.


Sent from my iPad using Tapatalk

 

[wwhitney]

But the exception to 220.87(1) does refer to kWh indirectly, in that it refers to average power over a 15 minute interval. Average power over 15 minutes = 4 * kWh used over that time period.

Any power measurement is going to be averaged over some time interval, so what's ambiguous is that 220.87(1) doesn't refer to the sampling interval for the "maximum demand data".

Cheers, Wayne

 

[electrofelon]

I would have to say no. But if you were to accuse me of splitting hairs, I would likely plead guilty.

220.87 is all about “power,” (i.e., KW, the rate at which energy is being used), not the total amount of energy that is used in a given amount of time (i.e., KWH).

Here is an extreme example: A panel has only one single load. It uses 36 KW of energy, runs constantly for 5 minutes, then turns off for 10 minutes. It repeats this cycle every 15 minutes. In each 15-minute interval, it would expend 3 KWH of energy. If you base your decision to add load to that panel on the 3 KWH value, then the system might be overloaded during the 5-minute run cycle of that one big load.

I like to use the following analogy: A driver made a 120-mile trip in 2 hours on a road that had a speed limit of 60 mph. He got a ticket for speeding. How is this possible?

But how is that different than 15 minute demand? Demand is an average power during that 15 minutes, not a peak. I am not seeing any difference, what am I missing?

P.S. when you said " It uses 36 KW of energy", you meant to say power

 

[tortuga]

I like to use the following analogy: A driver made a 120-mile trip in 2 hours on a road that had a speed limit of 60 mph. He got a ticket for speeding. How is this possible?

Wow that is a fantastic example now we have a Charlies rule #2.

 

[don_resqcapt19]

But how is that different than 15 minute demand? Demand is an average power during that 15 minutes, not a peak. I am not seeing any difference, what am I missing?

P.S. when you said " It uses 36 KW of energy", you meant to say power

The 36 kW load for 5 minutes is a 12kW demand for the 15 minute demand window, but only 3 kWH of usage for that same 15 minute window.
The numbers I posted in post 6 would show 3 in the quantity column.

 

[wwhitney]

The question still remains over what time period the "demand data" that 220.87(1) calls for should be averaged. I do not believe that it would be appropriate or expected to use one sample per cycle (60 per second), as that sampling rate would capture all sorts of transients, such as motor start ups, etc.

What sampling rate does a POCO provide when it makes demand data available, and what sampling rate do demand-based tariffs use? What sampling rate would the smart meter mentioned in post #6 provide if the "15 minute demand" register was activated? Presumably a sampling rate higher than a 15 minute average, as that data is already provided by the kWh register.

Since 220.87(1) Exception refers to a 15 minute average power, it's still a reasonable presumption that a 15 minute average power would suffice for 220.87(1). A related question is what part of the exception is exceptional. If the exception is only with respect to length of time for which you need demand data (1 month vs 1 year), then using the same sampling period in each case would be reasonable.

Cheers, Wayne

 

[electrofelon]

The 36 kW load for 5 minutes is a 12kW demand for the 15 minute demand window, but only 3 kWH of usage for that same 15 minute window.
The numbers I posted in post 6 would show 3 in the quantity column.

But it's the same thing isn't it You have to multiple the 3KWH times 4 because that is KWH, not "kilowatt-15 minutes".

 

[charlie b]

But how is that different than 15-minute demand? Demand is an average power during that 15 minutes, not a peak. I am not seeing any difference, what am I missing?

Perhaps we are both missing something. I agree with you that it is not the instantaneous maximum. But it is not the average either (or rather, not quite the average). Looking at the wording again (i.e., applying Charlies’s Rule), I now see that it says to use the “highest average kilowatts reached and maintained (my emphasis) for a 15-minute interval.” Putting the word “highest” next to the word “average” tends to muddy the waters. Here is one way to interpret that phrase:

Suppose the load for a given 15-minute interval was as follows: 3 KW for 3 minutes, 4 KW for 3 minutes, 5 KW for 3 minutes, 6 KW for 3 minutes, and 7 KW for the last 3 minutes. That is an average of 5 KW. But the load was not sitting at 5 KW for the entire 15-minute period. Nor was it sitting at 4 KW, 6 KW, or 7 KW. The only load value that was (now quoting from the exception) “maintained for a 15-minute interval” was 3 KW. What I mean is that the load was 3KW or more for the entire 15 minutes, and it was not 4, 5, 6, or 7 KW for the entire 15 minutes.

This reasoning brings to my mind the ancient riddle that asks how many months have 28 days. The correct answer is all of them. Some don’t stop at 28, but they all make it to 28 days before the next month begins.

What, then, should be assigned as the load for the 15-minute loading cycle I describe above? Using a strictly literal interpretation of the words, my answer would be 3 KW. But I don’t think that is the intent. As a design engineer, I would be inclined to go with 7 KW in the interests of conservatism. However, if I were reviewing someone’s calculation and noticed that they used 5KW, I don’t think I would cite that as a design error.

 

[don_resqcapt19]

Perhaps we are both missing something. I agree with you that it is not the instantaneous maximum. But it is not the average either (or rather, not quite the average). Looking at the wording again (i.e., applying Charlies’s Rule), I now see that it says to use the “highest average kilowatts reached and maintained (my emphasis) for a 15-minute interval.” Putting the word “highest” next to the word “average” tends to muddy the waters. Here is one way to interpret that phrase:

Suppose the load for a given 15-minute interval was as follows: 3 KW for 3 minutes, 4 KW for 3 minutes, 5 KW for 3 minutes, 6 KW for 3 minutes, and 7 KW for the last 3 minutes. That is an average of 5 KW. But the load was not sitting at 5 KW for the entire 15-minute period. Nor was it sitting at 4 KW, 6 KW, or 7 KW. The only load value that was (now quoting from the exception) “maintained for a 15-minute interval” was 3 KW. What I mean is that the load was 3KW or more for the entire 15 minutes, and it was not 4, 5, 6, or 7 KW for the entire 15 minutes.

This reasoning brings to my mind the ancient riddle that asks how many months have 28 days. The correct answer is all of them. Some don’t stop at 28, but they all make it to 28 days before the next month begins.

What, then, should be assigned as the load for the 15-minute loading cycle I describe above? Using a strictly literal interpretation of the words, my answer would be 3 KW. But I don’t think that is the intent. As a design engineer, I would be inclined to go with 7 KW in the interests of conservatism. However, if I were reviewing someone’s calculation and noticed that they used 5KW, I don’t think I would cite that as a design error.

It is my understanding that the demand register in the utility meter would give you 5 KW for that load and show 1.25 kWH of usage. Where this is used, I don't believe they are doing any calculations, but are getting the data from the utility, or in the case of the 30 day exception, from equipment temporarily installed that works the same way as the utility demand meter.

 

[wwhitney]

It is my understanding that the demand register in the utility meter would give you 5 KW for that load and show 1.25 kWH of usage.

If that's correct, and if that's the demand data that is used for 220.87(1), then the answer to the OP is yes.

Cheers, Wayne

 

[wwhitney]

I now see that it says to use the “highest average kilowatts reached and maintained (my emphasis) for a 15-minute interval.” Putting the word “highest” next to the word “average” tends to muddy the waters. Here is one way to interpret that phrase:

I think you are misinterpreting the word "highest". It is the same as the word "maximum" in "maximum demand". I.e. the sentence tells you to take create one average number for each 15 minute interval, for at least 30 days (so at least 4 * 24* 30 numbers), and take the highest of those numbers.

Cheers, Wayne

 

[tortuga]

I'll have to check what our data logger does, I have not used it in a while, I think it measures current and voltage of each phase and records one sample per second for 30 days.

 

[ron]

I think the language in 220.87 is chosen (and possibly poorly) to allow use of the 15 minute measurement made by the utility