Sizing service off electrical bill

[SparkyEst.]

I am going to ask quick question here. Currently I have a customer who used to run their building with a boiler system, they demoed over half of that building to leave a gymnasium and open library area. Now they are looking to move the library into an auditorium, but have to demo the whole space, which includes a 1600A 208v 3 phase service. They would like to move the service from the location to a new location, which is feasible 100% with some long days ahead during a change over. I do question though that the 1600A service is a bit overkill as they have simplified their system to mostly gas RTU's and minimal loads outside of that. They provided me a copy of their electrical bill for the meter, to see what the actual usage is for the past year in KWh, but I'm a little unsure of how to calculate that out to actual usage per day. The goal would be to get the into a 1200A I-line to eliminate stand up switch boards which could save them a "couple" thousand if not ten thousand or more. Can anyone provide input on how to calculate a service size off the existing loads presented by the utility paper work? Does this make sense?

 

[d0nut]

You can't use kWh to size a service. You need the actual kW demand from the utility company.

 

[SparkyEst.]

You can't use kWh to size a service. You need the actual kW demand from the utility company.

I have the report that shows 51.2kW for the month with the most usage....

 

[SparkyEst.]

Which is why I am questioning such a large service, because everything I am getting is around 180A at 208v 3 phase. 1600A was an original install with over 50% of that installation eliminated. Thanks,

 

[jaggedben]

Yes if you've got data for maximum demand in KW over one year, you can use that. Don't have my code book with me but I think 220.84 or thereabouts.

 

[Hv&Lv]

Agreed. Pull a years worth of bills or call the POCO and ask for the yearly max demand
220.87

 

[curt swartz]

I don't think the OP realizes the difference between Demand and KWH's. Trying to size from KWH's is not possible.

 

[electrofelon]

I have the report that shows 51.2kW for the month with the most usage....

That might be it, you just want to be sure that is the maximum demand for the year. Demand is usually recorded in 15 minute intervals and is the average power used during that interval. I have seen it referred to on a bill by several different names such as demand, kw demand, kw usage, or just KW.

 

[Hv&Lv]

That might be it, you just want to be sure that is the maximum demand for the year. Demand is usually recorded in 15 minute intervals and is the average power used during that interval. I have seen it referred to on a bill by several different names such as demand, kw demand, kw usage, or just KW.

its the Maximum...
The mechanical meters used to have drag hands on them that always showed maximum for whatever period until we manually reset them. Generally monthly when the meters were read.

 

[electrofelon]

its the Maximum...
The mechanical meters used to have drag hands on them that always showed maximum for whatever period until we manually reset them. Generally monthly when the meters were read.

I guess it could vary by utility, but I believe for most utilities 15 minute demand is indeed the average over that time period, not the absolute peak.

 

[electrofelon]

I guess it could vary by utility, but I believe for most utilities 15 minute demand is indeed the average over that time period, not the absolute peak.

Just a further thought: If absolute demand was used, there wouldn't be any need for the 15 minute period, it would just be the peak usage during that billing cycle (unless POCO wants load profile data for their planning, or there are different demand charges for different time periods).

 

[Hv&Lv]

I guess it could vary by utility, but I believe for most utilities 15 minute demand is indeed the average over that time period, not the absolute peak.

Correct..

I didn’t read “during that period” and jumped the gun

Had a discussion this week trying to explain demand metering to a customer and they were thinking it was the average of the month...

sometimes trying to explain area under the curve can be hard to do for some customers..

 

[electrofelon]

Correct..

I didn’t read “during that period” and jumped the gun

Had a discussion this week trying to explain demand metering to a customer and they were thinking it was the average of the month...

sometimes trying to explain area under the curve can be hard to do for some customers..

So peak power for a given 15 minute intervals would be higher than the 15 minute demand (unless power use was constant). Agreed?

It seems that mechanical meters would not have the ability to record a 15 minute demand, just peak.

 

[Hv&Lv]

So peak power for a given 15 minute intervals would be higher than the 15 minute demand (unless power use was constant). Agreed?

It seems that mechanical meters would not have the ability to record a 15 minute demand, just peak.

Agreed.. again, jumped the gun..

the mechanical meters used a thermal coil on the pointer. As the current increased the coil heated, and the red pointer moved up, dragging the black pointer with it.
it wasn’t instantaneous movement. So a sudden spike in current for say 30 seconds wouldn’t do much to the red pointer as much as the high current for say 10-12 minutes...

 

[Open Neutral]

I was surprised when PSE said their newest "smart" meter could not tell 'demand' and they'd need a commercial account to install a demand meter.
I gotta wonder about the software designers on the project.

 

[GoldDigger]

I was surprised when PSE said their newest "smart" meter could not tell 'demand' and they'd need a commercial account to install a demand meter.
I gotta wonder about the software designers on the project.

Not surprising to me, as no utilities that I am aware of care a bit about the demand numbers except for commercial services where there is a demand surcharge. The smart meter manufacturers probably included the feature initially and then removed it (via software change) to allow them to upsell to a more costly model.

 

[Open Neutral]

But we were sizing the service needed; demand would have been useful data for both sides.

 

[Carultch]

They provided me a copy of their electrical bill for the meter, to see what the actual usage is for the past year in KWh, but I'm a little unsure of how to calculate that out to actual usage per day. The goal would be to get the into a 1200A I-line to eliminate stand up switch boards which could save them a "couple" thousand if not ten thousand or more. Can anyone provide input on how to calculate a service size off the existing loads presented by the utility paper work? Does this make sense?

Suppose you have a trip where you'll need to drive 3 hours at 30 mph on back roads, and 1 hour at 60 mph on the highway. How far did you drive? 150 miles. What was your average speed? 37.5 mph.

Consider a very primitive car that could only drive at 50 mph, would it be sufficient to drive this trip? No. You would exceed its rated speed, when you drive that hour on the highway. You need a car that is rated to drive the peak speed you'll experience. But you wouldn't be aware that your peak speed is 60 mph, unless you had more information than "the average speed will be 37.5 mph".

This is precisely the issue with sizing a service off of daily, monthly, or annual kilowatt-hours. It doesn't tell you the critical part of the full picture. If the kilowatts drawn from the utility peaks during the day, we need to anticipate how high the peak will be. Even if the load consumed during the average hour is significantly less.

Demand is defined as the maximum kilowatts consumed during any given 15 minute interval, which is a much more accurate representation for the peak "speed" of drawing kilowatt-hours of energy. Power (kW) is to energy (kW-hrs) as speed is to distance. You might think we'd be interested in the instantaneous peak power, and there are plenty of loads that cause brief spikes of high power consumption to happen. However, spikes that last less than a minute are not significant enough to overload distribution equipment and wiring. A 15-minute basis for tracking full load power is a much better representation of how much ampacity your service would need, because it takes time to heat up wires, and it takes time for small amounts of overload to trip breakers. A brief 450A spike on a 400A breaker might go unnoticed, but a sustained 450A load for 15 minutes would likely trip this breaker.

 

[Fred B]

Suppose you have a trip where you'll need to drive 3 hours at 30 mph on back roads, and 1 hour at 60 mph on the highway. How far did you drive? 150 miles. What was your average speed? 37.5 mph.

Consider a very primitive car that could only drive at 50 mph, would it be sufficient to drive this trip? No. You would exceed its rated speed, when you drive that hour on the highway. You need a car that is rated to drive the peak speed you'll experience. But you wouldn't be aware that your peak speed is 60 mph, unless you had more information than "the average speed will be 37.5 mph".......

I love this illustration, really makes simple explanation, I've had customers asked me before why not size from kWh. I knew you can't, but struggled with a simple explanation of why. Even bigger issue comes from GCs who'll want to argue about it. Can I steal this illustration to use with my customers?

 

[Carultch]

I love this illustration, really makes simple explanation, I've had customers asked me before why not size from kWh. I knew you can't, but struggled with a simple explanation of why. Even bigger issue comes from GCs who'll want to argue about it. Can I steal this illustration to use with my customers?

Glad I could help with an explanation. Feel free to use it as you see fit.