NEC 220.87 Multiple Loads

[Tom Sawyer]

Currently three generators are providing power to three different loads on the property. None of the three loads are connect together. We are replacing the generators with all PG&E. When doing a 30-day load study on each generator, do we use the peak average power over the 30 day period regardless of synchronicity or do we use the peak average power that represent the highest total power during a given synchronized time?

For example:

Gen 1 = 100 KVA (peak apparent power) at 5:15 PM
Gen 2 = 25 KVA (peak apparent power) at 2:00 PM
Gen 3 = 74 KVA (peak apparent power) at 9:15 PM

Total peak power = 194 KVA x 1.25 = 242.5 KVA for the service.

OR

Gen 1 = 100 KVA (pap) at 5:15 PM
Gen 2 = 10 KVA (ap) at 5:15 PM
Gen 3 = 14 KVA (ap) at 5:15 PM

Highest total peak power for a synchronized time = 124 KVA x 1.25 = 155 KVA for the service.

 

[rbalex]

This is a bit tough because neither example is actually “correct.” Technically, you want the maximum demand of the simultaneous sum of the three loads whenever it occurs.

The first example is extremely (possibly excessively) conservative but should be acceptable. PG&E will probably opt for a smaller service anyway.

This isn’t a “classic” 220.87 application in the first place. Assuming the maximums were accurate, I might consider [100x1.25+ 25 + 74] kVA (i.e., 224kVA total) as a better approximation. This is sort of a hybrid of 220.87 and 220.14(C), with its references to 430.22 & 430.24. Since I don’t know the actual system it’s a WAG.

 

[Tom Sawyer]

220.87 - Power logger vary by a few minutes

220.87 - Power logger vary by a few minutes

Thank you Robert,

We are using Fluke 1735's which gives min, avg, and max KVA every 15 minute. According to 220.87 the highest KVA averaged over 15 minutes in the entire 30-days (2,880 readings) would be the demand. I'm calling that peak average KVA because the Fluke 1735 also has max KVA which is not averaged and can be very high when motors are starting.

The Flukes are not exactly synchronized (simultaneous) they vary by a few minutes from each other. Using the simultaneous numbers would be an advantage but not sure if it will pass design review.

I'm an electrical engineer, but not had the above experience.

I'm confident that the design would be fine using the simultaneous summation method, but not sure if the authority would agree.

Would this situation fall into the category of "best engineering judgment" ?

Do you know of any precedence?

Any help is most appreciated.