NEC 220.87

A

anbm

Senior Member
Location
TX
Occupation
Designer
Electrical panel A feeds both panels B and C.
30 days metering record peak amp. reading on panel B = 100A.
30 days metering record peak amp. reading on panel C = 150A.

Without metering panel A, the estimate max. panel A peak reading will be:

a) (100 + 150) x 125% = 312.5A

b) (150 x 125% + 100) = 287.5A

c) No answer until panel is metered?

Sounds like c is the correct answer? lol
 
The peak amp method can be very restrictive. Most accept the average peak over a 15 minute window. If your peak measurement is short, it will over state your load.
 
anbm said:
Electrical panel A feeds both panels B and C.
30 days metering record peak amp. reading on panel B = 100A.
30 days metering record peak amp. reading on panel C = 150A.

Without metering panel A, the estimate max. panel A peak reading will be:

a) (100 + 150) x 125% = 312.5A

b) (150 x 125% + 100) = 287.5A

c) No answer until panel is metered?

Sounds like c is the correct answer? lol
Click to expand...
I'd say 'A'
I'd also run one of the other calcs.
 
anbm said:
Electrical panel A feeds both panels B and C.
30 days metering record peak amp. reading on panel B = 100A.
30 days metering record peak amp. reading on panel C = 150A.

Without metering panel A, the estimate max. panel A peak reading will be:

a) (100 + 150) x 125% = 312.5A

b) (150 x 125% + 100) = 287.5A

c) No answer until panel is metered?

Sounds like c is the correct answer? lol
Click to expand...


A or C in my opinion. If panel A only feeds panels B & C then that meets the code.

Metering Panel A would most likely result in a lower overall number.
 
anbm said:
Electrical panel A feeds both panels B and C.
30 days metering record peak amp. reading on panel B = 100A.
30 days metering record peak amp. reading on panel C = 150A.

Without metering panel A, the estimate max. panel A peak reading will be:
Click to expand...
If you mean your best guess of what peak amp reading you'd get on panel A if you actually metered it for 30 days, the answer is "no more than 250A." If panel B's peak and panel C's peak happened to occur during the same recording interval (and on the same leg), then you'd see 250A. If not, you'd see a lower value.

But I'm not commenting on how you could use NEC 220.87 either with or without the additional reading.

Cheers, Wayne

P.S. Edit: if the recording meters were in place for the same 30 day period and synchronized, and they recorded enough information (currents and phase angles for each conductor for each recording interval), then you could reconstruct exactly what the recording results on panel A would have been. Even if you don't have the phase angles, you can get an upper bound by assuming the panel B and panel C currents are in phase.
 
Last edited:
P.P.S. The synchronization issue is a confounder, and actually means that two unsynchronized recorders on the same conductors will give you different peak readings. E.g. if for 5 minutes intervals you'd see 0A, 100A, 100A, 0A, then a 10 minute interval reader might see anywhere between 50A and 100A.

So that effect adds some uncertainty to my "not more than 250A" answer.

Cheers, Wayne
 
jim dungar said:
Just like I would assume the demand meters employ a sliding window with a 15 minute demand.
Click to expand...
You mean something like recording values every minute, and then the output is the highest sum you get from adding 15 consecutive minutes? Or even values every second, and looking at sums of 900 consecutive values.

That's good to know, and would basically eliminate my concern from post #6 (the effect of non-synchronization is very reduced). Do you know what the underlying recording interval typically is?

Thanks,
Wayne
 
Elect117 said:
A or C in my opinion. If panel A only feeds panels B & C then that meets the code.

Metering Panel A would most likely result in a lower overall number.
Click to expand...
Agree, however, under a) but when Panel B reaches max. panel C may not and vice versa... but it is worst case scenario.
 
When I was a helper I got paid to go on name plate Easter egg hunts thru a 30 unit complex. We had headlamps, rubber gloves, wipes and a few spray cleaners.
The reason was the loadcalc had to be done and the cost difference between the one where they go off utility peak and a 30 unit 'optional' was worth paying two helpers for a few days to bug 30 tenants and wipe off their nameplates. They say it saved something like 50 grand.
I wont pretend to know how or why that made any sense.
 
anbm said:
Agree, however, under a) but when Panel B reaches max. panel C may not and vice versa... but it is worst case scenario.
Click to expand...

If option A plus any additional load is less than the board then just use it as the load calc and get things moving.

If it puts you over, then go with option C and install the meter on panel A.

Or use one of the other calculations methods in 220.
 
You must log in or register to reply here.
Top