Estimating existing capacity of a panel and upstream panels

Status
Not open for further replies.
W

William Teller

New User
Location
Texas
Occupation
Electrical Engineer
I occasionally design renovations to existing facilities. Sometimes I have accurate as-built drawings that show the existing load of a panel and sometimes not. If you count 80 percent of the capacity of each branch breaker that is connected, you often come up with an existing calculated load that is far greater than the main breaker of the panel. Now if you have a main panel and disconnect that is rated at 1000A and have 4 sub distribution panels that are 600A and the load is more or less evenly distributed, then I would think that you could divide the peak demand load (in this case 425A at the service.) by 4 to estimate how much load is on each sub distribution panel. But, I was told you can not do this by a senior electrical engineer. I was told that if you add load to a panel, even if it is say 13 amps on an unused 20A 120A/1P breaker you would have to assume: 1) all of the peak demand load on the service is on the distribution panel that feeds that panel and that most of that load is on the branch panel you are updating. Or, 2) you have do a load analysis of that upstream distribution panel and every panel that is fed from it. Does that seem reasonable to you.

FYI, doing a 30 day load measurement is often not practical and the same senior engineer says that you have to do a load measurement of the panel and upstream distribution panel for a whole year to have meaningful results.

I appreciate any wisdom on this mater.
 
Looking at or adding up breaker ratings is meaningless. Dividing a demand figure evenly or proportionately among downstream panels is also meaningless.

That said, many panels will be lightly loaded, and I personally consider myself experienced enough to often add load without doing a load study or load calcs.

IMO the NEC is far too concerned with overloads. In the rare event it actually happens, I don't see what the big deal is.
 
I would agree with the senior. If you want as reasonably true answer you would need a load study on each panel for a length of time that includes seasonal changes, As others have stated, if all wiring and OCP is properly sized, if you misjudge you end up with a tripped breaker.
 
Before saying if a result is meaningful, you need to decide what your important metrics are. Not playing with words here, but rather saying that before you spend time doing a load calculation or measuring a demand load, you need to figure out how accurate that measurement needs to be to do the job you are about to do.

In the example the OP gave, you have a 1000A service with 4 600A sub-feeds. The measured demand load on the service is 425A. While I believe that the hunch that the demand load 'divides equally', consider what it means if you needed to add a 50A circuit to one of the panels. Assuming the worst case that _all_ of the load is on one of those 600A panels, and it still isn't overloaded, and could take the additional 50A.

The reason that the demand load doesn't divide equally is because of the way load diversity works. Each feeder might have large peak loads, significantly exceeding 425A/4, but if the different feeder peaks never overlap, the service load doesn't get too large.

If you really want an accurate measurement of load, then I agree with the senior engineer: you need a long term measurement on each feeder that you are evaluating. But in general I agree with the others here, that a really accurate load measurement is more than necessary to get the job done safely.

-Jon
 
Agree with all those above. Adding up breakers is not representative of the load on a system. In commercial applications we typically see 30%-60% (of service size) loading.

For example, different load types require different OCP ratings. Recepts is lesser than motors which is lesser than fire pumps. Not to mention, connections could be made for system redundancy and only switched during equipment failure.

Oddly enough OSHPD (CA's Office of Statewide Health Planning and Development) which often times exceed requirements outlined in the NEC, allows for 72hr load measurements. On at least one non-OSHPD project we've received city approval on using 3day load info based on OSHPD standards.

This is only practical if all concerning loads are running during that time. It would not be advised to take measurements in a facility on M-W, if most of their processes are completed Th-Su.
 
For most situations locally when you are dealing with services of 1,000 amps or more we can get a demand load and records from POCO.
 
Status
Not open for further replies.
Top