tallgirl
Senior Member
- Location
- Glendale, WI
- Occupation
- Controls Systems firmware engineer
Multiwire Branch Circuit?
- Thread starter new_ee
- Start date
- Status
- Location
- Lockport, IL
- Occupation
- Semi-Retired Electrical Engineer
The webmaster disabled hyperlinks as a temporary measure for combatting SPAM. He has put into place a process that, so far, seems to be sucessful in preventing the advertisements and worse topics that have been posted over that past month or two. He is working on getting the hyperlinks re-enabled. It should not be long before you can use them again.Smart $ said:
ramsy
NoFixNoPay Electric
- Location
- LA basin, CA
- Occupation
- Service Electrician 2020 NEC
Smart $ said:
Am I looking at a panel schedule?, if so I like it.
Code:
A B C
2.5 2.5 2.5
2.5 1.5 1.5 =
5kva 4kva 4kva
/ 277 / 277 / 277 =
18A 14A 14A
Well, not quite. Panel schedules [generally] do not have two entries per pole (re: your 2.5's and 1.5's), and approximated kva values are just that... approximated.ramsy said:
A quick way to get a more accurate value for unbalanced line-to-line loads is with this formula (computer formatted) derived from the Law of Cosines:
- sqrt(a^2 + b^2 + a*b)
Using 5000VA and 3000VA as an example...
sqrt(5^2 + 3^2 + 5*3) = 7
7kVA / 480V = 0.01458333333kA, or 14.58333333A
In following this thread, one can see that charlie's 15.6A is a bit low for A-phase conductor amperage, and your 14A is a just a tad low for B- and C-phase amperages.
ramsy
NoFixNoPay Electric
- Location
- LA basin, CA
- Occupation
- Service Electrician 2020 NEC
Smart $ said:
That looks familiar, and much simpler than vector graphs. I checked the NEC to see if I could use it without Engineering supervision. I didn't find any NEC references, but there is a NFPA 70 substitute in Appendix D, Ex. No. D5(a & b), which was calculated from 2 legs of a 3? 208Y, as VA / #legs / V? [/ PF] = I?
My previous attempt to match your vector results with a panel schedule was a similar method, but without rounding down the B & C phase are 14.44A vs 14.58A for the vector.
Whether the NEC-engineering panels' use of Ex. D5 suggests this approximation error is negligible or not --relative to vectors-- if permitted to represent unbalanced loads this way, I find that both 3? and 1? panel schedules are much simpler to build with simple division.
Using all 3 phases, I could not make sense of which 2 load variables using sqrt(a^2 + b^2 + a*b) matched which of the 3 phases. Perhaps with a 3-pole OCPD sized for the highest one it doesn't matter, but life seems easier for all L-L, L-N, un/balanced loads if conductor sizing is permitted with that panel-schedule method I did, originally derived from KingPd.
ramsy
NoFixNoPay Electric
- Location
- LA basin, CA
- Occupation
- Service Electrician 2020 NEC
charlie b said:
Does this mean “current between two phases” = “current between three phases.”
With a perfectly balanced 3? load, are Amps or kVA between any 2 legs the same as all 3 legs total?
Last edited:
tallgirl
Senior Member
- Location
- Glendale, WI
- Occupation
- Controls Systems firmware engineer
charlie b said:
But ... but ... he asked about attachments. If we can't have more than 15 messages, I'm not thinking we're going to be getting attachments back.
The [img][/img] tags still work. Here's a Multi-branch no-wire circuit for y'all --
I was going to write and attach a Java program to draw vector diagrams like what Smart used to post, but now that we can't do attachments ...
Last edited:
Well these graphics work. They are just a couple examples of Ronald Colemans work.
http://home.comcast.net/~ronaldrc/wsb/home.html
Roger
http://home.comcast.net/~ronaldrc/wsb/home.html
Roger
Afraid not.ramsy said:
That particular example simply divided the lighting load, which individually are line to neutral loads, among the two phases.
That's because the method of calculation assumes the loads to be balanced (1:1) and uses a 1.732 adjustment factor, or 0.866 per load. Being unbalanced at 5:3, the adjustment factor calculates to be 0.875 per load.
The portion of Example 5 you referred to has nothing to do with unbalanced loads. However, it does have a portion dealing with potential unbalanced loads in the service to meter bank feeder section. The procedure compensates by doubling the number of ranges, applying the appropriate demand factor, halving the demand per phase, then multiplying by three to arrive at a system VA. IMO, the NEC is saying the error is not negligible by doubling the number of ranges. The rest of the procedure is SOP.
The phase mutual to the loads. For example, if a represents load AB and b represents load AC, the result correlates to A-phase.
Yes, the circuit should have conductors and OCPD sized to the phase with the highest draw. This was my point for entering this discussion to begin with. As for the "panel schedule method", with branch circuits and feeders uprated 125% I believe you can do that safely :smile:
ramsy
NoFixNoPay Electric
- Location
- LA basin, CA
- Occupation
- Service Electrician 2020 NEC
Smart $ said:
My thanks to you & charlie for entering this discussion and developing the topic to this level. I am permitted to check the sqrt(a^2 + b^2 + a*b); and if cost-effective discrepancies exist with approximations, now I will know when to consult the required engineering supervision.
- Status