cdcengineer
Senior Member
I ran into a situation with an existing installation today. Odd service that has definitely been modified since it was first constructed.
The service consists of a 1,200 A, 208Y/120V, 3-phase main service disconnect which is hot sequence metered - all good so far. There are (3) sets of 500 kcmil CU conductors running into a tap box where there is a parallel set of 500kcmil AL feeders serving a 600 A switchboard (SWBD). The other tapped set consists of 500 kcmil CU which serves a 400 A panelboard. All of this equipment is less than 5' away from each other.
The first red flag I noted was that the (3) sets of 500kcmil CU were run in two conduits (violating 300.5(I) & 300.20). Not all phase and neutral run in same raceway. This got us looking closer. We then realized that the (3) 500 kcmil CU were not large enough for the 1200 A main. We couldn't identify this at first because the main was not accessible.
The calculated load for the building is just over 400 Amps. I proposed a solution that we remove one set of the (3) 500 kcmil CU leaving 760 Amps. We eliminate the tap situation serving only the 600 Amp SWBD. We would then re-feed the 400 A panelboard from a spare 400 A bucket in the SWBD.
All seemed good-to-go. However, I would never specify a 1200 A main ahead of a 600 A rated bus. The loads are less than 600 A, but if this was a new installation, we would buy a 600 A main. FYI - the SWBD consists of (2) 200A/3P & (1) 400A/3P fusible buckets.
Throwing out the extraneous info in the questions, the real question boils down to whether it is allowed to serve a 600 Amp bus from a 1200 A main? I know per exhibit 230.26 the sum of the OCP can exceed the ampacity of the service conductors. I first thought that we would violate 705.12 (120% rule), but this is not applicable.
I'm wondering if anyone can massage this brain cramp? I would never try and do something like this from the get-go, but because it's existing and they're trying to leave as much existing as possible, what is the gotchas I'm not seeing? I know there would be potential for loads to exceed the 600A bus which would not be protected against via OCP, but that could potentially happen in a tap scenario if we had two 600 A SWBD tapping say (3) SETS OF 600 kcmil CU feeders served from a single 1200 A main.
Any input is greatly appreciated as always.
DC
The service consists of a 1,200 A, 208Y/120V, 3-phase main service disconnect which is hot sequence metered - all good so far. There are (3) sets of 500 kcmil CU conductors running into a tap box where there is a parallel set of 500kcmil AL feeders serving a 600 A switchboard (SWBD). The other tapped set consists of 500 kcmil CU which serves a 400 A panelboard. All of this equipment is less than 5' away from each other.
The first red flag I noted was that the (3) sets of 500kcmil CU were run in two conduits (violating 300.5(I) & 300.20). Not all phase and neutral run in same raceway. This got us looking closer. We then realized that the (3) 500 kcmil CU were not large enough for the 1200 A main. We couldn't identify this at first because the main was not accessible.
The calculated load for the building is just over 400 Amps. I proposed a solution that we remove one set of the (3) 500 kcmil CU leaving 760 Amps. We eliminate the tap situation serving only the 600 Amp SWBD. We would then re-feed the 400 A panelboard from a spare 400 A bucket in the SWBD.
All seemed good-to-go. However, I would never specify a 1200 A main ahead of a 600 A rated bus. The loads are less than 600 A, but if this was a new installation, we would buy a 600 A main. FYI - the SWBD consists of (2) 200A/3P & (1) 400A/3P fusible buckets.
Throwing out the extraneous info in the questions, the real question boils down to whether it is allowed to serve a 600 Amp bus from a 1200 A main? I know per exhibit 230.26 the sum of the OCP can exceed the ampacity of the service conductors. I first thought that we would violate 705.12 (120% rule), but this is not applicable.
I'm wondering if anyone can massage this brain cramp? I would never try and do something like this from the get-go, but because it's existing and they're trying to leave as much existing as possible, what is the gotchas I'm not seeing? I know there would be potential for loads to exceed the 600A bus which would not be protected against via OCP, but that could potentially happen in a tap scenario if we had two 600 A SWBD tapping say (3) SETS OF 600 kcmil CU feeders served from a single 1200 A main.
Any input is greatly appreciated as always.
DC