Philip
Hi!
Let us agree that we disagree on the items discussed below.
Stu
Your comment about the importance of field cancellation makes sense to me. However, doesn't the fact that the parallel (currentee sharing) conductors for each phase (whether 2 or 4 or more) would be in sepatate 4-conductor (3 phases + neutral) cables... in separate conduits... increase the likelihood that the conductors sharing the current of each single phase would be of unequal lengths?
Isn't the reason for the de-rating: the unequal current sharing effect that I identified in my earlier posts?
Is it purely academic for the case of 4 cables (each with 4 conductors: 1 per phase + neutral)... sharing the current that you mentioned above?
Stu
The first question is one of those you can't really answer. It all depends on WHO is installing the feeders. If you have a cracker-jack ole time electrician who takes pride in his work, then its very likely the install will be neat and have beautifully made conduit bends, etc., and the wires/conductors will be very close in length. If they are longer or shorter by a few feet due to conduit bends, it really won't make any significant difference in terms of voltage drop or other criteria.
Additionally, an install with the length issues you suggest (eg. feeder 10% shorter) would very likely exhibit other "heads up" issues with respect to the overall quality of the job. The electrician would surely leave work in place with other significant issues. Bottom line: you're discussing means and methods.
The second issue of sharing the current, OTOH, is misleading because the other phases are lagging behind (assuming we're discussing "A" phase) at either 120 or 240 degrees at some non-peak value. The peak of "A" phase at 100A for instance will mean that the other two phases are not at peak and at some other value.
In any event, differences in feeder conductor lengths in multi-wire bundles won't matter that much as the resistances are extremely small in the first place.