Distribution Transformer Diversified Loading

Sizing a distribution transformer on connected load oversizes it enormously, and the term that closes the gap is the coincidence factor. Diversity strengthens as the group grows: two houses on one transformer coincide badly, thirty houses hardly coincide at all, so the factor falls from near 1.0 at a single customer toward an asymptote somewhere around 0.4 to 0.6 for a large residential group. Each utility carries its own curve, derived from its own metered data, and NO CURVE IS SHIPPED HERE for the same reason no clearance table is: a factor that is right for one system's housing stock and climate is wrong for another's, and a shipped curve would be believed. The two names for this quantity are swapped constantly and both are reported: the coincidence factor is diversified over connected and is at most one, the classic diversity factor is its reciprocal and is at least one. Distribution transformers are also allowed to run past nameplate for a few hours, because their thermal time constant is measured in hours -- a pot that peaks above 100% for two evening hours and sits at 40% overnight can have an entirely ordinary loss of life. That is why loading is reported against both the continuous rating and an entered short-time rating, and why load factor is worth carrying beside peak: peak alone says nothing about how long the peak lasts. What this does not do is decide whether a given overload is acceptable. Loss of life needs a full load cycle against an ambient profile through IEEE C57.91, and a short-time rating used without that study is a guess with a number on it. Electric vehicle charging and electric heat break residential diversity assumptions badly, and are exactly the case where an old factor misleads: several vehicles on one pot charge on the same timer, which is coincidence rather than diversity. One transformer, one customer class, one factor supplied by the reader. Secondary voltage drop is a separate calculation. The utility's transformer loading guide, IEEE C57.91, and the transformer manufacturer's ratings govern.

diversified demand = customer count x individual customer peak x coincidence factor; loading = diversified demand / rating; the classic diversity factor is the reciprocal of the coincidence factor; customers to the continuous rating = rating / (individual peak x coincidence factor).

The diversified-demand relation as standard distribution practice, by name, with IEEE C57.91 named for loading beyond nameplate. NO DIVERSITY CURVE IS SHIPPED: the factor is entered from the utility's own metered data. Loss of life is not computed. The utility's transformer loading guide, IEEE C57.91, and the transformer manufacturer's ratings govern.

Two multiplications and a division on a coincidence factor the reader supplies; no utility diversity curve is reproduced.

Estimate. AHJ and licensed professional govern.

Field names used by the API: customers, individual_peak_kva, diversity_factor, continuous_rating_kva, short_time_rating_kva, average_demand_kva, connected_kva, diversified_kva, continuous_loading_pct, customers_at_continuous, headroom_customers, classic_diversity_factor, load_factor

• The coincidence factor is an input, not a curve it is specific to a utility's housing stock and climatethe utility's metered load data
• Loss of life is not computed it needs a full load cycle against an ambient profileIEEE C57.91
• One customer class EV charging and electric heat break residential diversity assumptionsthe utility's transformer loading guide