Allocation of losses in distribution systems with embedded generation

Deficiencies in present-day loss allocation practices are demonstrated using as an example the substitution method applied in England and Wales to evaluate the impact of embedded generation (EG) on losses. Two new loss allocation schemes are proposed; one based on the allocation of marginal losses and the other on the allocation of total losses. Loss allocation factors computed from the proposed schemes are specific to location and time of use. This is especially important for EG whose impact on losses varies in time and space. A notable feature of the proposed loss allocation coefficients is that they can be positive or negative and therefore can recognise the presence of counter-flows such as those due to the presence of EG. Application of the proposed loss allocation schemes is illustrated on a real network based 265-node generic distribution system model incorporating EG.