Effects of Correlation Between Failures and Power Consumption on Customer Interruption Cost

This paper evaluates the correlation between failures and power consumption for a distribution systems operator. This is done in order to scrutinize whether the commonly used assumptions of constant failure rate and constant power consumption is reasonable to use for reliability calculations. The studied entity is energy not supplied, which is assumed to be a good estimate of how customer interruption costs are affected. Three different aspects are studied with respect to energy not supplied; seasonal variations, daily variations and repair time variations (as a function of the hour of day). The conclusion is that by using constant failure rates, repair rates and power consumption the approximation of customer costs becomes somewhat low, i.e. by 7 % for the studied case. This result indicates that the assumptions of constant failure rates, repair rates and power consumption are quite sufficient for at least the actual case study. -i.e. since this error probably is significantly smaller than other types of errors, for example customer outage costs estimates. Nevertheless, having performed these calculations the current results should be applied to further modeling of the studied network