An empirical procedure for the temporal aggregation of electric utility marginal energy costs

An empirical procedure is set forth for aggregating hourly marginal energy costs into seasonal and time-of-day subperiods. It is shown that cluster analysis can be successfully applied to hourly marginal energy costs to derive seasonal and time-of-day aggregates. The procedure will systematically yield a set of aggregates that are the most distinct and homogeneous possible. The use of the aggregates as guidelines should enhance the design of economically efficient rates and the rigorous evaluation of demand-side management programs. While the results are highly utility specific, the methodology can be generalized across utilities and production cost calculation methodologies. Among the more interesting findings of the application for the City of Austin, Texas, was the indication that greater accuracy was likely to be forthcoming from a third season than from a third time-of-day period.<
>