Econometric Estimation of a Cost Function of the Power Distribution Grid

Abstract

Most economic analyses on efficient pricing for the electricity grid use strong assumptions on the cost function; often assuming that network costs are fixed. However, estimated tariff consequences on the network and users still remain highly dependent on initial hypotheses; even if sensitivity studies are carried out more and more frequently.Inversely, most investment decisions by Distribution System Operators (DSO) are performed on a case-by-case basis; due to local planning constraints such as:- Diversity of local situations (topography, in-place technology, local requirements)- Incremental putty-clay character of grid investments (historical path dependence)- Uncertainties on demand (climatic variation, changes in electricity demand, etc.)In this paper, we present a comprehensive econometric analysis of the total stock of network equipment in France - by local area - in relation to local consumption patterns and to the geographical spread of users. The issue is to identify costs associated with the different services provided by the distribution grid: (1) network access, (2) subscribed capacity (kW) and (3) energy volume (kWh). By explaining long run networks costs with customers demand for network services, we provide a new and robust foundation for future work on distribution tariffs.