A theoretical and empirical analysis of district heating cost in Denmark

Policy-makers in Denmark have expressed a willingness to use benchmarks as a means to determine how much each district heating firm should reduce its cost. A reliable cost model should be based on a theoretically consistent specification, but no such model exists for vertically integrated district heating (DH) utilities. The first purpose of this paper is to address that gap, focusing on heat-only utilities. The second purpose is to use this specification to estimate its parameters using data from Denmark. The estimated parameters reveal that the Danish district heating utilities are subject to substantial economies of density, i.e. costs go down as the ‘network length’-to-‘heat production’ ratio decreases. We calculate two different inefficiency values for each utility: one conservative, where the utility-specific effects are treated as natural cost variation, and one liberal, where the same effects are treated as inefficiency. Those values can be interpreted as upper and lower bounds of the true inefficiency. Of the 634 observations included in the estimation, 10 have inefficiency values below 0.9 under both the conservative and the liberal assumptions.