Market demand: To construct fluctuating market demand, we depart from hourly mar- ket prices (from the European Energy Exchange (EEX)25) and hourly quantities consumed (from the Union for the Co-ordination of Transmission of Electricity (UCTE)26) for the year 2006. We chose the value of b in line with other studies on energy markets. Most studies that estimate demand for electricity27 find short run elasticities between 0.1 and 0.5 and long run elasticities between 0.3 and 0.7.28 The relevant range of prices is around P = 100 €/MWh and corresponding consumption is approximately Q = 50 GW. In our empirical analysis we thus use the slope b = 0.0055 which corresponds to an elasticity of around 0.4.
The computed intercepts θ for each of the 8760 of the year are sorted, their frequency- distribution is reported in figure 5. In order to satisfy the smoothness required to match
the theoretical framework we fit a Weibull distribution29
The fitted distribution exhibits fatter tails than the distribution of observed intercepts. Those fatter tails could be motivated by the uncertainty about levels of demand at the time of investment additionally to the fluctuation of demand.
Fuel-prices at plant €/MWh
Cost of CO2-Cert. €/MWh
c Production Cost in €/MWh
Overnight Investment in €/kW
Annual fixed cost in €/kWa
Free CO2 allocation in €/kWa
k Investment Cost in €/kWa
Table 1: Cost of Production c and
Cost of Investment k.
25 26 27 28
See www.EEX.com See www.UCTE.org See, for example, Lijsen (2006) for an overview of recent contributions on that issue. E.g. Beenstock et al. (1999), Bjorner and Jensen (2002), Filippini Pachuari (2002), Booinekamp (2007),
and many others. 29The Weibull distribution is given by F (θ) = 1
( θ β )
and it’s density by f(θ) = β
( θ β )
= 2 condition (b) of lemma 5 is satisfied.