Measuring Performance of Long-Term Power Generating Portfolios - Green Energy and Efficiency: An Economic Perspective - page 403

Power generation (yearly average)

160

140

120

Coal
Nat. Gas
Nuclear
Wind
Hydro
Pmp. Sto.

100

80

60

40

20

0

GG SP AG CB

Scenario

Fig. 6 Yearly average production by power technology under each scenario

3.4 Power Generation

Investments in power generation face a broad set of risks which affect competing
technologies differently. The model solves for the generation level of several
technologies and the amount of power served in each period. Hence it is possible to
compute the cumulative power produced, and also a number of statistics of the
underlying distribution. Figure 6 displays the role played by each technology on
average under each scenario.
Figure 6 suggests that the AG portfolio delivers the most even levels of power
generation in terms of the major technologies. CB has the most uneven portfolio
from this viewpoint. Other renewables (hydro, biomass,

… ) and non-renewables

(pumped storage, oil) play a minor role in any case.
Combined cycle gas turbines are set to become the major producers in the SP
generating portfolio (less so in the GG portfolio). This is consistent with the relatively
low development of renewable and low-carbon energy and the delay in meeting the
environmental target. However, this situation is in sharp contrast with that in the CB
portfolio. Indeed, it is here where gas-based generation reaches its minimum. Instead,
nuclear stations appear as the major providers in the CB scenario.
We can relate these production levels to their respective capacities installed. This
sheds light on the effective load factor of each technology which in turn affects their
pro

tability. 11 Figures 7 , 8 , 9 and 10 show the results under each scenario.
Under the three future energy scenarios coal has a higher share in power gen-
eration than in capacity installed; however both shares are almost equal in the CB

11 In models where optimal dispatch takes place on an hourly basis the underlying model is able
to determine the effective number of operating hours ( ENOH ). The load factor equals ENOH /
8,760. For instance the model in Delarue et al. [ 12 ] determines technology speci c load factors by
optimization. In our case, such a direct calculation cannot be made. Instead, we can calculate the
effective electricity output from each technology in a given period and the maximum possible
output in that period. Dividing the former by the latter we could get an indirect measure of
technology speci c load factors similarly by optimization.