Page 113 - Advances in Renewable Energies and Power Technologies

86 CHAPTER 3 Forecasting of Intermittent Solar Energy Resource

system when ISRES power is integrated” [35], as “comprising variability costs and
uncertainty costs” [38], or also as “additional costs that are required in the power
system to keep customer requirement (voltage, frequency) at an acceptable reli-
ability level” [39] in the literature.
Some references are listed below which will give us an order of magnitude of
integration costs or variability costs. Three reviews on integration costs gave a qua-
sisimilar range of 0e6 V/kWh [28,35,40]. The cost of variability depends on the
technology and is estimated to be 6.16e8.47 V/MWh for solar PV, 3.85 V/MWh
for solar thermal, and about 3.08 V/MWh for wind systems [41]. For wind energy
systems, integration costs between 1.57 and 4.22 V/MWh [42,43]. From data ob-
tained by independent systems operators, the integration cost for wind generators
were found in the range of 0.34e6.46 V/MWh [44]. The subhourly variability
cost for 20 wind plants was 5.93 V 0.86 V per MWh in 2008 and 2.81
V 0.37 V per MWh in 2009 [38].

4.2 FORECASTING AND INFLUENCES ON PRODUCTION COST
ISRES energy forecasting reduces the uncertainty of variable renewable generation.
It helps grid operators more efficiently to commit or decommit generators to accom-
modate changes in ISRES generation and react to extreme events (ISRES production
or load consumption unusually high or low, reducing too the curtailment). Forecasts
allow to reduce the amount of operating reserves needed for the system, reducing
costs of balancing the system. The forecasting error is a significant parameter in
the integration costs [45] and the lack of a good forecasting implies to use larger en-
ergy reserves, which cannot be used for other utilizations [46].
With such forecasts, grid operators can schedule and operate other generating ca-
pacity efficiently, reducing fuel consumption, operation and maintenance costs, and
gas emissions as compared with simply letting variable generation “show up” [47].
Here are a few remarkable examples that already demonstrate the cost
improvement:
• one percent point improvement has produced a profit of two million Danish
crowns [48].
• six million US$ savings 1 one year as a result of forecasting [49].
• Xcel Energy in reducing its mean average errors in forecasting from 15.7% to
12.2% saved 2.5 M$ [50].
• For GE [51]: the use of production forecasts reduced operating costs by up to
14%, or 5 billion $/year, that is, a reduction of operating cost of 12e20 $/MWh.
An improvement of the forecasting reliability has sometimes a significant influ-
ence on the integration cost:
• a 1% Mean Absolute Error (MAE) improvement in a 6-h-ahead forecast induces
a reduction of 972 k$ for 6 months (0.05% of the total system cost) and a
decrease of wind curtailments by about 35 GWh [52].