Wind speed variability and portfolio effect – A case study in the Brazilian market

Abstract

Abstract The profitability of a wind power project is evaluated during its development phase through a complete site assessment. At this stage, the LCoE (Levelized Cost of Energy) is determined considering many variables, such as wind turbine model, project layout, energy production, CAPEX (Capital Expenditure), OPEX (Operational Expenditure) and financial costs. However, in recent years, the seasonality and variability of the wind farms energy production have been gaining importance in this process in some markets, due to a migration of the new wind power projects from the regulated market to the free market. In the free market, the PPAs (Power Purchase Agreements) have short-term balance payments, different from the regulated market. In Brazil, payments in the regulated market are performed monthly based on the long-term AEP (Annual Energy Production) expected values, with annual and quadrennial balance payments. This paper focuses on an extensive case study of Brazil, where the energy spot prices vary constantly between 10 USD/MWh and 100 USD/MWh, exposing the wind farms owners to very relevant financial risk when trading energy based on long-term bilateral contracts. The assumed risks can become critical in cases of periods with extreme meteorological anomalies (wind speed variations below the respective long-term expected values), especially in situations where a simultaneous drought period occurs, eventually raising the market spot price to its ceiling value. A SPE (Specific Purpose Entity) account must be properly dimensioned and maintained, similar to a working capital, in order to the wind farm to be able to operate through these low wind speed periods without any capital call to the wind farm’s controllers. This represents additional financial costs in the form of trapped cash. This study analyses the wind speed variability in NE-Brazil, where most of the wind farms in the country are concentrated, and analyses how the portfolio effect contributes to the reduction of this variability. This has been performed making use of the data provided by 4 met masts that surround the NE Brazilian territory (with distances between them ranging from 459 km to 724 km) and combining the obtained values to theoretical turbine power curves of different MW platforms (2, 3, 4 and 6 MW). The total period of available measurement sums more than 29 years of data. The results showed that the variability was higher in periods of lower average wind speed and that the analyzed location with the highest variability was the one close to the coast and with higher altitude. Additionally, the combinations that provided the best portfolio effect were those that included the location in the SW, furthest from the coast and with higher altitude. The different turbine platforms did not present a relevant difference in terms of resulting variability on the energy production nor portfolio effect, but when considering the cumulative impact of the meteorological anomalies (larger periods of wind speeds above or below the long-term expected values), the larger the wind turbine, the larger was the resulting exposure.