With the increasing integration of the European electricity market and renewable energy sources, the effective management of cross-zonal transmission capacities has become ever more important. Electricity flows between market areas are constrained by network capacities, where congestion results in price differences between zones and consequently generates congestion income. For the Slovenian transmission and distribution system operator ELES d. o. o., this income represents an important source of financing for the development of the power system as well as for the calculation of network tariffs in accordance with European legislation.
This master’s thesis focuses on the analysis and forecasting of congestion revenues arising from the allocation of cross-zonal transmission capacities in the Republic of Slovenia, both on an annual and within-year basis. The purpose of the thesis is to develop a methodological framework and forecasting tool based on the analysis of network conditions, market structures, and price dynamics across countries within the European power system. Three main correlations were identified, upon which three forecasting models for different types of revenues were developed. The model for forecasting revenues from annual long-term transmission rights is based on historical positive electricity price spreads and available capacities between neighboring zones. The model for forecasting revenues from monthly long-term transmission rights relies on the average monthly revenues of Slovenia over the previous two years. The model for forecasting revenues from the day-ahead market coupling, on the other hand, is built on expected electricity price differences between Hungary, Germany, and France, in relation to past congestion revenues in Slovenia. The expected price spreads were derived from historical price differentials and concluded trades in futures products.
The forecasting results demonstrate a good alignment with realized revenues in previous periods. The smallest errors occur in the forecasts of annual revenues, followed by monthly revenues, while the largest deviations appear in the forecasts of day-ahead market coupling revenues, as these are directly dependent on the current network conditions. With the possibility of intra-year updates of forecasts using realized data and the preparation of revised projections until year-end, forecasting errors can be further reduced. For the purpose of improving forecast accuracy, it would be necessary to incorporate a more comprehensive fundamental analysis, since the current models primarily rely on historical trends and averages, while market fundamentals are only indirectly considered through futures-based forecasts.
|
