Impact of Cross-Border Flow Forecast Alignment on Flow-Based Market Coupling Efficiency

The developed model is based on a test network, retrieved from Barrios et al. It is adjusted to fit the purpose of this analysis. Transformers are not considered but all 220 kV and 380 kV lines are. The conventional power plants are provided by the dataset. The types of renewable energy capacities and their assignments to nodes is retrieved visually from Barrios et al.. The resulting network is depicted in Fig. 14. Fig. 16 provides an annotated overview of the network with an indication of all nodes and lines. Zones 1\u20133 are considered as the flow-based market coupling area. The adjacent zones represent consolidated zones, giving the possibility of trade with the FBMC area. This trade, however, is not modeled through FBMC but with static bilateral trade capacities (NTCs), which is the current way the CWE region conducts trade with all neighboring countries not part of FBMC. Weather data for renewable energy sources of France, Belgium and Germany are used for zones 1\u20133 (see below). It is important to note, however, that the generation capacities of zones 1\u20133 do not represent these countries. In terms of installed renewable energy capacities, zone 1 is dominated by conventional generation with small solar and wind capacities. Zone 2 is a solar-dominated zone and zone 3 shows large capacities of wind energy. For the case studies, a scenario with higher variable renewable energy source (vRES) is considered (Fig. 15). In zone 1, all renewable capacities are doubled, zone 2 further expands solar energy and reduced lignite capacities and zone 3 expands onshore wind as well as solar energy.22 The capacity factors for renewable energy sources are retrieved from SETIS [31] for the year 2015. Zones 1\u20133 are assigned the capacity factors of France, Belgium and Germany, respectively. The generation capacities for the remaining three zones are not provided. Here, the generation capacity structures of Italy, the Netherlands and Poland are used to mimic these zones Import\/Export 1\u20133, respectively. The capacities are obtained from ENTSO-E [14]. They are scaled down to power plant fleets with sizes similar to zone 1\u20133. The resulting generating capacities are depicted in Fig. 15. While node-specific load time series are provided for zones 1\u20133 (year 2012), the load time series for the remaining zones are created artificially. Here, the aggregated load time series of zones 1\u20133 is taken as a basis and scaled so the ratio of maximum zonal load to zonal installed capacities matched the real ratio of the countries. Concretely, the ratio of maximum load to installed capacity in the zone Import\/Export 1 matched the ratio of Italy. This has the effect, that in every zone a \u201crealistic\u201d load time series is created and the maximum load never exceeds the conventional generating capacities. Despite all attentiveness, uncertainty of input data from the external sources cannot be completely excluded. This includes the test network, originating from Barrios et al. [8], on which the derived model is based. At present, no weaknesses are known concerning this data source. Furthermore, weather data comes from the open access platform SETIS [31], consequently, the accuracy of the capacity factors depends on the quality of the dataset. Considering that the test network is hypothetical, the impact of possibly existing uncertainties on the results is low. Moreover, the installed capacities of adjacent zones are taken from ENTSO-E [14]. Since those capacities do not emerge from time series, the related uncertainties can be considered rather low.