Sensitivity analysis on flow-based capacity calculation

Flow-Based Market Coupling (FBMC) is being adopted across the Nordic electricity market, following its earlier implementation in other European markets, to allocate cross-zonal transmission capacity more efficiently than the previously used Net Transfer Capacity (NTC) method.Flow-Based (FB) capacity calculation, an integral part of the FBMC process, relies on several uncertain inputs, including load and renewable generation forecasts, Generation Shift Key (GSK) strategies, and topology changes. However, no systematic ranking of the inputs that most strongly influence the FB domain has previously been reported for the Nordic electricity market.

Without such a ranking, Transmission System Operators (TSOs) cannot effectively prioritize improvements in forecasting accuracy, modeling efforts, or operational practices. This thesis develops a method to identify the most influential uncertainties by applying a global sensitivity analysis to the FB capacity calculation. A two-stage approach (combining a screening and a ranking stage) was used to quantify how individual load, wind generators, GSK strategies, and topology errors affect the Remaining Available Margin (RAM) and Power Transfer Distribution Factors (PTDF) across all monitored network elements. The Morris method was used for initial screening to eliminate less important inputs, and the Sobol’ variance method was then applied to establish the final ranking.

The results indicate that topology changes/error and non-slack GSK inputs have the greatest impact on FB domain outputs, with topology error emerging as the most influential factor for both the RAM and Power Transfer Distribution Factor (PTDF) rankings. Load forecasts contribute substantially to RAM variance but have a negligible effect on PTDF variance. It also shows that the load forecast for a specific zone (i.e., the zone that could drive the flow in the transfer corridor) will be more important than the others. Wind generation forecasts have a moderate impact on RAM and a relatively minor effect on PTDF variance, mainly through their GSK participation. The methodology and ranking framework developed in this thesis provide a systematic approach for identifying key areas for improvement, supporting TSOs in enhancing FB capacity calculation processes, especially in the Nordic electricity market.