Dynamic safety thresholds as a measure to mitigate short-term load forecasting (STLF) errors for distribution system flexibility services

Short-term load forecasting (STLF) includes a level of uncertainty that complicates STLF applications in distribution management systems. From the distribution system operator's (DSO) perspective, forecast errors can lead to sub-optimal network control decisions with adverse impacts on asset safety or economic objectives. This paper examines MAE- and RMSE- optimized STLF and generation-dependent dynamic safety thresholds for peak load forecasts. It assesses their accuracy and effectiveness in handling forecast errors for substation congestion management in distribution networks.