Cost-optimal implementation of energy storage systems to mitigate congestion and increase self-consumption in future Dutch low-voltage networks

Abstract

Using preliminary versions of Advanced Scenario Management – Phase 2 (ASM-2) hourly supply and demand profiles on neighborhood scale for three 2030 energy transition scenarios, this study answers the research question “to what extent can energy storage systems (ESSs) serve as a cost-optimal mitigation option to address congestion and increase self-consumption in 2030?”. As ASM-2 profiles are on neighborhood scale, ESSs are evaluated on neighborhood scale as well, and only congestion in distribution transformers is taken into account. In addition to ESSs, PV-curtailment and grid reinforcement are assessed as additional mitigation options. To determine implementation and control strategies for the three mitigation options taken into account, an optimization study was performed using Mixed Linear Integer Programming in Gurobi (Python), for three different perspectives of ownership and managing ESSs: a collective of prosumers, distribution system operators and a combination of these. Results show that for the prosumer perspective it is likely that ESSs will serve as a cost-optimal mitigation option to increase self-consumption by 2030. For the DSO perspective, it is highly unlikely that ESSs will serve as a cost-optimal mitigation option to address congestion. However in the combined perspective, the potential for self-consumption increase provides reasonable possibilities to mitigate congestion ‘along the way’. The actual applicability of ESSs is however heavily dependent on ESS price developments and the further advancement of rooftop-PV. While this is taken into account in this study, it is recommended that further research adopts a monitoring approach with regard to these parameters. As DSOs are currently not explicitly allowed to own and operate ESSs, and prosumers are not allowed to own ESSs collectively on neighborhood scale, changes in laws and regulations would be needed to facilitate this. Lastly, because of (run)time constraints, the geographical scope for this study was restricted to neighborhoods in the Province of Utrecht. However, using the optimization script written for this study, a full scale version will be published as part of the ASM-2 results.