Broadening the Scope of Analysis for Peer-to-Peer Local Energy Markets to Improve Design Evaluations: An Agent-Based Simulation Approach

Local energy markets (LEM’s) have emerged as an important factor in the development of energy systems which coordinate distributed and privately-owned generation resources. LEM’s may enable energy transactions between physically or virtually connected users (consumers and prosumers); local power system load balancing, increased energy system resilience, and local socioeconomic improvement may be supported. We present an agent-based LEM simulation, modeling a prominent real-world LEM using a double-sided auction-based design (CDA), and demonstrate a novel approach to LEM design analysis. A range of fixed local-level supply-demand ratios are evaluated independently to increase results’ generalizeability, and individualized agent results are considered in the analysis. Previous results suggesting CDA may improve local outcomes compared to a baseline market are confirmed. However, potential design issues related to energy affordability for consumers and sale profitability for prosumers are also observed. Similarly, while local-level results show maximized market efficiency, agent-level results suggest considerable matching success disparities in the local marketplace, despite the use of homogeneous agents with well-tuned learning models. Further data analysis shows that agents’ day-to-day outcomes may also vary considerably under CDA. Results may provide decision support to system engineers and policy-makers, and may support LEM design and analysis work.