Long-term planning of Low-Voltage networks using reference network Models: Slovenian use case

• The k-medoids algorithm is applied to classify LV networks into representative clusters. • The great diversity of low-voltage networks among different DSOs. • The proposed reference network models reduce computational complexity. • Scenarios following NECP are built for PV, HP and EV penetration from 2025 to 2034. • The need for targeted investments to ensure network stability and sustainability. The increasing penetration of distributed energy resources (DERs), electric vehicles (EVs), and heat pumps (HPs) presents significant challenges for low-voltage (LV) distribution networks, requiring advanced planning methodologies to ensure grid reliability and cost-effectiveness. However, existing studies primarily focus on individual network simulations, which are computationally intensive and lack scalability. Moreover, most research relies on synthetic network models rather than real-world distribution system operator (DSO) data, limiting practical applicability. This study addresses these gaps by developing a Reference Network Model (RNM) tailored to the Slovenian LV distribution system. The first objective is to establish reference radial network models based on real DSO data, enabling simulation generalization across the entire distribution network. Using k-medoids clustering, LV networks are categorized into representative groups, facilitating efficient analysis without exhaustive individual network simulations. The second objective is to develop a generalization methodology that extrapolates simulation results from reference networks to the entire LV distribution system. Unlike conventional RNM applications, this approach integrates real-world Slovenian DSO data and incorporates scenario-based reinforcement planning to address the evolving impact of DERs, EVs, and HPs. A key result is cost-benefit analysis, which evaluates investment requirements and operational savings, offering insights for policymakers and DSOs to optimize network planning. Simulation results indicate that most required reinforcements will focus on LV line upgrades, particularly in regions with long feeders and high demand growth. The findings demonstrate that the proposed methodology significantly reduces computational burdens while maintaining high accuracy in predicting network reinforcement needs, making it a scalable and practical tool for long-term distribution system planning.