Optimising Renewable Energy Communities

Abstract

The energy transition in the Netherlands, characterised by the increase of renewable energy sources (RES) and the electrification of key sectors, has placed strain on the existing electricity grid. Challenges such as volatile electricity prices, curtailment of excess renewable energy, and grid congestion highlight the limitations of the current infrastructure, particularly in urban low-voltage (LV) networks. Here, transformers get congested due to limited capacity during peak hours. Expanding the grid is often infeasible due to high costs, long timelines, and spatial constraints. However, Renewable Energy Communities (RECs) have emerged as a promising solution to these challenges by optimising the use of the existing infrastructure within local contexts.

The challenge in improving affordability, sustainability and security is that the goals are contradicting, also referred to as the Energy Trilemma. Therefore, this thesis addresses the research question: “How can Dutch urban Renewable Energy Communities be designed and operated to enhance energy affordability, sustainability, and security?”. To answer this, a multi-objective linear programming (MO LP) model was developed using the Calliope software framework in Python. The model optimizes the design and operation of RECs across three energy trilemma dimensions: affordability, sustainability, and grid security. It incorporates solar photovoltaic (PV), battery energy storage systems (BESS), and grid interactions.

The research underscores the inherent trade-offs in balancing the energy trilemma. Affordability-driven scenarios minimize costs through extensive grid reliance, increasing emissions and transformer congestion, while sustainability- and grid-security-focused scenarios emphasize self-consumption, reducing both but incurring higher costs and curtailment. Maximizing solar PV capacity cuts CO₂ emissions but leads to substantial curtailment without sufficient storage or trading. BESS mitigate imbalances by shifting energy flows in time, yet grid dependence remains unavoidable, especially in winter when PV output is low.

There is thus no universally optimal REC design, effectiveness depends on stakeholder priorities. However, key insights hold across all scenarios: Dutch urban RECs can enhance affordability, sustainability, and security with approximately 750 kW of solar PV per 200 prosumers and 200 kW MV and LV batteries for hourly balancing. Designing RECs this way, could offer a more efficient solution for mitigating urban grid congestion than defaulting to grid expansion.

Despite its contributions, the study has limitations. The model simplifies grid interactions by focusing solely on the LV grid and transformer congestion, excluding medium-voltage (MV) and high-voltage (HV) dynamics. Behavioural feedback on market dynamics, such as the impact of widespread REC adoption on electricity prices, or the diminishing business case of batteries, is also not captured. These limitations underscore the need for future research to expand the model’s scope, address emerging technologies, and integrate multi-layered grid interactions that include feedback systems.

All findings of this thesis are open source. The model that was developed to answer the research questions can be accessed at:

https://github.com/Tomdebruin/MO-LP-Energy-Community-optimisation