Long‑term power system planning whether for large interconnected grids or for Non‑Interconnected Systems (NIS) traditionally relies on two separate processes: estimating the generation mix through hourly supply demand simulations, and assessing high‑voltage network flows to identify congestion and reinforcement needs. The complexity of combining these analyses generally prevents the use of a single unified model, making consistency between studies both necessary and difficult.

The accelerated deployment of renewable energy sources is reshaping this framework. In several NIS, electrical networks are approaching operational limits, increasing congestion risks. This new context requires a methodological shift in which long‑term planning integrates a coupled assessment of supply demand equilibrium and network constraints. Flexibility options such as strategic storage, local curtailment, and targeted reinforcements are then systematically evaluated to ensure robust planning and maintain coherence across prospective and power system studies.

The research proposes an integrated approach to NIS network planning, combining balancing studies with network analyses. Strategic storage is first sized to ensure supply–demand equilibrium and can also contribute to mitigating network constraints. If constraints persist, additional storage can be assessed and compared with conventional grid reinforcement to identify the most effective way to resolve remaining congestion.

The approach follows a sequence of successive steps. At each step, the placement and sizing of flexibilities are determined using an industrial optimization tool developed by EDF, dedicated to long-term and medium-term energy supply and demand balance studies for NIS power systems. This tool aims to calculate, over long- and medium-term horizons, operating generation plans that ensure the satisfaction of electricity demand while respecting both system and network constraints with an hourly resolution.

The model is formulated as a mixed-integer linear optimization problem, modeling production units through their installed capacities, technical limits, and availability. Demand trajectories and system service requirements (such as reserves) are enforced as constraints. The optimization ensures demand satisfaction and minimizes total operating costs in a context without market prices.

Beyond classical supply-demand formulations, the tool integrates a nodal high voltage network model coupling balance with the high-voltage grid. The network is represented by nodes and lines, using Power Transfer Distribution Factors (PTDF) matrices and maximum admissible flows on each line. This allows the tool to compute optimal solutions that are simultaneously feasible from a supply-demand balance and from a network perspective including the system resilience on any N-1 line contingency.

First non-decisional tests were conducted on Guadeloupe’s network, followed by an application for La Réunion‘s grid, focusing on renewable integration within a long term, ten year-horizon.

Findings highlight that this integrated methodology marks a major step forward in NIS planning. By embedding network constraints early in the design process and rigorously assessing tailored flexibility solutions, it reinforces coordination between supply–demand and reinforcement studies. This shift improves system resilience and provides a replicable framework for managing high‑RES challenges in NIS.