Shortest Path Algorithm To Assess New Levers & Flexibilities In Long Term Sizing Distribution Networks

The illustrative cases consider a DN inspired by the MV CIGRE network composed of two

HV/MV primary substations, which undergoes the connection of multiple medium voltage

(MV) producers every 5 years over a period of 15 years. During multi-stage scenarios, the DSO must optimize the successive connections of new producers in order to minimize the overall cost of reinforcement and connection. The DSO has several planning methods, allowing it to consider future connections and their uncertainties to determine the optimal long-term network.

Four planning methods are compared: two approaches are based on deterministic scenarios, one blind approach does not consider planning, and the last approach, which aims to be innovative, uses probabilistic scenarios and sizes the network using the notion of constraint violation risk.

The probabilistic approach being new, the article aims to describe its methodology, and its innovative aspects compared to deterministic approaches. It also presents its relevance as well as directions for future improvements. The main results of the article are:

• A well-conducted LTP avoids substantial transformer reinforcement costs.

• The threshold effects in planning: optimizing sizing without planning for additional hosting capacity makes the network vulnerable to unexpected DGs. An approach that incorporates the risk of exceeding capacity is necessary to balance optimization and robustness.

• The probabilistic approach positions itself as an intermediate solution compared to the deterministic approaches. It integrates uncertainties via an adjustable “risk rate” (𝜌). Thus, provided that threshold effects are avoided, it can provide a certain degree of robustness. Paris Session 2026

August 23 to 28

Palais des Congrès, Paris, France