IEEE Access (Jan 2024)

Regionalized Generation Expansion Planning: Integrating Spatial Constraints

  • Radhanon Diewvilai,
  • Kulyos Audomvongseree

DOI
https://doi.org/10.1109/ACCESS.2024.3488006
Journal volume & issue
Vol. 12
pp. 163856 – 163882

Abstract

Read online

A typical power system consists of a network of generation, transmission, and load spanning a wide area, with varying spatial characteristics crucial for realistic problem-solving. Traditional generation expansion plans (GEP) often treat the generation system as a single area, specifying types and sizes of new power plants to meet future demand without designating specific locations. This approach leads to generation-demand imbalances, unnecessary transmission expansions, and other issues. Incorporating regional constraints like local demand, tie-line capacity, and available resources is essential in GEP. Considering the entire transmission network can address spatial characteristics but presents challenges due to extensive data preparation and computational complexity. This paper proposes a GEP approach that accounts for spatial characteristics such as primary energy sources, renewable energy potential, and feasible locations for future units. With the proposed method, the power system is divided into multiple zones, each represented by a single bus connected by interzonal transmission lines. This zonal approach simplifies the transmission model by focusing on interzonal data, making it more practical for actual power systems. An area-based reliability index is used to evaluate each area’s reliability level, aiding in the suitable placement of future generation units. The proposed method was tested using Thailand’s latest power development plan, PDP2018 revision 1. Results show that accounting for spatial characteristics alters the generation expansion plan. Additionally, new units are distributed across the system to maintain area reliability. Improved computational efficiency of this proposed method allows for addressing uncertainty by solving multiple scenarios with varying input data and probabilities.

Keywords