Optimal Reactive Power Compensation in Distribution Networks with Radial and Meshed Structures Using D-STATCOMs: A Mixed-Integer Convex Approach

• Published in: Sensors (Basel, Switzerland) Vol. 22; no. 22; p. 8676
• Main Authors: Garrido, Víctor Manuel, Montoya, Oscar Danilo, Medina-Quesada, Ángeles, Hernández, Jesus C
• Format: Journal Article
• Language: English
• Published: Switzerland MDPI AG 10.11.2022; MDPI
• Subjects: distribution static compensators; Electric power distribution; Energy; Load; Matrix representation; mixed-integer convex optimization; Operating costs; optimal power flow approximation; Optimization algorithms; Optimization techniques; Power flow; Programming environments; radial and meshed distribution networks; Reactive power; Software; Germany; optimal power flow approximation; mixed-integer convex optimization; radial and meshed distribution networks; distribution static compensators
• ISSN: 1424-8220; 1424-8220
• DOI: 10.3390/s22228676

This paper deals with the problem regarding the optimal siting and sizing of distribution static compensators (D-STATCOMs) in electrical distribution networks to minimize the expected total annual operating costs. These costs are associated with the investments made in D-STATCOMs and expected energy losses costs. To represent the electrical behavior of the distribution networks, a power flow formulation is used which includes voltages, currents, and power as variables via incidence matrix representation. This formulation generates a mixed-integer nonlinear programming (MINLP) model that accurately represents the studied problem. However, in light of the complexities involved in solving this MINLP model efficiently, this research proposes a mixed-integer convex reformulation. Numerical results regarding the final annual operating costs of the network demonstrate that the proposed mixed-integer convex model is efficient for selecting and locating D-STATCOMs in distribution networks, with the main advantage that it is applicable to radial and meshed distribution grid configurations. A comparative analysis with respect to metaheuristic optimizers and convex approximations confirms the robustness of the proposed formulation. All numerical validations were conducted in the MATLAB programming environment with our own scripts (in the case of metaheuristics) and the CVX convex disciplined tool via the Gurobi solver. In addition, the exact MINLP model is solved using the GAMS software.