Open AccessSecurity‐constrained transmission expansion planning using linear sensitivity factors
Author(s) -
Mehrtash Mahdi,
Kargarian Amin,
Rahmani Mohsen
Publication year - 2020
Publication title -
iet generation, transmission and distribution
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.92
H-Index - 110
eISSN - 1751-8695
pISSN - 1751-8687
DOI - 10.1049/iet-gtd.2019.0844
Subject(s) - computer science , sensitivity (control systems) , scalability , mathematical optimization , electric power system , linear programming , ac power , transmission (telecommunications) , power balance , constraint (computer aided design) , voltage , graph , algorithm , power (physics) , mathematics , theoretical computer science , electronic engineering , telecommunications , engineering , physics , geometry , quantum mechanics , database
Formulating power flow equations with linear sensitivity factors (LSFs) reduces the number of variables and constraints, and consequently, the computational burden of power systems’ optimisation problems. This study proposes a transformative, computationally efficient model for transmission expansion planning (TEP). While the existing TEP models use bus voltage angles, the proposed TEP takes advantages of LSFs to formulate an optimisation. LSFs allow to omit voltage angles from the formulation and replace all nodal power balance constraints by one equivalent constraint. Thus, the proposed model includes less number of variables and constraints compared with the classical angle‐based model. These features significantly reduce computational costs of TEP and enhance its scalability, especially for large‐scale systems. Load and generation uncertainties are modelled using a data‐driven approach, and N − 1 security criteria are taken into account to ensure system security. All equations under normal and N − 1 conditions are considered using data of the complete network graph. Simulation results show that the proposed model provides the same results as the conventional angle‐based model while being much faster (more than 58% based on the authors’ case studies) and computationally more efficient.