Probabilistic Transmission Expansion Planning Under Renewable and Load Uncertainty

This paper presents a novel probabilistic transmission expansion planning framework that leverages the unscented transformation (UT) to address uncertainties in renewable generation and load while significantly reducing computational effort compared to traditional Monte Carlo simulation. It represents the first integrated application of UT for modeling multiple uncertainty sources simultaneously in transmission expansion planning. The proposed probabilistic framework was validated using two test systems: the Garver system and the IEEE 24-bus system. For the Garver system, the unscented transformation approach produced results identical to those of Monte Carlo simulation while requiring 237 times less computational effort (49,500 vs. 11,250,000 load flow evaluations). For the IEEE 24-bus system, the framework effectively modeled load and generation uncertainties, achieving confidence levels of 91.56% and 100.00% for target thresholds of 90% and 100%, respectively, demonstrating consistency with Monte Carlo simulation results. Statistical analysis of seven metaheuristic algorithms revealed that Harmony Search and Differential Evolution performed best for the Garver system, while Grey Wolf Optimizer excelled for the IEEE 24-bus system. The efficiency improvement enables more complex analyses, such as multi-stage planning and N-1 contingency evaluation, that are often prohibitive with traditional Monte Carlo approaches, while maintaining solution accuracy and reliability requirements.