Content area
The power system is experiencing a significant transition from conventional energy sources to renewable resources, driven by environmental concerns, economics, and the emergence of various new technologies. In this context, the need for updated planning tools to effectively identify investment becomes increasingly important.
This dissertation is focused on three aspects of expansion planning: contingency probability estimation; model reduction; and expansion planning.
For the first one, contingency probability estimation, a Markov model is proposed to characterize the possible states of a system that can cause any of the various contingency scenarios referred to as P1-P7 within the industry reliability standard TPL-001-4 developed and published by the North American Electric Reliability Corporation. Based on this model, we also developed a computational platform called Contingency Probability Estimation Tool. Planners can utilize the obtained result to incorporate reliability-related contingency considerations into their expansion planning process.
Coordinated expansion planning applications that optimize both transmission and generation investments are typically compute-intensive, and their use on industry-grade network models, typically having tens of thousands of buses, is not possible without compromising modeling fidelity in various ways. One way to reduce compute-time is to reduce the network size, but none of the suggested reduction methods to date were able to replicate the full model with sufficient fidelity. To address this, we developed a six-step network reduction method aimed at reducing the size of industrial-scale networks. The effectiveness of this method is validated through its application to the network of the Midcontinent Independent System Operator (MISO).
Once the reduced network is generated, we utilize it to conduct Adaptive Coordinated Expansion Planning (ACEP), enabling us to devise a prudent investment decision strategy that accounts for multiple future scenarios. In contrast to traditional stochastic programming, ACEP provides users with the ability to customize robustness by trading off core investments that are made under all scenarios with the investments that are made only when the future is revealed. In this context, robustness refers to the system’s ability to maintain reliable performance across various scenarios without being overly sensitive to uncertainties.
In the approach developed, after ACEP is solved on the reduced network, generation investments are translated to the full network model, and transmission expansion is run to refine the transmission investment part of the solution. ACEP was applied to the reduced MISO network, and results were obtained that validated the method.
