(434e) Optimization for Infrastructure Planning of Reliable and Carbon-Neutral Power Generation Systems: Application to California ISO

Ensuring high reliability is a critical issue that must be addressed alongside reducing CO₂ emissions in modern power systems, where the share of renewable generation is increasing. Power system reliability refers to the probability of power systems or a power grid, consisting of multiple generators, supplying uninterrupted electricity to meet customers' load demands [1]. One way to improve power system reliability is to add generators in parallel to power plants, preventing complete failure of each power plant. Additionally, adding storage, such as batteries, to renewable generators helps provide electricity during low renewable source availability or high load demand. In our recent works [2-3], we proposed an optimization model that optimizes the size and operation of reserve systems and main systems by considering the dual role of parallel generators in power system reliability. We identified that the most cost-effective approach to maximizing power system reliability is to maintain the current energy infrastructure without shifting conventional generation facilities to renewable generation facilities. However, this approach conflicts with current energy policies for sustainable energy systems, which for instance, aim to establish 100% renewable electricity systems. This means a limitation of our previous work is that practical constraints, such as minimum renewable generation share and CO₂ emission limits, were not considered.

In this work, we extend the Generalized Disjunctive Programming (GDP) model proposed in our previous works [2-3] to account for the impact of renewable generation and carbon capture systems on power system reliability, and CO₂ emissions. The model optimizes the number and size of parallel dispatchable generators, renewable generators, batteries, and carbon capture systems to maximize power system reliability while minimizing total cost and CO₂ emissions. Specifically, the probability of parallel generator failures is used to estimate power plant availability rigorously based on the number of parallel generators. The GDP model, consisting of Boolean and continuous variables, algebraic equations, and logic propositions [4], also incorporates load shedding penalties into the objective function. We present multiple scenarios with different constraints to assess how the optimal infrastructure changes with environmental constraints and target reliability levels. We demonstrate the value of the proposed GDP model by examining case studies from the California ISO, which aims to establish carbon-neutral power systems by 2045 [5].

References

[1] Cho, S., Li, C., Grossmann, I.E., “Recent advances and challenges in optimization models for expansion planning of power systems and reliability optimization”, Computers and Chemical Engineering, 165, 107924 (2022).

[2] Cho, S., Grossmann, I.E., “An optimization model for expansion planning of reliable power generation systems”, Computer Aided Chemical Engineering, 51, 841-846 (2022).

[3] Cho, S., Tovar-Facio, J., Grossmann, I.E., “Disjunctive optimization model and algorithm for long-term capacity expansion planning of reliable power generation systems”, Computers and Chemical Engineering, 174, 108243 (2023).

[4] Grossmann, I.E., and Trespalacios, F., “Systematic Modeling of Discrete-Continuous Optimization Models through Generalized Disjunctive Programming,” AIChE Journal, 59, 3276-3295 (2013).

[5] PSE Healthy Energy, “California Peaker Power Plants: Energy Storage Replacement Opportunities” (2020)