|
Multi-objective stochastic unit commitment with dynamic pricing and life cycle assessment for sustainable power systems |
|---|---|
| รหัสดีโอไอ | |
| Title | Multi-objective stochastic unit commitment with dynamic pricing and life cycle assessment for sustainable power systems |
| Creator | Dararith Tiv |
| Contributor | Aussadavut Dumrongsiri, Advisor |
| Publisher | Thammasat University |
| Publication Year | 2568 |
| Keyword | Multi-objective optimization, Life cycle assessment, Wind power integration, Power system losses, Dynamic pricing, Price-based demand response, Stochastic unit commitment, Mixed-integer linear programming, Duck curve mitigation |
| Abstract | Driven by the need to manage electricity cost volatility under high renewable energy penetration, modern power systems require planning tools that can coordinate generation scheduling, environmental impacts, and demand-side flexibility. This research proposes an enhanced Multi-Objective Stochastic Unit Commitment (SUC) framework that balances economic efficiency with environmental sustainability through the integration of Life Cycle Assessment (LCA) metrics. Unlike traditional models that treat demand as fixed, this study employs a closed-loop, iterative Mixed-Integer Linear Programming (MILP) framework in which stochastic operational cost signals are converted into practical dynamic Time-of-Use (TOU) tariffs for different customer groups. The proposed pricing mechanism is positioned as a practical tariff adjustment controller rather than a direct replacement for theoretical shadow-price-based marginal pricing.The methodology extends existing literature by incorporating a hybrid renewable portfolio of solar and wind generation, while rigorously accounting for a 12% system power loss factor to ensure grid realism. To handle uncertainty, the model utilizes 200 probability-weighted scenarios generated via Monte Carlo simulation and Kernel Density Estimation. Furthermore, a dynamic reserve requirement is implemented, mandating reserves equivalent to 3% of total demand plus a 5% margin for wind forecast error to maintain system reliability.The framework was validated using a modified Thai power system through a base-case analysis and customer price elasticity sensitivity analysis. Key findings demonstrate that the proposed pricing mechanism successfully reshapes the aggregate system demand, achieving a 3.8% reduction in the critical evening peak and a valley-filling effect of up to 2.8%. Sector-specific analysis revealed that the Commercial sector achieved the highest peak reduction of 12.1%, while the Residential sector increased off-peak usage by 3.9% in response to low-cost signals.A sensitivity analysis of customer price elasticity confirmed a clear inverse relationship between consumer flexibility and system expenditure, with high-flexibility scenarios yielding operational cost savings of up to 7.82%. Conversely, reduced elasticity leads to a 5.82% increase in financial burden, highlighting the importance of demand-side response for economic efficiency. The algorithm reached a stable economic equilibrium within 8 to 10 iterations across the tested cases. Because each case solves a 200-scenario stochastic MILP model within an iterative pricing loop, the computational time per complete sensitivity case ranged from approximately 1.35 to 1.73 hours, which remains practical for day-ahead planning studies.In addition, the computational performance of the multi-objective iterative mechanism was analyzed to assess its practical viability. Across the simulated base case and elasticity sensitivity cases, the model achieved stable tariff convergence within 8 to 10 iterations. The full nine-case sensitivity campaign required approximately 13.5 hours of computation, reflecting the combined burden of stochastic scenario representation, unit-commitment binary decisions, storage constraints, reserve requirements, and the closed-loop demand response update.The outcomes of this study provide significant implications for utility operators and policy makers by providing a robust mechanism to enhance grid stability through “green” demand-side management. By enabling a dynamic link between stochastic operational costs and consumer prices, the framework successfully mitigates the “duck curve” phenomenon and facilitates the integration of hybrid renewable portfolios. Specifically, the model offers a pathway to reduce total operational costs by creating financial incentives that align consumer behavior with periods of high solar and wind availability.Future research directions include the exploration of peer-to-peer (P2P) energy trading and the integration of electric vehicles (EVs) as mobile storage assets. Additionally, the framework could be scaled to larger, interconnected power systems to investigate the impact of inter-regional transmission constraints on dynamic pricing stability. |
