Operational power plant scheduling with flexible carbon capture: A multistage stochastic optimization approach. (2nd November 2019)

• Record Type: Journal Article
• Authors: Zantye, Manali S.; Arora, Akhil; Faruque Hasan, M.M.
• Is Part Of: Computers & chemical engineering. Volume 130(2019)
• Publication Date: 2019-11-02
• Subjects: Flexible carbon capture -- Stochastic dynamic programming -- Multi-stage optimization -- Scheduling under uncertainty -- Surrogate models; Chemical engineering -- Data processing -- Periodicals; 660.0285
• DOI: 10.1016/j.compchemeng.2019.106544
• Languages: English
• ISSNs: 0098-1354

Abstract

Highlights: A systematic approach to increase the profitability and decrease the cost of power plant operations in the presence of active regulatory constraint on carbon emission under uncertainty. A multi-stage stochastic programming algorithm for power plant scheduling with flexible carbon capture. Analysis of the effect of regulatory constraint on CO2 emission on the profitability under market uncertainty. Surrogate model-based continuous expressions that eliminate need for discretization of state space and reduces computational effort. Comparison of the results incorporating price uncertainty with a base case. Graphical abstract: Abstract: To mitigate CO2 emissions, it is often suggested that power plants deploy carbon capture systems. However, high cost is an impediment for the deployment of these systems. To counter this, a power plant can be integrated with a flexible capture unit which varies its load with fluctuating electricity prices. In this work, a multi-stage stochastic programming approach is applied to optimally schedule power production and carbon capture operations to maximize daily profit in a market with uncertain hourly electricity prices. Low-complexity surrogate models are developed for optimal action policy at each stage, which reduce the computational complexity of estimating profit for different price scenarios. The expected value of perfect information obtained is within 25% of the maximum achievable profit while meeting the CO2 emission constraints. Moreover, the profitability improves by 40% compared with the deterministic case assuming expected values of stochastic parameters. This demonstrates the quality of the stochastic solution.