(240g) Energy Demand Management in Process Systems Subject to Time-Varying Electricity Prices: A Decomposition-Based Approach

As the availability and pricing of electricity vary during the day due to intermittency in renewable resources or real-time electricity tariff policies, the chemical industry, especially those with power-intensive processes, needs to rethink operating strategies to maximize economic viability. The implementation of demand response (DR) strategies, which focuses on demand side activities and has been in the center of all pragmatic energy policy decisions, is the key to minimizing production costs and maximizing profits. The economic potential of DR for industrial processes has been acknowledged in numerous studies (see, for example, [1]-[3]).

In developing DR strategies, the target lies in scheduling the production rates, production times and inventory levels at each time step to achieve specific objectives. Therefore, process control-related activities are closely linked to the responsive demand augmentations given that control actions involve real-time manipulation of selected process variables to hold key product quality and production rates near their desired target values. Moreover, as the plant operating schedule varies, process control facilitates the dynamic transition periods between different operating modes. Therefore, it becomes imperative to develop design and operating methodologies that can effectively integrate production scheduling and control functions (e.g., see [4]-[5] for some recent comprehensive reviews of the literature on the integration of scheduling and control.

In a prior work [8], we posed the DR problem through a mixed-integer nonlinear programming (MINLP) formulation where optimal operating policies (modes) were found in response to varying electricity prices and intermittent on-site renewable generation. A key assumption in that work was that the control policies were predetermined and no attempt was made to optimize the transition periods. By relaxing that assumption, the goal of the present work is to offer a new DR formulation for the integration of scheduling and control subject to time-varying electricity prices. The formulation simultaneously takes into account the production scheduling with respect to minimizing production cost, the optimization of controller parameters to realize the objective of reducing the time and lost product during transitions, and the integration of scheduling and control to achieve a more economical operating policy. Instead of formulating one complex and unsolvable MINLP problem, we propose a novel decomposition scheme that generates smaller sub-problems that can often be solved more easily. The proposed strategy is implemented within a receding horizon optimization framework, in which only the first transition is carried out with optimally tuned controller parameters and a subsequent scheduling problem is then reformulated with the new updated storage level corresponding to the optimized transition time. A case study involving a continuous chemical reactor is used to demonstrate the contributions and efficacy of the developed approach.

References:

[1] M. Paulus and F. Borggrefe, â??The potential of demand-side management in energy-intensive industries for electricity markets in Germany,â? Applied Energy, 47, 1492-1503, 2011.

[2] S. Mitra, I. E. Grossmann, J. M. Pinto and N. Arora, �Optimal production planning under time-sensitive electricity prices for continuous power-intensive processes,� Comput. Chem. Eng., 38, 171-184, 2012.

[3] Y. Wang and L. Li, â??Time-of-use based electricity demand response for sustainable manufacturing systems,â? Energy, 63, 233-244, 2013.

[4] S. Engell S and I. Harjunkoski, â??Optimal operation: Scheduling, advanced control and their integration,â? Comput. Chem. Eng., 47, 121-133, 2012.

[5] M. Baldea and I. Harjunkoski, â??Integrated production scheduling and process control: A systematic review,â? Comput. Chem. Eng., 71, 377-390, 2014.

[6] C. Tong, A. Palazoglu, N. H. El-Farra and X. Yan, â??Energy demand management for process systems through production scheduling and control,â? AIChE Journal, 61, 3756-3769, 2015.