(52c) A Column Generation Approach to Multiscale Capacity Planning for Continuous Power-Intensive Processes

Zhang, Q. - Presenter, University of Minnesota
Flores-Quiroz, A., University of Chile
Pinto, J. M., Linde plc
In light of highly time-sensitive electricity prices in deregulated electricity markets, power-intensive industries have the increasing need to shift load to low-price time periods in order to maintain cost-competitive operations. An industrial plant's ability to shift load is primarily given by its operational efficiency and dynamic behavior, as well as its production and storage capacities. Therefore, careful capacity planning is required for assessing the value of additional flexibility and planning for expected long-term changes in demand. Here, the main challenge lies in the simultaneous consideration of decisions at two different time scales, namely the long-term decisions related to the capacity expansion strategy and the short-term operational decisions.

The multiscale capacity planning problem for continuous power-intensive processes was first introduced by Mitra et al. (2014). The objective is to find the optimal long-term capacity expansion plan, which involves adding or modifying process units and storage capacity over the course of multiple years. The resulting mixed-integer linear program (MILP) is very large in size and hence difficult to solve.

In this work, we apply a column generation approach to solve an extended deterministic version of the model proposed by Mitra et al. (2014). The Dantzig-Wolfe decomposition takes advantage of the block-diagonal structure of the problem and allows us to obtain the solution by solving multiple independent MILPs, one per season, in an iterative fashion. We demonstrate the effectiveness of the proposed algorithm in an extensive computational study as well as in a real-world industrial case provided by Praxair.


Mitra, S., Pinto, J. M., & Grossmann, I. E. (2014). Optimal multi-scale capacity planning for power-intensive continuous processes under time-sensitive electricity prices and demand uncertainty. Part I: Modeling. Computers and Chemical Engineering, 65, 89–101.