(72t) An Optimization Framework for the Day-Ahead Scheduling of District Cooling Systems | AIChE

(72t) An Optimization Framework for the Day-Ahead Scheduling of District Cooling Systems


The main goal of most district cooling systems is to centrally produce and distribute chilled water at a minimum cost while satisfying demand. Chilled water is produced by production plants and distributed to buildings equipped with energy transfer stations (sub-stations). After the heat transfer takes place, water is returned to the production plants where the chilled water is regenerated. District cooling is essential for sustainable economic and urban development.  The reason for this is that by centralizing the chilled water generation, district cooling systems can increase system efficiency, reduce air pollution, decrease emissions of refrigerants, and manage electricity demand on-site and to the electric grid as a whole. 

Chillers are the main components of the chilled water production plants. The production cost of chilled water plants that use electrically powered chillers is deeply sensitive to electricity price variability [4]. To minimize the production cost, thermal energy storages are often used [1,3]. The main idea is to store chilled water in the thermal energy storage when the electricity price is low and discharge it when the electricity price is high. Although this operating rule may seem trivial to implement, multi-modal electricity price curves, thermal energy storages with limited capacity as well as multi-period constraints on chillers present a challenge to the operators who aim at developing the day-ahead operating schedule of the cooling system at a minimum cost.

In this work we propose an optimization framework that deals with this situation. We show that by combining the latest results on piece-wise linearization techniques [2] with disjunctive programming strategies [5] we are able to generate an optimization model which can be solved fast and efficiently as a mixed-integer problem. We describe how this framework is implemented and successfully used to calculate the day-ahead optimal schedule of the thermal energy storage and chillers at the Thermal Energy Corporation, a district energy company located in Houston, TX.

[1]    W. Bahnfleth and W. Joyce. 1994. Energy use in a district cooling system with stratified chilled water storage. ASHRAE Transactions. Volume 100, pp 1767-1778

[2]    B. Geißler, A. Martin, A. Morsi, L. Schewe. 2012. Using Piecewise Linear Functions for Solving MINLPs. The IMA Volumes in Mathematics and its Applications. Volume 154, pp 287-314

[3]    W. J. Cole, K. M. Powell and T. F. Edgar. 2012. Optimization and advanced control of thermal energy storage systems. Chemical Engineering Reviews. Volume 28, pp 81–99

[4]    K. M. Powell, W. J. Cole, U. F. Ekarika, T. F. Edgar. 2013. Optimal chiller loading in a district cooling system with thermal energy storage. Energy. Volume 50, pp 445–453

[5]    J. P. Ruiz, J. Wang and C. Liu. 2013. Logic-based outer-approximation method for the security constrained unit commitment problem. IET Generation, Transmission & Distribution. Volume 7, pp 1210-1218