Demand flexibility evaluation for building energy systems with active thermal storage using model predictive control

Published

June 29, 2022

Author(s)

Guowen Li, Yangyang Fu, Amanda Pertzborn, Zheng O'Neill, Jin Wen

Abstract

Model Predictive Control (MPC) has been demonstrated to be an efficient way to reduce building operating costs, especially for buildings with thermal storage systems, by changing the power demand profiles. Different parameter settings of MPC have also been shown to have significant influence on building power usage, which may therefore influence building demand flexibility. In this study, we estimate how MPC parameters such as the prediction horizon (PH) can influence building demand flexibility. A virtual high-fidelity building testbed was created in Modelica based on actual measurement data from a chiller plant with an ice storage tank system. Then the virtual system was randomly perturbed to generate training data for the MPC models. The MPC was formulated as a nonlinear programming problem and solved using a global optimization solver. We found that MPC can reduce operating costs by 15.8 % and reduce the peak power demand by 9.6 % compared with rule-based storage-priority control. The building demand flexibility initially increases as the PH increases and then reaches its plateau when the PH is longer than 20-hours. Comparison of the building demand flexibility will provide insights into choosing the suitable MPC formulation for a grid-interactive efficient building.

Proceedings Title

ASHRAE Transactions

Volume

128 Part 2

Conference Dates

June 25-29, 2022

Conference Location

Toronto, CA

Conference Title

2022 ASHRAE Annual Conference

Pub Type

Conferences

Download Paper

Local Download

Keywords

active thermal energy storage, model predictive control, hvac, demand flexibility

Citation

Li, G. , Fu, Y. , Pertzborn, A. , O'Neill, Z. and Wen, J. (2022), Demand flexibility evaluation for building energy systems with active thermal storage using model predictive control, ASHRAE Transactions, Toronto, CA, [online], https://tsapps.nist.gov/publication/get_pdf.cfm?pub_id=933865 (Accessed December 10, 2024)