Skip to main content
Thesis defences

PhD Oral Exam - Jennifer Date, Building, Civil and Environmental Engineering

Model-Based Enhanced Operation of Building Convective Heating Systems & Active Thermal Storage


Date & time
Tuesday, October 5, 2021 (all day)
Cost

This event is free

Organization

School of Graduate Studies

Contact

Dolly Grewal

When studying for a doctoral degree (PhD), candidates submit a thesis that provides a critical review of the current state of knowledge of the thesis subject as well as the student’s own contributions to the subject. The distinguishing criterion of doctoral graduate research is a significant and original contribution to knowledge.

Once accepted, the candidate presents the thesis orally. This oral exam is open to the public.

Abstract

This thesis presents an experimental and theoretical study of a reduced-order modelling methodology and dynamic response of convectively heated buildings, their respective space heating peak demands, and active thermal storage. A methodology was developed for the generation of control-oriented building models which can then be used within model predictive control (MPC) or other model-based control strategies for eventual implementation into the building HVAC industry that can satisfy occupant comfort and improve building-grid interaction. A methodology to identify and evaluate MPC strategies is presented to improve a building's energy flexibility or building-grid interaction capabilities. The focus is on the winter operation of archetypal buildings found in Quebec, mainly low-mass and low-rise buildings. Also, there is an emphasis on modelling building thermal mass and a dedicated thermal storage device. In this work are several case studies, where measured operational data from the buildings are used for model development, calibration, and to compare against in simulation studies. The chosen buildings are typical construction with convective-based heating systems commonly seen throughout Quebec: a detached low-mass house, a low mass retail building, and a warehouse (with incorporated active thermal storage device).

Incorporating the principles of building energy flexibility, together with on-site energy storage devices and advanced or optimized control strategies is essential for optimizing energy use and matching demand with the availability of electricity from the grid at critical times. The development of a suitable control-oriented thermal model is crucial for improving building energy flexibility and building-grid interaction. In this work, the focus was put on the grey-box modelling approach for thermal modelling of buildings with convective heating systems and modeling for thermal energy storage devices. The two applications for reduced-order thermal modelling (buildings and dedicated active thermal energy storage devices) require different modelling approaches for control applications. It was observed that many available case studies -- which are representative of typical Quebec buildings -- have adequate sensor points and data for creating a low-order model suitable for control, and grey-box modelling has strengths from physics-based modelling, while also incorporating optimization in the model development and calibration. As buildings change, in terms of operation, physical elements, or occupancy schedules, a grey-box model can be modified to account for these changes.

The study on a residential building outlined a methodology for multi-level control-oriented modelling for buildings with several zones and multiple floors. This multi-level approach allows the user to “zoom in and out” so that models at each control level remain manageable, easy to calibrate and easy to physically interpret. A global low-order grey-box model (1R1C) was developed and used to rapidly calculate the thermal load of the building, while a very detailed benchmark floor-level model was developed and used for verification and MPC-based simulation studies. For the development of specific control algorithms for each zone, an adequate simplified zone-level model must be developed. It was found that if zone-level accuracy is of importance, one must incorporate into the model the thermal mass of the structure between zones.

In the second case study -- a smaller commercial bank building -- implementation of MPC was presented for a conventional building (a building of basic construction, systems, and technologies) to reduce the yearly utility bill and avoid the summer peak load penalty. The cost function aimed to minimize the utility rate during each prediction horizon while meeting upper and lower indoor temperature constraints. Through a parametric study, it was seen that using a longer control horizon (greater than six hours), produced better results for this building. A cost savings of 25% on the yearly electric utility bill and a peak power reduction of 38% were achieved, by implementing a new optimized temperature schedule for the building every 12 hours. The main difference between the typical operation temperature schedule and the optimized setpoint schedule is a preheating of the building in the few hours prior to the start of occupation. With the new optimized operation, the cost per square meter for the bank would decrease from $30.19/m2 to $22.57/m2, or a yearly savings of $7.62/m2.

The last case study considered in this thesis comprises a 1650 m2 warehouse facility equipped with a dedicated active high-temperature thermal energy storage device. A general methodology was presented for the development and analysis of control-oriented models for the enhancement of operation of an electric thermal storage device (ETS) and energy use within a building. The developed methodology for implementing MPC strategies for space heating to a warehouse zone equipped with a dedicated active thermal storage device was presented. The goal was to predict and maximize the building energy flexibility the building could provide to the electric grid by evaluating the newly developed Building Energy Flexibility Index (BEFI) for different strategies. MPC with active thermal storage was shown to increase BEFI and provide energy flexibility to the grid during peak times and can perform superior to manual BAU control. As an example, a BEFI of 55% to 100% is achieved when the notification from the utility to the customer is 12 hours ahead of a 6AM Demand Response (DR) event. Depending on the objective function, this means that the average demand during the critical times can be reduced by an amount between 36 kW (55%) and 65 kW (100%). It was also observed that the utility cost to the customer can be reduced by 12-30% when compared to the BAU operation with ETS.

Back to top

© Concordia University