Control of Residential Space Heating for Demand Response Using Grey-box Models

Authors

Rasmus Elbæk Hedegaard
Aarhus University
Keywords: Reduced order models, Model predictive control, Demand response, Smart-meter consumption data, Market-driven demand response, Practical weather data acquisition

Synopsis

Certain advanced control schemes are capable of making a part of the thermostatic loads of space heating in buildings flexible, thereby enabling buildings to engage in so-called demand response. It has been suggested that this flexible consumption may be a valuable asset in future energy systems where conventional fossil fuel-based energy production have been partially replaced by intermittent energy production from renewable energy sources. Model predictive control (MPC) is a control scheme that relies on a model of the building to predict the future impact on the temperature conditions in the building of both control decisions (space heating) and phenomena outside the influence of the control scheme (e.g. weather conditions). MPC has become one of the most frequently used control schemes in studies investigating the potential for engaging buildings in demand response. While research has indicated MPC to have many useful applications in buildings, several challenges still inhibit its adoption in practice. A significant challenge related to MPC implementation lies in obtaining the required model of the building, which is often derived from measurements of the temperature and heating consumption. Furthermore, studies have indicated that, although demand response in buildings could contribute to the task of balancing supply and demand, suitable tariff structures that incentivize consumers to engage in DR are lacking. The main goal of this work is to contribute with research that addresses these issues. This thesis is divided into two parts.

The first part of the thesis explores ways of simplifying the task of obtaining the building model that is required for implementation of MPC. Studies that explore practical ways of obtaining the measurement data needed for model identification are presented together with a study evaluating the suitedness of different low-order model structures that are suited for control-purposes.

The second part of the thesis presents research on the potential of utilizing buildings for demand response. First, two studies explore and evaluate suitable incentive mechanisms for demand response by implementing an MPC scheme in a multi-apartment building block. These studies evaluate two proposed incentive mechanisms as well as the impact of building characteristics and MPC scheme implementation. Finally, a methodology for bottom-up modelling of entire urban areas is presented, and proved capable of predicting the aggregated energy demand of urban areas. The models resulting from the methodology are then applied in an analysis on demand response.

Cover for Control of Residential Space Heating for Demand Response Using Grey-box Models
Published
March 26, 2019

Details about this monograph

ISBN-13 (15)
978-87-7507-447-1
doi
10.7146/aul.317.215