Active building modelling and control
Achieving net-zero carbon emissions by 2050 will entail significant changes to the way electrical energy is generated, transmitted and used. Future electricity systems will be characterised by substantial volumes of intermittent renewable generation that will displace conventional plants, thus increasing the need for back-up reserves and ancillary services. Besides, decarbonising the heating and transport sectors through electrification will lead to increases in peaks that are disproportionately higher than the corresponding increases in annual electricity demand. In this new reality, leveraging operational flexibility will be a critical challenge for developing a cost-efficient net-zero carbon system.
The Active Building Centre (ABC) is funded by UKRI as part of the Transforming Construction Challenge through the Industrial Strategy Challenge Fund (ISCF). The ABC project is led by Swansea University and includes industrial and academic partners from across the UK. The ABC focusses on net-zero carbon buildings with topics spanning from development of new technologies, through the optimal operation of buildings, to societal challenges.
The ABC project aims to transform the construction industry to facilitate the UK’s commitment to net-zero carbon emissions by deploying smart distributed energy resources and energy management system that support the system integration of low-carbon technologies. One of the key objectives of this work is to identify the system benefits and quantify the value of Active Building technologies. A range of technologies will be analysed, including electric vehicles, smart appliances, smart heating with short and long-duration thermal storage systems, electricity storage, and distributed generation such as solar PV. Those technologies will provide flexibility by modulating the local energy consumption or generation and providing system services to support both local and national electricity system. The whole-energy system analysis for AB is illustrated in the picture below.
The work conducted at Imperial involves the enhancement of the integrated whole energy system model to incorporate the AB concept and its technologies. The model is used to:
- Evaluate the impact of a UK-level rollout of Active Buildings solutions in the context of the overall future decarbonisation of UK’s energy supply,
- Assess alternative design strategies of ABs driven by whole-system objectives (instead of driven by the local objective). The conflicts and synergies between using AB flexibility for local and national level management will also be analysed.
Optimal technology sizing is required to ensure that the system design suits the dwelling under consideration in terms of efficiency, cost and flexibility. Simultaneous optimisation of design and operation of a system is called co-design. The effect of a control strategy on the design process is frequently overlooked. Incorporating the behaviour of a controller in the loop leads to a potentially computationally demanding problem. Ignoring the control behaviour, on the other hand, may lead to an ill-suited set of technologies for a given situation.
Model predictive control (MPC) has excellent potential for improving energy usage in buildings and increasing their flexibility and adaptation capability to support the whole energy system since it enables straightforward integration of state and input constraints. The resulting optimal control problem is challenging from the perspective of integrating the model of the residential building and predictive control as well as from the perspective of numerical methods used to solve the resulting optimal control problem. One of our objectives is, therefore, to study transcription methods that would provide an alternative to widely used shooting or collocation methods, as well as investigate decomposition methods for large scale optimisation.
Another type of optimal control problem that we work on is related to robust control and optimisation under uncertainty. The objective is to provide an optimal control framework that would be robust to uncertainty in the model, yet at the same time would satisfy the requirements from the occupants of the building. Typical approaches to handle uncertainty in predictive control result in large scale optimisation problems, and our goal is to provide quick and efficient methods of solving this kind of problems. For that purpose, we study active set methods which enable a reduction of the size of the optimisation problem.
In particular, the optimal control approach will enable the integration of the operation of residential buildings and services provided by the grid on both local and national levels.
Work package: Predictive Control and Active Agents
Dr E. C. Kerrigan (EEE&Aero Department) – email@example.com - work package leader
Prof N. Shah (ChemEng Department) - firstname.lastname@example.org
Dr E. Atam (EEE Department) - email@example.com
Dr P. Falugi (EEE Department) – firstname.lastname@example.org
Dr E. O’Dwyer (ChemEng Department) – email@example.com
Dr M. Zagorowska (EEE Department) – firstname.lastname@example.org
Work package: Energy Systems and Infrastructure Analysis and Modelling
Prof G. Strbac (EEE Department) – email@example.com - work package leader
Dr D. Pudjianto (EEE Department) - firstname.lastname@example.org
Dr X. Zhang (EEE Department) – email@example.com
Dr S. Giannelos (EEE Department) – firstname.lastname@example.org
Dr M. B. J. Woolf (EEE Department) – email@example.com
L. Mauricette (EEE Department) - firstname.lastname@example.org
Work package: Business Models, Analytics and Incentivisation
Prof Richard Green (Business School) - email@example.com - work package leader
Dr Joachim Geske (Business School) - firstname.lastname@example.org