Optimising heating and cooling of smart buildings

Abstract

This thesis is concerned with optimization techniques to improve the efficiency of heating and cooling of both existing and new buildings. We focus on the thermal demand-side and we make novel contributions to the optimality of both design and operational questions. We demonstrate that our four novel contributions can reduce operations cost and consumption, optimize retrofit and estimate relevant parameters of the built environment. The ultimate objective of this work is to provide affordable and cost-effective solutions that take advantage of local existing resources.

This work addresses four gaps in the state-of-the-art. First, we contribute to current building practice that is mostly based on human experience and simulations, which often leads to oversized heating systems and low efficiency. The results in this thesis show the advantages of using optimization approaches for thermal aspects in buildings. We propose models that seek optimal decisions for one specific design day, as well as an approach that optimizes multiple day-scenarios to more accurately represent a whole year.

Second, we study the full potential of buildings’ thermal mass and design. This has not been fully explored due to two factors: the complexity of the mathematics involved, and the fast developing and variety of emerging technologies and approaches. We tackle the mathematical challenge by solving non-linear non-convex models with integer decisions and by estimating building’s thermal mass. We support rapid architectural development by studying flexible models able to adapt to the latest building technologies such as passive house design, smart façades, and dynamic shadings.

Third, we consider flexibility provision to significantly reduce total energy costs. Flexibility studies often only focus on flexible building loads but do not consider heating, which is often the largest load of a building and is less flexible. Because of that, we study and model a building’s heating demand and we propose optimization techniques to support greater flexibility of heating loads, allowing buildings to participate more efficiently in providing demand response.

Fourth, we consider a building as an integrated system, unlike many other modelling approaches that focus on single aspects. We model a building as a complex system comprising the building’s structure, weather conditions and users’ requirements. Furthermore, we account for design decisions and for new and emerging technologies, such as heat pumps and thermal storage. Optimal decisions come from the joint analysis of all these interconnected factors.

The thesis is structured in three parts: the introduction, the main body and the conclusions. The main body is made by five chapters, each of which focuses on one research project and has the following structure: overview, introduction, literature review, mathematical framework description, application and results section, conclusion and future works. The first two chapters discuss the optimization of operational aspects. The first focuses on a single thermal zone and the second in two connected ones. The third chapter is a continuation of the first two, and presents an approach to optimize both operations and design of buildings in a heat community. This approach integrates the use of an energy software already in the market. The fourth chapter discusses an approach to find the optimal refurbishment of an existing building at minimum cost. The fifth chapter shows an inferring model to represent a house of a building stock. We study the common case where the house’s data is lacking or inaccurate, and we present a model that is able to estimate the required thermal parameters for modelling the house using only heating demand.