Automated shading systems have the potential to substantially reduce building energy consumption, increase occupant exposure to natural daylight and reduce visual and thermal discomfort. The performance of automated solar shading systems, however, greatly depends on how these screens or slats are operated. Conventional control strategies for automated shading systems, that tend to follow binary close-open approaches, are ill-equipped to respond to the large variety of environmental conditions and shifting performance goals in managing the indoor climate of office buildings. Consequently, they inadequately satisfy the visual comfort requirements of occupants and are frequently associated with user dissatisfaction.
Recently, a series of enabling developments has led to promising comfort-driven control strategies being proposed that seek to maximise the admission of daylight by closing shading devices only to the extent that is necessary to prevent daylight glare or thermal discomfort. If these strategies are to be deployed successfully at scale, however, there are several challenges that need to be overcome. For the development and application of such comfort-driven automated shading strategies there is a need for detailed insight into how control and design parameters can be leveraged to influence building performance trade-offs. Additionally, there is a need for generically applicable and scalable workflows for the development of control strategies and their successful deployment in specific buildings.
This doctoral dissertation investigates how computational analyses and optimisation can be used to support the development and application of comfort-driven shading strategies. More specifically, the objective of this research is to develop and test a computational framework for performance evaluation and optimisation of advanced automated shading strategies. Firstly, this framework consists of a virtual test bed (VTB) aimed at analysing the performance of (i) advanced shading controls, (ii) materialisation and shading design features, and (iii) applications of dynamic solar shading systems within performance-driven façade design processes. Secondly, this framework encapsulates a set of computational support methods, aimed at performance analysis, optimisation, and quality control in the application of the VTB.
Because of the fragmented façade design and delivery process, the design and control parameters that define the building performance effects of automated shading systems are specified by various decision makers at different positions in the supply chain. In addition, these design and control parameters vary in their physicality and scale. The requirements for the computational framework are therefore investigated through an iterative process involving four application studies. This process includes feedback from stakeholders in the shading industry. In each application study, the computational framework is developed and tested by applying it to analyse and optimise a series of design and control aspects that are specific to a particular type of shading system and decision-maker perspective.
A literature review identifies the needs and possibilities for applying building performance simulation (BPS) effectively to support decision making in this domain. Based on the identified needs and possibilities, and the findings of the four application studies, a set of requirements for the computational framework were obtained.
The main contribution of the research is the developed computational framework. This framework consists of a VTB for analyses and optimisation of automated shading systems and a set of computational support methods that facilitate the effective use of this VTB. The VTB is designed for multi-domain and multi-scale simulation of automated shading systems. It employs a co-simulation approach between high-resolution domain specific tools using middleware software. These domain specific tools are used for transient thermal building simulation, daylighting and glare simulation, and control system simulation. Additionally, the VTB uses a stepped approach where sub-system simulations are used to describe the physical behaviour of shading materials, shading devices, and the overall fenestration system. The emergent physical behaviour predicted by subsystem simulations at the lower levels of scale is used to define input parameters for models that describe the fenestration system at a higher level of abstraction and predict performance effects on the building level. The VTB also includes multiple features for modelling energy systems at varying levels of detail.
Each VTB module is individually verified and validated throughout the four application studies. In addition, the four application studies show that the modular structure of the VTB allows it to be configured to fulfil various simulation objectives and describe a variety of shading systems. Furthermore, these studies illustrate fit-for-purpose application of the VTB and provide guidance in the selection of model complexity and resolution. Through these studies, this research contributes to the knowledge on performance modelling of complex fenestration systems.
The analyses and optimisation support methods that are developed in this research use statistical classification techniques to identify high performance sensor configurations, detection algorithms and control parameters. The support methods contribute to the body of knowledge on the simulation-based development of advanced shading strategies. Their novelty can be found in the beneficial trade-off between (i) their replicability, (ii) their effectivity in finding control strategies that optimally exploit non-intrusive and non-ideal sensors, and (iii) the time, effort, and skill that are required of developers in their application. Many existing approaches tend to perform particularly well in only a subset of these three aspects. Because the support methods place due emphasis on all three aspects, however, they allow the creation of control strategies that are potentially more scalable in the current context.
The computational framework has gone through usability testing in a broad variety of representative applications that illustrate the most significant ways of it. The application studies address: optimisation of design and control aspects on the five most relevant levels-of-scale (i.e.: sensor deployment strategy, control strategy, shading material, shading system, façade configuration), various types of shading systems (e.g., sun-tracking roller shades and vertical blinds), and the perspectives of various decision makers that are positioned at varying places within the façade design and delivery process (e.g., control developers and façade designers).
Finally, this research contributes insights into the causal relationships between solar shading design and control features and building performance effects. These insights give reason to reconsider the often thermally driven approach to the design of facades, the specification of solar shading devices and the development of control approaches.