The browser you are using is not supported by this website. All versions of Internet Explorer are no longer supported, either by us or Microsoft (read more here: https://www.microsoft.com/en-us/microsoft-365/windows/end-of-ie-support).

Please use a modern browser to fully experience our website, such as the newest versions of Edge, Chrome, Firefox or Safari etc.

Multi-objective Optimization of Fenestration Design in Residential spaces. The Case of MKB Greenhouse, Malmö, Sweden

Author

  • Iason Bournas
  • Ludvig Haav

Summary, in English

This thesis investigates the optimization of fenestration design for multi-family apartments, considering the heating demand, daylight autonomy as well as overheating. A literature review was conducted to situate the thesis focus within the broader academic field of façade optimization, and a specific apartment located in the city of Malmö was chosen as the study object. The results presented are the outcome of climate-based daylight modelling (CBDM) simulations and dynamic thermal modelling (DTM) simulations, all of which were integrated in a single script definition within the visual programming environment of Grasshopper (2016). A significant part of the study involved the use of an optimization algorithm, to assess multiple fenestration designs based on their daylighting and heating performance. The optimum window position, size and shape were assessed as a function of the achieved daylight levels, the energy required for heating, the impact of solar gains and the amount of overheating time for the studied spaces. Overall, it was shown that the objectives of heating and daylighting are in conflict in the Swedish context, when the aim is to satisfy both luminous and thermal needs. In addition, it was shown that the window-to-wall ratio is not sufficient as information regarding the building performance, as different geometrical aspects of windows and their position can lead to different results for the same glazing area.

Summary, in English

In 1929 the prominent architect Le Corbusier said: “The history of architectural material… has been the endless struggle for light… in other words, the history of windows.” Today, architects should consider not only the amount of light that can be admitted through the façade, but also the amount of heat that will subsequently exit the building, as windows constitute the weakest thermal barrier of a buildings envelope. In addition, large glazing areas should be avoided as they can result in overheating issues during summer. To satisfy all objectives at the same time, an optimization of the window geometry should be conducted where multiple window designs are evaluated.

If different aspects of windows are to be considered (number, shape, size, position) for a facade, the increase of complexity and design alternatives can lead to a high number of possible solutions. In this thesis, the amount of possible design solutions was higher than seven million. Therefore, when faced with a large number of iterations, practitioners traditionally make simplifications and only examine a limited number of designs. Genetic algorithms can help find solutions that can satisfy both energy and daylighting goals, avoiding assessing all possible designs, but focusing only on the optimum ones. This type of facade optimization has been investigated numerous times and in general, the related work has shown that thermal and luminous needs can be satisfied with more than one facade solutions. Following this logic, this thesis provides architects the tools to identify a range of optimum solutions to select among them. The intention is to guide design choices based on simulation results processing and evaluation, rather than on architectural intuition or conventional wisdom.

A significant part of this study involved the use of a genetic algorithm to generate and assess different window geometries in an efficient, automated workflow. Simulations are typically used in a step-wise manner, with the designer suggesting a solution and subsequently simulating it by the use of relevant software. Genetic algorithms can accelerate the optimization process by automatically generating designs that are assessed based on user-defined goals. A preliminary part of this thesis investigated the optimum modelling methodology in order to avoid long simulation times and, at the same time, have accurate results. Having a fast and accurate model is crucial when the intention is to evaluate multiple designs.

The study object was the MKB Greenhouse apartments located in Augustenborg, Malmö, as an example of low energy residential spaces. Parameters of the study included windows shape, size, position and orientation. It was shown that the optimum amount of glazing area depends on the orientation and on specific goals. If heating is to be optimized, south openings should be maximized, but if overheating is the defining parameter, then northern windows should be larger.
The results indicated that the amount of glazing does not provide sufficient information on the building performance, in terms of daylighting. It was shown that highly placed small windows can actually admit more daylight than windows of a double size. This was the result of light penetrating deeper in space when entering at a high level. In addition, the heating energy consumption is reduced when placing windows on surfaces that receive more solar irradiation, while the electric lighting use can be more effectively improved when spreading the openings throughout the envelope.

Significant focus was given to the processing and presentation of the results. The intention was that the user could easily access and visualize different window designs, along with information about their performance on heating, daylighting and overheating. In this way, the proposed method can be deployable in an every-day scenario, where the environmental performance of proposed solutions can be dynamically illustrated. This can facilitate the cooperation between design teams and other parties, such as interested developers, clients, environmental specialists etc.

Overall, it was shown that current software can be scripted to create an automated simulation environment to conduct state-of-the-art optimizations. This approach can lead to a good set of solutions for designers to work with in the initial design stage. Although conventional architectural wisdom is hard to be substituted by a genetic algorithm, this process can be implemented in order to acquire information about designs that will perform better in the future, and about designs that should better be avoided. The goal here is for this approach to be followed in the initial design stage, which can shape the final building performance dramatically.

Publishing year

2016

Language

English

Document type

Student publication for Master's degree (two years)

Topic

  • Technology and Engineering

Keywords

  • overheating time
  • heating
  • daylighting
  • fenestration
  • optimization
  • residential

Supervisor

  • Marie-Claude Dubois (Docent)

Scientific presentation