Development of a simple framework to evaluate daylight conditions in urban buildings in the early stages of design

Abstract

The thesis proposes a simple method to aid urban designers in the daylighting aspect of the decision-making process in the early stages of design when the outline of the city is defined. As input to this simple method, complex simulations of the urban canyon structure were made. Paper I reports on a study on the simplified representation of the structure of the streets and buildings in cities. From the results in this study, a simple 4-step method was developed to evaluate facades in an urban context based on daylight simulations in which the densities of the urban building layout, external surface reflectances, and facade window areas were varied. The method developed was based on a CIE overcast sky, so it did not consider the effect of building orientation, geographical location, changes of sky distribution, or the time of day on its results. In Paper II , climate-based daylight simulations of the urban structure were introduced. The climate-based simulations consisted of annual simulations of the daylight conditions. These simulations took into account the enormous variations in daylight illuminances during the year as well as geographical location, facade orientation and user occupancy patterns. For these simulations, the luminance distribution of the sky was described by the Perez all-weather sky-model, based on hour-by-hour input of direct and diffuse irradiance from weather data files. This meant that the weather data files used had an important impact on the simulations. In Paper III , the impact of different weather data sets for a given location and of the time-steps applied was therefore investigated. Occupancy patterns also have an important impact on climate-based daylight simulations. The effect of applying occupancy proles of varying complexity was investigated in Paper IV . In the urban planning stage of design, you usually know very little, or nothing, about the future occupants of the building, so it is useful to know whether detailed simulations of occupancy profiles, as opposed to using simplified assumptions, would increase the value of the simulation. The general hypothesis to be evaluated in this thesis was that: Simple models for the early stages in the design of urban building structures can represent the complexities of daylighting without loss of important characteristics. This hypothesis was investigated through the work reported in the four papers appended. The research showed that the simple method can be expanded to include more complex aspects if the simplifications include the important parameters. To transform complex urban simulations into the simple 4-step representation, it was necessary to include simulations with i) various facade reflectances, ii) rooms located on different floors, iii) buildings with different orientations, and iv) buildings in different geographical locations. So when the right framework is set-up, the general hypothesis was confirmed. The more complex model of the urban structure used weather and occupancy data. The research showed that it made very little difference to the simulation outcome if different weather data files were applied for a given location. Furthermore, simulation with hourly mean irradiance values, as opposed to 1-min resolution, also made little difference to the simulation outcome. This means that simulating the urban structure based on hourly-mean values is sufficient. From the investigations made on the impact of occupancy profiles, it was found that applying an absence factor, as opposed to simulating the dynamic presence of occupants, also made little difference to the simulation outcome. In other words, the complex input in terms of the presence of occupants can be reduced to an absence factor.