Conference Year

2016

Keywords

Solar Irradiance Model, Diffuse Irradiance Models, Irradiance Models For Tilted Surfaces, Building Energy Simulation

Abstract

Building energy simulations require a detailed characterization of the boundary conditions in order to solve the air heat balance problem. While for dry and dew bulb temperatures and wind speed, hourly profiles can be successfully developed from raw data measurements, to be included in weather files, beam and diffuse solar irradiance on the external surfaces of the building envelope are not directly known. Measured global solar irradiance is rarely available for the possible orientations of buildings surfaces and when the distinction between the beam and diffuse irradiance components is recorded, it is only for horizontal measurements. Consequently, solar irradiance models are necessary to elaborate hourly beam and diffuse irradiation profiles on the various elements of the building envelope. We can distinguish the different models into two groups: horizontal diffuse irradiance models – to calculate the diffuse component on the horizontal plane starting from the global measurement, and irradiance models for tilted surfaces – to estimate the irradiance collected on tilted surface with a given orientation. Even if some models try to exploit some mathematical expressions to describe the phenomenon, most of them rely mainly on statistical correlations derived from empirical data. Since the samples of solar irradiance measurements used for the models development have been collected mostly in North America and European localities, these models are far from providing an adequate worldwide representativeness and none can be precisely defined as the best one. As observed in previous works in the literature and other authors’ works, the adoption of a pair of models for horizontal diffuse irradiance and for the irradiance on tilted surfaces introduces an uncertainty on the estimation of this specific boundary conditions in building energy simulation. However, different localities, building configurations and kind of energy need – heating and cooling, have different sensitivities to this suboptimal characterization of the solar irradiance. Thus, for some cases, the choice on the couple of models can affect significantly the simulation outputs and influences the robustness of the design solutions. Moreover, also the time discretization of the outputs can be relevant on the quantification of the uncertainty brought by the choice of solar irradiance models. Further expanding previous analyses, in this contribution, 22 horizontal diffuse irradiance models have been coupled with 12 irradiance models for tilted surfaces in order to develop hourly profiles of solar irradiance for 5 European localities (Berlin, Vienna, Trento, Rome and Messina). The 264 alternatives have been used as input in TRNSYS for the simulation of the hourly energy performances of a set of 72 simplified residential buildings, built varying parametrically insulation and thermal inertia of opaque components, windows surface and orientation and solar heat gain coefficient of glazing. The distributions of hourly heating and cooling energy needs along the year and for the different configurations of the sample have been studied in order to identify the building features enhancing the uncertainty due to the solar irradiance modelling on short-term outputs.

Share

COinS