Abstract
The world’s population growth and the current urbanization trend have brought more and more attention on the energy needs of cities. In particular, buildings are already responsible for a large share of cities’ energy consumption and future urban development must be carefully planned to meet the carbon neutrality goals that have been set worldwide. In this context, Urban Building Energy Modeling is the most promising technique for supporting the sustainable design of new urban areas, as well as driving energy and retrofitting policies at city scale. However, performing building energy simulations at urban scale is very computationally demanding. For this reason, in this work, it is presented a novel simplification algorithm named “shoeboxing” aimed at speeding up urban simulations while retaining accurate results at fine time and spatial scales. The proposed procedure, specifically developed to properly deal with buildings of complex shape, simplifies any building shape into a representative shoebox by means of a set of geometrical indicators. The urban context, buildings’ self-shadings and adjacencies are also incorporated in the simplified models. The goodness of the proposed method has been assessed by comparing the simulation outputs of detailed and simplified models at stand-alone and urban level. For this purpose, fictional buildings and districts have been developed in order to validate the approach as broadly as possible and highlight its strengths and possibilities for improvement. The procedure has been capable of reducing the simulation time up to almost 36 times, while limiting the hourly and annual prediction accuracy loss in any climatic condition considered. Given the outcomes obtained at building level, the algorithm has been used for three different applications to state the positive impact it could have in speeding-up also Building Energy Modeling and methods such as Multi-Objective Optimization, and in promoting the employment of Building Performance simulation in standard common practice. Thanks to these applications, the shoeboxing algorithm showed to be suitable also to employed for building-level analysis.