Exploring data flows for building modelling at urban level - Case study of Gothenburg residential buildings

Abstract

Building sector was acknowledged as a priority to reduce energy consumption and GHG emissions due to its high potency in energy reduction. With the increasing complexity, especially with respect to the energy supply and demand, building stock needs an update of knowledge to understand the building stock performance better. At the city level, building stock model is a tool to assess the energy performance, therefore assisting the formation of a proper strategy to reduce the energy demand. This thesis aims to evaluate the information flow between different building models used on the urban scale and to explore mechanisms for continuous update of the modelling inputs. The data flow for building stock modelling in Gothenburg residential buildings is developed by integrating the dataset from EPC, Land survey and property map. The archetypes are constructed from historical architecture data and BETSI database to classify the individual building information data. The integrated dataset along with assigned archetype was screened and modelled on ECCABS (Energy, Carbon and Cost Assessment for Building Stocks). ECCABS is a building stock model with a bottom-up perspective that calculates energy use based on the physical properties of the buildings. Two inputs are modelled in ECCABS based on their assigned archetypes; 1) Typology based on historical architecture data and 2) Typology based on BETSI database. The modelling result on total energy delivered was then validated with the measured data taken from Energy Performance Certificate (EPC). The results show that the total energy demand calculated in ECCABS performed better in the input with BETSI archetype (R2 = 0,92) compared to historical data input archetype (R2 = 0,71).