Using artificial intelligence models and degree-days method to estimate the heat consumption evolution of a building stock until 2050: A case study in a temperate climate of the Northern part of Europe

Abstract

The energy use in buildings is highly influenced by outdoor temperature changes. In the contest of nowadays climate change, its impact on the energy sector is important and needs to be assessed. This study investigates how the heat consumption (HC) of the existing regional building stock, located in a temperate climate in the Northern part of Europe (Belgium), will be influenced by future climate changes. First, the Seasonal Auto-Regressive Integrated Moving Average (SARIMA), Long Short-Term Memory (LSTM) and Gated Recurrent Unit (GRU) models are used to predict the temperature until 2050 from historical data. Second, the UK Met Office equations are applied for computing the heating degree-days (HDD) considering the base temperature of 15°C. Finally, the HC of this building stock is projected until 2050 using the degree-days (DD) method. The decrease in HDD is about −11.76% from 2012 to 2050. The HC reduction, calculated at the regional scale, is reaching −8.82 %, −10.00%, and −11.26% for respectively residential, tertiary, and industrial buildings. The calculated HC is mapped on municipality, urban region, and province scales. The produced maps will help decision-makers set up efficient energy management strategies. The used methods can be replicated in other regions with the same kind of data.

Keywords

• AI modelling
• ML and DL models
• Heating degree-days estimation
• Heat consumption forecasting
• City scale