A machine-learning model for the prediction of aggregated building heating demand from pan-European land-use maps

Abstract

Aggregated building energy demand is a useful indicator for urban energy planning. It can be used by planners and decision-makers to identify clusters of high energy demand in a given urban area and efficiently plan, for example, district heating networks. Various data sources exist at the pan-European level describing land use and built areas. Combined with statistical data, such maps have been used in previous research for estimating building energy performance aggregated at the hectare level, using engineering assumptions. In this paper, we show that large-scale land-use maps alone can be used for predicting annual building energy demand with an accuracy comparable to the one of previous engineering models. We hence present a preliminary method based on Convolutional Neural Networks at different spatial resolutions. The resulting model was trained and tested in an area of about 170 km1 in Geneva (Switzerland) using a local annual heating demand dataset comprising 16239 buildings. On a 300-m aggregation tile, the obtained mean error (14.3%) is significantly reduced compared to the one of a simple linear model (37.2%). Using solely land-use data, we also achieve similar results for a 100-m tile as those of an engineering model from the literature.

Potential for district heating networks from waste heat: an assessment tool and its application to sewage treatment plants in the Canton of Zurich

“…This model extends the model of the maximum heat power presented in [2], by calculating hourly demand values based on the difference of temperature between indoors and outdoors. The inputs for the simulations are the building construction period and usage, as provided by the Swiss Register of Buildings and Dwellings and retrieved via DHgeN [2]. By neglecting internal gains, solar gains and thermal inertia this should be considered as a conservative value of building energy need.…”

“…All substations within r ′ are connected in a graph created using DHgeN [2]. To this end, we consider the street network in the area as well as a 25x25 m grid.…”

“…the euclidean distance between two points • the graph distance, i.e., the length of the shortest path in the graph between two points • the weighted graph distance, i.e. the graph distance multiplied by a weight which is dependent on the type of edge (either from a street segment or a grid edge) as computed by DHgeN [2].…”

“…The substations are then connected to the heating station following their cluster rank and, subsequently, by graph distance until their gross cumulative annual heat demand reaches the available heat potential of the cluster heat source. The final network is then created based on the selected substations using DHgeN [2] so as to plan the actual network layout (Figure 2).…”

“…Li et al [1] developed a method for the optimal spatial location of different heat sources in the context of geothermal potential. Previous work focused on finding the optimal network layout to connect a pre-defined number of consumers given urban constraints [2].…”