Search

In this paper we focus on a critical component of the city: its building stock, which holds much of its socio-economic activities. In our case, the lack of a comprehensive database about their features and its limitation to a surveyed subset lead us to adopt data-driven techniques to extend our knowledge to the near-city-scale. Neural networks and random forests are applied to identify the buildings' number of floors and construction periods' dependencies on a set of shape features: area, perimeter, and height along with the annual electricity consumption, relying a surveyed data in the city of Beirut. The predicted results are then compared with established scaling laws of urban forms, which constitutes a further consistency check and validation of our workflow. [ABSTRACT FROM AUTHOR]

Urban expansion, driven by population and economic growth, has been a major contributor to increased levels of energy consumption across the globe. Beirut, Lebanon's capital, is no exception in facing a surge in its demand for electricity as it expands. However, with frequent power outages, underpinning a largely problematic power sector, Beirut's demand for electricity is becoming a real hurdle that impedes the city's economic growth and development. This paper introduces a near-city-scale building energy model, BEirut Energy Model BEEM, which estimates the building stock's electricity consumption in two different districts in Beirut. The methodology uses rule-based expert data for an archetypal classification of the buildings based on their functions and periods of construction with their corresponding attributes including the number of floors, number of apartments, and bimonthly electricity consumption to generate a 3D model for 3630 buildings coupled to the hourly weather conditions and topographic map, which is then simulated in EnergyPlus. The predicted consumption of 2311 buildings is then calibrated with actual available metered data, to adapt the model to Beirut's occupancy and users' behaviors. Calibrated results are mapped to reveal the spatiotemporal distribution of energy peak demands which provide insights for future interventions. An analysis of the spatial distribution of electricity use demonstrates a spatial clustering that underlies urban energy demand which can be used for smart grid zoning. [ABSTRACT FROM AUTHOR]