Exploring the influence of urban context on building energy retrofit performance: A hybrid simulation and data-driven approach

Cities are an integral part to meeting the world's sustainable energy goals. Specifically, retrofits have been implemented to improve energy efficiency and reduce carbon emissions in the buildings sector. Recent simulation, reduced-order, and data-driven approaches have been used to predict the current energy consumption of urban buildings. However, these efforts are limited in their ability to evaluate potential impacts of future retrofits as they are unable to account for inter-building energy interactions that can influence urban building energy performance. To overcome these limitations, we extend a previously developed hybrid data-driven urban energy simulation (DUE-S) model that leverages building energy simulations and deep learning models by now predicting the impact of various building energy retrofits on multiple spatiotemporal scales across a city. We evaluate this approach on a case study of 29 densely co-located buildings in downtown Sacramento, California, USA. Our results indicate that accounting for urban context can compound the impact of retrofits on individual buildings by up to 7.4% as they also influence the electricity use of their surroundings. Finally, we show how DUE-S can provide insights on how to select buildings for retrofit that captures a potential compounding energy savings effect. We develop a greedy optimization algorithm that minimizes the number of required retrofits needed to achieve maximal energy savings across an urban study area. As a result, this work underscores how a flexible urban energy prediction model such as DUE-S can help inform energy-related decisions for a variety of urban-minded stakeholders including architects, engineers, planners, and policymakers.

Automated summary

Cities are crucial for achieving global sustainable energy goals, with recent efforts focusing on retrofitting buildings to improve energy efficiency. However, current methods are limited in evaluating future impacts. To address this, researchers extended a hybrid data-driven urban energy simulation model and developed an optimization algorithm to maximize energy savings through retrofitting buildings in urban areas.