Web-Based Visualization and Prediction of Urban Energy Use from Building Benchmarking Data

Published in Bloomberg Data for Good Exchange 2015, 2015

Recommended citation: Kontokosta, C., Tull, C., Marulli, D., Pingerra, R., & Yaqub, M. (2015) Web-Based Visualization and Prediction of Urban Energy Use from Building Benchmarking Data. Presented at Bloomberg Data for Good Exchange 2015 New York, NY. https://www.researchgate.net/profile/Christopher_Tull/publication/282781435_Web-Based_Visualization_and_Prediction_of_Urban_Energy_Use_from_Building_Benchmarking_Data/links/561c6ab808ae6d17308b1843.pdf

Abstract

New York City has pledged to reduce its greenhouse gas emissions by 80 percent by the year 2050, and 60 percent of these reductions will need to come from the buildings sector. Unfortunately, the wide and rapid adoption of energy conservation measures is hindered by the lack of granular, comprehensive, and easily accessible energy usage data for buildings. To increase the volume of available building energy data, New York City’s Local Law 84 requires large buildings to disclose their energy consumption.

This paper details two ongoing projects to increase both the availability and comprehensiveness of building energy data. The first is a web-based visualization tool which allows users to understand patterns of energy consumption in individual buildings and across the city. The second project attempts to generalize from disclosure data by creating a predictive model of annual energy consumption for each building in the city. Building-level predictions are then validated against aggregate zip code-level data from local utilities.