Two years after the U.S. Department of Energy’s Oak Ridge National Laboratory provided a model of every building in America, business partners are using the tool for tasks ranging from designing energy-efficient buildings and cities linking energy efficiency with real estate value and risk. International companies such as Google and SmithGroup share the benefits by making the resulting data publicly available. Since the buildings sector accounts for 40% of US energy consumption, increasing its efficiency is vital to national decarbonization goals.
Dozens of companies have requested data from ORNL’s Automatic Building Energy Modeling, or AutoBEM, software suite, project manager Joshua New said. He and his team developed AutoBEM using high-performance computing to process layers of imagery data with information about individual buildings: their size, use, construction materials, heating and cooling technologies and d other attributes.
“The unifying theme is to create a digital twin of our nation’s buildings,” New said. “We can simulate market-relevant ways to reduce energy consumption and offset with renewable sources.”
The software simulated the energy consumption of 123 million structures, representing 98% of the US building stock. New’s team is updating the software this year for even more detail and accuracy.
Google is using AutoBEM to enhance its free Environmental Insights Explorer tool, launched in 2018 to help cities around the world recognize greenhouse gas sources and reduction opportunities. Saleem Van Groenou, product manager for Environmental Insights Explorer, said Google wants to incorporate more accurate energy efficiency simulations for buildings.
“Oak Ridge has much deeper expertise than we do in building energy systems and modeling management and action,” Van Groenou said. “We can now help cities focus more on what changes need to be made, and then track the impact of those changes over time.”
Google is combining its treasure trove of building data with ORNL’s ability to scale energy models and develop machine learning algorithms, Van Groenou said.
Google is one of the big five companies that contribute data, labor time and equipment to AutoBEM partnerships. The participation of ORNL researchers is funded by the DOE.
Most AutoBEM users focus on existing buildings, but SmithGroup, an international architecture and engineering firm, takes the approach of building efficiency from the forefront.
“Our interest in AutoBEM and our collaboration with the lab stems from an urgent need to expand the work we are doing in response to climate change,” said Stet Sanborn, who oversees the ORNL collaboration for SmithGroup. “The number of buildings we need to touch and the rate at which we need to do it exceeds what one individual could do in their lifetime. And we have to do it in the next five years. He pointed out that AutoBEM’s ability to run 200,000 energy models in less than an hour is equivalent to the output of one employee working full time for 365 years.
For SmithGroup, ORNL simulated every possible combination of design parameters, building types, and U.S. climate zones. This information was used to train an artificial intelligence tool, essentially allowing the company to pre-simulate the energy impact of each design possibility for any building.
AutoBEM also incorporates climate change scenarios identified by the Intergovernmental Panel on Climate Change, modeled by ORNL’s Climate Change Science Institute. This feature caught the attention of partner LightBox, which offers a real estate information mapping and analysis platform.
“As a leader in the commercial real estate and geolocation industries, we’re delivering new datasets that are critical to understanding new risks,” said Zach Wade, LightBox’s vice president for data science. . “LightBox plans to use AutoBEM to model the long-term energy and operating costs of buildings and to help understand and report greenhouse gas emissions, providing valuable information to real estate investors, brokers, lenders and banks, appraisers, engineers and environmental consulting firms.”
LightBox and some other partners will in turn bring benefits to AutoBEM, providing datasets such as building footprints, interior details, property parcel boundaries and financial information to improve future simulations.
Additionally, partners such as SmithGroup and Google have committed to sharing datasets created using AutoBEM. “The whole market needs to change, and that’s where the relationship with AutoBEM becomes extremely important,” SmithGroup’s Sanborn said. “We don’t want to hold a secret sauce or limit everyone’s ability to drive efficiency in response to what is truly a climate emergency.”
Other AutoBEM partners include glass manufacturer Cardinal Glass Industries and Bentley Systems, an infrastructure engineering software company. Cardinal Glass, which supplies window manufacturers, uses the tool to understand the energy performance of various window types in different regions and climate scenarios compared to other efficiency improvements. Bentley Systems is exploring how to leverage city-scale digital twins and building energy models to optimize building design and decarbonization.
“The biggest surprise has been the degree of interest from companies and the range of modeling or data analysis they are asking for,” New said.
AutoBEM development, expansion, and collaborations are funded by the DOE, including the Office of Electricity, Office of Energy Efficiency and Renewable Energy Building Technologies, and the National Nuclear Security Administration. The research team leveraged supercomputing resources at Argonne National Laboratory.
UT-Battelle operates the Oak Ridge National Laboratory for the U.S. Department of Energy’s Office of Science. The largest supporter of basic physical science research in the United States, the Office of Science works to address some of the most pressing challenges of our time. For more information, please visit energy.gov/science.
Disclaimer: AAAS and EurekAlert! are not responsible for the accuracy of press releases posted on EurekAlert! by contributing institutions or for the use of any information through the EurekAlert system.