Researchers at Cornell University have developed a new software tool that can simulate the energy use of buildings in a small city within minutes on a standard laptop computer. By running simulations, the tool helps policymakers prioritize the most cost-effective decarbonization strategies.
Using the city of Ithaca, New York, as an example, the urban building energy model quickly mapped more than 5,000 residential and commercial buildings along with their baseline energy consumption. The model simulated investments in home weatherization, electric heat pumps, and rooftop solar panels, while also factoring in financial incentives. The insights generated will inform the city’s efforts to achieve carbon neutrality by 2030.
The researchers noted that the tool's automated workflow, accessibility, and accuracy — without requiring advanced computing power — make it extremely valuable for small cities that lack decarbonization resources and expertise. The new model currently supports counties surrounding Ithaca and can be further scaled up to larger cities or even entire states.
Timur Dogan, Associate Professor in the Department of Architecture and Design Technology at the College of Architecture, Art and Planning and first author of the paper "Bottom-up Urban Building Energy Modeling to Assess Electrification Measures for Space Heating" published in the Journal of Building Performance Simulation, stated: "We are very excited that this technology is scalable and efficient enough for even a state to map out energy consumption and retrofit potential — whether from a carbon, financial, or efficiency perspective. From a small city to a county or a state, this can really influence policy and how priorities are set."
According to data from the United Nations Environment Programme, the building sector accounts for more than 37% of global carbon dioxide emissions. "Digital twins," including urban building energy models, are key to enabling data-driven decarbonization decisions. However, their complexity and cost have limited their application and scalability. For example, Oak Ridge National Laboratory's "Simulating America" initiative uses supercomputers to process massive amounts of data to build models or test scenarios, which is impractical for most cities.
Dogan noted that his lab's models achieve similar results in a very short time, sufficient to guide planning, and can explore all feasible scenarios through iterative simulations. This is accomplished through a series of small innovations, including fast physics-based simulation engines using low-order energy models and machine learning techniques to describe buildings and infer their unknown properties.
These models integrate publicly available data such as geospatial maps, tax records, building permits, and census data, along with energy usage data from local utility companies. For Ithaca, this data was provided by New York State Electric & Gas through an agreement with Avangrid.
The Ithaca case study revealed several "compelling and sometimes counterintuitive insights" relevant to cities of similar size. For example, replacing gas furnaces with heat pumps would increase operating energy costs for many buildings, indicating that this measure should be combined with home weatherization and rooftop solar panels to make the transition economically attractive. Although large commercial buildings may appear to be the primary focus, when financial incentives are considered, the priority shifts to retrofitting multifamily residential buildings as the most cost-effective option.
Dogan said: "These models allow you to flag properties that are worth investigating further. Out of 5,000 buildings, we were able to identify 'these 100 buildings are clearly the ones that should be prioritized.'" Rebecca Evans, Ithaca's Director of Sustainability, said the city welcomes continued collaboration with Dogan's lab. She noted: "The technology Timur's team is building is innovative and scalable, and has already been applied in large city and county projects. Urban building energy models have the potential to dramatically reduce the capital required to identify buildings suitable for electrification, integrate them into investment portfolios, and create attractive investment opportunities."
Dogan also stated that while America's largest cities may have dedicated staff and budgets for decarbonization challenges and analysis, more accessible and scalable models — such as the one developed by his team — have the potential to help many smaller cities that would otherwise act blindly. "Together, these cities will have a huge impact, and we hope to provide them with information they can use to make the right decisions."
