Restaurants have long been scored and graded by health departments, and now, this concept has been extended to score and grade buildings on energy efficiency.
New York was the first city in the country to pass legislation requiring buildings (commercial and residential) that are over 25,000 square feet to post energy grades at public entrances. The law will go into effect in 2020.
Currently, grading consists of calculating an ENERGY STAR score (based on guidelines from the U.S. Environmental Protection Agency). However, researchers Sokratis Papadopoulos and Constantine E. Kontokosta published a paper, “Grading Buildings on Energy Performance Using City Benchmarking Data,” in the Journal of Applied Energy. The study proposes a GREEN grading system that incorporates machine learning to improve accuracy.
Papadopoulus is a Ph.D candidate and graduate research assistant in the Department of Civil and Urban Engineering and Center for Urban Science + Progress at New York University. Constantine Kontokosta is an associate professor of Urban Science and Planning; and Director of the Urban Intelligence Lab in the Department of Civil and Urban Engineering at New York University.
What the GREEN process entails
“The GREEN system assigns grades to buildings based on their energy performance compared to buildings with similar characteristics,” says Papadopoulos. “Essentially, we train a machine learning model to predict a building’s energy use intensity using various physical, occupancy, and qualitative predictors that are known to influence energy use,” he says. “Then, we grade it from A-D, based on the deviation between the actual energy use and the model-predicted one.”
Examples of grading buildings
“Think about a property with a higher unit density compared to the average building in New York City, and let’s say it consumes 150 kBtu/sq.ft. annually,” Kontokosta says. “When compared to the entire building stock, its energy consumption is high, and one might assume the building is not energy efficient. However, given its more intense occupancy, it is expected that it would consume more energy.”
Kontokosta says their model takes this, and other building attributes, into account, and compares this property with others that have similar unit densities. “If high density buildings consume on average 180 kBtu/sq.ft., for example, our building under the GREEN grading system would receive a better grade than when simply compared to the entire sample.”
What types of buildings are graded?
“In our paper, we validated our method on the large residential building stock of New York City (i.e. buildings with area larger than 50,000 sq.ft.), but our method is generalizable to any building typology as long as there are appropriate data to train our models,” Papadopoulos explains. Some of the types of properties in New York City can be graded are offices, hospitals, hotels, K-12 schools, resident homes and dormitories, supermarkets and grocery stores and worship facilities.
Advantages of grading buildings on sustainability
“Grading buildings on energy performance – in a rigorous and unbiased way – helps to reduce information asymmetries and to transform the energy efficiency market,” Kontokosta says. “Owners can be aware of their building’s performance compared to peers, which we hope will nudge them to become more efficient,” he says. “From the end users’ perspective, grades add transparency to the energy costs associated with a property and renters/buyers are now able to value energy efficiency in their real estate decisions.”
So, for example, if companies are looking for space to rent, a building with a high energy grade would be more attractive than one with a lower score. In a city like New York, which has some of the largest electricity costs in the country, these grades are especially attractive. And in such a competitive market, building owners will be incentivized to work on improving their grades.
Terri Williams is a freelance journalist with bylines at The Economist, USA Today, Yahoo, the Houston Chronicle, and U.S. News & World Report. Connect with her on Twitter or LinkedIn.