How Buildings Can Get Smarter about Carbon

By Michael Samuelian | 12/20/22

Buildings in New York City are by far the largest contributors of greenhouse gasses (GHG) in the city, responsible for nearly 70% of the city’s emissions. To remedy this, in 2019, the NYC Council passed Local Law 97 (LL 97) which will regulate the amount of carbon that buildings can emit through the creation of a series of increasing yearly caps on carbon emissions. According to the U.S. Green Building Council, “The law (Local Law 97) is the most ambitious building emissions legislation enacted by any city in the world.” LL 97 is indeed ambitious, and it is well-intentioned, but it is also flawed. The main problem is how it measures carbon emissions, utilizing a very blunt (yearly) assessment of building performance, a benchmarking tool called ENERGY STAR. 

New data science tools can help building owners and the City get more informed about building performance. A team of engineers and data analysts from JB&B, researchers from Cornell Tech’s Urban Tech Hub and leading NYC building owners recently collaborated on the development Benchmark 8760, a new tool that explores how hourly data can be deployed to help measure and improve building performance.

What is Local Law 97?

The law will affect every building in New York City greater than 25,000 square feet and sets increasingly stringent limits on carbon emissions per square foot starting in 2024 and 2030. This regulation will cover more than 50,000 buildings and nearly 60 percent of the city’s building area.

LL97 requires a 40 percent citywide emissions reduction by 2030 from a 2005 baseline. And by 2030, the Law is expected to reduce carbon dioxide and carbon dioxide equivalents by approximately 6 million tons if owners comply. It is estimated by the NYC Department of Buildings that these caps would cut building’s carbon by more than 25% from today, which would be equivalent to the city of San Francisco’s entire citywide emissions.

These are ambitious and worthy goals, but we need more advanced tools to help us better understand the performance of buildings if we are going to aggressively and intelligently reduce our carbon emissions in New York City and beyond. In order to make this well-intentioned regulation more effective, we need more data. The project asked a relatively straightforward question:

What if occupancy data was measured relative to emissions data in office buildings?

Project Goals

The goal of this study was to use granular data to better and more effectively model the energy usage of commercial office buildings in NYC. Benchmarking is used to measure a building’s energy to similar buildings. ENERGY STAR is the industry standard for benchmarking commercial buildings, but it only measures buildings on an annual basis.  There are multiple variables that are lost when a building’s energy usage is reduced to a single yearly number, namely density, time of use and weather data. 

Time of use carbon is not considered in ENERGY STAR benchmarking. We can and should be benchmarking hourly building data and we have the tools of data science to do it. 

Today, ENERGY STAR cannot compare 2 buildings with differing density. And density is varying widely in a city where work from home has taken hold for nearly half of all office workers. And as we saw during covid, many buildings, while empty, still used nearly as much energy as if they were fully occupied.

Working closely with leading commercial building owners, the team identified 10 buildings to participate in this study that integrated hourly temp and occupancy data with emissions data. A new platform would be developed from the data called Benchmark 8760.

Collecting the Data

The team had to develop new standards for measuring occupancy levels since the sources of occupant data was so varied. Some buildings used lobby cameras with computer vision and some used turnstile data. In only one example did the team need to install new sensors to count building occupants. 

According to James Coleman, a data analyst from JB&B working on the project “The hardware wasn’t hard – the software development was challenging”. A new data standard had to be developed that could translate heterogenous raw data into more standardized forms.  It was also challenging to get data out of a building according to Coleman, since today most data lives “inside” the building due to cyber security concerns.

Lesson Learned

One of the key findings from this pilot project was that people counting is common in most buildings, but owners and building managers need to move toward a more standardization and consistent reporting systems if the data is going to be useful.

Like all projects that rely on data collection, protecting individuals privacy was a central concern. Rigorous and secure efforts were put in place to protect building occupants’ personal information. And in the case of this project, only the number of occupants in a building at any given time is of interest. Every effort should be made to deploy mechanisms without any personally identifiable information (PII) captured.

We recommend that the industry focus on developing a new benchmarking platform that integrates hourly data and building performance into emissions calculations before Local Law 97 goes into effect. This new platform would be like ENERGY STAR portfolio manager, but it would have the capability of integrating occupancy, time of day and weather data. Ultimately, we can get much more predictive with regard to energy usage, insight that will be crucial for a greener grid powered largely by renewables. 

This project was developed as open source and data is made public for research purposes. The 8760 team is developing public facing data analytics tools to showcase the public data. We also believe that once building owners move toward getting more intelligent with regard to hourly building occupancy, whole new research areas, analytics and products will get developed. The commercial office sector is the lifeblood of the New York City economy, now is the time to get more informed about our less carbon-dependent future. 

Greening the Grid

New York State has ambitious renewable energy goals. It is planned that 80% of the state’s electric grid comes from renewables by 2050. As we transition to a greener grid, which is less dependent on fossil fuels and more dependent on solar and wind power, our time of day usage of power becomes more central to the resiliency of the grid. Load shifting will be crucially important as we move towards decarbonizing the grid. This problem will only get greater as we move towards full electrification and dramatically increasing demand on the grid.

In NYC it is estimated that nearly 70% of carbon emissions are coming from buildings. Buildings and buildings systems must get smarter with regard to energy usage. The 8760 platform is just the beginning of our vision to develop a prototype that measures energy and carbon at a more granular level than simply annual data. 

In the all-electric future that many predict, we will need buildings that work more closely with the grid – taking a closer look at when energy is being used. 

There are 8760 hours in a year and there is no reason why we cannot gain a better understanding of how our buildings are operating and generating GHG emissions for each of those hours. We need to get smarter and more sensitive to how and when we use energy and Benchmark 8670 is an important step towards getting stakeholders to work together to decarbonize our building systems and do our part to enable the transition to a less carbon dependent future. We cannot have a greener future if we don’t get smarter about how, where and when we use energy.