Heating and cooling an office used to be really easy. People would show up in the morning and leave in the evening, Monday through Friday, every week of the year except holidays. Plus, back then, office managers (or anyone in the office, for that matter) only cared about keeping everyone comfortable. Hot and cold calls were the enemy, and energy costs were a rounding error.
Now the world has changed. Workers come and go in the office as they please, leaving entire parts of the office empty on certain days. Corporations want much more than just comfortable workers, they have sustainability mandates, carbon targets, and impact statements. Tenants now prioritize workspaces that can help them do the most with the least amount of resources.
All these changes make heating and cooling a building much harder. Building managers and engineers who run these complex systems have to figure out how to lower energy usage with variable occupancy while still keeping everyone from complaining about the temperature. The solution that many are turning to is a technology that can help their HVAC systems be much smarter. By using real-time occupancy sensors and machine learning, buildings can effectively find the optimal heating and cooling schedules, no matter how irregular the conditions are. Buildings can opt to stop heating or cooling areas with no or low occupancy, adjust how much outside air can be pushed in, or delay the start or stop times of their equipment based on what is happening at that very moment.
Most building teams already know about the usage of real-time data and advanced analytics, but the question still remains: how much can it actually save? Installing these systems can be costly and time-consuming, so many building owners doubt that they will create a substantial return on their investments. A recent study by Berkeley Labs and the Better Buildings Initiative is helping to quantify exactly how much can be saved.
The study was done on a 300,000-square-foot office constructed in 1965 situated just blocks from the White House in Washington, DC, owned by The Tower Companies. The 12-story building has two chillers totaling 850 tons, two primary chilled water pumps, four boilers, and three hot water pumps. The building was chosen for a few reasons, because it was older so it was a good representation of much of the country’s building stock, and because it already had building management software installed, so it wouldn’t need to be completely upgraded to be able to use the advanced smart building software that would be needed to make the system more efficient.
After an extensive RFI process, Prescriptive Data’s Nantum OS software was chosen. The study was done over a two-year time frame, which turned out to be a very unique period to do so. “The intention wasn’t to use an empty building, but COVID had other plans,” said Aaron Brondum, Vice President at Prescriptive Data. “The lower occupancy actually turned out to be a good thing because it taught us a lot about how a building can adapt to having fewer people in it and helped create a baseline for how much energy buildings use even with no one in them.”
As the machine learning software learned more about how the building would adapt to different conditions, it was able to recommend changes to how the systems were run. “We were able to do things floor by floor, so if one floor didn’t have anyone, we were able to experiment with running the fans less, or not at all,” Brondum said. “From there, engineers could take that knowledge and use it on the other floors.”
The study concluded that adapting the way that the heating and cooling systems ran saved 6 percent of the whole building’s operating costs. The largest impact came from how the software was able to adjust the start and stop times for the heating and cooling systems. “People create quite a bit of heat, so if you have a lot of them in one part of the building, you can often reduce the amount of external heat that you need to add to that area,” said Brondum.
Overall the study highlighted how smart buildings can make a real impact on the country’s carbon footprint. “Meeting the Biden Administration’s decarbonization goals will require significant reductions in the carbon footprint of existing buildings,” said scientist Jessica Granderson, Berkeley Lab’s interim division director of the Building Technology and Urban Systems Division. “New smart building technologies, such as those evaluated in the study, are needed to meet those goals.”
Saving energy is only the beginning. As building operating systems get more sophisticated, they can also adjust not just to reduce how much energy is used but to create a strategy for when it is used. While not part of this study, the software is now able to consider how much energy costs and how much renewable energy is being produced at the moment and adjust accordingly. “We can adjust when to heat and cool a building depending on how much renewable energy is available, which can reduce costs, lower our carbon footprint, and make our grid more resilient,” Brondum continued.
Adding these extra inputs creates a calculation that only a computer can do effectively. The days of “set it and forget it” heating and cooling schedules are over. Even though the additional considerations of energy cost savings and sustainability make it much harder to keep a building comfortable, smart building technology is helping empower building engineers to do exactly that. Installing smart building systems will always have an upfront cost, there is no way around that. But now, thanks to this recent study, building owners can see exactly how much they stand to save and realize that installing the new technology is well worth the cost.
MORE ABOUT OUR SPONSOR | NANTUM OS BY PRESCRIPTIVE DATA
Nantum OS by Prescriptive Data is an award-winning platform optimizing buildings’ operational performance while saving energy, reducing carbon emissions, and lowering costs without sacrificing occupant health or comfort. Combining historical data with predictive analysis and real time occupancy, Nantum OS enables buildings to hit their ESG goals via actionable insights.