This AI Software Saved New Yorkers $5 Million in Heating Costs



Emissions from existing buildings are responsible for about 70% of New York City’s climate emissions, with space heating as the dominant source. Yet most of the city’s multifamily buildings still rely on central steam boilers that cycle on and off when the outdoor temperature drops below a certain threshold, regardless of indoor conditions. The result is a system that leaves many residents sweltering in the dead of winter, wasting fuel and money while releasing unnecessary greenhouse gases.

Completely overhauling and modernizing a central boiler system — many of which date to the early 1900s — and installing a building-scale heat pump could address many of these issues. But that’s an expensive, complex, and disruptive endeavor that many building owners either can’t afford or simply don’t want to undertake. And while heat pump startups such as Quilt and Gradient are making inroads in single-family homes and individual apartment units respectively, neither is working to optimize the operations of existing steam boilers, which remain the dominant heating source for New York’s apartment stock.

That’s where Entech, a 30-year-old building energy management company, comes in. The company’s platform has long used indoor sensors to monitor the performance of central boilers and help them run more efficiently. Last year, however, the company revamped its software to incorporate artificial intelligence. The new system, called S2, autonomously monitors 20-plus sensors installed throughout the buildings where it operates, adjusting heating cycles with greater precision while continuously tracking the overall health and performance of boiler room operations.

On Wednesday, the company announced the results from the S2’s first year of operations: Across 401 New York City apartment buildings, the platform slashed emissions by nearly 25%, avoiding more than 16,000 metric tons of carbon pollution and generating over $5 million in savings for property owners.

Previous iterations of the company’s tech relied on preset rules such as, “When it’s 55 degrees [Fahrenheit], you need a shorter cycle, and when it’s 20 degrees, you need a longer cycle,” Heather Zoberman, Entech’s director of product development, explained to me. Those settings dictated how long a boiler turned on and how long it stayed off. With AI, however, the company can measure how quickly individual units are actually heating up and adjust performance in real-time.

For a company that spent decades focused on incremental improvements to boiler operations, it’s a meaningful shift. “Now we have the ability to do flame modulation — so a higher flame, a lower flame— based on the load, based on the building temperatures,” Zoberman told me. The same level of granular control applies to the fans and pumps that move heat through the building, too. “A little bit slower fan, a little bit lower flame is really where you get those savings that add up,” she said. According to Entech, those savings are typically passed onto the residents, with the average tenant saving roughly $200 on heating costs last year.

While building owners are happy to see these savings too, many are turning to Entech primarily to comply with the New York City Council’s Local Law 97, which requires buildings larger than 25,000 square feet to cut emissions 40% by 2030 compared to 2005 levels, and reach net zero emissions by 2050.

The nonprofit housing developer and operator Breaking Ground, for example, builds supportive housing for low-income and formerly homeless New Yorkers, and has been doing so for decades. It adopted Entech’s new boiler control system just six months ago to comply with the emissions law. While Breaking Ground’s deputy VP of facility operations, Lorenzo Torres, didn’t have exact savings figures on hand, he said the system has saved the organization “a lot of money,” largely by enabling staff to remotely identify equipment issues such as leaks and temperature fluctuations without having to send anyone to the building and before they develop into expensive headaches.

“We do have a work order system, but data is only as true as the person that’s entering the data,” Torres explained. Thus if a tenant misidentifies an issue or fails to file a work order in the first place, Breaking Ground might assume everything is running efficiently. By contrast, “the S2 controller actually is able to, with conviction, let us know that there is an issue with the boiler,” he said.

What Entech’s system still can’t do is solve the problem of unit-level temperature variation. Factors such as floor level, window exposure, and radiator placement mean some apartments will naturally run hotter or colder than others. But because Entech primarily operates in apartment complexes with central boilers, it can still only make adjustments at the building level Because of this, its system could be a complement to something like a smart radiator, which can control how much heat each apartment receives.

Now, Entech is looking to expand beyond New York. Boston is a natural next market, Zoberman told me, given its stringent building emissions requirements. Chicago is also on the company’s radar, thanks in part to incentives from the natural gas utility People’s Gas, which can help offset the cost of energy efficiency upgrades. The company’s ambitions extend beyond just geographic expansion, however — it’s also broadening its platform to monitor and optimize central cooling systems and other electrified technologies such as heat pumps and mini splits.

It looks like it should have plenty of room to run. Additional jurisdictions from Washington D.C. to St. Louis are increasingly adopting hard caps on building emissions, while dozens more now require annual energy-use reporting — often a first step towards more stringent regulation.

Popular

Latest News