Schools Cut Energy Bills With Guaranteed-Savings Performance Contracts, but Battery Storage Strains the Model
K-12 school districts in the United States face rising energy costs, and a growing number are absorbing them through a financing structure that requires no capital outlay. Louis Maltezos, co-president of the energy services firm Ameresco, told Utility Dive that more than 30 percent of the energy a typical district consumes is wasted, and that a district of 4,000 students can save roughly $160,000 a year through efficiency measures alone.
The pressure is rising. Maltezos points to utilities raising rates to fund grid buildout, a trend he ties in part to growing data center load. A public school district cannot vote its way out of a power bill, and every increase lands on an operating budget that is already committed elsewhere.
The vehicle. The mechanism Ameresco is selling is not a battery or a rooftop array. It is a contract. Under an energy-savings performance contract, the energy service company designs, finances, and builds the upgrade, then guarantees a level of savings and is repaid out of the verified reduction in the utility bill. The district commits no upfront capital. Just as important for a public institution, the structure allows negotiated procurement rather than the traditional bond-and-ballot cycle, which compresses the timeline from years to months.
The project. Ameresco’s new program at New York’s Mount Sinai School District shows the standard shape. The work, valued at more than $10 million, covers LED lighting, energy management system upgrades, high-efficiency transformers, and rooftop solar. It began in May 2026 and runs through the end of 2027.
The segment. Schools, hospitals, and municipal buildings are the original home of the performance contract, the cluster the industry has long called the MUSH market. They are also, with few exceptions, buildings without a yard. A suburban district can put panels on a gym roof, but a four-story urban school, a city hall, or a hospital wing has no outdoor pad for a battery cabinet and no appetite for one near an occupied building. If on-site storage has a path into these buildings, that path runs through the performance contract, because that is how these owners buy energy infrastructure at all.
The guarantee. An ESPC works because the savings can be promised. A high-efficiency transformer loses less energy than the one it replaced, every hour, regardless of weather or schedule. An LED draws a known fraction of the wattage it displaced. The energy service company can model the reduction, measure it, and stand behind a dollar figure for fifteen or twenty years because the physics do not vary. The savings are deterministic.
Demand-charge savings from a battery are not. A storage system reduces a commercial bill by discharging during the building’s monthly peak, shaving the maximum demand interval that sets the charge. The value depends on the battery being charged at the right moment, dispatching into the actual coincident peak, and doing so every billing period for the life of the contract. Miss the peak by one interval in one month and the guaranteed saving for that month does not materialize. The saving is conditional, and the condition is operational performance, not a fixed property of the hardware.
Measurement. Performance contracts are settled against a measurement-and-verification standard. The most widely used in the field, the International Performance Measurement and Verification Protocol, is general industry practice rather than a feature of the Ameresco projects described here; its methods range from isolating a single retrofit with dedicated metering to comparing whole-building consumption against a baseline. Efficiency measures fit these options cleanly. A storage asset that earns its return by reshaping a demand curve is harder to verify, because the counterfactual, what the peak would have been without the battery, has to be reconstructed every month from interval data, and because a warm spring or a schedule change can move the peak independent of how the battery performed.
That is the underwriting problem. An energy service company guaranteeing a battery’s demand-charge savings is not guaranteeing a hardware specification. It is guaranteeing dispatch reliability and, implicitly, that the utility’s rate structure will not change underneath the contract. Both are risks the lighting retrofit never carried.
Storage as a line item. This explains why storage tends to appear in these announcements as a phrase rather than a funded measure. The Utility Dive piece lists onsite solar, storage, and performance contracts together as tools districts can use, but the funded scope at Mount Sinai is lighting, controls, transformers, and solar. Solar generation has a deterministic, well-modeled output curve and a long record of bankable measurement precedent. Storage, valued on demand-charge avoidance, does not yet have the same standardized guarantee language. The binding constraint is contractual, not technological: whether a contractor will write a guaranteed number under the battery.
What could change it. The same infrastructure that complicates the guarantee could eventually underwrite it. Interval metering is now widespread, and the granular data that makes a demand-charge baseline hard to reconstruct by hand is exactly what a software model needs to verify one automatically. As dispatch optimization matures and a body of measured demand-charge performance accumulates, the savings become more defensible, and an energy service company can price the residual risk rather than avoid it. Standardized measurement methods for storage, if the industry converges on them, would do for batteries what existing protocols did for lighting.
Working in the other direction is rate design itself. The demand charges that make storage valuable are being restructured in many jurisdictions, from narrower peak windows to daily and interval-based measurement. Every such change raises the value of a well-dispatched battery and, at the same time, the risk of guaranteeing its savings over a twenty-year term. The economics improve and the guarantee grows harder in the same motion.
The institutions with the strongest physical case for indoor storage, the buildings with no room for a cabinet outside, are buying their energy upgrades through the one procurement channel built entirely around a promise that storage, for now, is the hardest measure to make.
Sources