New York City Ties Local Law 97 Battery Deductions to Measured Energy Rather Than Rated Capacity

The New York City Department of Buildings assumes a building battery is 85 percent efficient. That figure is fixed for calendar years 2024 through 2029, and it is load-bearing: it sits inside the method the city has published for converting stored electricity into a greenhouse gas emissions deduction under Local Law 97.

Buildings Bulletin 2025-014 is the document that turns a statutory allowance into arithmetic. Local Law 97 has permitted emissions deductions for clean distributed energy resources, energy storage among them. What was missing was a published way to calculate one.

The two roles. A Host is the site where the system physically sits. An Off-taker is a building consuming energy from a remote system. According to the analysis published by the law firm Hodgson Russ, a covered building may claim a deduction as either, or as both.

Two provisions do most of the work.

The bulletin permits an adjusted Total Emissions Spread methodology in place of the earlier Time of Use approach, and that framework carries the 85 percent round-trip efficiency assumption for 2024 through 2029.

Separately, for a Host claiming partial capacity, the deduction is calculated from measured consumed energy during the prior reporting year rather than from rated system capacity. Off-site storage does not disqualify a building, and deductions may be allocated by subscription and energy usage rather than by physical co-location.

The price of a ton. Local Law 97 penalizes emissions above a building’s cap at $268 per metric ton. That figure is what a deduction is ultimately worth: every ton subtracted from a covered building’s reported emissions is $268 of penalty exposure removed, provided the building is over its cap in the first place. A deduction mechanism has limited value to a building already clearing its limit. Its value concentrates in the buildings that do not.

The efficiency assumption has an end date. The 85 percent figure covers calendar years 2024 through 2029. It is fixed through the current compliance period and unspecified beyond it. Anyone underwriting a twenty-year asset against this mechanism is underwriting a few years of published assumption and the remainder against future rulemaking. The predictability that makes the deduction financeable today is bounded by a date already visible on the calendar.

Measured energy displaces nameplate. The design choice inside the partial-capacity rule has the longest reach. A deduction keyed to measured consumed energy in the prior reporting year is a deduction earned by cycling. A system sized generously but dispatched rarely earns in proportion to what it actually moved, not in proportion to what it could have moved.

That reverses the incentive structure of most storage procurement. Capacity is what a buyer purchases, what a warranty covers, and what a proposal leads with. Under the bulletin, capacity is an upper bound rather than a claim. A smaller unit cycling daily can out-earn a larger unit held in reserve, and the reporting-year lookback means the earning has to have already happened before it can be claimed. The deduction is retrospective, not projected.

It also imposes a metering obligation. A claim built on measured consumed energy requires measurement of a quality that survives review, which is an operational cost that a nameplate-based deduction would not have carried.

The subscription language opens a market. Permitting deductions to be allocated by subscription, and permitting a building to claim as an Off-taker without hosting hardware, means a covered building can hold Local Law 97 compliance value from a battery it does not own and cannot see. That is a familiar commercial shape, with one difference from a straightforward credit purchase: a storage subscription pays out on metered performance in a prior year. The compliance instrument carries delivery risk. A subscriber whose remote system underperforms discovers the shortfall in the reporting cycle, after the year in which it could have been corrected.

A San Diego meter asks a different question of the same hardware. SDG&E’s rate summary for medium commercial customers, effective June 1, 2026, prices Schedule AL-TOU-M secondary service at a $30.34 per kilowatt non-coincident demand charge plus a $38.46 per kilowatt summer maximum on-peak demand charge, a combined summer exposure of $68.80 per kilowatt. Footnote 3 of that summary states the non-coincident charge “shall be based on the higher of the Maximum Monthly Demand or 50% of the Maximum Annual Demand,” so one unmanaged annual peak sets a floor under twelve subsequent bills.

Set the two regimes side by side and the divergence is a dispatch problem. A demand charge rewards discharging when a building’s own load is highest. A peak-period emissions deduction rewards discharging when the grid is dirtiest. Those hours overlap often enough to justify a project on both, and not reliably enough for a single control strategy to maximize either. Optimizing against a carbon cap and optimizing against a kilowatt ratchet are separate objective functions that happen to share a battery.

The California ratchet sharpens the conflict. Because the non-coincident charge is set by the higher of monthly demand or half the annual maximum, a battery that stands down on a summer afternoon to chase an emissions-favorable discharge window risks establishing a peak that prices the following twelve bills. The demand-charge objective is unforgiving in a way the deduction is not: a missed deduction costs one year of subtraction, while a missed peak costs a year of floor.

What the bulletin changes. Building storage has been sold on rate arbitrage, with carbon compliance offered as an unpriced extra. New York has now published a methodology, an efficiency constant, and a reporting-year basis for it. Whether that produces claimed deductions at any scale will not be visible until the Department of Buildings reports on filings made under the bulletin, and the answer will depend less on the arithmetic than on how many covered buildings find themselves above their caps when the current compliance period closes.

Sources