NYC Local Law 97

Buildings as power plants, run by AI.

BLD Power starts with Local Law 97 compliance software, then funds a battery and electrification package at 0 CAPEX. AI runs Power and Grid — peak shaving, dispatch, and bids onto the New York grid.

The problem

$268 a ton over the cap.

The emissions penalty is $268 per metric ton CO₂e over the applicable cap, every year. Same building, tighter 2030 cap, larger bill. Miss the May 1 filing and a separate $0.50/sf/month clock starts after the June 30 grace.

2024–2029 · Midtown East office — sample

$55,530

207.2 t over cap × $268

2030 cliff · same emissions

$350,437

$6,371,578 implied at 5.5% · commercial example

The stack

Compliance. Power. Grid.

Compliance is the file. AI operates Power and Grid: it forecasts load, peak-shaves the Con Edison bill, and bids stored capacity while holding a backup reserve. Building staff and the board stay in control and can set limits.

01 · Compliance

Annual subscription software

Exposure, pathway, filing calendar, and an owner PDF every year. About $0.03/GSF, floor $2,400.

  • Emissions versus the published occupancy cap
  • 2024–2029 versus the 2030 cliff on one page
  • Verified savings from AI dispatch land in the ledger
Product detail

02 · Power

Battery + electrification, 0 CAPEX

We fund the battery and electrification at 0 CAPEX. AI peak-shaves the Con Edison bill and coordinates electrified loads. The price is a share of those verified electric-bill savings and, when paired with electrification, a share of the verified LL97 penalty reduction shown in the filing.

  • 50/50 until capital is recovered, then 30 / 70 host
  • AI discharges against the monthly peak kW and expensive hours
  • Staff set limits; a backup reserve stays in the battery
Product detail

The grid

Earn from bills, fines, and the grid.

The AI decides when to bid stored capacity into a virtual power plant on Value Stack, Con Edison demand response, and aggregated wholesale bids, while protecting a backup reserve for the building. This site does not publish specific grid-revenue figures. A battery alone does not reduce LL97 reported emissions. Electrification does. The battery absorbs peak demand so that work is affordable.

How the AI works

AI runs Power and Grid.

It forecasts the building’s load, peak-shaves the Con Edison bill, bids stored capacity, and writes verified results into the Compliance ledger. Building staff and the board stay in control and can set limits.

Forecasts the building’s load

Interval meter data, weather, and occupancy patterns. The forecast is what the battery and electrified loads run against.

Peak-shaves the Con Edison bill

Con Edison bills demand on the monthly peak kW. The AI discharges to cut that peak and shifts load off expensive hours.

Charges, discharges, holds, and bids

It decides when to charge, discharge, or hold, and when to bid stored capacity into Value Stack, Con Edison demand response, and aggregated wholesale bids. A backup reserve stays in the battery for the building.

Coordinates loads and the filing

Heat pumps and electric hot water are scheduled so electrification does not spike the peak. Verified savings and emissions data feed the Compliance ledger and the LL97 filing.

How it works

Three steps to a binding offer.

01

48-hour feasibility

Coverage, occupancy coefficients, and $268/ton exposure for 2024–2029 and the 2030 cliff. The first pass is an estimate, including whether Power is in scope.

02

Site walk

We count Fire Code control areas on site. Indoor Li-ion is generally 400 kWh per control area. FDNY and DOB review the specific room.

03

Binding offer

Compliance can start as an annual subscription. Power, if offered, is a 0-CAPEX energy services agreement — 50/50 until recovery, then 30/70. Host CAPEX stays at zero. Once live, AI runs Power and Grid within limits the owner sets.

Proof

The $268-per-ton math.

Open the public demo

48-hour first pass

Forty-eight hours to a first pass.

Address, GSF, and a contact. We return exposure against $268/ton and whether Power is in scope.

Request feasibility