Con Edison Moves to Make Its Pricing Pilot Permanent, and the Rate Levels Are Not Yet Public

Con Edison has asked the New York Public Service Commission to stop new enrollment in its Innovative Pricing Pilot and to convert the pilot tariffs into permanent rate classes. The vehicle is Case 26-E-0476, which would amend the utility’s Electric Tariff P.S.C. No. 10. The comment period opened on August 5 and closed on October 5, 2026.

What the Commission’s public notice establishes is narrow. The filing would convert Riders Z and AA, along with Rate I, into new Smart Energy Plan rate classes. Smart Energy Plan bills price peak and off-peak demand separately on a per-kilowatt basis. The pilot has covered small-business accounts. The notice leaves the demand-charge levels and the covered customer classes to the filing itself, which sits in the Department of Public Service docket and was not reviewed for this article. Everything that follows about rate design is therefore structural rather than numerical.

The structure. A two-period demand charge bills the maximum kilowatt drawn in each window separately, rather than assessing one non-coincident maximum across the whole month. That distinction changes the behaviour a tariff rewards. Under a flat non-coincident demand charge, a customer is billed on the single highest interval of the billing period, whenever it occurs, and avoiding it requires controlling load around the clock. Under a peak and off-peak split, the expensive interval is confined to a defined window, and a controllable asset only has to defend that window.

It also changes the cost of recharging a battery. Under a flat per-kilowatt-hour delivery rate, pulling power overnight to refill a storage system costs energy and nothing else. Under a two-period demand structure, that same recharge sets a billable off-peak kilowatt. The economics of peak shaving under such a tariff are a spread between the peak and off-peak per-kilowatt rates, not the peak rate alone, and the gross peak-shaving figure overstates the net result by whatever the off-peak charge recovers.

One condition qualifies that. The off-peak charge attaches to the maximum kilowatt in the off-peak window, which means a site whose existing overnight load already exceeds the charging power adds nothing to its off-peak maximum by charging. Buildings with flat overnight baseloads, continuous refrigeration, data closets, around-the-clock ventilation, keep more of the spread. Buildings that go dark at night pay for the recharge in full. None of this can be priced without the filed tariff leaves.

What is unknown. The two variables that determine whether the conversion matters commercially are the demand-charge levels in the proposed classes and the customer classes those levels attach to. Neither is in the public notice. A concentrated peak-period charge with a low off-peak counterpart produces a wide spread and a strong signal; a narrow spread produces very little. The peak window definition is the third missing input, because the width of the peak period determines how much stored energy an asset needs to cover it. Until the filed leaves are read, the direction of the rate design is known and its magnitude is not.

The contrast with the repricing cases. Two other active US rate proceedings change what a kilowatt of demand costs, rather than who is billed in kilowatts. Florida Power and Light’s business rate sheets, effective October 1, 2026, carry a $12.70 per kilowatt base demand charge on GSD-1, the schedule covering the 25 to 499 kilowatt class, alongside a $33.71 monthly base charge and a 2.825 cent per kilowatt-hour base energy charge; the time-of-use variant prices on-peak demand at $11.90 per kilowatt. Those rates sit inside the four-year plan the Florida Public Service Commission approved on November 20, 2025, which added $945 million of annual revenue starting January 1, 2026. FPL’s published insert states that the PSC-approved tariff governs where the two disagree.

Arizona Public Service filed in June 2025 for a $579.52 million net base rate increase, 13.99 percent, and raised the request to $611.3 million, 14.75 percent, in its April 2026 rebuttal. The hearing phase has concluded. An administrative law judge’s recommended opinion and a commission vote remain. Consumer Reports filed an opposition letter on October 6 citing a 16 percent residential increase and 1,035 customer signatures, and the Arizona Attorney General has testified for roughly 3 percent. APS has told the commission it will seek separate treatment for large industrial users including data centres.

Both proceedings adjust price levels within existing rate structures. The New York filing proposes a structural change: moving a set of customers onto a billing basis that separates peak from off-peak kilowatts permanently, in place of a pilot.

Why the conversion is the interesting part. Demand charges in the United States have historically been a large-building instrument. FPL’s GSD-1, at a 25 kilowatt floor, is a representative shape: the per-kilowatt charge begins where loads are big enough to justify the metering and the customer attention. A pilot-to-permanent conversion in a utility territory as dense with small commercial accounts as New York City tests whether that instrument scales downward, and whether customers who have never had a reason to think about a 15-minute interval respond to being billed on one.

The rate-design question sitting underneath is whether a demand signal aimed at small accounts produces load-shifting or simply produces higher bills for customers without the equipment to respond. A battery, a thermostat schedule, or a controllable compressor converts a two-period charge into an avoidable cost. Without one, the charge is a fixed feature of the bill. That asymmetry is the standard objection to demand charges in small classes, and the filing in 26-E-0476 is where Con Edison’s answer to it would be found.

The timeline. The comment window closed on October 5. The Commission is now deciding whether the pilot structure becomes permanent, with no implementation date publicly set. Anyone modelling storage economics in the Con Edison footprint has a defined research task rather than a conclusion: retrieve the filed tariff leaves from the DPS docket, read the proposed per-kilowatt levels and the peak window definitions, and confirm which service classifications the new Smart Energy Plan classes cover. The direction of travel is on the public record. The numbers are not.

Sources