Presenter outlines locational-pricing effects of New York ISO—dynamic reserve proposal

A presenter at a New York ISO technical conference described a design for a "dynamic reserve" procurement mechanism that would be integrated into the ISO—security-constrained unit commitment for day-ahead and real-time markets and said the implementation phase is ongoing. The presentation listed co-authors Kush Khanna (Hitachi Energy), Matthew Musto (Replix Power), Kenjin Yupitjie and Suzanne Varghese (New York ISO).

The presenter said dynamic reserve explicitly models reserve deployment to resolve post-contingency transmission-flow limit violations. Under the proposal, local and systemwide reserve requirements would be represented by paired constraints: one that defines post-contingency flows for contingencies and a second that models reserve procurement needed to eliminate violations. The presenter noted a slide typo and corrected a sign in the procurement equation shown on the slides.

In the talk the presenter defined several locational marginal prices: LBMP for generator energy, LORP for operating reserve and a locational price for the load. He said each locational price is a linear combination of system-level constraint multipliers and scheduled prices multiplied by coefficients for the relevant quantity. One consequence, he said, is that a load's locational price can be viewed as a combined charge for both energy and reserve, and increases in load may raise both energy supply and local dynamic-reserve requirements.

Using a two-area illustrative example, the presenter described a nested subarea served through two transmission lines. In the example the system load was $1,500 and the subarea load $500; generators outside the subarea (S1—S3) and inside (P1—P2) submitted different energy and reserve bids. The presenter walked through contingency scenarios: loss of the largest generator (S3, 800 megawatts) that the example met with 10-minute reserve from S1 and P1; combined outages (for example S2+S3, 1,400 megawatts) allocated across available reserves; and a line-loss scenario in which a remaining line—s flow limit was increased from 200 to 300 megawatts.

The presenter said these constraints can produce non-intuitive results: superficially cheaper energy bids do not always lower total system cost if they require higher reserve procurement to meet reliability constraints. In the example, a small outside-load increase required fractional adjustments across generators and reserve schedules and raised total system cost by $32 in the scenario shown.

On congestion rent, the presenter defined congestion rent as the product of a transmission-constraint shadow price and its flow limit summed across transmission constraints. He said that, compared with a static-reserve market design (where congestion rent equals the ISO—net energy revenue), dynamic reserve introduces an additional term related to ISO payments to generators for reserve provision. "There—s an extra term," the presenter said, explaining that the example shows congestion rent under dynamic reserve can be lower than under static reserve, but he added he would not guarantee that will be true in every case.

During Q&A, Derek Wax (OCNC) asked why congestion rent under dynamic reserve appeared always lower in the presenter—s slides. The presenter reiterated the additional reserve-payment term and said the example demonstrates the effect but cautioned against treating it as a universal rule. An online participant asked whether the work had been published; the presenter answered that it had not.

The session closed after the Q&A. The presentation focused on modeling choices and example outcomes; no policy votes or formal decisions were taken at the session.