Demand Response Programs

How demand response pays, and where the money actually comes from

Grid operators have two ways to keep supply and demand balanced on the worst afternoon of the year: build generation that runs for a handful of hours annually, or pay consumers to use less during those hours. The second is dramatically cheaper, and demand response is the market that formalises it. A commercial facility commits an amount of curtailable load, gets paid a capacity payment for standing ready, and gets paid again for the energy actually shed when an event is called.

The capacity payment is the part that matters for most commercial participants, because it arrives whether or not an event ever occurs. That also explains the penalty structure: you are being paid for a promise, so the programs measure whether the promise was kept, and under-delivering during a called event can wipe out a year of payments. Committing what you can genuinely shed rather than what you theoretically could is the whole discipline.

The programs differ substantially by market

In ERCOT, the primary commercial routes are the ancillary services markets and Emergency Response Service, with a shorter notification window and a market that has repeatedly demonstrated it will call on them. Texas also has the largest four-coincident-peak exposure in the country, which means the same curtailment capability serves double duty — earning demand response revenue and reducing the transmission charge that four summer intervals set for the following year. For most Houston and Dallas industrial sites, the second is worth more than the first.

In PJM — covering Pennsylvania, Ohio, New Jersey, Maryland, Illinois, Delaware and Washington DC — capacity clears through an auction held years ahead, and demand response bids into it as a resource alongside generation. The commitment is correspondingly firmer and the testing more rigorous. NYISO runs its own suite of programs with zonal value that differs sharply between New York City and upstate, and ISO-NE covers Massachusetts, Connecticut, New Hampshire, Rhode Island and Maine, where winter reliability rather than summer peak increasingly drives the value.

Demand response also works in territories with no retail supplier choice at all, which is worth stating plainly because it is where the misconception costs most. Austin, San Antonio and much of California have municipal or regulated utilities running their own demand-side programs. If you cannot switch suppliers, this and utility tariff optimization are usually the two largest levers you have.

Which loads are worth committing

• Cold storage and refrigerated warehousing. The best fit in commercial demand response. Thermal mass means compressors can be idled for an hour with no product risk, and pre-cooling before an event turns the building itself into storage.
• Manufacturing with schedulable processes. Batch operations, grinding, compressed air and non-critical auxiliary systems shed cleanly. Continuous processes in chemical plants and metals production usually cannot, though even there the auxiliary load is often larger than expected.
• Office buildings and retail chains. HVAC setpoint drift of two degrees for an hour is invisible to occupants and material in aggregate. For chains, the value is in the fleet: no single store qualifies, and two hundred stores together do.
• Distribution centers and agricultural operations. Charging, conveyance, irrigation pumping and grain drying are all schedulable, and the seasonal concentration of agricultural load often lines up with peak events.
• Sites with on-site generation. Standby generators and CHP allow load transfer rather than reduction, which lets facilities that cannot curtail — hospitals and some data centers — participate anyway, subject to air permit limits on run hours.

Getting the number right before you enrol

The work that decides whether demand response is worth doing happens before enrollment: an interval-data review to find how much load is genuinely sheddable and for how long, a written curtailment sequence naming specific equipment, and a comparison of what the same capability is worth as peak load management rather than as program revenue. Frequently the answer is both, and occasionally the answer is that the capacity payment is smaller than the coincident-peak charge the same curtailment would avoid, in which case enrolling would be the more expensive choice.

Demand response also interacts with your supply contract. Curtailment changes your load shape, and a tight bandwidth clause negotiated without accounting for it can turn program revenue into a swing penalty — a genuinely avoidable outcome that contract negotiation handles by writing the expected curtailment into the forecast. Efficiency work cuts the same load permanently, and a written energy strategy is what keeps the three from being optimized against each other.