Demand response (DR) programs pay commercial buildings to reduce electricity demand when the grid is stressed. The payment can be substantial — $50–150/kW of committed curtailment capacity per year in most Midwest utility DR programs — and the obligation is typically modest: be available to curtail for 4–6 hours on 10–15 event days per year, with 2-hour advance notice.
On paper, this looks like free money for any building with HVAC flexibility. In practice, many buildings that enroll in DR programs fail to curtail effectively because their HVAC systems aren't pre-positioned to absorb a demand reduction without immediately degrading occupant comfort. Understanding the difference between a building that can deliver DR capacity and one that can't — and how thermal forecasting changes that calculation — is what this post is about.
How DR Program Mechanics Work
The structure of DR programs varies by utility and ISO, but the core mechanics are consistent. Buildings commit a certain amount of curtailable demand capacity — expressed in kW — and receive a capacity payment for that commitment regardless of whether events are actually called. When the utility or ISO calls an event, the building receives a dispatch signal (typically via email, automated API, or phone call) with the required curtailment amount and the event duration. The building reduces demand accordingly, and the utility meters verify performance against a calculated baseline.
The baseline calculation — how the utility determines what your demand "would have been" absent the DR event — is one of the trickier aspects of DR participation. Most utilities use an average of the 10 most recent non-event, non-weekend days as the baseline. If your building has been actively reducing demand through demand management before the DR event, your baseline may already be lower than your historical average, which reduces your apparent curtailment and can affect both performance payments and future capacity commitments. This interaction between baseline-day demand management and DR performance is worth understanding before optimizing aggressively for non-event demand reduction.
Performance penalties apply when a building fails to deliver committed curtailment. Penalty structures vary widely — some utilities assess per-kW penalties for shortfalls; others reduce future capacity payments; some allow a certain number of non-performance events per year before penalties apply. Knowing your program's penalty structure is essential for setting a realistic curtailment commitment level.
The HVAC Pre-Positioning Problem
HVAC is the largest controllable load in most commercial buildings, making it the natural target for DR curtailment. But HVAC curtailment has a fundamental constraint: the building's thermal state when the DR event starts determines how long curtailment can be sustained before comfort degrades.
A building at 72°F with full cooling running at 100% staging can absorb a 50% HVAC reduction for perhaps 20–40 minutes before space temperatures rise above the comfort limit, depending on outdoor temperature, occupancy, solar load, and building thermal mass. A building that has been pre-cooled to 69°F with substantial thermal mass buffer can sustain the same 50% reduction for 90–120 minutes before temperatures rise above comfort limits.
The difference — 30 minutes of comfortable curtailment versus 2 hours — is entirely determined by the building's thermal state at the start of the DR event. And that thermal state is determined by what the HVAC system did in the 2–4 hours before the event. A building that received a 2-hour DR dispatch signal and used those 2 hours to pre-cool has dramatically more curtailment capacity than a building that received the same signal and started curtailment from a warm state.
This is the core reason why thermal demand forecasting and DR program participation are natural complements. A 72-hour thermal forecast that identifies high-DR-risk days (extreme summer heat events, grid stress conditions) 2–3 days ahead gives the facilities team advance preparation time well beyond the 2-hour dispatch notice. The forecast essentially extends the effective lead time for DR pre-positioning.
Types of DR Products and Which Buildings Qualify
DR programs are not uniform — different programs have different performance requirements that suit different building types. The most common program types in Midwest utility territories:
Interruptible service programs: The building's service can be interrupted (load fully curtailed) during emergency events. These programs have high capacity payments but require the ability to shed all or nearly all HVAC load on short notice. They're appropriate for buildings with standby generation or buildings where the occupancy allows for full shutdown (data centers with backup power, certain industrial processes). Most occupied commercial office buildings aren't good candidates.
Load control programs: The utility installs a demand response switch on specific HVAC equipment (typically rooftop units or cooling towers) and cycles that equipment off during events. These programs work well for buildings with multiple discrete HVAC units where cycling 30–50% of units off doesn't cause immediate comfort problems. The building operator typically doesn't have direct control — the utility operates the cycling switch directly.
Price-responsive demand programs: Buildings receive real-time or day-ahead price signals and are expected to voluntarily reduce demand during high-price hours. No hard commitment, no penalty for non-performance, but also lower capacity payments. These programs are appropriate for buildings that want demand reduction flexibility without firm performance obligations.
Capacity programs (often ISO-administered): Buildings commit specific kW capacity to the forward capacity market operated by the regional ISO (MISO, SPP, PJM). These programs have the highest payments but also the most rigorous performance verification, including metered baseline comparison and potential penalties for non-performance during mandatory test events and actual emergency events.
We're not saying every building should pursue capacity market participation — the performance verification requirements and penalty exposure are real, and some buildings simply don't have the thermal flexibility to commit reliably. We're saying that buildings with strong thermal mass and forecast-based pre-positioning capability can participate confidently where others can't.
Aggregators and the Enrollment Process
Most commercial buildings access DR programs through demand response aggregators — third-party companies that enroll multiple buildings into DR programs, manage dispatch signals, and handle program compliance on behalf of building owners. Aggregators typically earn a percentage of the capacity and performance payments they generate, passing the remainder to the building owner. The split varies by program and aggregator, but 70/30 or 60/40 (building/aggregator) splits are common in Midwest DR programs.
The enrollment process for most programs requires a baseline data submission (12 months of 15-minute interval meter data), a technical assessment of which loads are curtailable and by how much, and a commitment to install automated curtailment controls or accept utility-installed load control equipment. For buildings that are already integrating a forecast system with BMS setpoint control, the incremental work to add DR dispatch signal response is modest — the BMS integration for forecast-based pre-conditioning and the BMS integration for DR curtailment response use the same infrastructure.
Enrollment timelines for capacity programs can be long — MISO capacity market forward auctions happen 3 years ahead of the commitment year, and enrollment deadlines are fixed. For a building that wants to participate in the next capacity auction, starting the enrollment and technical assessment process at least 6 months before the auction deadline is necessary. Price-responsive and load control programs have more flexible enrollment timelines and can typically be activated within 30–60 days of application.
Measuring and Verifying DR Performance
After a DR event, the utility or aggregator calculates whether the building delivered its committed curtailment. The measurement methodology is specified in the program rules and typically uses the established baseline calculation method (10-day average or similar) to determine what the building "would have" consumed, then compares actual consumption during the event to that baseline.
Instrumentation requirements vary — most programs require interval meter data at 15-minute resolution, which commercial meters already provide. Some capacity programs require additional sub-metering of specific curtailable loads to verify that the demand reduction came from DR-eligible equipment rather than from coincidental load reduction unrelated to the event.
Maintaining DR performance records — event dates, actual vs. committed curtailment, baseline calculations — is worth doing systematically regardless of whether the program requires it. This data tells you whether your pre-positioning strategy is working (buildings that pre-cool before events consistently outperform their committed curtailment) and provides documentation for any program performance disputes.