Perform Event Stage Impact Measurements (M&V)

Measurement & Verification (M&V) Impact Measurement quantifies the actual hourly effect of a program event—typically the change in energy use or demand—using a method that is repeatable, auditable, and statistically robust. This feature applies a Difference-in-Differences (DiD) regression technique to estimate program-level reductions for individual events; most commonly used for direct device control programs where interval meter data is available for enrolled participants. The DiD Regression technique analyzes the change in electricity consumption between a group of customers participating in a demand response program (referred to as the Treatment group) compared to a similar group not participating (referred to as the Control group), both before (pre-treatment period) and after (treatment period) the demand response event, allowing the outcome to isolate the causal effect of the program on electricity usage by accounting for any pre-existing trends in consumption patterns across the groups.

This process leverages existing entity grouping to automatically assign participating devices to a Treatment Group and define a preferred Control Group by creating a “dummy” program made up of non-participating customers. Several new objects have been introduced including a new Measuring Component set, a new Aggregator measuring component type, a M&V Usage Subscription, and an M&V Usage Transaction entity to allow associating the treatment group to the control groups, capturing the details for triggering the M&V Impact estimation process and storing the calculated values. You can also configure the number of days after the current business date post which the M&V Impact estimation process will be triggered. In addition, a new maintenance object has been introduced to capture the summary of event data at the service point level for program participants.

As the first step, the application determines the relevant program and identifies qualifying program events within a configured date range. For each event, it retrieves constituent data for both the pre-treatment and treatment periods, loads it into a staging area, and filters out records based on criteria such as event outcome and the number of participating devices for the same service point. Once the dataset is finalized, the process aggregates data across entity groups and applies DiD regression to compute the “difference of differences”: (1) the Treatment Group’s change from pre-treatment to treatment, and (2) the difference between the Treatment and Control Groups during the treatment period. The regression results are then recorded on an M&V transaction for each Treatment/Control group pairing.

After the DiD analysis is complete, separate batch processes produce the following outputs: (1) an Event Impact Extract for Program Participants, which calculates an average impact for each Treatment Group and applies it to all devices in that group behind a metered service point, (2) another Event Impact Extract for the control group service points’ consumption interval data for the same treatment/event period to serve as the comparison baseline and (3) a Statistics Extract, which generates a Regression and Coefficient Data file containing the regression-specific details for each Treatment/Control group combination.

This provides a reliable, defensible way to quantify the true hourly load reduction delivered by program events by comparing participants to a well-matched control group, minimizing distortion from weather and broader system conditions. These insights support better forecasting and target setting, smarter dispatch and enrollment strategies, and earlier identification of under-performing segments or operational issues—ultimately improving program effectiveness, cost-efficiency, and the quality of reporting to internal leadership and external stakeholders.

Steps to enable and configure

To enable and configure this feature, refer to the Configuring M&V Options section in the Administrative User Guide for more information.