Pacific Gas & Electric Company (PG&E), one of the largest dual-fuel utilities in the US, serves over 5 million electric customers and 4.4 million natural gas customers.[1] One of the programs PG&E offers to help encourage customers to reduce summer electricity usage and the load on the power grid in California is their SmartRate™ Critical Peak Pricing program.

SmartRate™ is a rate plan that offers PG&E customers lower electric rates from June 1st through September 30th, except during the hours of 2:00 p.m. to 7:00 p.m. on event days (referred to by PG&E as SmartDays™). SmartRate™ is designed to help reduce load on the electric grid on days when resources are constrained. Typically, PG&E calls SmartDays™ when there is high demand and exceptionally hot weather. This type of resource has become increasingly more relevant as California faces increasing grid demand, particularly due to increased temperatures and climate events.

Opinion Dynamics completed a load and bill impact evaluation of PG&E’s SmartRate™ Critical Peak Pricing program to help PG&E forecast future program performance. Of particular interest to PG&E was understanding which customer segments drive impacts as well as the distributional impacts across customers to inform program delivery and support customer targeting.

Opinion Dynamics employed a quasi-experimental design to estimate load impacts and identify matched comparison groups with similar characteristics to event days as a baseline. We generated load impact ex post estimates for each hour of each of the fourteen SmartRate™ events, as well as ten-year ex ante forecast.

Some of the most informative results that help to maximize program cost-effectiveness through targeted program marketing comes from knowing which segments of the population are most responsive to SmartRate™ price signals, as well as when and for whom to call events. Consequently, Opinion Dynamics provided PG&E load impact data in aggregate as well as by customer segments (e.g., technology type, low-income, local capacity area).

We based our ex-ante impacts for future years on PG&E’s forecasts of program enrollment, development of weather-sensitive reference loads, and projections of per- participant load impacts that are estimated based on ex-post load impacts . Our analysis uncovered steadily declining load impacts and customer enrollment in the program. In response to these findings, Opinion Dynamics incorporated a series of additional analyses within the evaluation to contextualize the declining trend in load impacts and identify the drivers of declining load impacts and their implications on future load impacts.

Upon completion of our evaluation, Opinion Dynamics provided PG&E with ex-ante load impact projections for the ten years following the evaluated year, supporting program planning efforts. Additionally, in our analysis we found that participants changed their energy consumption behavior on both event days and non-event days. In other words, the rate isdesigned to provide a price signal to use more energy on non-event days and only to reduce energy usage during peak hours on event days. The price signal to reduce energy during event hours was strong enough, however, that some participants lowered their consumption during peak hours on all days—responding more as if they were on a time-of-use rate rather than a critical peak pricing rate. The size of the spillover varied by customer segment and while not large on a per-participant basis, resulted in potentially 2 MW of additional load impacts in aggregate.

Opinion Dynamics also provided PG&E with results from the multi-level modeling analysis as well as findings that newer enrollees contribute smaller load impacts compared to customers who enrolled earlier. We developed customer education, behavioral messaging, and targeting strategies enabling PG&E to leverage this research for future program years.


[1] Pacific Gas and Electric Company. “Company Profile.” Last updated: May 2, 2021. https://www.pge.com/en_US/about-pge/company-information/profile/profile.page