Selected publications available in Adobe Acrobat
format.
If, after reviewing the report, you need additional information on a study, you can check the
EM&V Contacts list and contact the person responsible for
the area associated with the study.
Publication Type: Energy Efficiency; Demand Response; Self Generation
Category: Impact Evaluation; Program Design; Evaluability Study
Author: Demand Side Analytics
Sector: Residential
Implementer: PG&E
Title:
Early M&V Report for Program Year 2023 Energy Efficiency Summer Reliability Program
Abstract: This Early M&V report estimates savings from PG&E’s 2023 Energy Efficiency Summer Reliability Program (EESRP) and documents an accuracy assessment of Population Normalized Meter-based Energy Consumption (Population NMEC) methods to estimate savings (no savings were claimed for EESRP). The analysis compared hourly whole-building data and 15-minute battery discharge data and clarifies impacts of battery discharge on energy use beyond weather, season, and time of day. EESRP's objective was reducing Peak (5 pm-9 pm) and Net Peak (7 pm-9pm) demands Aug-Oct 2023, 7 days a week. PG&E designed and administered EESRP per CPUC Decision 21-12-011 (12-8-2021) which orders the IOUs to prepare for extreme weather in the summers of 2022 and 2023 in response to an emergency proclamation issued by the governor (7-30-2021) and directs the CPUC to “free up energy supply to meet demand during extreme heat events and wildfires" and "to expedite deployment of clean energy resources.” The proclamation directed energy agencies to act to achieve energy stability.
Marketed as Peak Power Rewards, PG&E collaborated with solar and battery provider Sunrun to launch EESRP as part of a PG&E strategic plan to optimize energy usage during high demand hours. Key findings include:
• 8,483 customers enrolled. Energy savings: 10,349 MWh and 10,563 MWh as measured by battery discharge data and hourly interval data respectively.
• Average estimated per customer aggregated peak reduction impact: 1.68 kW (battery discharge data) and 1.69 kW (interval data).
• SolarEdge and Delta inverter sites showed lower peak savings (0.70 kW) as compared to Tesla inverter sites (3.96 kW) calculated using hourly meter data. This difference results from SolarEdge batteries having experienced a pre-period intervention and underscores need to ensure baselines use uniform conditions. The report explores outcomes of calculations to estimate energy savings using ”battery end use” data (as measured by Sunrun inverters) and “whole home” data (measured using PG&E’s net meters) and resulted in key findings:
• Temperature and Time-of Week (TTOW) and a Difference-in-Difference (DiD) model with controls were most effective for both data sources.
• Adding battery end-use data enhanced model accuracy and precision compared to methods using only whole home data.
• Models incorporating Peak metrics outperformed methods with annual metrics, suggesting potential benefit in adopting alternative metrics of accuracy for battery programs.
• Large sample size (>5K) allowed for robust estimations, meeting FSU targets with a range of savings (3%-15%).
• Tesla batteries showed greater savings compared to SolarEdge and Delta. TTOW, while effective generally, was insufficient to capture battery behavior during atypical conditions in the baseline period.
• Individual-matched controls DiD method demonstrated effectiveness in the whole home evaluation; although it didn't provide value in a pre-post analysis as it didn't capture the entirety of the impact due to subset of participants exhibiting consumption patterns not observed in control group.
• Recommendations include revising baseline construction to consider undisturbed load patterns; incorporating battery end-use data for more accurate baselines and model evaluation, and adding additional right-hand variables including Solar Irradiance and Cloud Cover to bolster Population NMEC analysis models for battery programs. EESRP was not renewed for 2024.
2023 Load Impact Evaluation for Pacific Gas & Electric’s Emergency Load Reduction Pilot
Abstract: This study quantifies the load impacts of the Residential and Non-Residential Emergency Load Reduction Program pilot. The study focuses on two primary research questions: What were the 2023 demand reductions due to dispatch operations? What is the magnitude of future dispatchable load reduction capability for 1-in-2 and 1-in-10 weather conditions?
The pilot was rolled out in 2021 upon direction by the Commission to expand the state’s portfolio of emergency reliability resources beyond those available in CAISO capacity markets and utility specific load modifying resources such as Critical Peak Pricing. Events are triggered by the CAISO in response to extreme grid stress, and event reductions are settled via a $2/kWh payment, determined using baseline settlement rules. Nine non-residential ELRP events were called in PY 2023, with different subgroups being dispatched for specific events. The average PY 2023 weekday 6pm to 9pm event did not produce meaningful load reductions when evaluated across all non-residential ELRP subgroups. Seven A.4 residential ELRP events were called in PY 2023, and the average event produced 18.1 MW of aggregate load reduction. No A.6 residential events were called.
Program Year 2021 Southern California Edison Agricultural & Pumping Interruptible Program Load Impact Evaluation
Abstract: This study analyzes the performance of Southern California Edison’s Agricultural & Pumping Interruptible (AP-I) Program during the 2021 demand response season and forecasts the impact of the program under a range of years and weather scenarios. The Agricultural & Pumping Interruptible (AP-I) program is a longstanding demand response program in Southern California Edison (SCE)’s territory. In exchange for a monthly bill credit, customers agree to participate in DR events with no notice. Event participation included 964 enrolled customers for the only event of 2021. For this event day, where all participating customers are dispatched, the program provided an average of 28.77 MW (59.6%) of load shed. Including only the full event hours (6 pm to 8 pm), the aggregate impact was 36.12 MW (74.7%). AP-I impacts are determined by the percent of installed switches being successfully dispatched. Over the ex ante forecast horizon, the switch paging success rate is expected to grow, with additional investment in upgrading switches and improving the paging network during this time. As enrollment stays constant and the switch paging success rate increases over the next ten years, aggregate August Peak Day impacts will increase over time, ranging from 29.94MW in 2022 (SCE 1-in-10) to 32.80MW in 2032 (CAISO 1-in-10).
Program Year 2021 Southern California Edison Agricultural & Pumping Interruptible Program Ex Post Load Impacts
Abstract: This table generator is an appendix of the Program Year 2021 Southern California Edison Agricultural & Pumping Interruptible Program Load Impact Evaluation. It presents the aggregate and per-customer ex post load impact results for SCE's Agricultural & Pumping Interruptible Program across multiple customer segments, including results for each individual event and the average event in 2021.
Program Year 2021 Southern California Edison Agricultural & Pumping Interruptible Program Ex Ante Load Impacts
Abstract: This table generator is an appendix of the Program Year 2021 Southern California Edison Agricultural & Pumping Interruptible Program Load Impact Evaluation. It presents the aggregate and per-customer ex ante load impact results for SCE's Agricultural & Pumping Interruptible Program across multiple customer categories for varying weather scenarios, months, and years.
Program Year 2021 Southern California Edison Real Time Pricing Program Load Impact Evaluation
Abstract: This study analyzes the performance of Southern California Edison’s Real Time Pricing Program in 2021 and forecasts the impact of the program for a range of weather conditions and day types. The Real-Time Pricing Program offers commercial and industrial customers the opportunity to react daily to price signals and reduce loads when prices are high. The RTP program delivered 2.37MW during the 4-9pm window on Hot Summer Weekdays: a 7.6% impact. As RTP prices are the highest on these days relative to the otherwise applicable tariff (OAT), ex post impacts are predictably higher on Hot Summer Weekdays, while impacts decline in Moderate and Mild Summer Weekdays. While there is no statistical difference in consumption between High Cost and Low Cost Winter Weekdays, there is a reduction in consumption during the weekend peak on High Cost Weekends compared to Low Cost Weekends. RTP enrollments are expected to decline over time, from 103 in 2021 to 84 enrolled customers in 2032. Program load impacts of approximately 2.5MW during the 4pm-9pm hours are projected.
Program Year 2021 Southern California Edison Real Time Pricing Program Ex Post Load Impacts – Price Model
Abstract: This table generator is an appendix of the Program Year 2021 Southern California Edison Real Time Pricing Program Load Impact Evaluation. It presents the aggregate and per-customer ex post load impact results for SCE's Real Time Pricing Program across multiple customer segments, including results for different months and day types in 2021.
Program Year 2021 Southern California Edison Real Time Pricing Program Ex Post Load Impacts – Event Model
Abstract: This table generator is an appendix of the Program Year 2021 Southern California Edison Real Time Pricing Program Load Impact Evaluation. It presents the aggregate and per-customer ex post load impact results for SCE's Real Time Pricing Program across multiple customer segments, including results for different months and day types in 2021.
Program Year 2021 Southern California Edison Real Time Pricing Program Ex Ante Load Impacts – Program-Level Price Model
Abstract: This table generator is an appendix of the Program Year 2021 Southern California Edison Real Time Pricing Program Load Impact Evaluation. It presents the aggregate and per-customer ex ante load impact results for SCE's Real Time Pricing Program across multiple customer categories for varying weather scenarios, months, and years.
Program Year 2021 Southern California Edison Real Time Pricing Program Ex Ante Load Impacts – Portfolio-Level Price Model
Abstract: This table generator is an appendix of the Program Year 2021 Southern California Edison Real Time Pricing Program Load Impact Evaluation. It presents the aggregate and per-customer ex ante load impact results for SCE's Real Time Pricing Program across multiple customer categories for varying weather scenarios, months, and years.
Program Year 2021 Southern California Edison Real Time Pricing Program Ex Ante Load Impacts – Program-Level Event Model
Abstract: This table generator is an appendix of the Program Year 2021 Southern California Edison Real Time Pricing Program Load Impact Evaluation. It presents the aggregate and per-customer ex ante load impact results for SCE's Real Time Pricing Program across multiple customer categories for varying weather scenarios, months, and years.
Program Year 2021 Southern California Edison Real Time Pricing Program Ex Ante Load Impacts – Portfolio-Level Event Model
Abstract: This table generator is an appendix of the Program Year 2021 Southern California Edison Real Time Pricing Program Load Impact Evaluation. It presents the aggregate and per-customer ex ante load impact results for SCE's Real Time Pricing Program across multiple customer categories for varying weather scenarios, months, and years.
Program Year 2021 Southern California Edison Summer Discount Plan Program Load Impact Evaluation
Abstract: This study analyzes the impact of Southern California Edison's Summer Discount Plan program for a range of weather conditions and dispatch hours. Summer Discount Plan is a voluntary demand response program that provides incentives to residential and non-residential customers who allow SCE to manage the use of their air conditioner when grid conditions require additional resources. The impacts were evaluated using a quasi-experimental design where a matched control customer was identified for each participant. The load impacts were calculated via difference-in-differences by comparing the energy use of participants and the control customer during event and hot non-event days. The SDP program has approximately 180,000 residential customers enrolled and includes nearly 204,000 control devices and 774,000 tons of air conditioner load. Approximately 84% of residential customers elect the higher incentive option, allowing SCE to curtail air conditioner demand (100% cycling) during SDP demand response events. On the commercial side, there are approximately 7,700 customers enrolled with about 69,000 control devices and nearly 350,000 tons of air conditioner load. Roughly 65% of customers elect the higher incentive, accounting for 62% of the total commercial air conditioner load. During the system peak day, the SDP program reduced demand by 166 MW on the first and only event hour. Compared to prior years, 2021 was a substantially cooler year, with lower air conditioner loads, and lower SCE system demand. During normal (1-in-2) August peak day planning conditions, participants can reduce demand by 166 MW across the five-hour 4:00–9:00 PM peak window. In practice, program resources are dispatched by grid location, with varying event times and under different weather conditions.
Program Year 2021 Southern California Edison Summer Discount Plan Program Residential Ex Post Load Impacts
Abstract: This table generator is an appendix of the Program Year 2021 Southern California Edison Summer Discount Plan Program Load Impact Evaluation. It presents the aggregate and per-customer ex post load impact results for residential customers in SCE's Summer Discount Plan Program across multiple customer segments, including results for each individual event and the average event in 2021.
Program Year 2021 Southern California Edison Summer Discount Plan Program Commercial Ex Post Load Impacts
Abstract: This table generator is an appendix of the Program Year 2021 Southern California Edison Summer Discount Plan Program Load Impact Evaluation. It presents the aggregate and per-customer ex post load impact results for commercial customers in SCE's Summer Discount Plan Program across multiple customer segments, including results for each individual event and the average event in 2021.
Program Year 2021 Southern California Edison Summer Discount Plan Program Residential Ex Ante Load Impacts
Abstract: This table generator is an appendix of the Program Year 2021 Southern California Edison Summer Discount Plan Program Load Impact Evaluation. It presents the aggregate and per-customer ex ante load impact results for residential customers in SCE's Summer Discount Plan Program across multiple customer categories for varying weather scenarios, months, and years.
Program Year 2021 Southern California Edison Summer Discount Plan Program Commercial Ex Ante Load Impacts
Abstract: This table generator is an appendix of the Program Year 2021 Southern California Edison Summer Discount Plan Program Load Impact Evaluation. It presents the aggregate and per-customer ex ante load impact results for commercial customers in SCE's Summer Discount Plan Program across multiple customer categories for varying weather scenarios, months, and years.
Program Year 2021 Southern California Edison Smart Energy Program Load Impact Evaluation
Abstract: This report documents the 2021 load impact evaluation of the Southern California Edison’s (SCE) Smart Energy Program (SEP). SEP is a residential demand response (DR) program that utilizes Wi-Fi connected smart thermostats to reduce air conditioning load in participating households during peak hours. SCE retained Demand Side Analytics (DSA) to conduct the SEP load impact evaluation for 2021. The primary objectives of this report are to document the findings of an ex post (after the fact) load impact study for 2021 events and provide ex ante (forward looking) estimates of SEP peak demand reduction capability over the next eleven years (2022 to 2032) under various weather conditions. SCE dispatched SEP on eight days during PY2021 between June and September. On three of these days, there were multiple events called, usually in response to need later in the day. In total, there were eleven events with ten being a result of self-scheduling in the day-ahead market and one for reliability purposes. SCE cited both types of dispatch triggers for the 9/9/2021 event. Current forecasts expect enrollment to increase to 63,114 customers by that 2032. The estimated load impact of SEP in 2022 ranges from 63.8 MW to 69.8 MW during hour ending 17. Estimated impacts decline across the RA window and range from 15.1 MW to 17.4 MW in hour ending 21. Average impacts for the five-hour RA window range from 30.9 MW to 35.6 MW.
Program Year 2021 Southern California Edison Smart Energy Program Ex Post Load Impacts
Abstract: This table generator is an appendix of the Program Year 2021 Southern California Edison Smart Energy Program Load Impact Evaluation. It presents the aggregate and per-customer ex post load impact results for SCE's Smart Energy Program across multiple customer segments, including results for each individual event and the average event in 2021.
Program Year 2021 Southern California Edison Smart Energy Program Ex Ante Load Impacts
Abstract: This table generator is an appendix of the Program Year 2021 Southern California Edison Smart Energy Program Load Impact Evaluation. It presents the aggregate and per-customer ex ante load impact results for SCE's Smart Energy Program across multiple customer categories for varying weather scenarios, months, and years.
2022-2023 Load Impact Evaluation of Pacific Gas and Electric’s Smart Thermostat Control Pilot
Abstract: This report presents the outcomes of Pacific Gas and Electric's (PG&E) Smart Thermostat Control Pilot conducted during the summers of 2022 and 2023. The Pilot comprised two components: demand response (DR) events and daily automated time-of-use (TOU) rate plan optimization. A key focus of this Pilot was the enrollment and effectiveness of various smart thermostat brands and the impact of temperature-based demand reduction strategies. At the close of 2023, the Pilot observed a significant market dominance of Nest thermostats (72.4%), followed by ecobee (26.0%), Emerson (1.6%), and newly introduced Honeywell Home thermostats (0.4%).Because ecobee thermostats offered the most effective functionality for TOU automation, the Pilot focused on the effects on those thermostats. The Pilot revealed that around 58% of ecobee users utilized TOU automation consistently in 2022 and 47% by the end of 2023. On high-load days, the smart thermostats demonstrated an average demand reduction of 0.13 kW per site during TOU control hours, with the effect diminishing over longer periods, especially during the net load peak hours (7–9 PM).
The Pilot underscores the variance in impacts due to geographic dispatch and temperature conditions. In 2022, the highest demand reduction was observed on September 6, a day of extreme heat and high system load, highlighting the correlation between temperature and DR effectiveness. The 2023 data further reinforced this, showing the most substantial impacts in hot regions such as the Central Valley and Sierras. The Pilot concluded that 90% of the variation in dispatchable demand reduction can be attributed to weather conditions, the duration of the event, and the time of day.
If, after reviewing the report, you need additional information on a study, you can check the
EM&V Contacts list and contact the person responsible for
the area associated with the study.