Today’s Agenda

PG&E Presentation Template

Estimating Key Parametersfor the Retail Plug-Load Portfolio (RPP) Program:

Recommended Methodological Approaches

EM&V, Residential Programs, Products and Channel Teams

October 23, 2014#Residential PLA Program ##Todays AgendaOur ObjectivesRPP Program ReviewGeneral Estimation PrinciplesAdapting Parameter Inputs for the E3 Calculator: Key Considerations for Market TransformationProposed Methods to Estimate NTGRs, EULs, and IMCsOur Requests to the Cal TFResidential PLA Program #Our ObjectivesPresent recommended approaches to estimate values for three key savings parameters for Cal TF approval:Net-to-gross ratios (NTGRs)Incremental measure costs (IMCs)Effective useful life (EUL)

Present recommended approaches to adapt parameter inputs for use with E3 calculator for Cal TF approval

Seek Cal TF approval for UEC estimation methodology (presented at last months meeting)Residential PLA Program #RPP Program ReviewProgram Approach:Utilize retailer engagement to increase the demand and supply of more energy efficient appliance and consumer electronic productsMeasures: RPP offers a portfolio of consumer energy efficiency measures that are delivered through the retail channel partnersIncentive Structure:Incentives are paid to participating retailers for each program-qualified product sold.Evaluation Framework:While this mid-stream program is primarily a market transformation program, it contains elements of a resource acquisition program.

Residential PLA Program #RPP Logic Model

Residential PLA Program #General Estimation PrinciplesFor products in DEER:Use EUL and IMC values found in DEERFor products NOT in DEER:Use best available data for EUL and IMC estimates.

NTGRs for resource acquisition programs:DEER values reflect 3-5 year timeframes, and are applicable to downstream resource acquisition programs.NTGRs for market transformation programs:DEER values are not applicable, NTGRs must be derived through other methods that involve long-term forecasting. Residential PLA Program #Adapting Inputs for the E3 Calculator: Market Transformation ConsiderationsGross Savings:The total of the stream of savings for all of the measures adoption forecasted over the projected initiative term times each measures EUL. TRC costs:Input the stream of administrative and incremental measure costs that match the time frame of the savings impacts.NTG Inputs:Use the ratio of forecast total market change minus the forecast of the baseline changes divided by the total market change for the same time period used for both the costs and savings.Residential PLA Program #Net-to-Gross Ratios (NTGRs)NTGRs represent net program load impacts divided by gross program load impacts.NTGR is sometimes used to convert gross measure costs to net measure costs.The relationship between program-driven and naturally-occurring adoptions evolves over the life of the program.This dynamic must be captured and entered into the B/C calculator.

Hypothetical exampleResidential PLA Program #NTGRs: Our RecommendationPreferred: Estimate NTGRs over time using an analogic diffusion model

For each product category, long-term forecasts over the life of the RPP Program can be usedScenarios (using the Bass diffusion model) with and without the RPP Program can be forecasted over the life of the programThe resulting differences in these two scenarios represent the expected net impacts of the RPPThe net impacts divided by the gross impacts yield the NTGRs

Alternative: Use a Delphi panelIdeal but costly: Combine the twoResidential PLA Program #Incremental Measure Cost (IMC)Incremental measure cost is defined as the difference in the cost of an efficient measure and a baseline measure attributable to the difference in efficiency.Methods must be developed to capture the fact that IMCs decline over the life of the programHistorical trends of IMCstraditionally observed to decline over timewill be reviewed for measures with characteristics similar to those in the RPP suite (e.g., CFLs, DVD-Rs, transistor radios, desktop PCs, laptop PCs, lower-volume laser printers, and multi-function devices).Declining IMCs must be entered into the B/C calculator over the life of the program.Residential PLA Program #Using Webcrawlers to obtain IMC information

Many retail-focused webcrawlers track price trends for specific products, such as pricegrabber.comThe IOU C&S team recently developed a custom webcrawler to determine IMC for residential lighting products.The webcrawler collected a wide range of product features, then the team conducted a multiple regression analysis to better understand how product features contribute to pricing. The team is considering expanding this approach to additional productsResidential PLA Program #IMCs: Our RecommendationWe recommend using webcrawlers to estimate IMC.For each product, we would identify key product features, including retail price and Energy Star qualification status.Using these data, we could then develop a regression analysis to estimate the incremental cost of an energy efficient product (hedonic price models)Data collection can be easily replicated to identify changes over time. Retailer price data also holds promise.Price data provided by retailers for each model sold along with key product characteristics could be used to estimate the IMC using hedonic price models Targeted in-store shelf surveysWhile they have very limited value, given that the RPP is based on a dynamic product mix that is expected to change over time, they can be used to selectively calibrate webcrawler results.Residential PLA Program #Effective Useful Life (EUL)The effective useful life is an estimate of the median number of years that a measure installed under a program is still in place and operable.

While EULs could conceivably change over the over the life of the program, such changes might be difficult and costly to track reliably.

Consistent with retention and persistence studies, the focus of RPP EUL estimates is to estimate the effective useful life for installed equipment, since this is what is producing program-generated savings.Residential PLA Program #EULs: Our RecommendationFirst choice: Use available research data For products that are federally-regulated, EUL should be based on research data used in the standards proceeding. Where this is not a federal standard, we will use best available research (e.g., EPA, manufacturers or other research).

Delphi panel Use a formal consensus approach for generating EUL estimates from a group of experts.

Ideal but more costly: Combine available research data and Delphi panel.Residential PLA Program #Our Requests to the Cal TFDo you agree with our recommended approaches to estimating these three parameters?

Would you like any additional information about the proposed estimation approaches?Do you have any changes you would recommend to our proposed methods?Are there any additional approaches that you could recommend?Do you have suggestions about how to input parameter estimates into the E3 calculator?

Residential PLA Program #Next StepsApproval of the UEC methodologyApproval of methodologies for estimating key parametersRPP work paper draft reviewReview of parameter estimatesRPP work paper approval

Residential PLA Program #Questions?#Residential PLA Program #