Residential applliance data, assumptions and methodology for end-use forecasting with EPRI-REEPS 2.1

Abstract

This report details the data, assumptions and methodology for end-use forecasting of appliance energy use in the U.S. residential sector. Our analysis uses the modeling framework provided by the Appliance Model in the Residential End-Use Energy Planning System (REEPS), which was developed by the Electric Power Research Institute (McMenamin et al. 1992). In this modeling ' framework, appliances include essentially all residential end-uses other than space conditioning end-uses. We have def'med a distinct appliance model for each end-use based on a common modeling framework provided in the REEPS software. This report details our development of the following appliance models: refrigerator, freezer, dryer, water heater, clothes washer, dishwasher, lighting, cooking and miscellaneous. Taken together, appliances account for approximately 70% of electricity consumption and 30% of natural gas consumption in the U.S. residential sector (EIA 1993). Appliances are thus important to those _esidential sector policies or programs aimed at improving the efficiency of electricity and natural gas consumption.This report is primarily methodological in nature, taking the reader _Jaroughthe entire process of developing the baseline for residential appliance end-uses. Analysis steps documented in this report include: gathering technology and market data for each appliance end-use and specific technologies within those end-uses, developing cost data for the various technologies, and specifying decision models (both the functional form and equation parameters) to forecast future purchase decisions by households. Our implementation of the REEPS 2.1 modeling framework draws on the extensive technology, cost and market data assembled by LBL for the purpose of analyzing federal energy conservation standards. The resulting residential appliance forecasting model offers a flexible and accurate tool for analyzing the effect of policies at the national level.The proliferation of models and model parameters and the lack of detailed documentation in the end-use forecasting area have been sources of considerable confusion for practitioners and policy-makers alike. By making explicit the data and assumptions behind our analysis, we seek to clear up some of this confusion and move the energy efficiency debate beyond simply the numerical results of models.

Residential sector end-use forecasting with EPRI-Reeps 2.1: Summary input assumptions and results

“…Unless otherwise stated, exogenous inputs are taken from the U.S. Department of Energy's Annual Energy Outlook (US DOE 1995a). All other inputs are taken from the detailed data compilations documented in and Hwang et al (1994), except where noted.…”

“…As described in Hwang et al (1994) and , and summarized in Table 4.2, a multinomial logit equation is used to choose among different generic technologies. Each generic technology is characterized by utility functions based on household characteristics, technology characteristics, and exogenous variables.…”

“…This report summarizes the REEPS model framework and inputs described in Hwang et al (1994) and , and presents the results of the baseline forecast for each end-use.…”

Evaluating machine learning for predicting next-day hot water production of a heat pump

“…Many other related approaches uses only model-based approaches [6], [7] or artificial neural networks [8], [9], [10]. Furthermore, in contrast to related work, we predict the individual household water production for the next-day and spread over the whole day while they only do long-term predictions.…”