Abstract

Evaluating the extent to which an individual building consumes energy in excess of its peers is the first step in initiating energy efficiency improvements. Energy benchmarking offers such an assessment, albeit simplified. A widely used energy benchmarking tool is the Environmental Protection Agency's (EPA) Energy Star® Portfolio Manager (PM), which is based on the Commercial Buildings Energy Consumption Survey (CBECS) database. This tool relies on standard linear regression models between building energy use intensities (EUIs) and certain building shell characteristics and equipment types. The statistical significance of these models is rather poor, and they only include continuous building variables. In an effort to improve these models, we have investigated the use of a state-of-the-art, tree-based ensemble learning methodology, the random forest (RF) algorithm, to identify the most influential CBECS parameters that impact building EUI for medium office and school building types. Surprisingly, none of the building features turn out to be really influential. Many of the CBECS building features found to be ranked relatively high are not the ones included in the PM models, and the resulting RF models were only marginally better than the linear regression PM models. These findings cast doubts on the veracity (i.e., the accuracy and completeness) of the data contained in the CBECS database. More careful appraisal is warranted given the widespread use of this database for extracting meaningful generalized correlations between energy use and various building and related characteristics. The poor statistical significance of these correlations has important implications for ongoing building energy conservation policy instruments, such as energy reporting and disclosure legislation being enforced by numerous U.S. cities.

Original languageEnglish (US)
Title of host publicationASHRAE Transactions
PublisherAmer. Soc. Heating, Ref. Air-Conditoning Eng. Inc.
Pages17-28
Number of pages12
Volume121
ISBN (Print)9781936504961
StatePublished - 2015
Event2015 ASHRAE Winter Conference - Chicago, United States
Duration: Jan 24 2015Jan 28 2015

Other

Other2015 ASHRAE Winter Conference
CountryUnited States
CityChicago
Period1/24/151/28/15

Fingerprint

Energy utilization
Managers
Benchmarking
Linear regression
School buildings
Office buildings
Environmental Protection Agency
Stars
Energy efficiency
Energy conservation

ASJC Scopus subject areas

  • Mechanical Engineering
  • Building and Construction

Cite this

Kaskhedikar, A., Reddy, T. A., & Runger, G. (2015). Use of random forest algorithm to evaluate model-based EUI benchmarks from CBECS database. In ASHRAE Transactions (Vol. 121, pp. 17-28). Amer. Soc. Heating, Ref. Air-Conditoning Eng. Inc..

Use of random forest algorithm to evaluate model-based EUI benchmarks from CBECS database. / Kaskhedikar, Apoorva; Reddy, T Agami; Runger, George.

ASHRAE Transactions. Vol. 121 Amer. Soc. Heating, Ref. Air-Conditoning Eng. Inc., 2015. p. 17-28.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Kaskhedikar, A, Reddy, TA & Runger, G 2015, Use of random forest algorithm to evaluate model-based EUI benchmarks from CBECS database. in ASHRAE Transactions. vol. 121, Amer. Soc. Heating, Ref. Air-Conditoning Eng. Inc., pp. 17-28, 2015 ASHRAE Winter Conference, Chicago, United States, 1/24/15.
Kaskhedikar A, Reddy TA, Runger G. Use of random forest algorithm to evaluate model-based EUI benchmarks from CBECS database. In ASHRAE Transactions. Vol. 121. Amer. Soc. Heating, Ref. Air-Conditoning Eng. Inc. 2015. p. 17-28
Kaskhedikar, Apoorva ; Reddy, T Agami ; Runger, George. / Use of random forest algorithm to evaluate model-based EUI benchmarks from CBECS database. ASHRAE Transactions. Vol. 121 Amer. Soc. Heating, Ref. Air-Conditoning Eng. Inc., 2015. pp. 17-28
@inproceedings{b5260222151748b6ac721e532026d13f,
title = "Use of random forest algorithm to evaluate model-based EUI benchmarks from CBECS database",
abstract = "Evaluating the extent to which an individual building consumes energy in excess of its peers is the first step in initiating energy efficiency improvements. Energy benchmarking offers such an assessment, albeit simplified. A widely used energy benchmarking tool is the Environmental Protection Agency's (EPA) Energy Star{\circledR} Portfolio Manager (PM), which is based on the Commercial Buildings Energy Consumption Survey (CBECS) database. This tool relies on standard linear regression models between building energy use intensities (EUIs) and certain building shell characteristics and equipment types. The statistical significance of these models is rather poor, and they only include continuous building variables. In an effort to improve these models, we have investigated the use of a state-of-the-art, tree-based ensemble learning methodology, the random forest (RF) algorithm, to identify the most influential CBECS parameters that impact building EUI for medium office and school building types. Surprisingly, none of the building features turn out to be really influential. Many of the CBECS building features found to be ranked relatively high are not the ones included in the PM models, and the resulting RF models were only marginally better than the linear regression PM models. These findings cast doubts on the veracity (i.e., the accuracy and completeness) of the data contained in the CBECS database. More careful appraisal is warranted given the widespread use of this database for extracting meaningful generalized correlations between energy use and various building and related characteristics. The poor statistical significance of these correlations has important implications for ongoing building energy conservation policy instruments, such as energy reporting and disclosure legislation being enforced by numerous U.S. cities.",
author = "Apoorva Kaskhedikar and Reddy, {T Agami} and George Runger",
year = "2015",
language = "English (US)",
isbn = "9781936504961",
volume = "121",
pages = "17--28",
booktitle = "ASHRAE Transactions",
publisher = "Amer. Soc. Heating, Ref. Air-Conditoning Eng. Inc.",

}

TY - GEN

T1 - Use of random forest algorithm to evaluate model-based EUI benchmarks from CBECS database

AU - Kaskhedikar, Apoorva

AU - Reddy, T Agami

AU - Runger, George

PY - 2015

Y1 - 2015

N2 - Evaluating the extent to which an individual building consumes energy in excess of its peers is the first step in initiating energy efficiency improvements. Energy benchmarking offers such an assessment, albeit simplified. A widely used energy benchmarking tool is the Environmental Protection Agency's (EPA) Energy Star® Portfolio Manager (PM), which is based on the Commercial Buildings Energy Consumption Survey (CBECS) database. This tool relies on standard linear regression models between building energy use intensities (EUIs) and certain building shell characteristics and equipment types. The statistical significance of these models is rather poor, and they only include continuous building variables. In an effort to improve these models, we have investigated the use of a state-of-the-art, tree-based ensemble learning methodology, the random forest (RF) algorithm, to identify the most influential CBECS parameters that impact building EUI for medium office and school building types. Surprisingly, none of the building features turn out to be really influential. Many of the CBECS building features found to be ranked relatively high are not the ones included in the PM models, and the resulting RF models were only marginally better than the linear regression PM models. These findings cast doubts on the veracity (i.e., the accuracy and completeness) of the data contained in the CBECS database. More careful appraisal is warranted given the widespread use of this database for extracting meaningful generalized correlations between energy use and various building and related characteristics. The poor statistical significance of these correlations has important implications for ongoing building energy conservation policy instruments, such as energy reporting and disclosure legislation being enforced by numerous U.S. cities.

AB - Evaluating the extent to which an individual building consumes energy in excess of its peers is the first step in initiating energy efficiency improvements. Energy benchmarking offers such an assessment, albeit simplified. A widely used energy benchmarking tool is the Environmental Protection Agency's (EPA) Energy Star® Portfolio Manager (PM), which is based on the Commercial Buildings Energy Consumption Survey (CBECS) database. This tool relies on standard linear regression models between building energy use intensities (EUIs) and certain building shell characteristics and equipment types. The statistical significance of these models is rather poor, and they only include continuous building variables. In an effort to improve these models, we have investigated the use of a state-of-the-art, tree-based ensemble learning methodology, the random forest (RF) algorithm, to identify the most influential CBECS parameters that impact building EUI for medium office and school building types. Surprisingly, none of the building features turn out to be really influential. Many of the CBECS building features found to be ranked relatively high are not the ones included in the PM models, and the resulting RF models were only marginally better than the linear regression PM models. These findings cast doubts on the veracity (i.e., the accuracy and completeness) of the data contained in the CBECS database. More careful appraisal is warranted given the widespread use of this database for extracting meaningful generalized correlations between energy use and various building and related characteristics. The poor statistical significance of these correlations has important implications for ongoing building energy conservation policy instruments, such as energy reporting and disclosure legislation being enforced by numerous U.S. cities.

UR - http://www.scopus.com/inward/record.url?scp=84938850662&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84938850662&partnerID=8YFLogxK

M3 - Conference contribution

SN - 9781936504961

VL - 121

SP - 17

EP - 28

BT - ASHRAE Transactions

PB - Amer. Soc. Heating, Ref. Air-Conditoning Eng. Inc.

ER -