Multifamily Envelope Leakage Model

O. Faakye; D. Griffiths; Wyatt Metzger; Wyatt Metzger

Multifamily Envelope Leakage Model

O. Faakye, D. Griffiths, Wyatt Metzger, Wyatt Metzger

Consortium for Advanced Residential Buildings

Research output: NREL › Subcontract Report

Abstract

The objective of the 2013 research project was to develop the model for predicting fully guarded test results (FGT), using unguarded test data and specific building features of apartment units. The model developed has a coefficient of determination R2 value of 0.53 with a root mean square error (RMSE) of 0.13. Both statistical metrics indicate that the model is relatively strong. When tested against data that was not included in the development of the model, prediction accuracy was within 19%, which is reasonable given that seasonal differences in blower door measurements can vary by as much as 25%.

Original language	American English
Publisher	National Renewable Energy Laboratory (NREL)
Number of pages	33
State	Published - 2015

Bibliographical note

Work performed by Consortium for Advanced Residential Buildings, Norwalk, Connecticut

NREL Publication Number

NREL/SR-5500-62675

Other Report Number

DOE/GO-102015-4499

Keywords

blower door testing
Building America
CARB II
empirical model
envelope leakage
guarded
multi-linear regression models
multifamily apartments
predicting fully guarded test results
residential
residential buildings
single-family attached housing
unguarded

Access to Document

https://www.nrel.gov/docs/fy15osti/62675.pdf

Cite this

@misc{1d11cdfcfcef469b8f14658f4003fbee,

title = "Multifamily Envelope Leakage Model",

abstract = "The objective of the 2013 research project was to develop the model for predicting fully guarded test results (FGT), using unguarded test data and specific building features of apartment units. The model developed has a coefficient of determination R2 value of 0.53 with a root mean square error (RMSE) of 0.13. Both statistical metrics indicate that the model is relatively strong. When tested against data that was not included in the development of the model, prediction accuracy was within 19%, which is reasonable given that seasonal differences in blower door measurements can vary by as much as 25%.",

keywords = "blower door testing, Building America, CARB II, empirical model, envelope leakage, guarded, multi-linear regression models, multifamily apartments, predicting fully guarded test results, residential, residential buildings, single-family attached housing, unguarded",

author = "O. Faakye and D. Griffiths and Wyatt Metzger and Wyatt Metzger",

note = "Work performed by Consortium for Advanced Residential Buildings, Norwalk, Connecticut",

year = "2015",

language = "American English",

publisher = "National Renewable Energy Laboratory (NREL)",

address = "United States",

type = "Other",

}

TY - GEN

T1 - Multifamily Envelope Leakage Model

AU - Faakye, O.

AU - Griffiths, D.

AU - Metzger, Wyatt

N1 - Work performed by Consortium for Advanced Residential Buildings, Norwalk, Connecticut

PY - 2015

Y1 - 2015

N2 - The objective of the 2013 research project was to develop the model for predicting fully guarded test results (FGT), using unguarded test data and specific building features of apartment units. The model developed has a coefficient of determination R2 value of 0.53 with a root mean square error (RMSE) of 0.13. Both statistical metrics indicate that the model is relatively strong. When tested against data that was not included in the development of the model, prediction accuracy was within 19%, which is reasonable given that seasonal differences in blower door measurements can vary by as much as 25%.

AB - The objective of the 2013 research project was to develop the model for predicting fully guarded test results (FGT), using unguarded test data and specific building features of apartment units. The model developed has a coefficient of determination R2 value of 0.53 with a root mean square error (RMSE) of 0.13. Both statistical metrics indicate that the model is relatively strong. When tested against data that was not included in the development of the model, prediction accuracy was within 19%, which is reasonable given that seasonal differences in blower door measurements can vary by as much as 25%.

KW - blower door testing

KW - Building America

KW - CARB II

KW - empirical model

KW - envelope leakage

KW - guarded

KW - multi-linear regression models

KW - multifamily apartments

KW - predicting fully guarded test results

KW - residential

KW - residential buildings

KW - single-family attached housing

KW - unguarded

M3 - Subcontract Report

PB - National Renewable Energy Laboratory (NREL)

ER -

Multifamily Envelope Leakage Model

Abstract

Bibliographical note

NREL Publication Number

Other Report Number

Keywords

Access to Document

Fingerprint

Cite this