skip to main content
OSTI.GOV title logo U.S. Department of Energy
Office of Scientific and Technical Information

Title: Actual and Estimated Energy Savings Comparison for Deep Energy Retrofits in the Pacific Northwest

Abstract

Seven homes from the Pacific Northwest were selected to evaluate the differences between estimated and actual energy savings achieved from deep energy retrofits. The energy savings resulting from these retrofits were estimated, using energy modeling software, to save at least 30% on a whole-house basis. The modeled pre-retrofit energy use was trued against monthly utility bills. After the retrofits were completed, each of the homes was extensively monitored, with the exception of one home which was monitored pre-retrofit. This work is being conducted by Pacific Northwest National Laboratory (PNNL) for the U.S. Department of Energy Building Technologies Program as part of the Building America Program. This work found many discrepancies between actual and estimated energy savings and identified the potential causes for the discrepancies. The differences between actual energy use and modeled energy use also suggest improvements to improve model accuracy. The difference between monthly whole-house actual and estimated energy savings ranged from 75% more energy saved than predicted by the model to 16% less energy saved for all the monitored homes. Similarly, the annual energy savings difference was between 36% and -14%, which was estimated based on existing monitored savings because an entire year of data is not available.more » Thus, on average, for all six monitored homes the actual energy use is consistently less than estimates, indicating home owners are saving more energy than estimated. The average estimated savings for the eight month monitoring period is 43%, compared to an estimated savings average of 31%. Though this average difference is only 12%, the range of inaccuracies found for specific end-uses is far greater and are the values used to directly estimate energy savings from specific retrofits. Specifically, the monthly post-retrofit energy use differences for specific end-uses (i.e., heating, cooling, hot water, appliances, etc.) ranged from 131% under-predicted to 77% over-predicted by the model with respect to monitored energy use. Many of the discrepancies were associated with occupant behavior which influences energy use, dramatically in some cases, actual versus modeled weather differences, modeling input limitations, and complex homes that are difficult to model. The discrepancy between actual and estimated energy use indicates a need for better modeling tools and assumptions. Despite the best efforts of researchers, the estimated energy savings are too inaccurate to determine reliable paybacks for retrofit projects. While the monitored data allows researchers to understand why these differences exist, it is not cost effective to monitor each home with the level of detail presented here. Therefore an appropriate balance between modeling and monitoring must be determined for more widespread application in retrofit programs and the home performance industry. Recommendations to address these deficiencies include: (1) improved tuning process for pre-retrofit energy use, which currently utilized broad-based monthly utility bills; (2) developing simple occupant-based energy models that better address the many different occupant types and their impact on energy use; (3) incorporating actual weather inputs to increase accuracy of the tuning process, which uses utility bills from specific time period; and (4) developing simple, cost-effective monitoring solutions for improved model tuning.« less

Authors:
; ; ;
Publication Date:
Research Org.:
Pacific Northwest National Lab. (PNNL), Richland, WA (United States)
Sponsoring Org.:
USDOE
OSTI Identifier:
1057836
Report Number(s):
PNNL-21870
BT0101000
DOE Contract Number:  
AC05-76RL01830
Resource Type:
Technical Report
Country of Publication:
United States
Language:
English
Subject:
energy efficiency; energy modeling; residential retrofits; actual versus estimated savings

Citation Formats

Blanchard, Jeremy, Widder, Sarah H, Giever, Elisabeth L, and Baechler, Michael C. Actual and Estimated Energy Savings Comparison for Deep Energy Retrofits in the Pacific Northwest. United States: N. p., 2012. Web. doi:10.2172/1057836.
Blanchard, Jeremy, Widder, Sarah H, Giever, Elisabeth L, & Baechler, Michael C. Actual and Estimated Energy Savings Comparison for Deep Energy Retrofits in the Pacific Northwest. United States. https://doi.org/10.2172/1057836
Blanchard, Jeremy, Widder, Sarah H, Giever, Elisabeth L, and Baechler, Michael C. 2012. "Actual and Estimated Energy Savings Comparison for Deep Energy Retrofits in the Pacific Northwest". United States. https://doi.org/10.2172/1057836. https://www.osti.gov/servlets/purl/1057836.
@article{osti_1057836,
title = {Actual and Estimated Energy Savings Comparison for Deep Energy Retrofits in the Pacific Northwest},
author = {Blanchard, Jeremy and Widder, Sarah H and Giever, Elisabeth L and Baechler, Michael C},
abstractNote = {Seven homes from the Pacific Northwest were selected to evaluate the differences between estimated and actual energy savings achieved from deep energy retrofits. The energy savings resulting from these retrofits were estimated, using energy modeling software, to save at least 30% on a whole-house basis. The modeled pre-retrofit energy use was trued against monthly utility bills. After the retrofits were completed, each of the homes was extensively monitored, with the exception of one home which was monitored pre-retrofit. This work is being conducted by Pacific Northwest National Laboratory (PNNL) for the U.S. Department of Energy Building Technologies Program as part of the Building America Program. This work found many discrepancies between actual and estimated energy savings and identified the potential causes for the discrepancies. The differences between actual energy use and modeled energy use also suggest improvements to improve model accuracy. The difference between monthly whole-house actual and estimated energy savings ranged from 75% more energy saved than predicted by the model to 16% less energy saved for all the monitored homes. Similarly, the annual energy savings difference was between 36% and -14%, which was estimated based on existing monitored savings because an entire year of data is not available. Thus, on average, for all six monitored homes the actual energy use is consistently less than estimates, indicating home owners are saving more energy than estimated. The average estimated savings for the eight month monitoring period is 43%, compared to an estimated savings average of 31%. Though this average difference is only 12%, the range of inaccuracies found for specific end-uses is far greater and are the values used to directly estimate energy savings from specific retrofits. Specifically, the monthly post-retrofit energy use differences for specific end-uses (i.e., heating, cooling, hot water, appliances, etc.) ranged from 131% under-predicted to 77% over-predicted by the model with respect to monitored energy use. Many of the discrepancies were associated with occupant behavior which influences energy use, dramatically in some cases, actual versus modeled weather differences, modeling input limitations, and complex homes that are difficult to model. The discrepancy between actual and estimated energy use indicates a need for better modeling tools and assumptions. Despite the best efforts of researchers, the estimated energy savings are too inaccurate to determine reliable paybacks for retrofit projects. While the monitored data allows researchers to understand why these differences exist, it is not cost effective to monitor each home with the level of detail presented here. Therefore an appropriate balance between modeling and monitoring must be determined for more widespread application in retrofit programs and the home performance industry. Recommendations to address these deficiencies include: (1) improved tuning process for pre-retrofit energy use, which currently utilized broad-based monthly utility bills; (2) developing simple occupant-based energy models that better address the many different occupant types and their impact on energy use; (3) incorporating actual weather inputs to increase accuracy of the tuning process, which uses utility bills from specific time period; and (4) developing simple, cost-effective monitoring solutions for improved model tuning.},
doi = {10.2172/1057836},
url = {https://www.osti.gov/biblio/1057836}, journal = {},
number = ,
volume = ,
place = {United States},
year = {Mon Oct 01 00:00:00 EDT 2012},
month = {Mon Oct 01 00:00:00 EDT 2012}
}