1. Prof. Chris Hendrickson
Using AMI data to inform efficiency decisions Assessing the value of information in residential building simulation
Brock Glasgo
Engineering and Public Policy
Prof. Inês Azevedo
Prof. Chris Hendrickson
Civil and Environmental Engineering

2. Device-level data analytics to guide policy
Chapter 1 – Using AMI data to characterize residential energy use
Chapter 2 – Assessing the value of information in residential building simulation
Validating existing baselines and tools
Chapter 3 – Simulating a transition to direct current in homes
Chapter 4 – Expert assessments on the future of direct current in buildings
Evaluating an intervention
This paper I’m going to talk about is one of four chapters…

3. Energy code analysis and design
Developed under the DOE Building Energy Codes Program by Pacific Northwest National Lab (PNNL, 2012)
Uses DOE’s EnergyPlus software
Simulate code changes across the US
Provides decision support to:
The DOE to inform their consensus ruling on proposed changes to the International Energy Conservation Code (IECC)
State and local jurisdictions considering adopting a newer energy code
The reason we’re interested in building simulation tools…

4. Energy code compliance
Building energy codes are changing
Require higher levels of efficiency
Use different formats to present requirements
Have increased in scope to include more end uses
New paths are needed to effectively implement and verify compliance
One current compliance pathway: RESNET’s HERS scoring system
Suite of software tools
HERS scores cited directly in IECC codes
Soon to be replaced with a single EnergyPlus-based tool
That’s one way these tools are being repurposed. The second, and more recent…

5. Research questions
How accurate are the air conditioning, ventilation, and other load energy consumption estimates in EnergyPlus models specified with the results of a typical energy audit?
What building and occupant characteristics are most important for accurate building energy simulations?
What do these results say about the use of EnergyPlus as a tool for residential energy code analysis and compliance verification?
So that leads us to the research questions…

6. Monitored electric data
Modeling method
PNNL prototypes
Prototype editor
Home models
Audit records
Monitored electric data
Local weather data
Survey records
The way we try to answer those questions…
Occupancy data
Images: Mathworks, 2017; DOE, 2017; PNNL, 2012

7. Detailed home data Provided by the Pecan Street Research Institute
Sample contains ~700 volunteer homes with measured data at the circuit level at 15-minute intervals
(106) homes meet data validation criteria
Single-family detached homes
Complete energy audit records
Survey records
Monitored (2015):
Whole-home use
AC condensing unit
Central air fan
Not a representative sample
Average household income > $130k/yr
The detailed home data we use…

8. Occupancy estimation
Device use reflects occupant activity, not occupancy
Sleeping or inactive hours show up as unoccupied
Occupancy estimated by finding morning wakeup and evening arrival
We estimate home occupancy…
The problem with this method…

9. Results: Whole-home
EnergyPlus simulations generally overestimate whole-home consumption
Errors are greatest for the most efficient homes
CVRMSE range from 40% to over 300%
RECS mean
So now looking at results, first we’ll look
We’ll first look at whole-home consumption first…

10. Results: AC condensing unit
Condensing unit consumption contributes to net annual consumption overestimate
Errors are greatest for homes with lowest consumption
CVRMSE range from 60% to over 1000%
RECS mean
As I said we also have device-level result…

11. Autosized condenser capacity
Sensitivity analysis
Floor area
EER age correction
Default EER
Infiltration
HVAC set points
Orientation
Ceiling height
Window type
Window area
Attic R-value
And lastly here we see the results of the…
The axes here…
Occupancy
Residents
Lighting
Brick exterior
Autosized condenser capacity

12. Conclusions
Our EnergyPlus simulations do not accurately estimate actual, occupied energy consumption at the device or building level
None of the (106) models meet ASHRAE calibration tolerances at the hourly level
Annual consumption estimates can be accurate, but hourly interval data can have large errors
Largest differences in modeled and actual consumption occur in the most efficient homes and systems
More detailed behavioral, operational characteristics are needed for accurate modeling
We quantify the effect of default vs. actual values for 15 common model parameters
With that, we can draw some conclusions

13. Discussion
Results provide context for the use of EnergyPlus in single-family homes
New application for building simulation, more research is needed
Commercial sector can serve as a guide
As more datasets like Pecan Street’s become available, more modeling and validation will be possible
DOE Building Energy Codes Program
Use of simulation tools reflects a transition to a systems approach to code design
Building energy code focus has been, and continues to be, on building envelope: ceilings, walls, windows, floors, foundations
Prototypes are not representative of many homes’ actual performance, but savings estimates from incremental changes are likely reasonable
These results, first of all…
As far as the DOE’s…

14. Discussion
RESNET’s EnergyPlus compliance verification tool needs to be validated against actual building performance
Results here, and previous research in the commercial sector show that simulations do not accurately estimate actual energy consumption in occupied buildings
Tool has not been released, so details are unclear
Lighting and other appliance efficiencies?
Programmable thermostats?
Actual building infiltration?
Need to establish assumptions and standard conditions that fairly value efficiency measures and limit the effects of model inaccuracy
And lastly…

15. Acknowledgements
Advisors and co-authors Inês Lima Azevedo & Chris Hendrickson
Center for Climate and Energy Decision Making
The Pecan Street Research Institute
Phillips and Huang Family Fellowship in Energy