Better Measurements, Better Models: Advances in Measuring and Predicting Thermal Performance
As 2-D and 3-D modeling to predict building performance becomes more widespread, we need to consider what data are feeding those models. At this point, modeling software has become quite sophisticated, and many users don’t know how the underlying algorithms were developed. Most do not appreciate the simplifications made to represent building and assembly geometry, to describe indoor and outdoor climate, or to characterize material properties. Too frequently, we do not calibrate complex models with appropriate measured data. An example of the type of improved measurement that will allow models to be better calibrated comes from data generated using the RDH-Building Science Laboratories (RDH-BSL) Advanced Double-Guarded (ADG) Hot Box.
History of Modeling
The earliest performance models ignored thermal bridging and used a single R-value to represent the thermal performance of enclosure elements. For simplicity, models did not include full hourly calculations—practitioners used degree days or binned data to describe outdoor (weather) conditions. The earliest calculations were completed by hand, but as computers became commonplace and calculation speed increased, it became possible to complete full hourly simulations (8,760 hours per year) of simple building energy models.
Some of the more-complete early models also included air leakage—typically assuming a single fixed value—ACH. More-complex hourly models used one rate for infiltration (naturally occurring air leakage) and another for scheduled mechanical ventilation.
In time, we acknowledged that air leakage is too complicated to express in one or two constant numbers. It depends on wind speed and direction, configuration exposure of the building, airtightness and distribution of air leaks, and mechanical systems. This acknowledgment led to new approaches, such as the Alberta Air Infiltration Model, to predict hourly air infiltration. More recently, there has been a strong trend toward recognizing the impact of thermal bridging. In the mid-1990s, researchers at Oak Ridge National Laboratory (ORNL) introduced the idea of Clear-Wall R-values to account for regularly spaced framing members, such as studs, and Whole-Wall R-values to account for more-complex framing related to windows, building geometry, and so forth. The ORNL researchers used a combination of physical testing and computer models to develop their approach to accounting for thermal bridging. This approach gained favor, and many in the industry adopted the term Effective R-value to mean the R-value of insulation adjusted for regular thermal bridging (due to studs, z-girts, clips, and so on).
A more-contemporary ASHRAE research project, ASHRAE 1365-RP, has since improved upon the standard industry approach. In this project, researchers used a complex 3-D computer model to consider a wide range of thermal bridging scenarios. Point and linear thermal transmittances account for nonrepetitive thermal bridges, and can be overlaid on the homogenous enclosures represented by Clear-Wall or Effective R-values in existing building energy models.
Most recently, the industry has begun to acknowledge the temperature dependence of thermal properties.
What do all of these developments mean for computer models? Whether they are modeling 2-D or 3-D heat transfer (e.g., THERM, HEAT3, etc.), air movement (e.g., AIM, CONTAM, etc.), hygrothermal performance (e.g., WUFI), or full building energy consumption (e.g., E-Quest, or Energy Plus), many models in the building industry are based on, and calibrated to, measured data. However, as the models and their subjects get more complicated, it becomes more difficult to assess the combined effect of many parameters, so we simplify by reducing the number of cases studied or by looking at factors in isolation. For example, with respect to opaque wall assemblies, how does the combination of thermal bridging, air movement through the assembly, and temperature dependency relate to real-life performance?
The ADG Hot Box was designed to help answer these questions.
The ADG Hot Box
The ADG Hot Box was originally designed for the Thermal Metric Project, a multiyear, multipartner investigation led by Building Science Corporation (BSC). The goal of this project was to contribute to the development of an alternative thermal performance metric. A baseline set of walls was developed and tested, producing a set of reference values for use in future research and testing. These reference values are publicly available in the Thermal Metric Summary Report (Building Science, 2013). The ADG Hot Box was designed and built based on American Society for Testing and Materials (ASTM) standards, but it improved on standard hot box design in several critical ways. For example, it can
- test higher R-value enclosure assemblies;
- expose enclosure wall samples to realistic temperature differences while maintaining the interior temperature at normal room temperatures; and
- measure the impact of imposed airflow at a given pressure difference across the specimen in both directions.
Put simply, the design of the ADG Hot Box allows multiple factors (temperature, airflow, wall design, etc.) to be controlled at once; hence, it is possible to consider how a combination of factors affect performance. A more detailed explanation of the hot box can be found in Schumacher et al. (2013).
Outcomes of Testing
Many factors that have been shown to be significant when considered in isolation were also found to be important when considered in the more-comprehensive tests of the Thermal Metric Project. A comparison of previously developed R-value metrics demonstrates the impact of thermal bridging. As seen in Figure 1, R-values for three of the Thermal Metric walls were significantly higher for the Center-of-Cavity metric (which does not account for thermal bridging) than they were for the Clear-Wall metric (which does). These Center-of-Cavity and Clear-Wall R-values are calculated (rather than measured) using published material thermal conductivities (or R-value per inch). Test results from the Hot Box show the impact of thermal bridging, and when corrected for material property differences, agree well with calculated values such as the Clear-Wall R-value (see Figure 2). Accounting for thermal bridging brings results of the models noticeably closer to actual data.
Testing also highlighted factors that were not frequently acknowledged in discussions of thermal performance. For example, results showed that the thermal conductivity of materials is dependent on temperature, and as a result, the R-value of the wall is likewise dependent on temperature. Figure 2 shows the measured R-values for reference walls 6, 7, and 8, when sealed airtight (using continuous, tightly fitted, and sealed polyethylene films on the indoor and outdoor sides) and tested over a wide range of temperature conditions. At lower temperatures, the thermal performance of the assemblies was typically better than predicted by the models, while at higher temperatures, the opposite was true. Some of the more-comprehensive modeling software (e.g., WUFI, EnergyPlus) can account for this phenomenon, but other commonly used programs (e.g., THERM) can’t.
Even when software has the ability to account for temperature-dependent conductivity, the relationship between conductivity and temperature is often assumed due to a lack of measured material data. These assumptions are problematic, because different materials can exhibit different patterns. For example, porous, air-filled insulation materials tend to have steeper slopes (i.e. R-value/in. goes up faster as temperature decreases) when compared with closed-pore, refrigerant-filled insulation. There are also major exceptions, such as some polyisocyanurate insulations that exhibit a sharp decrease in R-value per inch as temperatures approach and go below freezing. The Thermal Metric Project demonstrated that measured temperature dependency data could be important to accurate modeling.
The Thermal Metric Project also investigated air leakage. Figure 3 shows the measured R-values for the three reference walls at an outdoor temperature of 0°F, under three conditions: when they are sealed airtight, when they are not sealed but no overall pressure difference is induced, and when they are subjected to a 10 Pa pressure difference to induce infiltration.
Air leakage always increases the total heat flow through the building enclosure. However, air interacts with the materials in an assembly, changing the temperature field in and through the assembly. The total measured heat flow is usually different and most often is less than predicted by the most commonly used air leakage models. These results suggest that airflow is a more complex factor in thermal performance than is commonly assumed, and indicate a need for further research.
Building Science Corporation. Thermal Metric Summary Report. Somerville, Massachusetts: BSC, 2013.
Schumacher, C. Polyisocyanurate Short Report, Information Sheet 502, 2013.
Schumacher, C., and D. Ober. Thermal Metric: The End Is Near. Westford BSC 17th Building Science Symposium. Westford, Massachusetts, 2013.
The Bottom Line . . . for Now
So is thermal performance more than the sum of its parts? In some cases, results suggest that calculations based on theoretical models and limited testing can lead to accurate predictions of assembly R-value. For example, the Thermal Metric Project results support the use of simple calculations to assess the impact of thermal bridging. This outcome is helpful, as it allows practitioners to use modeling results based on previous data with more confidence.
In other cases, the ADG Hot Box and the testing of the Thermal Metric Project produced new information and/or highlighted the importance of addressing less well-known issues. For example, temperature dependence was shown to have a significant impact on the assembly R-values. An increasing number of models are capable of accounting for temperature dependence, although we need to measure and document more material property data to support the use of these models.
Airflow, and its interaction with solid materials as it moves through an assembly, is complex. Future research will likely continue to examine the impact of workmanship on air movement, energy use, and moisture deposition.
Following the conclusion of the Thermal Metric Project, RDH Building Science Laboratories has continued to use the ADG Hot Box for product development and advanced thermal performance testing. We believe there is more to learn about thermal performance by considering multiple factors and by collecting precise measured data.
Enter your comments in the box below:
(Please note that all comments are subject to review prior to posting.)
While we will do our best to monitor all comments and blog posts for accuracy and relevancy, Home Energy is not responsible for content posted by our readers or third parties. Home Energy reserves the right to edit or remove comments or blog posts that do not meet our community guidelines.