Solar Heating Skews Daytime Infrared Wall Insulation Tests, Simulations Show

Estimating how well a building wall insulates is harder than it looks when the sun is up. A new simulation study published in Architectural Research shows that solar radiation heating an exterior wall surface, together with heat stored inside the wall and released hours later, can push infrared-based estimates of thermal transmittance far from the design value, with errors that swing from strong overestimates at night to large underestimates during daytime warming. The findings suggest that checking walls for insulation quality with a handheld thermal camera during the day requires more than simply waiting for a cloudless moment without direct sun.

The study, conducted by Ahhyun Song and Minjae Shin, used transient heat-transfer simulations rather than field measurements, which allowed the researchers to switch solar radiation on and off while holding every other weather condition identical. Three exterior wall assemblies were modeled with the same design U-value of 0.581 W/(m2·K) but very different heat capacities: 39.4, 469.4, and 689.1 kJ/(m2·K). The first was classified as a lightweight wall and the other two as heavyweight walls. Insulation thickness was adjusted in each assembly so that thermal transmittance was matched, isolating the effect of thermal mass alone, a comparison the authors note has been difficult in previous field studies where U-value and heat capacity varied together.

The simulations were run in Physibel VOLTRA version 9.0, with each wall modeled as a 1.0 m by 1.0 m planar assembly and a 10-minute computational time step. Outdoor conditions came from TMYx weather data for Seoul, Republic of Korea, the indoor air temperature was held at 20 °C, and December served as a warm-up period before January results were analyzed. For each wall, one no-solar reference case was compared with four orientation-specific solar-exposure cases covering east, south, west, and north orientations, producing fifteen simulation cases in total. The exterior convective heat-transfer coefficient was derived from meteorological wind speed using the Ito relationship, and longwave radiation exchange accounted for sky and ground radiation environments.

The U-value estimates were calculated with the outdoor infrared thermography method proposed by Albatici and Tonelli, which divides the sum of radiative and convective heat transfer between the exterior wall surface and the outdoor environment by the indoor–outdoor temperature difference. Because the method relies directly on the exterior surface temperature, anything that heats that surface independently of conductive heat flow through the wall, such as sunshine, distorts the result. Conventional guidance has therefore recommended measuring during winter nights or before sunrise, with an indoor–outdoor temperature difference of at least 10 °C, wind below 1 m/s, and no solar exposure. But as the authors point out, heat absorbed during the day can continue to affect the surface temperature long after sunset, so a sun-free moment does not guarantee a quasi-steady thermal state.

The simulated solar-induced surface temperature deviation followed the temporal pattern of wall solar radiation and was consistently larger under high-solar conditions, defined as the three valid days with the highest daily accumulated global horizontal irradiation, than under low-solar conditions. The lightweight wall responded fastest and most strongly: under high-solar conditions, the maximum median surface temperature deviation reached roughly 15 °C for the east-facing wall and 19 °C for the south-facing wall. The two heavyweight walls showed smaller peak deviations, but their deviations declined more gradually after solar radiation began to fall, reflecting heat stored within the wall mass. North-facing walls received relatively little solar radiation and showed comparatively small deviations in all cases.

Thermal time lag, calculated as the difference between the daily peak times of wall solar radiation and of the surface temperature deviation, varied with wall type, orientation, and solar condition. The most striking case appeared for south-facing walls under high-solar conditions: the lightweight wall responded nearly synchronously with the solar peak, while the two heavyweight walls peaked approximately 1.3 hours after it and remained elevated as solar radiation declined. Under low-solar conditions the lags ranged from 0.3 to 1.2 hours across wall types and orientations. Small negative values of −0.2 hours for the lightweight wall were interpreted as an approximately synchronous response rather than a meaningful thermal lead, given the 10-minute temporal resolution and the broad peak profiles.

Even with solar radiation switched off entirely, the estimated U-value error varied substantially with the time of day. Under the no-solar reference condition, the median relative error for the lightweight wall was approximately −10 percent during nighttime, −17 percent during the pre-sunrise period, and −41 percent during daytime. The heavyweight walls showed a different pattern: median relative errors of approximately 78 and 62 percent at night and 90 and 93 percent before sunrise, but approximately −156 and −153 percent during daytime warming. The authors attribute the daytime underestimates to the delayed surface-temperature response of the high-heat-capacity walls, which kept their exterior surfaces below the outdoor air temperature during warming, producing a negative convective heat-transfer term and, in some cases, very low or non-positive apparent U-values.

When solar exposure was included, the solar-induced relative error rose with increasing wall solar radiation for all orientations, and the rise was generally steeper for the lightweight wall, particularly at moderate to high radiation levels. Spearman rank correlations between current wall solar radiation and the solar-induced relative error were positive for every wall type and orientation, ranging from approximately 0.63 to 0.93, with the lightweight wall showing the strongest correlations in most orientations and north-facing heavyweight walls the weakest. The authors emphasize that because the data were recorded at 10-minute intervals and were temporally dependent, these coefficients were treated as descriptive indicators rather than evidence of causality. Considerable scatter within individual radiation bins, especially for the heavyweight walls, indicated that the same instantaneous solar load can produce different errors depending on the wall’s transient thermal state.

That scatter pointed to the role of stored heat. When the researchers repeated the correlation analysis using time-shifted wall solar radiation, delayed by the independently estimated thermal time lag, the correlations generally improved for the heavyweight walls. Under high-solar conditions the correlation increase for the two heavyweight walls reached 0.165 and 0.150 for the south-facing orientation, the largest of any case, with smaller increases for east, west, and north orientations. Under low-solar conditions the correlations increased for all wall types and orientations, with gains of 0.040 to 0.100 for the heavyweight walls and 0.010 to 0.039 for the lightweight wall. The changes for the lightweight wall under high-solar conditions were small and inconsistent, ranging from −0.034 to 0.035, consistent with its nearly synchronous response.

The authors conclude that both current solar exposure and the delayed thermal response of a wall should be considered when interpreting daytime infrared-based U-value estimates, particularly for walls with higher heat capacity, and that field measurement periods should be selected with attention to wall thermal characteristics and recent thermal history rather than current sunshine alone. They also note the study’s limits: the analysis covered a limited range of weather conditions and wall types, and future work should extend the analysis to broader conditions, quantify relationships with outdoor air temperature, wind speed, and heat capacity, and develop prediction or correction models validated with field measurements.

Subject of Research: How solar radiation and wall thermal mass affect daytime infrared thermography estimates of exterior wall U-values

Article Title: Solar-Induced Surface Temperature Deviation and Thermal Time Lag in Daytime IRT-Based U-Value Estimation

Article References: Song, A., & Shin, M. (2026). Solar-Induced Surface Temperature Deviation and Thermal Time Lag in Daytime IRT-Based U-Value Estimation. Architectural Research, 28(1), Article 20. https://doi.org/10.1007/s44515-026-00022-0

Image Credits: AI Generated

DOI: 10.1007/s44515-026-00022-0

Keywords: infrared thermography, U-value, building envelope, solar radiation, thermal time lag, wall heat capacity, heat transfer simulation, energy efficiency, insulation assessment, exterior walls, thermal mass, building retrofit