Can one think about sun shining on roofs and walls in terms of delta-T?
Any resources to help think about this would be great.
Many discussions of heating vs cooling include a statement something to the effect of, “in winter the delta-T can be 80°F or more, but in summer even on a 100°F day the delta-T is only 30°F”.
Sure, that makes sense, but what about when sun is shining on a surface (roof or wall)? Can we still think about that in terms of delta-T? ie something like, “The sun heats the outside of the wall/roof to ~140°F, so the delta-T is something like 70°F” (for periods of the day!) (I’m making those numbers up, but thermometers do show temps that high in the mid-afternoon sun here in Austin, TX). Or does the radiant vs conduction stuff render that pointless?
Or let me ask this another way: why aren’t the parts of the walls that are hit with direct sun (in cooling climates) treated differently in insulation materials than other walls? Yes, they are with windows/window-loads, but not the rest of the wall. Either in terms of R-value or in terms of thermal mass (accounting for the fact that the elevated energy those walls receive is sometimes short lived during the day)?
Apologies if that makes no sense 🙂
GBA Detail Library
A collection of one thousand construction details organized by climate and house part
Replies
>"Or let me ask this another way: why aren’t the parts of the walls that are hit with direct sun (in cooling climates) treated differently in insulation materials than other walls?"
Because it would unduly burden the code with complexity that would surely confuse the builders.
Custom high performance builders will often use WUFI or BeOpt and other tools to tweak the insulation specs, but it's usually not "worth it" to add the complexity of different R-values for different cardinal orientations or even different roof slopes & shading factors, even when they DO opt for fine-tuning the glazing spec for different orientations.
In a cooling climate the SHGC and shading factor & orientation of a window make a bigger difference on average & peak cooling loads than going from R15 to R25 in a wall assembly, or going from R30 to R50 in a roof. In Austin a south facing window with a relatively high SHGC can represent even less of a cooling load than a l0w-gain window on the west side due to the external reflection at the high summertime sun angles on the south side, but it's even better if southerly roof overhangs can shade those windows. It's more difficult to mitigate against unwanted gain on the east and west sides with overhangs due to the lower sun angle, so the SHGC (and window size) is going to matter more on those sides of the house than the south face.
The duration of the peak delta-Ts on sun drenched roofs and walls is pretty brief compared to the duration of indoor to outdoor air delta-Ts. The solar reflectivity index (SRI) of the roofing material, as well as the steepness of the pitch can dramatically affect those peak temperature differences too. Higher pitched roofs convection-cool to something closer to the outdoor air temp than flat roofs, and high SRI roofs can be several 10s of degrees cooler than a low SRI roof even on flat roofs.
Thanks Dana, awesome as always :) And yes, code complexity bad, makes sense.
No doubt that west facing walls and windows take a beating in the afternoon (and east in the morning, although doesn't feel as problematic due to air temps). And west striking sun happens same time as people get home and fire up the internal loads (TV, oven, opening doors etc).
Sounds like it would make sense to treat the western facing walls more like a roof, thinking about their SRI etc. I also wonder whether a bottom/top vented rainscreen gap on the western walls might not convection cool better than might be expected (hey, a wall is like a really steep roof, right? :).
And perhaps it would make a big difference to use an insulation (or wall structure) with higher thermal mass on the western walls (absorbing the temporary heat during drenching). I think that's what's going on in this demonstration (the video, not the graphs): https://foursevenfive.com/blog/head-to-head-heat-test-eps-vs-gutex-vs-mineral-wool/
If anyone had a link to a BeOpt or WUFI analysis that dealt with some of these issues, I'd be very interested.
>"Sounds like it would make sense to treat the western facing walls more like a roof, thinking about their SRI etc. "
Not really- a wall has HUGE convective-cooling capacity compared to any roof. A west facing wall doesn't get nearly as many hours of direct sun as a south facing roof pitch too.
A roof is not just a tilted wall. Both the daytime sun exposure and night sky radiation differences are pretty stark, as is the amount of direct wetting & dew a roof gets relative to a wall. It's more different than just apples & pears.
While putting them into code is a problem, they are worth keeping in mind. For example, shade trees make a difference.