Of course I’m going to love a presentation that included differential equations! Fellow engineers Cameron Munro (Passive Analytics) and Harley Truong (Logikhaus) put their heads together shortly before the conference to envisage what form a new as-built measure for the performance of a building’s thermal envelope might take—the equivalent of a blower door test for testing airtightness. This is a great research question as knowing how much heat a building is leaking through its envelope is an important measure of performance and comfort.
Cameron Munro. Image (c) APA
At the moment, Passivhaus certifiers rely on heaps of documentation to assess the performance. There is a gold standard as-built evaluation: it’s called a co-heating test and we helped design one for FMI’s Net Zero House.
The catch is co-heating tests are expensive to set up, logistically difficult and take lots of time. The test runs for two weeks, the outside weather needs to be within certain parameters, the dwelling must be devoid of people and lots of heaters and fans are required to maintain a consistent indoor temperature. Not cheap! Or easy. So co-heating tests don’t get done outside of specific, formal research situations.
Harley started with some clever slides and questions for the audience that showed the need for accurate and understandable markers of performance, to help stimulate informed demand for homes with better thermal performance.
Cameron talked about the kind of alternative test they were looking for, a metric that was intuitive, easily measured, repeatable, universally applicable and a tool for valid comparisons. First option: diurnal temperature range, quickly discounted because of multiple problems, not least that a tent in a perfect climate will ace this test.
Next was a decrement factor, which requires explanation, is less intuitive and can’t be compared to anything in PHPP.
However, the decrement factor does come with equations! Image (c) APA.
Enter suggestion #3, a thermal time constant. This value is measured in hours and indicates thermal lag: how slowly the internal temperature follows the outdoor temperature. A high-performance home might have a decrement factor of 60 hours, twice as much as a Code-minimum home. This measure is only valid for airtight buildings though, because it’s driven by transmission heat loss (not ventilation-related losses). The main advantage is it can be linked to two values in the PHPP file, offering the tantalising possibility of a predictive number that can be compared to an as-built result.
I’m sorry to hose down expectations of some quick fix here, but this is really just the very beginning of the research. I looked at the graphs more and quizzed Cameron and Harley afterwards. The immediate problem is the uncertainty bars are too wide to be useful, there’s too much overlap between high-performance homes and Code-minimum ones to be able to make valuable distinctions. I don’t think it’s feasible to assume no internal heat gains either, not in an occupied house. It costs about $40K to run a co-heating test on a specific building and you still get a cloud of data points that produce reasonably large error bars. It’s not realistic to hope for results that approach this in utility based just on looking at data and making assumptions.
It was fun to discuss it though. It left me wondering whether to approach this problem from the other direction. How can a co-heating test be made cheaper, easier and faster but still produce results that are reliable and accurate enough? Could it be possible to kick the occupants out just for one night for instance, and to pack the fridge freezer contents into chilly bins and turn it off for 24 hours? To get reliable data, it’s crucial to remove living, breathing, heat-producing mammals from the site of investigation and unplug their machines that generate heat.
Related reading:
The results of the FMI Zero House co-heating test are discussed at length in Fenestration, FMI’s magazine. Autumn 2024 edition, beginning on page 38. The most relevant graphs are on page 46.