Against the backdrop of all the recent noise about increasing Building Code minimum R-values, let’s look back at how the minimum standard (or typical values) have changed over time. This is a question I am asked regularly. The methodology varies a bit over time, but this excellent article authored by Nigel Isaacs while he was at BRANZ is the best summary I know of.
Short answer: the earliest constructions that sheltered people in Aotearoa, made of raupō or rammed earth, actually did a pretty good job.
So did houses built between 1890 and 1910, according to Nigel: “Good craftsmanship and thoroughly seasoned timber gave a tight construction, double-hung sash windows provided well-controlled ventilation, and open fireplaces gave background ventilation. As a result, houses were warm, dry, and usually well ventilated.”
I do challenge that assessment. Perhaps that new-fangled construction method worked then in the context of open fires constantly tended by a housekeeper, ie because of constant heating input.
But it anyway went downhill from there, in part Nigel argues because of green timber and shabbier craftsmanship. “The traditional 1920s wall section (weatherboard with rough timber lining and an airtight cavity), [compared] with the 1950s walls (brick veneer and lath and plaster) … found that the R-value had fallen by 50% from R0.6 to R0.3–nearly as good as a tent.” [emphasis mine]
Historically, this country has justified really poor levels of insulation and resultant poor health purely on the basis it keeps capital costs low. The rationale for the latest increases to H1 is saving 40% on heating demand for homes. However, the rationale makes no mention of the health benefits that come from a warmer, drier home. It confounds me that the policy makers have valued those benefits at zero.
The other interesting point to note is that the 2003 PAS recommendation for wall insulation, R2.6, is still higher than the revised H1/AS1 requirements for housing (R2.0).
Nigel’s article is also available in a downloadable PDF format here. The table is a very good reference, so I am reprinting that below with all due credit.