Total Construction Duration to date

Monday, 9 December 2013

It's All About The Air Barrier!

Building Science Corporation has just released their report (available here ) that confirms that if you give insulation a chance by air sealing the assembly, all insulation will perform the same from a thermal transfer stand point.


The report highlights the increasing level of importance attributed to getting the details right as you increase the performance of an assembly by adding more insulation. Loosing 10% of your nominal value when you have a R15 assembly is a lot less critical than when you have a R50 assembly.

In order to complete this study, BSC had to design and build a hot box that significantly improved upon past designs.  

Key improvements include the ability to: 
  • test higher R value enclosure assemblies (which have lower heat fluxes),
  • 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
This new and improved hot box was able to measure and confirm the effects of thermal bridging, the performance of insulation at various temperature gradients (for instance some polyisocyanurate insulations used in the TM Research Project exhibited a sharp increase in thermal conductivity - and decrease in Rvalue/in. - as temperatures approach and go below freezing), and the effects of air movement through part or all of the assemblies.

An interesting outcome of the air leakage measurements:

"Air leakage always increases the total heat flow through the building enclosure. However, air interacts with the materials in an assembly as it travels through. This interaction changes the temperature field in the assembly and through an assembly. The Thermal Metric wall test results provide strong evidence of the interaction between conductive and convective heat flows. This interaction results in heat exchange between the air and the materials inside the wall assembly and the total measured heat flow will be less than predicted by the commonly used discrete air leakage model"

Also of interest was the observation that "All of the reference test wall assemblies were subjected to significant temperature differences: up to 50°C or 90°F in the winter tests and up to 40°C or 72°F in the summer tests. Natural convective looping was not noted in any of the wall assemblies."

One of the most important confirmations for me was that "All wall assemblies experienced a loss in thermal performance due to air movement through the assembly. This is true for all of the assemblies tested regardless of the type of insulation material used (e.g. cellulose, fiberglass, ocSPF, ccSPF, XPS)".  With the failure of an effective air barrier that I typically see on spray foam jobs, this research helps to confirm that without an effective AB, even the might spray foam has diminishing thermal resistance. The report also confirmed "spray foam insulation only seal areas where the spray foam is installed; significant leakage paths often remain at wood to wood connections"  This confirms my conviction that spray foam is highly over rated.  If it does not guarantee an air tight assembly (see 3.6.4 on pg 102 of the report to see air barrier failures that occurred even in these lab conditions), why use it when there are cheaper products of lower environmental impact available?  I see attempting to use insulation as an air barrier as being just as foolish as attempting to use a poly vapour barrier to double as an air barrier.  Both products are hard pressed to represent an effective and durable air barrier.

Finally, it was interesting to note that they were able to measure the various resistance to air flow that different insulation represent. In 1.1 the report confirms that wet sprayed cellulose has more resistance to air movement over various fibreglass bat products.  It is too bad that they did not test mineral wool as it would be interesting to see how it compares to WPC.  Maybe next time!

No comments: