Best Insulation for Homes - My Conclusions

There are several practical applications to take away from this experiment:

Insulation systems of equal R value perform similarly

This sounds like a "no-duh" statement but still I was surprised how little difference there was between insulation systems.  Even though I had purposely set each system to about the same R-value I have seen so many marketing pitches over the years I really expected there to be a significant difference between them, especially under depressurization.  The R-value rating system isn’t perfect but it it’s not bad either.   

This experiment also indicates that all insulation systems are equally effective, theoretically.  You may argue that theory has nothing to do with it, my images show everything performed about the same in the real world.  Let me explain.  I have infrared imaged far too many buildings to believe that my simple statement above explains it all.  Truth is, most of the failures I see are not in standard 8 foot high wall bays.  Most of the failures I see involve penetrations in the insulation like outlets or windows, or changes in the insulation plane like, sloped ceilings, dormers or kneewalls.  Practically speaking, some insulation systems handle the inconsistencies of a building better than others.  That said, my experiment says that, with meticulous installation, a person should be able to make any one insulation system perform as well as any other.  I’ve suspected that for a long time but I was still surprised to see it borne out in the experiment. 

I was pleased with the performance of the flash and foil system.  The manufacturer lists an equivalent R-value of 7.66 if the barrier was installed in a 2x6 wall bay with 2 inches of spray foam in the outside sheathing.  The key word is "equivalent".  A radiant barrier blocks heat primarily by reflecting infrared radiation (heat waves), this a totally different mechanism of blocking heat loss by using a thermally resistive material. Radiant barriers cannot be described through R-value, a conductive heat loss measure, unless they are installed in a system and the system is rated collectively. Bottom line: other than some minor thermal effects described below, the radiant barrier WITH spray foam performed as well as other systems with the same R-value.

The minor variances between insulation systems can be explained

I noticed that the studs showed up more distinctly in the bays with the radiant barrier than in the bays with just spray foam.  The radiant barrier is made of aluminum, excellent reflector of radiant heat but also second only to gold in thermal conductivity.  In other words, aluminum has essentially a negative R-value, it conducts heat better then it blocks it.   Wherever the aluminum is in contact with the stud it slightly reduces the R-value of the wood.  Radiant barriers need to be suspended out of contact with other materials in order to minimize their conductive downside.  Hence the baffles in the radiant barrier I used.  

The other minor difference in the radiant bay is that at the bottom of the wall bay it is slightly cooler than the rest of the bay.  I believe this is because, other than the thin aluminum baffles, the wall bay is empty until you hit spray foam.  Wherever you have an  air pocket a natural convection loop will form.  The interior side of the wall bay warms the air pocket slightly and the exterior side cools it slightly.  The cold dense air sinks to the bottom of the bay and warm expanded air floats to the top.  The more freely the air can move the more pronounced the convection loop.  Convection loops have been demonstrated many kinds of loose-fill insulation systems.  The greater the temperature difference from inside to outside the more pronounced the convection loop.  Even though the moving air is trapped inside the wall bay and has little direct connection to the outside air it still has the net effect of reducing the effective R-value below the rated R-value. Many insulation systems are designed to trap dead air, if the air is moving, even very slowly, the insulation decreases in R-value.

The test results seem to indicate that the flash and batt system performs better than flash and foil.  From my perspective, they performed about the same.  The flash and foil bay appeared to perform slightly worse than the flash and batt bay but it is unclear in my experiment if the flash and foil system underperformed because of the internal convection or the lower R-value.  

The studs also showed up prominently with the bay insulated by fiberglass but for a different reason than the flash and foil bay.  In all my other test bays the insulation was in intimate contact with the sides of the studs, in essence the insulation was glued to the studs.  In the fiberglass bay, the insulation was touching, but not adhered to, the studs.  Heat flows from where it’s hot to anywhere that is cooler as it seeks equilibrium, it will always take the path of least resistance.  Studs with insulation glued to the sides force heat to heat to transfer on a longer path, it must conduct all the way through the full thickness of the wood.  To go right or left would be a higher resistance pathway than the wood.   If there are slight gaps on either side of the stud, however, heat will conduct to the first location that is cooler and then radiate out.   All things considered, this is really a fine point in home heat transfer and has a minor impact on energy efficiency.  

Same three suspects show up again and again.

All the tested insulation systems were set to about the same R-value and they performed about the same, as we would expect.  That is good news, it means that one insulation system is interchangeable with another as long as you know the R-value per inch and you have enough room to fit in the desired total R-value. Unfortunately, the strength of an insulation system does not rest solely in its R-value.  You have to consider 3 things that can degrade the performance of your insulation:

1.      penetrations in your building shell let air go through or around your insulation

2.      materials with lower R-value than your insulation create thermal “bridges” to the outside

3.      internal convection loops slowly transfer warmth from the warm side to the cold side of your insulation 

Despite the strengths of each insulation system, one of the 3 suspects showed up in nearly every system I tested.  Air leakage is the most serious of the 3, thermal bridging and internal convection are lesser effects but important if you are trying to make a super efficient structure.   This test reaffirmed to me that you won’t beat the 3 suspects just by picking the "right" insulation system.  Beating the 3 comes through a combination of design, materials, workmanship, and inspection.   This test shows it really doesn’t matter which insulation material you choose as long as you account for the 3 suspects so that your system performs at the R-value you paid for.  And there’s more good news, designing your insulation system to beat the 3 suspects can help prevent insidious problems like mold, pipe freeze-ups, wood rot, and sheetrock shadowing.      

Go to Flash and Foil Results
Go to XPS and Foam Sealant Results
Go to Fiberglass Results
Go to Spray Foam Results
Go to Flash and Batt Results
Return to Best Insulation for Homes page from Best Insulation Experiment Conclusions page