Replies

1. GBA Editor
   Martin Holladay | Mar 16, 2009 01:08pm | #1

   Terry,
   As with most building science questions, the answer is, "It depends."

   1. If the insulation is spray polyurethane foam, there should be no degradation in R-value.

   2. If the insulation is cellulose, the degradation in R-value, if any, will be less than if the insulation consists of fiberglass batts.

   3. If the insulation consists of fiberglass batts, the degradation will be less for an attic floor than for a fiberglass-insulated kneewall with no air barrier on the back side of the insulation.

   4. The degradation in R-value will be minor in most cases, but will be greater as delta-T increases (that is, during very cold weather).

   5. Solving the problem by devising a way to install a top-side air barrier is more expensive than just making the insulation deeper.

   6. The bottom line: if you can't afford spray foam, use cellulose insulation. And install it thick — cellulose is cheap.

2. 
   Riversong | Mar 16, 2009 02:47pm | #2

   Terry,

   Where did you find that statement? An air barrier, to be effective, should be continuous (uninterrupted) and contiguous with the thermal envelope (insulation). A single air barrier membrane or assembly rarely surrounds the insulating elements.

   Many modern homes incorporate an interior air/vapor barrier as well as an exterior weather-resistant barrier (WRB) which typically is at least a partial air barrier.

   An unvented cathedral ceiling/roof assembly typically encapsulates the insulating element with air barriers, but a vented attic does not if the insulation is in the attic floor.

   Using vent baffles to direct ventilation air from the soffits to above the insulation helps reduce the "wind-wash" R-value loss.

   More important, though, in most old attics is the thermal bridging of the framing members through the insulation, air leakage paths from the conditioned space into the attic, and uninsulated and unsealed attic hatches.

   For instance, an attic floor framed with 2x6 16" oc with R-19 fiberglass batts would have an effective R-value due to thermal bridging of R-16.9 if the insulation was perfectly installed with no compression or voids (and this does not take into account air movement).

   A 10 sf uninsulated hatch in a 600 sf ceiling insulated to R-38 would have an effective R-value of 23.5 (38% reduction).

   So an efficiency stategy should include preventing wind-wash, sealing all penetrations, insulating and sealing the hatch, and adding more insulation (preferably cellulose) to stop thermal bridging.