Saw this in a newsletter recently, very relevant info:
A new study has been completed at NCTL Testing in York, PA with AWDI. The tests pitted ASTM 2112 Installation, "Caulk and Walk"™ installation and AWDI's R5 installation protocol. The tests and computations confirm and document the dramatic difference between installed performance of windows as compared to the promised performance determined in laboratory tests where the window alone is tested. These product tests become promises in literature, but unfulfilled too often in the home. The tests conducted at NCTL and in Germany in 2011; conform to ASTM recognized air and water tests and for installed windows and included the gap between the window frame and the opening frame. Final thermal results conform to best preactices as defined by NFRC.
Research work tied to the British Fenestration Rating Council A through G Window Standards, has also confirmed that sealant only installations reduced an "A" window to a "D" window performance. Sealant and insulation in the cavity fare better, but still reduce the overall energy efficiency to well below the performance attained in laboratory testing.
The testing at NCTL of AWDI's contention of reduction of in-service performance from promised performance, established inarguable results that show that standard installation reduces an R-5 window to an installed rating of barely over R-2.
Bottom Line: Government encouraged .30/.30 windows when installed in new construction lose more than 15% of the thermal performance promised, and in replacement, that efficiency loss is closer to 25%. People who buy .30/.30 windows hoping for the energy efficiency they promise - don't get anywhere near it.
The culprits? Failure to re-establish the wall/window weather barriers (there are 5), poor R-Values of sealants and trim, inconsistent application of cavity insulation, air leaks due to thermal bridges like shims and capping, and sealant failures, i.e., poor installation