Deconstructing the Window

How to Achieve Optimum Thermal Performance



Photo of Helen Sanders

Helen Sanders

General Manager

Technoform North America

Photo of Jörg Lenz

Jörg Lenz

Technoform Glass Insulation

Photo of Petra Sommer

Petra Sommer

Technoform Glass Insulation



Over the years, the industry has relied heavily on the increasing performance of low-e coatings to drive window U-factors (thermal transmittance) lower. However, the full performance of the window is driven by the frame and the edge of glass conductance, and neglecting the window perimeter can result in poorly performing fenestration systems that do not meet US national model energy codes, are uncomfortable to sit next to, and result in problematic condensation.

In order to help designers specify the appropriate fenestration components and performance, the factors that contribute to the U-factor and condensation resistance of a window are deconstructed. The sensitivity of edge of glass and window U-factors to different edge of glass design parameters, such as frame bite relative to the sight line, sealant height, spacer conductivity etc. are also examined. The data show that when reviewing the thermal performance of different edge of glass solutions, it is important to compare “like with like” relative to sealant height, frame bite and desiccant fill, since all these factors impact thermal performance. Also, it is critical to ensure that thermal simulations are done using edge of glass constructions that include the actual amount of sealant which will be used in the final project.

To meet the U-factor requirements of the newest US model building codes (IECC 2015 & 2018, ASHRAE Standard 90.1-2016), in climate zones 4 (Washington DC) and 5 (Chicago) fixed metal fenestration must now use a thermally broken frame and dual pane low-e insulating glass, plus one of the following: A warm edge spacer, argon gas fill, fourth surface low-e, or a higher performance (larger) frame thermal break. In climate zone 6 (Minneapolis) two of these additional strategies are necessary to achieve compliance with a thermally broken frame and dual pane low-e. In climate zones 7 and 8, all four of these additional strategies are needed to make a dual pane system work, or alternatively designers may select a high performance thermally broken frame with triple glazing.


With the increased focus on providing daylight and views for building occupants, as well as the need to reduce energy consumption and improved thermal comfort, a significant tension has been

Members Only


Figure 1: A schematic diagram of the frame and edge of glass showing the three different methods of heat transfer.

Heat is transferred at the window perimeter by conduction through

Members Only

U-Factor: Edge of Glass Sensitivity Analysis

The impact on linear thermal transmittance (\( \Psi \)) of sealant height (Fig.3), frame bite (Fig.5), desiccant and desiccant quantity has been calculated according to DIN ISO 10077-2 for an

Members Only


The analysis shows that the linear thermal transmittance of the edge of glass is significantly impacted by the amount of sealant used and the bite of the frame, and to

Members Only

Implications for Compliance with Energy Codes

The edge of glass, center of glass and frame performances all contribute to the overall performance of a window, and with the increasing stringency of energy codes on both sides

Members Only


It is important to take a balanced approach to picking strategies to meet the U-factor requirements. In order to make the center of glass performance have the most impact, improving

Members Only

Rights and Permissions

Arasteh, Dariush K., Selkowitz, Stephen, Apte, Joshua, and LaFrance, Marc, "Zero Energy Windows" LBNL report 60049, ACEEE Summer Study on Energy Efficiency in Buildings, 2006.

Culp, T., Code status report to Glass Association of North America, 2017.

International Energy Agency, Energy Efficiency in the North American Building Stock, Executive Summary, 2007.

Malekfazali, A., Effect of Window U-factor on Building Perimeter Zone Energy Use Intensity, to be published.