We were asked why the U-value of a glass is different when the glass is installed horizontally.
There are 3 heat transfer modes: conduction, convection, and radiation. The convection part is dependent on the glass tilt and it affects the glass U-value.
By default, we evaluate the U-value of a glass with the vertical tilt, which is the most common position of glasses. For a horizontally tilted glass, the U-value is significantly greater than the U-value of the same glass with the vertical tilt.
Besides the dependency on tilt, the U-value is also dependent on the glass height. Other thermal properties (e.g. SHGC) are dependent on the tilt too.
However, it does not mean that the glass U-value shall be evaluated with different tilts. There are primarily two applications:
Glass performance rating
Fenestration performance rating
For glass performance rating, it is sufficient to evaluate the glass U-value with the vertical tilt only. With this standardized tilt, fair comparisons can be performed conveniently.
For fenestration performance rating, the glass tilt is considered in the evaluation by default.
The method described in the post is only applicable to one single layer of coating. For a surface coated with two or more layers of coatings, the method can only measure the top layer, but cannot determine the apparent thermal conductivity of multiple layers.
To determine the apparent thermal conductivity, the calculation method needs to be used. In this post, we use a two-layer coating system as an example:
Measure the thermal conductivity of each layer individually (k1 and k2). Note: the samples need to be prepared individually too, as the method can measure the top layer only.
Calculate the overall thermal resistance: R = d1/k1 + d2/k2, where d1 and d2 are the thickness of each layer.
Calculate the apparent thermal conductivity: k = (d1 + d2)/R
For the calculation of the overall thermal resistance, our online ETTV U-value calculator can be used. The R-value result reported is the overall thermal resistance of all layers in the system.
Thermal conductivity, thermal resistance, and thermal transmittance are the 3 commonly used properties related to material or system thermal insulation performance. The 3 properties and their differences are explained below.
Thermal conductivity (K-value)
Thermal conductivity represents the ability of a material to conduct heat.
Thermal conductivity is also called K-value and its unit is W/(m⋅K). The smaller the thermal conductivity, the better the thermal insulation performance.
Listed in the table below are the typical thermal conductivity ranges of selected building material types.
Building material type
Typical thermal conductivity range
Insulation material (e.g. polyurethane/polystyrene foam, mineral wool)
0.02 – 0.04 W/(m⋅K)
Wood, plywood, and gypsum board
0.1 – 0.5 W/(m⋅K)
Coating (e.g. paint), plastics, and rubber
0.1 – 0.5 W/(m⋅K)
Concrete (light weight or heavy weight), brick, and tile
0.5 – 2.5 W/(m⋅K)
Metal (e.g. stainless steel or aluminum)
15 – 200 W/(m⋅K)
Thermal conductivity is typically for homogeneous materials. For inhomogeneous materials (e.g. concretes or composite panels), their average thermal conductivity is referred to as the apparent thermal conductivity.
Thermal resistance (R-value)
Thermal resistance represents the ability of a material layer to resist heat transmission. Thermal resistance is calculated as:
Thermal resistance is also called R-value and its unit is (m2K)/W. The greater the thermal resistance, the better the thermal insulation performance.
Thermal resistance is always in terms of one or multiple material layers with fixed thicknesses:
Material layer: thermal resistance is applicable to layer-by-layer structures only.
Example 1: for a wall system with 3 layers: concrete + rock wool insulation + plaster, there is a thermal resistance for each material layer (or multiple layers combined together).
Example 2: for studs, frames and fasteners, and other materials without layered structure, there is no thermal resistance for such materials
Fixed thickness: thermal resistance is dependent on thickness.
Example: the insulation performance of a thin insulation material could be worse than a thick conductive material (e.g. 1 mm thick polystyrene foam vs. 100 mm thick glass)
Thermal transmittance represents the ability of a wall, roof or fenestration system to transmit heat. Thermal transmittance is calculated as:
Thermal transmittance is also called U-value and its unit is W/(m2K). The smaller the thermal transmittance, the better the thermal insulation performance.
Thermal transmittance is always in terms of a complete wall/roof/fenestration system and air-to-air thermal transmission.
Complete wall/roof/fenestration system: thermal transmittance is applicable to a complete wall/roof/fenestration system only.
There is no thermal transmittance of an individual material layer unless the wall/roof/fenestration system is formed by 1 material layer only
Example 1: for a wall system with 3 layers: concrete + rock wool insulation + plaster, there is a thermal transmittance of the complete system only, but no thermal transmittance of each material layer.
Example 2: for a double glazing unit (DGU) system with 3 layers: glass + air gap + glass, there is a thermal transmittance of the complete DGU system only, but no thermal transmittance of each layer.
Air-to-air thermal transmission: thermal transmittance includes thermal transmission through both indoor and outdoor air layers.
The indoor and outdoor air layers adjacent to a wall/roof/fenestration system introduce some additional thermal resistance too (called surface film resistance). The additional thermal resistances by air layers are always included in thermal transmittance calculations.
Thermal transmittance is therefore always dependent on the environmental conditions (such as wall/roof/fenestration orientation, indoor/outdoor airflow speed, and indoor/outdoor surface emissivity).
Example: a wall system and a roof system made of the same materials may be with different thermal transmittances, due to the different airflow speeds along a vertical surface (for a wall) and a horizontal surface (for a roof).
OTM provides two online thermal transmittance (U-value) calculators:
Thermal conductivity (K-value) vs. thermal resistance (R-value)
Thermal conductivity is independent of thickness; thermal resistance is dependent on thickness.
Thermal conductivity is for homogeneous materials (or averaged for inhomogeneous materials); thermal resistance is for material layers.
Thermal resistance (R-value) vs. thermal transmittance (U-value)
Thermal resistance can be in terms of an individual material layer in a wall/roof system; thermal transmittance is always in terms of a complete wall/roof/fenestration system and there is no thermal transmittance of an individual material layer (unless the wall/roof system is made of a single material layer only).
Thermal resistance excludes the effect of indoor and outdoor air layers; thermal transmittance includes the effects of indoor and outdoor air layers (air-to-air thermal transmission).
Thermal resistance is independent of environmental conditions; thermal transmittance is dependent on environmental conditions.
The discussions above follow typical engineering practices. In the academic context, the practices could be different.
If we use double glazing unit (DGU) glasses as an example, the engineering practice is to use its thermal transmittance (U-value), instead of the other two, for performance evaluations, though in the academic context it is still correct to calculate the apparent thermal conductivity and thermal resistance of a DGU glass.
Listed in the table below are the standard NFRC winter and summer environmental conditions:
Outdoor air temperature
Outdoor wind speed (convection)
5.5 m/s (26 W/m2K)
2.75 m/s (15 W/m2K)
Outdoor sky temperature
Outdoor sky emissivity
Indoor air temperature
ASHRAE/NFRC inside model
ASHRAE/NFRC inside model
Indoor room temperature
Indoor room emissivity
These are the standard environmental conditions defined in the NFRC standards and widely used by the industry. It is possible to define local environmental conditions and use them in a specific region.
In the US, the NFRC winter U-value is used (called U-factor and with the imperial unit).
In Singapore, the NFRC summer U-value is typically used, and, in our test reports, both the winter and summer U-values are reported.
For most glasses, the summer condition U-value is smaller than the winter condition Value. The main reason is that the outdoor wind speed in the summer conditions is lower and it results in a lower outdoor side convective heat transfer rate.
We are now able to test wall system U-value on-site. The test method is based on ISO 9869-1, with some improvements for Singapore’s environmental conditions.
The measurement instrument setup is illustrated below:
The following 3 quantities are measured:
Indoor side wall surface temperature (by a temperature sensor)
Outdoor side wall surface temperature (by a temperature sensor)
Heat flux through the wall (by a heat flux sensor)
The instruments need to be deployed on-site for a few days. The thermal resistance (R-value) of the wall system is calculated from the averaged results. The thermal transmittance (U-value) of the wall system is calculated from the R-value and the surface film resistances defined in the BCA ETTV code.
For better measurement accuracy, a surface electric heater of the size 0.5 m x 0.5 m is attached to the indoor side of the wall system to increase the indoor/outdoor temperature difference across the wall system.