The Simon Skjodt International Orangutan Center, which opened at the Indianapolis Zoo in May 2014 features an 8,000-square foot extensive green roof. Architect: Browning Day Mullins Dierdorf. Photograph by Daniel Overbey.

Originally published on June 12, 2019 by BNP Media through the Building Enclosure Blog.

One of the consequences of blogging is the peer review of the internet. I am always happy to take comments. I value the dialogue because of the learning involved and it reaffirms that what I am offering via this platform is reaching (and hopefully helping) other professionals.

Recently, I received a message regarding a blog post I created seven years ago. I guess it is true that nothing on the internet ever goes away. The message questioned the propriety of my position regarding green roofs as a dedicated insulation layer in a roof assembly.

Back in June 2012, I stated that green roofs should not be considered as dedicated insulation layers and thus should not be counted as part of a roof’s total assembly R-value:

Completely dry growing media (the “dirt” of the green roof, so to speak) has a minimal R-value (typically about 0.5/inch). Plants need water to grow, so the green roof is designed to retain some degree of moisture. There is almost always some amount of water existing in a green roof and that moisture prompts a highly accelerated thermal energy transfer via advection. Thus, one can not really count a green roof assembly towards a roof’s total R-value.

I stand behind the technical basis of the statement and the blog post, but I do understand why I received some criticism. While I maintain that green roofs are not great insulators in a relative sense, my blog post was over-simplistic.

Let’s go a bit deeper into this issue.

Thermal performance of green roofs

There are several long-standing studies that consider the thermal performance of green roofs. Consider one such study from Carnegie Mellon University by Dyanna Becker and Daisy Wang (2011) in which the green roofs they studied showed improved thermal performance compared to nearby control roofs. According to the study:

The Hamerschlag Hall green roof was found to lose 26% less heat than the control roof in heating months. The Allegheny County Office Building green roof was found to lose 8.2% less heat than the control roof in heating months and gain 75% less heat than the control roof in cooling months. 

While such an impact on heat transfer is potentially beneficial to the energy performance of a project, it does not necessarily mean that green roofs are good insulators.

Thermal conductivity (k) is the heat transferred by conduction through a substance of a given thickness in a given time when a given temperature difference is applied to given area. It is the basic measure of the conductive heat transfer of a substance. The I-P units are Btu–inches per square foot per degree Fahrenheit temperature difference per hour (or Btu•in./ft²•hr•°F).

All materials exhibit some degree of thermal conductivity. Recall the difference between insulation and thermal mass. Insulation wants to impede conductive heat transfer. Conversely, thermal mass wants to allow conductivity it because the faster the material can import thermal energy from the collection point at its surface, the more efficient the thermal mass will be at storing and distributing heat. An inch of extruded polystyrene insulation board may have a thermal conductivity of 0.2 Btu•in./ft²•hr•°F, whereas an inch of concrete may allow 12 Btu•in./ft²•hr•°F.

The thermal resistance of soil used in Becker and Wang’s heat flux analysis was based on research from D.J. Sailor, et al, (2007), which indicated a fluctuation in thermal conductivity based on the moisture content of the green roof’s soil. The thermal conductivity of the green roof soil (in I-P units) ranged between 1.2 (at 0% saturation) and 2.84 (at 82% saturation). The soil would not serve as an especially good insulation material (the range of equivalent R-values would all be below R-1 per inch compared to extruded polystyrene at R-5 per inch), but it would not be nearly as conductive as typical concrete either. According to Sailor, a green roof’s soil performs somewhere in a range between oak and adobe in terms of conductivity.

A green roof is not insulation. But it can help your building in terms of energy performance.

When researchers cite the insulation qualities of green roofs, understand that they are not necessarily claiming that green roofs are great insulation materials. Rather, they are citing the energy benefits derived by several dynamic effects:

Shading effect: Foliage and soil will keep solar radiation off of  the surface of a roof. This effect is going to help during the cooling season in many climate regions.

Creating a thermal mass “buffer” against daily fluctuations: Green roofs will have the dynamic effect of reducing heat transfer across a roof assembly, thus blunting the effect of diurnal temperature swings during the peak of the heating and cooling seasons.

Retarding heat transfer by advection: The transfer of heat by the flow of a fluid (advection) is a major reason why one will find an inconsistent range of equivalent R-values listed for various green roof systems.

Causing evapotranspiration: During the cooling season, evaporation and plant transpiration (evapotranspiration) will occur as moisture trapped in the root zone of plants and droplets in the foliage layer evaporate and dissipate. As the moisture content of the soil is reduced, so too will this cooling effect.

There are many benefits to green roofs

I hope that my 2012 post has never misled anyone. If you are looking to increase insulation values, green roofs are not your best option. However, there are potential energy-performance benefits to be reaped by green roofs in addition to the benefits related to rainwater management and restored ecosystem services on a project site. Green roofs can make great financial sense and improve the performance of building projects.

Becker, D. and Wang D., 2011. Green Roof Heat Transfer and Thermal Performance Analysis. Carnegie Mellon University

Sailor, D.J. et al, 2007. Thermal Property Measurements for Ecoroof Soils Common in the Western U.S.. Energy and Buildings 40 (2008), pp. 1246-1251