ABSTRACT: Simple roof geometry is desirable in snow country design, but on some projects complex ge-ometry is required in order to meet the needs of function, aesthetics, or marketability. These projects are of all sizes, ranging from single-family homes to very large multi-use facilities. Unbalanced snow distribution on these complex roofs requires study of local weather data to formulate a basis for predicting the snow distribution on these roof geometries.

While wind tunnel, water flume and computer modeling can be helpful in determining drift formation, these processes are, in many instances, cost prohibitive for all but large projects. Furthermore, these techniques normally do not take into account the accumulations factor resulting from multiple storm events and freezing and thawing.

A design case study will be utilized to graphically present the methodology described.


In the United States, roof design snow load values are based upon codes and standards. These codes and standards normally utilize a 50-year mean recur-rence interval to determine acceptable snow loads. This means that there is only a 2% statistical chance that the code-based design snow load will be exceeded within the 50-year time frame. These design snow loads are therefore on the more extreme end of the scale for a given locality in order to protect the health, safety and welfare of the project and its occupants. However, complex roofs can develop snow loads beyond code limits.

Roof snow distribution (hence, load) is virtually never uniform - it is always in an unbalanced condition. These conditions may outweigh all others. To what extent the unbalanced roof snow conditions occur is typically determined by two sets of factors.

The first set of factors includes the project's roof shape, complexity and type. The more complex the roof design, the more likely large, unbalanced snow loads will occur on certain areas of the roof. The type of roof (i.e. cold roof/warm roof), roof projections (vents/fire/flues) and roofing material will affect the roof snow retention (hence loading) assuming a given roof design and compass orientation at a given site. Most of these roof design factors are controllable by the design team.

The second set of factors affecting roof snow distribution is the weather occurring at the site. In most situations the microclimate elements, making up the term "weather," cannot be altered – the design must conform to them. It behooves the design team to analyze the site’s past weather records to understand how these weather elements can exacerbate unbalanced snow distribution on the roof to where the nominal snow loads exceed the normal unbalanced code/standards design criteria. Failure to do so may have a very serous impact on the structure.

To illustrate the application of empirical weather analysis relating to unbalanced roof snow distribution on a complex roof, the following design case study is presented.

1 2 3 4 5 6

Snow Distribution on Complex Roofs
Presented at the Fifth International Conference on Snow Engineering, Davos Switzerland, July 2004
by Ian Mackinlay FAIA and Richard S. Flood AIA/CSI.

Case study building designed by the architectural office of Bohlin Cywinski Jackson, Seattle, Washington, U.S.A. The photograph in Figure 1B is by Ian Mackinlay. Illustrations for Figures 3, 4A, 4B, 6, and 7 are by Molly Blunden.