Recent tornadoes call attention to wind load provisions

While driving through Central Illinois recently, I was struck by the beauty of a wind farm. The turbines' vanes responded to the wind traversing the open plains with grace and dignity, providing a visual indication of the wind's movement that you could not otherwise see.

In that scenario, the wind was a positive element for all concerned, acting as a nondepletable resource for generating power onsite. But the opposite can also be true. The wind can destroy trees, signs and buildings. For this reason, we design buildings―including components and cladding such as windows―to withstand strong wind forces.

The art of designing building components to resist the wind is still evolving. As time goes on, we learn more about the wind and are better able to predict its effects.

Recently, the wind load provisions of ASCE 7 Minimum Design Loads for Buildings and other Structures were revised. ASCE 7 is the standard referenced in the International Building Code for the determination of design loads for buildings.

The revision to ASCE 7 reflects a change in the model used to predict wind loads. Previous editions relied on what was known as "the 50-year wind" as the wind speed to which buildings were designed to remain serviceable. The phrase, "the 50-year wind," meant there was a 2 percent probability that wind speed would occur in any given year. Given a long enough time period, that design wind speed would occur at 50-year intervals. For example, over a period of 1,000 years, we would expect that design wind speed to occur approximately 20 times.

The new edition of ASCE 7 is based on wind speeds with a much lower probability of occurring. The exact model to be used depends upon the potential threat to life safety. Those considered a low threat to life safety―such as agricultural buildings―are designed to resist a 300-year wind. Buildings whose collapse is a significant threat to life safety, such as those occupied by large groups of people, hospitals, or police and fire stations, are designed to resist a 1,700-year wind. All other buildings need to be designed to resist a 700-year wind, according to the new ASCE 7.

I was beginning to think I had my mind wrapped around this concept, when an EF-4 tornado hit the St. Louis airport (Lambert Field) on Good Friday. The designation of EF-4 means winds in excess of 168 mph. Recently, the revised ASCE 7-10 established design wind speeds of 120 mph as a 1,700-year wind in the St. Louis area. This is considerably less than the 168-plus mph wind actually experienced.

Alright, I thought. So a wind of extremely low probability hit Lambert Field. Such things can happen. But then I learned that an EF-4 tornado also hit St. Louis County, including Lambert Field, in 1967. In fact, since the late 1800s, a total of four EF-4 tornados have hit St. Louis County. On average, more than 200 tornados hit North America every April. Since 1855, more than 90 tornadoes "of significant size"―EF-2 or greater, with wind speeds in excess of 113 mph―have hit the eight-county area of Illinois that includes Chicago. The tornadoes that occurred across the South the end of April demonstrate these types of events do occur and the damage they can do. That hardly seems like a 1,700-year-return period to me.

I was wondering what I was missing with regards to understanding the new model, so I asked a couple of experts in the field.

What I was told is that even though four EF-4 tornados have hit St. Louis County in the past 100-plus years, the area affected by each one was so small that the specific locations they hit actually experienced only one tornado. In other words, the areas they hit did not overlap.

I suppose the same might be true of the 200-plus tornados that touch down in North America every April. North America is a pretty large area, and a tornado's path is often less than ½ mile wide. Plus, not all of those tornados reach EF-4 intensity.

But that doesn't change the fact that two EF-4 twisters have hit Lambert Field in St. Louis within the past 50 years: one in 1967 and one on Good Friday of this year. Perhaps they didn't destroy the same buildings, but they damaged the same airport. I must confess, I am not convinced of the adequacy of the modeling being used.

Lack of testing

Recently, I had the privilege of visiting the Institute for Building and Home Safety's new testing facility in Richburg, S.C., with attendees of the Southeast AAMA Spring Meeting. This facility can simulate wind speeds of up to 140 mph over a two-story building. These winds are representative of a Category 3 hurricane.

When asked, however, IBHS representatives admitted they did not have any immediate plans to study the effects of tornadoes at their new facility. The reason is the wind patterns caused by tornados are quite different from those caused by hurricanes. The facility was not built to simulate those types of wind patterns, and it's not fully understood what type of equipment or facility would be required to accurately replicate them. On a related note, the American Architectural Manufacturers Association recently released AAMA 512-11 Voluntary Specifications for Tornado Hazard Mitigating Fenestration Products. The specification uses existing test methods and other procedures to qualify windows and other glazed fenestration products for tornado hazard mitigation.

Over 30 years ago as an undergraduate at Northwestern, I had the opportunity to participate in a research project that focused on determining which locations would provide the wind speeds necessary to make wind-generated power at that site economically feasible. My role was to plot data by hand. At that time, plotting by hand was faster than using the computers available.

Perhaps in another 30 years, we will know how to replicate tornados in a testing facility. Perhaps we will be able to design buildings to resist these destructive storms. I hope so.