How St. Louis could fortify its buildings against the next tornado

This article first appeared in the Solutions newsletter. Click here to learn more and sign up.

Few videos of the May 16 tornado are scarier than the one taken from the top floor of The Hudson apartment complex on DeBaliviere at Pershing. You can watch it here.  In two seconds flat, the roof lifts away and the walls collapse.

It wasn’t an isolated event. Lots of roofs failed in the damage path of this EF 3 storm. At least one of its five fatalities—Deloris Holmes of The Ville—appeared to be caused by roof failure. 

And roof failure is bound to happen again in the next tornado unless St. Louis bolsters its building stock in the years to come. 

What allows for a modicum of hope here is that over the past decade and a half, federal researchers at the National Institute of Standards and Technology (NIST) have learned a great deal about why roofs fail during tornadoes and how to prevent it. They did this by studying the aftermath of the Joplin tornado.

That storm killed 161 people. About 84 percent of those deaths were “related to building failure,” NIST found. The research team discovered, through wind-tunnel experiments, data, and debris, that tornadoes foment an unusual mix of air pressures and wind angles to create “uplift”—a vertical force that buildings historically weren’t designed to resist.

One architectural feature that resists uplift is “continuous load path”: Fastening a building’s components from the roof to the foundation so that when uplift hits, the wind doesn’t tug on the roof alone and pry it off, but rather, tugs on an entire structure that’s anchored into the ground.

The NIST team’s findings on tornado resistance led to a new engineering standard, which then got incorporated into the most recent International Building Code. Thus for the first time, the IBC, a.k.a. “the model code,” had provisions for tornado loads. It required tornado-resistant features for buildings in two risk categories: Category III, which includes schools, nursing homes, and theaters, and Category IV, which includes firehouses, police facilities, and hospitals. The idea is to avoid mass-casualty events and also to preserve the emergency services’ ability to respond to a calamity.

And the storms being designed for in this model code are not the severest, but rather, the most common—that is, EF0 through EF2 storms, a range that encompassed about 97 percent of all tornadoes from 1995 to 2022, according to NIST. One reason for that focus is what the NIST team’s learned in Joplin: “Although the most powerful winds laid waste to buildings near the tornado center, more than 70 percent of the total damage path stemmed from lower wind speeds farther out.” In other words: Even if you design only for winds in the weaker half of the scale, you can still significantly reduce wreckage. 

The city of St. Louis’s building code is based on the IBC 2018, an older version that doesn’t have the tornado standards in it. Asked whether a code change was even under discussion at the Board of Aldermen, a spokesman for president Megan Green said: “Most of our attention has been on providing immediate and short-term relief.” And indeed, even the new model code, if adopted as written, wouldn’t have required The Hudson’s developers, Lux Living—yes, those guys—to make their structure resistant to tornadic uplift, because that newest code excludes from that requirement the vast majority of residential structures.  

But the model code is just that: a model. Cities need not adopt it wholesale. If St. Louis wished to go further and, say, require that all new apartment complexes have load paths of a certain strength, it could do so. It could also require the same of new single-, two-, or multi-family dwellings. Legacy housing stock is by far the norm in the city, so a code change affecting only new builds would take generations to solidify as a tornado shield—but over time, it could, just as after the Great Fire of 1848, the city slowly transitioned from wood to brick, thereby reducing the risk of infernos.

But would tornado-resistant features be worth the cost, given the infrequency of tornados in the city? NIST has put out a document showing a variety of seemingly inexpensive fasteners—metal straps, rods, and brackets—to strengthen load path in wood-frame homes. But one can imagine how, at scale, those costs could add up. Michael Powers, the real estate development manager at Habitat for Humanity Saint Louis, says that creating affordable housing is already a challenge as it is. “The conversation about resilience is a really good one,” he says, “but I’d worry about adding on any additional barriers to construction.” In his view, there should be much more focus right now on the wisdom and cost-effectiveness of home repair. Such work, which includes tuckpointing, could’ve strengthened some buildings against the May 16 storm, and right now, could help displaced residents return home.

As for retrofitting the city’s legacy brick abodes that weren’t damaged on May 16, I asked Tim Becker, a Tower Grove South resident and professional home inspector, how they were built. (He specializes in old homes.) He confirmed that they weren’t designed to resist uplift. “You have multiple fail points with these roofs,” he said. In theory, if you were already replacing your roof, then during that process you could fasten it down with hurricane clips (or “H clips”) without too much extra expense. But removing a roof when it’s not urgent? “That’s a hard sale for most people,” Becker says.

Perhaps it’s like insurance: nobody likes paying their premiums, and some people choose not to. But either way, by the time disaster strikes, you’ve already made your choice. St. Louis has passed into a period in which preparing for the next tornado remains an option—at least for now.