As I was Saying …

We've been installing roof systems on commercial buildings for a couple of centuries now, so you'd think we'd have reliable data to use when making roof system design decisions.

Well. Consider wind design—one of the most basic elements, presumably, in a roof system specification. The premise is fairly straightforward: We want roofs to stay on buildings during predictable high-wind events within a reasonable margin of error.

Instead, what do roof system designers typically do? They drop in an FM Global requirement for wind-uplift resistance, which, in many cases, is based only on the designer's interest in reducing the remotest possibility of liability. We see many roof systems in places like Indiana and Kentucky, therefore, designed to withstand hurricane-force winds.

And how do we test such roof systems, post-construction, to be sure they've been installed properly? We put a small-scale, negative-pressure assembly—politely referred to as a "bubble"—over sections of the installed roof system. Then, we generate negative pressure by taking air from the bubble and use the results to predict roof system performance in wind events.

As we might have imagined, the results have proved to be about as reliable as Tarot cards—without the fun. We know, for instance, that a 200-pound inspector can deflect a steel roof deck enough to cause test failure. Maybe what we need is an ASTM standard for body mass of roof inspectors who conduct bubble testing.

A recent NRCA survey of members who've experienced the bubble test indicates a failure rate of about 45 percent, including some failures that occurred under the watch of a full-time inspector. If 45 percent of the nation's roofs were blowing off, I suspect we'd be hearing about it.

It's time to apply real science to the endeavor. In the June issue, you learned about the Wall of Wind project being conducted at Florida International University, Miami, which is supported in part by funding from The Roofing Industry Alliance for Progress. It's a huge step in the right direction. There are similar steps being taken in other research arenas, as well.

As our industry becomes more complex, and as research and development departments feel the strain of a slowing economy, we'll need to continue to be sure we develop the right kinds of scientific analyses to make the best roofing decisions. That's easier said than done, to be sure, but has never been more important.

Bill Good is NRCA's executive vice president.