Nick Bradford created a mathematical model that provides a quicker, less expensive assessment of a home's wind-load capabilities. [Times photo: Fraser Hale]

How Times study was done

Click here to see the results of the Times study

Times staff writer

© St. Petersburg Times, published June 27, 1999

Click here for main story The long, lonely fight Insurance companies will still pay Repairs can cost thousands What you can do about your own house The experts Letter from Lennar Homes Summary of engineer, builder responses

"I'll bet you $5,000, if you pulled 10 house plans, not one would meet the wind-load requirements," engineer J.C. Russello told the Times in February. Russello was a consultant to a Hernando County resident who thought his home was poorly designed and constructed (see story, Page 9A).

The newspaper didn't take Russello up on the bet, but by mid-March, many wind-engineering experts had told the Times they thought there was a widespread problem with gable-wall designs outside of South Florida.

That's when the newspaper consulted Rajan Sen of the Department of Civil and Environmental Engineering at the University of South Florida. Sen recently sought a grant to study the effectiveness of the fixes that were supposed to make older gable-roofed homes wind-resistant. Sen referred the newspaper to one of his students, doctoral candidate Nick Bradford, a professional engineer who runs his own consulting firm in Tampa.

Evaluating a home's wind-load capabilities can take 10 hours per house. To accommodate the Times' desire to study a broad sampling of homes around the state, Bradford created a mathematical model that provides a quicker, less expensive assessment. It is also less precise than a full-fledged engineering analysis of each house plan.

Bradford worked with plans gathered by the Times from building departments in nine coastal counties in Florida. So that Bradford would not know whose plans he was reviewing, the newspaper removed the names of the builders and design professionals.

Bradford's methods yield wind-load numbers that can be compared with the Standard Building Code's 100- to 110-mph design standard. The 100-mph standard in the code presumes the house sits in an open field, where the wind is not deflected by trees or other houses. Because most houses are in subdivisions, where foliage and other structures work as buffers, a house designed to the code's 100-mph standard will likely withstand much higher winds.

In Bradford's analysis, each assumption he used to create his model favored the builders' designs. In fact, the consulting engineer for the Florida Board of Professional Engineers, James Owen Power, considered Bradford's assumptions too kind.

So did Russello, the Hernando engineer, who designed a house in Citrus County that was included in the Times study.

"There's no way these houses have the design capacity shown in the study," Russello said. "I would say the capacities shown in the study are all too high, including mine." (Bradford's analysis showed a design capacity of 178 mph for Russello's house. Russello said it was closer to 135 mph.)

Nevertheless, not a single engineer, architect or builder whose house design was found to be faulty agreed with the study.

Many echoed Stuart Miller, president of Lennar Homes, who wrote that the Times' conclusions were in "direct contradiction with long accepted and approved engineer(ing) practices."

However, when the engineers and builders were asked to provide mathematical calculations, showing how their design methods met the Standard Building Code's wind-load requirements, most declined.

Lennar Corp., which built more than 2,000 Florida homes in 1997, did provide two sets of calculations to the newspaper. Experts who reviewed the work on behalf of the Times said both designs fail to meet code.

At 2:30 p.m. Friday, Lennar informed the newspaper that it had presented calculations on one of the designs to James Mehltretter, one expert who had reviewed the Times analysis, and agreed with its conclusions. Mehltretter, after examining Lennar's calculations, said the home met code.

The calculations presented to Mehltretter were performed by engineer Lance Atkins. They were not the same calculations earlier provided to the Times, which had been performed by engineer Sam Greenberg.

Friday evening, engineer Ronald Zollo, who reviewed Lennar's plan with Greenberg's calculations, reiterated his finding that the home fails to meet code.

"Nothing's changed," Zollo said. "The bottom line is, are you willing to defend what you've concluded. And I'm willing to do that."

The calculations on the second Lennar home were reviewed by Power, at the request of the Engineering Board. Power found the home did not meet code and that the calculations did not reflect what is shown on the building plans.

One "detail is appreciably different from what is indicated on the Permit Drawings," Power wrote. The building plans also showed a masonry beam; the calculations are based on a concrete beam, Power wrote. A pre-cast concrete beam is stronger than a masonry beam.

Nick Nicholson, a Brooksville engineer, gave the newspaper four sets of computerized calculations on plans he'd certified for Cozy Homes, Carroll Custom Homes and Blue Stone, three North Suncoast builders.

Nicholson also provided a letter from Edward Hubert, the creator of the computer program Nicholson used in his calculations. Hubert said Nicholson's "calculations are correct and comply with the Standard Building Code."

Power, the Engineering Board's consultant, studied the plans at the board's request and found all four failed to meet code, Nicholson's computer reports notwithstanding.

"Let me give you an analogy," Power said. "Suppose you wanted to check the strength of a chain. There are a number of sophisticated ways to do that. But if you notice that a link is missing, it doesn't much matter how you analyze it. The chain fails.

"The link missing in this case is the bracing at the top of the gable-end walls."

Nicholson's attorney, Douglas Bevins, said the blueprints were not the only documents used to construct the house. There were also what he called "builder plans." Together, those documents show the homes meet the code.

Bevins said the newspaper was attempting to dramatize an "academic debate (about) the proper design of gable ends . . ." at the risk of Nicholson's reputation.

Many of the engineers whose plans were included in the analysis designed the gable walls as though they were like other walls -- that is, adequately braced, top and bottom. It's a standard assumption. Even Bradford used it to seal plans before he and Sen began to research the strength of gables. If that assumption was correct, as Bradford's model shows, all the houses in the study would have met or exceeded the code's wind-load requirements.

But it is just that assumption that is rejected by engineers and wind-load experts who evaluated Bradford's analysis at the Times' request and concurred with its conclusions.

The Florida Board of Professional Engineers, which licenses and disciplines engineers for the state, asked its consultant, Power, to assess the Times study. He, too, said its conclusions were sound.

© St. Petersburg Times. All rights reserved.