What Went Wrong > Miami Herald, December 20, 1992 > Page 2
SPECIAL REPORT: WHAT WENT WRONG
FAILURE OF DESIGN AND DISCIPLINE
JEFF LEEN, STEPHEN K. DOIG and LISA GETTER Herald Staff Writers
Section: SPECIAL SECTION
Like a latent fingerprint found at a crime scene, a clear pattern has appeared in the vast sprawl of destruction left by Hurricane Andrew.
The storm's deadly imprint emerged from a three-month Miami Herald investigation that used computers to analyze 60,000 damage inspection reports.
A computer created a color-coded map showing how 420 neighborhoods weathered the storm. When a map of estimated wind zones was superimposed over the damage, the pattern became unmistakable:
Many of the worst-hit neighborhoods were far from the worst winds.
The damage wasn't consistent -- some ravaged neighborhoods sat next to others with much less destruction.
The analysis turned up another startling fact: Newer houses did worse than older ones.
A lot worse, in fact. Houses built since 1980 were 68 percent more likely to be uninhabitable after the hurricane than homes built earlier.
The age pattern settles a debate that erupted after Andrew's winds died: What was responsible for most of the damage? Only the wind? Or shoddy construction, faulty design and flimsy materials?
For most newer homes, how they were built was more important than where they were located -- and thus how they were affected by the wind -- in determining the extent of destruction. In other words, man is to blame for a considerable part of the damage.
"We're talking about $100,000 to $150,000 losses which should have been $25,000 to $50,000 losses," said Dean Flesner, a State Farm vice president.
"We're talking about families whose lives have been totally destroyed because their home is uninhabitable, versus families who probably could have remained in the home while repairs were made."
Why did houses built since 1980 do so poorly in the storm? To find out, The Herald investigated scores of building and design failures, as well as the county's system for preventing them.
There was ample evidence of breakdowns in the construction and inspection safeguards meant to protect the public from exactly the sort of devastation dealt out by Andrew.
* A close examination of eight storm-damaged subdivisions built by some of Dade's largest developers revealed houses shot through with so many construction and design flaws they became easy targets for the hurricane.
* Building inspectors, faced with a boom in construction, were pressured to perform up to four times the number of inspections that should properly be done in a day.
"You don't build a bad house or a bad building except on purpose," said Harley Lasseter, 87, a member since 1959 of the county Board of Rules and Appeals, which oversees the building code. "They don't go up accidentally. Sorry workmanship and the supervision is bad, or someone has found some way to get around the code."
Builders angrily deny that shoddy construction was a significant factor in determining hurricane damage. For them, the hurricane alone was the culprit.
"Why are we still picking on the hurricane when we have so much rebuilding to do?" said William Delgado, executive director of the Latin Builders Association. "Why are we nitpicking on the little things? For how long is the media going to try to crucify the builders who really had nothing to do with a hurricane that exceeded code?"
But the computer analysis shows widespread destruction in areas where the sustained wind appears to have been far below the 120-mile-per-hour standard mandated by the South Florida Building Code, according to preliminary determinations by scientists working for the National Oceanic and Atmospheric Administration.
The findings are supported by an independent engineering study obtained by The Herald. The study, commissioned by a major insurance company, estimated maximum sustained wind above 120 mph in only a relatively small area south of Cutler Ridge encompassing Princeton, Naranja Lakes, Homestead and Florida City.
The insurance company's engineers inspected 121 houses in areas where the sustained winds were estimated below 120 mph and concluded that 70 percent had damage traceable to code violations.
"Up until the storm the criteria for acceptance was whatever you got by with on the last job because the inspector didn't catch it or there wasn't an inspector," said James Marks, a Coral Gables structural engineer who has inspected about 100 homes damaged by Hurricane Andrew for homeowners making repairs and insurance claims. "And that became the standard of the industry.
"The builders are far from being adequate. Why was this horrible construction done? Who allowed it and who permitted it to happen? Let's face it. It's there."
The Dade grand jury, assigned to investigate what went wrong after Andrew, came to the same overall conclusion:
"While we, as a community, have suffered greatly as an unavoidable result of Hurricane Andrew, this suffering was aggravated by the systematic failure of our construction industry and building regulation process," the grand jury stated last week in its final report. "Had the failures not existed, much of this suffering would have been prevented."
Much damage occurred in areas where wind was under 120 mph
After the storm, many victims could only stare at their ruined houses and wonder why their neighbors nearby fared so much better.
Bad construction or big wind? Poor design or weak building materials?
Experts rushed forward to explain. Meteorologists talked of 200 mph wind "streaks" that tore certain houses apart. Engineers saw numerous construction flaws in the rubble.
In each case, the expert was forced to draw quick conclusions from a scant sampling of data. No one knew the overall pattern. No one had hard data to back up scores of impressions from the field.
After a peak gust reading of 163 mph was registered at the National Hurricane Center in Coral Gables, many believed that a killer storm simply flattened South Florida, like a 160-mph rolling pin.
"My personal opinion, when you're hit with winds of up to 160 miles per hour, something's got to give," said Eduardo Roca, a general contractor and former Dade building inspector.
Metro Mayor Steve Clark: "There's nothing we could build today with home construction that would stand. I suspect the wind was over 200 miles per hour."
"This was not a killer storm as the weather bureau would have you think," said Marks, who has run his own engineering firm for 21 years. "The engineering committee on wind, who are the experts in the world, said it was 110 to 125 miles per hour. Almost all of the damage I saw occurred well below 120 miles per hour."
In fact, except for a few isolated areas, most of South Dade experienced nothing like sustained winds of 120 mph.
The National Hurricane Center has determined that Andrew's maximum sustained wind -- one minute average -- was 145 mph. The maximum peak -- two-second average -- is believed to have been 175 mph. But these maximums were restricted to small parts of the county.
"The higher the wind, the smaller the area it affects," said Peter Black, a scientist with NOAA's Hurricane Research Division on Key Biscayne.
Even the scientists can't agree on how strong the wind was
Still, the wind debate will continue long after all the hurricane debris has been cleared away. Part of the problem is that no one has any definitive wind data -- Andrew's wind profile was not captured on any Doppler radar that would show more precisely where the worst winds blew.
Instead, wind scientists must try to measure the storm's strength by analyzing a wide variety of partial radar images,
pressure readings and a patchwork of wind readings from a collection of anemometers that differ greatly in quality.
That process is still going on, and won't be completed until sometime next year.
Scientists have another way of measuring the strength of storms: carefully studying the damage on the ground, a process known as F-scale analysis. By looking at houses that lost their roofs or walls, the scientists can gauge the speed of the wind that did the damage.
NOAA-supported scientists Greg Forbes and Roger Wakimoto conducted such a study of South Dade after Andrew, along with Ted Fujita, the University of Chicago professor who invented F- scale analysis.
The analysis allowed them to map zones showing the differing intensity of Andrew's wind.
After the storm, another and quite different damage analysis was launched as a small army of Dade County building inspectors went door-to-door recording whether houses were habitable. The inspectors did not visit mobile homes, which were almost uniformly destroyed in the storm.
The massive inspection process is now nearly two-thirds complete. To conduct its computer study, The Herald purchased the completed data from the county, more than 60,000 inspection reports, and ran them through its own computers. The damage reports were merged by address against other county data, such as the property tax roll and the building and zoning database.
NEWER HOMES HIT HARD
Homes built since '80 more likely to be found uninhabitable
The Herald computer analysis compared the frequency of uninhabitable homes against all available variables: price, house size, lot size, and so on. When the computer printouts were examined, the only variable that showed a consistent pattern was the year the house was built.
To check that this age pattern wasn't caused by some quirk in the selection of houses for inspection, The Herald performed a test computer run using the unlikely assumption that every home south of Kendall Drive that hadn't been inspected so far was habitable. Even if that assumption were true, the computer showed homes built since 1980 still would be 63 percent more likely to be uninhabitable as older homes.
The Herald analysis took into account the wide range of wind speed across South Dade by looking at damage frequencies within each wind zone mapped by NOAA.
For instance, in the highest-wind zone, a relatively small elliptical region of a few square miles centered near Naranja Lakes where three-fourths of the homes were left uninhabitable, the age difference is relatively narrow: 84 percent of homes built since 1980 were left uninhabitable, versus 70 percent of those built earlier.
Significantly, the age pattern of inspected homes becomes clearer as you move away from the storm's strongest winds. In the zone where NOAA scientists determined the wind was weakest, between Kendall Drive and Southwest 184th Street, 33 percent of the inspected homes built since 1980 were uninhabitable, compared to only 10 percent of the older ones. That means the newer homes there were more than three times as likely as older ones to be heavily damaged.
Compare that striking difference, for instance, with the narrower damage gap for price: 35 percent of inspected homes assessed at under $100,000 were uninhabitable, compared to 21 percent of more expensive homes. Even that gap is affected by age: inspected homes built since 1980 and valued at more than $100,000 were more likely to be uninhabitable than were older and less expensive homes.
There are, of course, exceptions to every rule, instances where tiny variations in wind or design or materials led to the destruction of small groups of houses while homes on the same street went unscathed. Such small wind phenomena are extremely difficult to trace.
But wind patterns that devastate entire subdivisions over square-mile-sized areas are another matter entirely. Such a large-scale wind effect could be accomplished only by a extremely high sustained wind or a tornado. Both leave marks that are easy for scientists to trace.
To find how the wind affected larger areas, the computer rated subdivisions according to the percent of uninhabitable homes reported by inspectors. Using this method, The Herald was able to produce a map showing how 420 subdivisions fared in the storm.
Three areas had at least 90 percent of homes ruled uninhabitable
With two key anomalies, the map shows that, moving south, subdivisions generally sustained more damage the closer they were to the center of Andrew's eye.
Three areas were hardest hit, with 90 percent or more of the homes uninhabitable over square-mile-sized areas: the Naranja Lakes-Princeton area, nearest the storm's center, and the two anomalies: the area off Old Cutler Road south of the Deering Estate and the Country Walk area.
Naranja Lakes was no surprise. Shortly after the storm, Hurricane Center Director Robert Sheets declared it the high wind area. The NOAA scientists doing the wind-damage analysis agreed, rating the area F-3, meaning peak winds -- the "streaks" that have received so much attention -- gusted to more than 175 miles per hour, Black said.
This extreme high wind area was relatively small -- about three square miles, and the high-wind streaks within it were even
Continued on Page 3SR
Copyright 1992 Miami Herald