Nola.com: Report: Levee failure could not be foreseen
Gibbons Burke
<http://www.nola.com/archives/t-p/index.ssf?/base/
library-97/1142060616174100.xml&coll=1>
Report: Levee failure could not be foreseen
Task force says unique forces caused breach in
17th Street Canal levee
Saturday, March 11, 2006
By Bob Marshall
Staff writer
A unique combination of stresses that engineers
could not have predicted caused the 17th Street
Canal floodwall to fail and flood thousands of
homes and businesses during Hurricane Katrina,
according to an interim report of the task force
investigating the disaster for the Army Corps of
Engineers.
The report also points to soil subsidence that
left floodwalls and levees lower than design
specifications as contributing to the other
failures and breaches that helped flood 80 percent
of New Orleans and killed more than 1,100
residents in August.
Although independent analysts have blamed the 17th
Street Canal failure on faulty engineering,
including flawed soil investigations by local
firms, the Interagency Performance Evaluation Task
Force, composed of experts from academia and
industry as well as state and federal agencies,
said evidence points to forces that came together
in a combination unique to the science and thus
could not have been anticipated by the system's
design teams.
"I would say it's certainly going to come as a
surprise to many people, if not most people," said
Ed Link, University of Maryland professor and task
force project director.
The group said the causes of the London Avenue
canal floodwall collapses are not yet known and
emphasized that its findings are preliminary.
Bob Bea, a University of California professor who
is part of a National Science Foundation
investigation into the failures, said the task
force's explanation of the 17th Street Canal
breach is lacking.
"It's our jobs as engineers to anticipate the
failure points, and when that doesn't happen,
breakdowns like this occur," Bea said, emphasizing
that he is speaking only for himself and not the
NSF team. "The corps has a documented history
where they say, 'We couldn't have anticipated
this, therefore it was an act of God.'
"An experienced engineer knows he can't accept
that."
Four steps to hell
Interagency task force members said experiments
with sophisticated computer models show the 17th
Street Canal floodwall came down in a four-step
process:
-- As water in the canal rose to 10 feet -- an
unprecedented but not unplanned height -- the
pressure from the water and wind-driven waves in
the canal began to push, or deflect, the concrete
floodwall and its subsurface supporting steel
sheet piling away from the canal and toward
Lakeview.
-- The deflection created space between the wall
and the levee on the canal side.
-- Such flexing is expected by designers, as is a
small opening between the wall and the levee. But
what happened in this case, and was not expected,
was the separation extended the entire length of
the sheetpile wall to 17.5 feet below sea level.
Water rushed into this opening quickly, creating a
channel separating the floodwall from the levee on
the inside of the canal and allowing high water
pressure to travel directly down to the soil
layers beneath the wall.
-- The final blow came when a layer of clay about
15 feet below sea level that extended beyond the
toe of the levee began slipping toward Lakeview,
causing the levee to collapse and the wall with
it.
'Failure mechanism'
The fatal flaw in the weak soils beneath the
structure was not the now-notorious layer of peat
widely cited by independent analysts for months,
the task force said. In fact, the failure surface,
as engineers call it, did not occur under the
levee or canal, but at a level beneath the toe of
the levee and in the yards of homes adjacent to
the canal.
Link said task force tests showed the
soil-strength estimates done by local firm Eustis
Engineering when the walls were built proved to be
more conservative than actual results. Further, he
said there was no method of testing the plans for
a combination of forces that caused the collapse
-- called the "failure mechanism" by engineers.
"We've searched the literature and found nothing
that resembles this," he said. "I'm not saying
nothing exists, but so far we haven't found it."
There was disagreement on that point.
Bea said a 1986 corps study showed such
separations could occur.
"That report was done by the Vicksburg (Miss.)
research station for the New Orleans District, but
there's no evidence it ever made its way to the
(engineering) firms doing the work," said Bea, who
added that a full discussion of the report would
be in the National Science Foundation study to be
published next month.
Corps officials acknowledged the report, titled
"E-99 Sheet Pile Wall Field Load Test Report," but
disputed Bea's interpretation.
Neither the interagency task force nor the corps
dismissed the long-standing criticism that sheet
pilings should have been driven at least to the
bottom of the canals -- a standard engineering
practice -- rather than stopped at 17 feet. While
they agreed deeper pilings generally make stronger
walls, they have yet to run simulations to
determine whether deeper pilings would have
prevented this collapse given the other conditions
now known.
Sinking floodwalls
Soil subsidence levels in a region that was
largely marsh and swamp fewer than 100 years ago
is well known, but the rate of sinkage, which left
many structures below the heights built to guard
against storm surges, apparently took the panel by
surprise. For example, the Industrial Canal
floodwall that was built to 15 feet actually
measured just above 12 feet when Katrina hit, a
loss of 2.7 feet.
Task force teams "documented that many sections of
the levees and floodwalls were substantially below
their original design elevations, an effective
loss in protection," the report said.
Corps officials said the Bush administration has
budgeted almost $3 billion to repair and restore
all levees and floodwalls in the region up to
design heights during the next two years.
Louisiana State University professor Ivor van
Heerden, a member of the state team investigating
the failures, said he was not surprised by the
report and generally agreed with its findings. He
said the corps started using updated elevation
data only five years ago, even though the state
had been urging a change for years.
"So the fact that the corps have found some levees
lower than they should be reflects local
subsidence but also that they built them lower
than they should be because they would not update
their datum," van Heerden commented by e-mail. "It
is cheaper to make a wall 12.8 feet tall rather
than one 14 feet tall!"
Further, van Heerden wrote, "whether the fail
plane (on the 17th Street Canal) was in peat or
clay is really academic. The structure underwent
catastrophic structure failure, the same for the
two breaches on the London Avenue Canal."
Hassan Mashriqui, an engineer and storm modeler at
the LSU Hurricane Center, said he would be
cautious about estimates of wave forces inside the
17th Street Canal because his findings show that a
huge pile of debris that stacked up against the
Old Hammond Highway bridge across the canal
probably blocked much of that force.
. . . . . . .
Bob Marshall can be reached at
rmar...@timespicayune.com or (504) 826-3539.