Army Corps Admits Design Flaws in New Orleans Levees

Times Staff Writer

The Army Corps of Engineers acknowledged Thursday that design defects in the levees protecting New Orleans caused the majority of flooding during Hurricane Katrina and that the disaster would almost certainly trigger reforms in how the federal government protected the American public.

The corps said its 40-year effort to construct a hurricane protection system for southern Louisiana had resulted in a set of piecemeal projects that “was a system in name only,” a recognition that a wide range of errors, weak links and incomplete construction was at the heart of the massive damage that occurred Aug. 29.

The corps released a 6,000-page report written in the couched language of government engineers but which delivered a stunning set of findings about errors made in the design of storm walls and earthen levees that failed during Katrina.


The report found that four major breaches of I-walls, a type of concrete storm wall that sits on an earthen levee, caused 65% of the flooding in the New Orleans area.

Although the report’s summary never uses the words “design defect,” corps officials said they now accepted that their work had shortcomings and errors that were responsible, in large part, for the damage.

“We do take accountability,” Lt. Gen. Carl Strock, commander and chief engineer of the corps, said at a news conference, where he was joined by five Army generals, the federal coordinator for Gulf Coast reconstruction and an assistant Army secretary.

The corps, Strock said, “is deeply saddened and enormously troubled” by the disaster.

Strock said the remaining I-walls would eventually have to be replaced because they proved ineffective during Katrina. The primary breaches at the 17th Street, London Avenue and Industrial canals occurred when storm waters were still several feet below the tops of the walls, meaning they failed well below the maximum forces they had been designed to withstand.

The report also puts in the historical record a formal acknowledgment of the scope of the disaster, which killed 1,293 people in the New Orleans area.

“The flooding caused a breakdown in New Orleans’ social structure, a loss of cultural heritage and dramatically altered the physical, economic, political, social and psychological character of the area,” it said. “These impacts are unprecedented in their social consequence and unparalleled in the modern era of the United States.”


Although many of the technical findings had been released in preliminary studies by the corps, the report details significant new information:

* Southern Louisiana is sinking much faster than generally recognized, and levees were at substantially lower elevations relative to sea level than they were designed to be. In some cases, levees were 2 1/2 feet below their designed elevations. Moreover, as other federal agencies recognized the problem in recent years, the corps decided not to reexamine the issue.

* The city’s pumping system, the only way to remove water from below sea level, was not designed to operate during a major storm. Because most of the region’s pumps were inundated after Katrina, the corps was forced to use portable equipment that took 53 days to pump out the city, allowing the flood waters to saturate and destroy structures.

* Twenty-five percent of all the housing in the region was destroyed. In New Orleans, the proportion is believed to be higher. The destruction of homes accounted for 75% of the losses caused by Katrina, estimated by the corps at more than $20 billion. Outside estimates are much higher, exceeding $100 billion.

* A computer simulation showed that if the levees had not failed, the city still would have flooded because 14 inches of rain fell during a 24-hour period, and storm surges went over the levees. If the levees had held, the flooding would have been about one-third of what occurred.

Strock said the repaired sections of levees were now the strongest parts of the system. The repairs were supposed to be completed by June 1, but they are about two months behind schedule. Among ongoing projects is the installation of new pumps at the mouth of the 17th Street Canal. The enormous undertaking required heavy construction work on 169 miles of damaged or destroyed levees.

The report is full of technical language, including descriptions of design flaws as “overestimation of subsurface strength.” Corps officials said the reason for such language was not to reduce the effect of the findings, but rather to set a “nonjudgmental tone.”

Although the breaches caused most of the flooding, a significant part of the Katrina damage occurred because the storm was larger than the system was designed to withstand.

Although wind speeds had dropped sharply by the time the hurricane hit New Orleans, its power over the Gulf of Mexico had created the largest ocean surge to ever hit North America, the report said.

The surges along the Gulf Coast reached 28 feet, and waves of 14 feet came on top of the surges, creating a 42-foot wall of water. The ocean level was lifted by a drastic fall in the barometric pressure, one of the sharpest drops ever recorded in the region, Strock said.

The same failure mechanism was to blame for all of the I-wall breaches. As water rose against the walls, it caused them to tilt away from the canals and open a small gap along the base of the wall. Water rushed into the gap and cut the levee in half, significantly reducing its strength.

The water then penetrated deep into the foundation, causing instability in the case of the 17th Street Canal levee and seepage in the case of the London Avenue Canal levee. In both places, the water penetrating the foundation weakened the system so much that the earthen embankment shifted.

The report said the corps failed to recognize the potential for this failure. Outside experts, including a team of investigators from UC Berkeley, said in a report last week that the corps had ignored its own research that predicted such failures could occur.

In addition, the corps’ report said, its engineers had used erroneous estimates of the soil strength under the levees. They also used tests that were spaced too far apart and only took averages of the soil strengths. The averages did not matter because the levees failed in the weakest layers.

The report does not attempt to explain why flawed decisions were made. Reed Mosher, a senior geotechnical expert for the corps and one of the principal investigators, said the biggest ingredient in the errors was the technology in use at the time the walls were designed.

“We wouldn’t make the same mistakes today,” he said.

The UC Berkeley engineers said the technology and know-how existed to build stronger levees, but the corps had lost some of its expertise starting in the 1980s as it shifted emphasis to managing projects and let private engineers do the design.

Strock said he “would not make apologies” for focusing on project management and disputed the assertion that the corps had lost its technical competence. A second investigation led by the University of Maryland is scheduled to report next month on the institutional and cultural factors within the corps that may have led to the errors.

The corps’ investigation, known officially as the Interagency Performance Evaluation Task Force, was conducted by 150 experts and led by University of Maryland engineer Lewis E. Link. The report unveiled Thursday is a draft; a final version is due in September.

The findings will trigger significant changes in how the corps and other government agencies protect the public from storms and floods, Strock said.

The corps is responsible for 12,000 miles of levees across the U.S. It will start treating the levees with the same oversight as it does with dams, which have historically received a higher priority in terms of public safety.

Maj. Gen. Donald Riley, director of civil works for the corps, said reforms were also coming in how flood protection systems were designed and built.

In the future, he said, levees and other flood protection projects will have greater resiliency, so that even if storm waters overwhelm them, they will reduce the effects of floods and allow for a faster recovery.

The corps will abandon its use of an antiquated method of estimating storm forces, known as the standard project hurricane, Riley said. It was that model that led the public to incorrectly believe that the New Orleans levees could withstand a hurricane as powerful as Katrina.

In deciding whether to build levees, the corps has long used cost benefit analysis that considered the potential loss of property, not human life or the social value of a city. Now, he said, diverse factors will be considered when evaluating whether to build levees.

Riley said the corps also would be adopting a different system to evaluate the risks and consequences of its systems. Since Katrina, outside experts said the corps erred in risk analysis of its levees and should have built much larger margins of safety. Riley said the new system would use modern risk analysis and would become a national standard.

Strock acknowledged that the corps must prove to the public that it could provide a higher level of performance.

“For those who doubt us, words alone will not change minds,” he said.

But the agency has emerged from the disaster as a stronger organization, he said.

“We are not wringing our hands; we are going to work.”