Floodwalls in Swampy New Orleans `Like Putting Bricks on Jell-O'

When the Texas construction firm AquaTerra Contracting began work on an Army Corps of Engineers hurricane protection project on New Orleans' West Bank, it encountered a serious problem: Its floodwalls wouldn't stand up straight in the mushy soil.

AquaTerra workers tried driving steel sheet piling down to the 55-foot depth the design required for the walls' foundation, company CEO Clay Zollars said. But the piling began to lean inward.

Zollars said the corps decided to nearly double the depth of the steel foundation to 105 feet. That didn't work either.

"Before we completed the wall, it began to lean and sink also," Zollars said. "The pilings were inadequate. The corps corrected that by installing some additional reinforcing steel in the concrete, but the wall still is leaning."

The top of one section of the 10-foot wall is more than a foot off the vertical, he said. AquaTerra is seeking $5 million it says the corps owes it for the extra work on the $11.1 million contract. Corps officials won't comment on the case because of the dispute.

The problems illustrate one of the basic obstacles to building reliable levees -- or any heavy structure -- in south Louisiana: It's a swamp.

Questions about soil are at the heart of investigations into why some of New Orleans' levees breached during Hurricane Katrina. Investigators say poor foundation conditions almost certainly led to the breaching of floodwalls along the 17th Street and London Avenue canals in New Orleans, flooding large parts of the city. The walls broke without being topped, suggesting a design or construction flaw. Layers of soft soils and organic matter under the wall foundations were not able to withstand the underground pressures generated by high waters in the canal, engineers say.

"They were struck with a bad situation, and they made a poor choice with those floodwalls, trying to put a structural wall on plastic soils. It's like putting bricks on Jell-O. There isn't a lot of support," said J. David Rogers, a veteran forensic engineer at the University of Missouri-Rolla who specializes in dam and floodwall failures.

What is unclear is how the corps and its contractors went forward with designs that some engineers now say appear fundamentally flawed. A team of engineers at the University of California at Berkeley studying the levee failures said that the corps' design standards do not seem to have accounted for all the soil uncertainties, raising questions about the design of the entire levee system.

The challenges of building floodwalls in weak, wet soils are well-known to engineers. A corps design manual warns that "by their very nature, floodwalls are usually built in a flood plain which may have poor foundation conditions."

Unexpected problems with weak soil have cropped up before. The AquaTerra case resembles a 1990s dispute concerning the 17th Street Canal floodwall. Segments of that wall also tipped off-center when the concrete wall sections were poured, requiring additional work and sparking a legal tangle. As with AquaTerra, the corps left the leaning walls in place.

A slightly tilted wall wouldn't necessarily be a safety hazard, engineers say, and it's not clear if the 17th Street Canal floodwall's early problems are directly linked to its ultimate failure.

Mississippi delta soil is notoriously unpredictable, both in composition and the ways it responds to stress. It's squishy and wet, with alternating layers of sand, silt, soft clays and peat, imbedded with the odd shells and decaying organic matter such as cypress trees.

Engineers must figure out how to imbed stable structures in this gumbo that will remain upright and withstand occasional extreme pressures from hurricane storm surges, winds and waves.

To do that, they depend on a delicate balancing of the forces of friction and gravity.

Floodwalls, skyscrapers, homes and other structures are typically built on steel or concrete piles imbedded in the earth. They get some support from the bottom tip of the structure, the way legs hold up a table. But most of the work is done by friction. Pile foundations are held immobile by friction between the soil and the surface of the pile. Long piles offer more security because they have more surface area and generate more total frictional force. Multistory buildings in downtown New Orleans are anchored by concrete friction piles extending hundreds of feet below the surface.

But a foundation is only as strong as the soil it's built in, and in engineering terms, strength is the soil's ability to resist forces acting on it and remain in place.

The area around the breached canals was swamp before it was drained or filled to make way for the city's residential expansion in the early decades of the 20th century and after World War II. Before Katrina inundated it, it looked like any neighborhood. But a completely different landscape lurks just under the surface.

"If you fly over the LaBranche Wetlands (upriver from the city), you will see wet and dry areas, areas with vegetation and areas with none," said David Lourie of Lourie Construction, a New Orleans-based soil engineering firm. "If you imagine some of that occurring at depths of 50 or 100 feet underground, that's what we've got in New Orleans residential areas."

Forces acting on the swamp for hundreds of years before humans decided to make it livable deformed it in peculiar ways, Lourie said, creating an unpredictable underground terrain.

"Through the passage of time, changes in Gulf water levels, changes in river flows, some of those (soil) surfaces were eroded or cut away," he said. "There were natural variations in the surfaces. They weren't all flat like a tabletop. You can have variations block to block. ... On one block you are over the center of a channel, and you could be only a block away and not over the same channel."

That means the requirements to anchor a foundation can also vary block by block. That's why detailed soil testing is essential before building a levee, or any big structure, to identify exactly what's below ground.

Documents show that in the 17th Street and London Avenue floodwalls, original soil borings were done about every 300 feet. It's not clear if there were later surveys that collected more data, but investigators say the soil surveys could have missed spots of soil weakness, and that could have created unidentified weak points in the walls.

In designing a wall, engineers weigh not only structural questions, but also issues of expense versus the high cost of failure.

Floodwalls "must be designed for the most economical cross section per unit length of wall, because they often extend for great distances," a corps design manual says. "Added to this need for an economical cross section is the requirement for safety. The consequences of failure for a floodwall are normally very great since it protects valuable property and human life."

Engineers say that the corps standards required an unusually low safety factor for the floodwalls, perhaps a remnant of a time when most levees protected sparsely populated rural areas, not cities and suburbs. A higher safety factor would require stronger walls -- and cost more.

(John McQuaid can be contacted at john.mcquaid@newhouse.com. Bob Marshall contributed to this story.)

In order to keep a complex subject organized...

It is indeed a complex subject and you are doing great work in keeping it organized. I appreciate your efforts greatly!

Yes, that's a valid concern, seepage under the berm(or roadway). At least two solutions : a clay berm as a base instead of the silt/gumbo they have now, which means importing HUGE amounts of impervious clay on barges over many years(for 300 miles of levees around N.O,); or deep concrete walls, say 6 to 10 ft deep(8" thick), on both sides of the roadway(from 20 to 25 ft wide), basically a big post-tensioned box beam that has buoyancy in and of itself. Thus it would exert little weight-pressure on the gumbo as you'd have vertical dead man anchors holding it down against that positive buoyancy. That still leaves the problem of a road panel strong enough to carry semitrucks yet light enough to pop right up, via buoyancy and side piano hinges, in a tsunami wave, that's what I'm designing right now... The objective, in this instance, is a ring road entirely around a re-built New Orleans that provides a 20 to 25 ft high sea wall for any future hurricane storm surge; and it operates automatically by natural forces. W=P

Thanks. It doesn't do any good to draw up sketches, I invented this idea 4 years ago and sent it to Des Moines and FEMA : nothing. Then again to 9 coastal states transportation depts recently(NH/NJ/NC/SC/GA/FL/MS/OR/WA) : nothing but a minor "we'll forward it" from OR. Large bureaucracies are 80% human dead wood so I figure the only way thru the hardened shell of ignorance/inertia is with a scale model/video tape it in action, and send it to governors of coastal states. If they STILL want to drown in a hurricane storm surge or tsunami, hey that's fine w/me, I live at 3000 ft elevation here in MT. So, back to work on the FLOOD ROAD model...

Disclaimer: Opinions posted on Free Republic are those of the individual posters and do not necessarily represent the opinion of Free Republic or its management. All materials posted herein are protected by copyright law and the exemption for fair use of copyrighted works.