Why Did the 17th Street Canal Levee Fail? It ranks as one of the worst engineering disasters in history: The failure of the levee system to protect New Orleans from Hurricane Katrina. The latest word from the investigation into the failure points to more than one cause.

Why Did the 17th Street Canal Levee Fail?

Why Did the 17th Street Canal Levee Fail?

  • Download
  • <iframe src="https://www.npr.org/player/embed/5418521/5419261" width="100%" height="290" frameborder="0" scrolling="no" title="NPR embedded audio player">
  • Transcript

Waves from Katrina topple over one of New Orleans' levees. Courtesy Donald McCrosky hide caption

toggle caption
Courtesy Donald McCrosky

Waves from Katrina topple over one of New Orleans' levees.

Courtesy Donald McCrosky

Bob Bea, a University of California, Berkeley engineering professor, says a weak layer of soil under the 17th Street Canal's wall caused it to fail. Bea says the Army Corps should have built a better wall. David Kestenbaum, NPR hide caption

toggle caption
David Kestenbaum, NPR

Hear Part 1 of the Story

Audio will be available later today.

Katrina Multimedia

INTERACTIVE: The "Times-Picayune" newspaper has created a display that shows, hour-by-hour, how Hurricane Katrina overtook New Orleans.


* Flash Flood: Hurricane Katrina's Inundation of New Orleans, August 29, 2005



ANIMATION: NASA has created an animation from satellite images that show Katrina gaining strength before walloping into New Orleans.


* NASA Katrina Animation



VIDEO: A government investigation simulation shows how the force of Katrina's waters may have caused the canal wall to tilt, which opened up a gap between base of the wall and the earthen mound that anchored it. That added pressure on the base helped dislodge the wall.


* Levee Failure Simulation


In 1965, the Army Corps proposed a plan that might have stopped the 17th Street Canal breach from happening. The plan would have put two big flood surge gates at deep channels that let water into the lake. The plan was abandoned due to local opposition. Melody Kokoszka, NPR/U.S. Army Corps of Engineers hide caption

toggle caption
Melody Kokoszka, NPR/U.S. Army Corps of Engineers

In 1965, the Army Corps proposed a plan that might have stopped the 17th Street Canal breach from happening. The plan would have put two big flood surge gates at deep channels that let water into the lake. The plan was abandoned due to local opposition.

Melody Kokoszka, NPR/U.S. Army Corps of Engineers

Hear Part 2 of the Story

  • <iframe src="https://www.npr.org/player/embed/5418521/5419359" width="100%" height="290" frameborder="0" scrolling="no" title="NPR embedded audio player">

George Sins lived along the 17th Street Canal. When he returned to his house after Hurricane Katrina, he found squatters there: turtles, fish and crabs had taken up residence in his pool. David Kestenbaum, NPR hide caption

toggle caption
David Kestenbaum, NPR

The site of the 17th Street Canal levee breach in New Orleans. U.S. Army Corps of Engineers hide caption

toggle caption
U.S. Army Corps of Engineers

The site of the 17th Street Canal levee breach in New Orleans.

U.S. Army Corps of Engineers

Different teams investigating the New Orleans levee failures disagree about the role weak soil beneath the structures may have played in the failure of the levees. U.S. Army Corps of Engineers hide caption

toggle caption
U.S. Army Corps of Engineers

Different teams investigating the New Orleans levee failures disagree about the role weak soil beneath the structures may have played in the failure of the levees.

U.S. Army Corps of Engineers

The levee failures around New Orleans rank among the worst engineering disasters of all time. Some of the breaches can be blamed purely on the storm -- it was huge, after all.

But investigators are finding that other failures were due to engineering flaws. The London Avenue Canal and the 17th Street Canal both failed when water had only risen part of the way up the wall. Had those walls held, large sections of New Orleans might have stayed dry.

George Sins knows this well. He could see the wall of the 17th Street Canal from his kitchen -- when he had a kitchen. The wall was there at the back of his yard, atop a long mound of earth. The canal was behind it; it ran from Lake Pontrachain, through his neighborhood and on for about two miles. He never gave it much thought.

Inside the Investigation

Letter from Frederic Chatry, Chief, Engineering Division

New Orleans District, Army Corps of Engineers


The two independent teams say this letter from 1986 shows that the Army Corps knew averaging soil strengths was a bad idea. The letter discusses the design for a different canal, the Orleans Outfall Canal to the east of the 17th Street Canal. In it Frederic Chatry of the Army Corps says that the contractor designing the wall should not average the soil strengths but instead should assume a weaker value -- one chosen such that two-thirds of the measurements are higher than that value. In the margin of the letter someone has written "major change if we do this."


Read Chatry's Letter




Department of the Army U.S. Army Corps of Engineers

Engineer Manual 1110-2-2504 "Design of Sheet Pile Walls"



This manual lays out U.S. Army Corps guidance for constructing floodwalls. It says that because soils can change from one location to the next, engineers should assume the real soil strength is weaker than the average measurement. As in Chatry's letter, it recommends using a value chosen a third of the way from the weakest value. Bob Bea, an engineer working with one of the independent investigations of the levee failures says the Army Corps did not follow these guidelines in the design of the 17th Street Canal.


Read the Manual




Interim Report of the Interagency Performance Evaluation Task Force


This interim report from the government investigation into the levee failures lays out the team's analysis of why the 17th Street Canal wall failed. The team did not find a weak layer of soil beneath the wall in the area of the breach. (Critics point to weak soil as a key factor.) The report instead says the wall failed because it tilted outward, opening up a gap between the wall and the earthen support -- a gap which allowed water to press against the full face of the wall. The report identifies a second flaw: a weak layer of clay near the levee under neighboring backyards that contributed to the wall's collapse.


Read the Report




Letter from National Science Foundation-Sponsored Team


The NSF team here argues that the Army Corps might have foreseen the possibility of a gap opening up between the wall and the earthen-levee support. And it says the Corps should have known about the weak clay. The team does not address one key issue: The official investigation found no weak soil layer; the NSF team says the layer existed and should have been spotted.


Read the NSF Letter




Letter from the External Review Panel to General Carl A. Strock, Chief of the U.S. Army Corps of Engineers

March 23, 2006


This letter comes from a blue-ribbon independent panel appointed to oversee the government's investigation into the failure. The charter for the panel (called the External Review Panel) told them to stick to the mechanics of what had gone wrong, but the panel felt compelled to go beyond that. Chairman David Daniel says his panel found "a disturbing pattern." The original designers he said, "played things too close to the edge" in the design of the 17th Street Canal.


Read the Letter

Sins and his neighbors trusted the wall had been well built. The U.S. Army Corps of Engineers had a reputation for overbuilding things, and the local levee boards sometimes complained that the Corps' standards were too high.

A Televised Water World

When hurricane Katrina threatened, Sins and his wife did the prudent thing and evacuated. After the storm hit, he had one of those weird modern moments: He saw his house on television. News cameras on helicopters were flying over the 17th Street Canal. He could see the circular window on his second floor, but the rest of the house was underwater.

The cameras showed a huge gash in the 17th Street Canal, about four doors down from his home. The canal connected to Lake Pontchartrain a few more blocks down, and water from the lake was pouring into the neighborhood through the gash. And it wasn't stopping. New Orleans was filling up, becoming an extension of the lake. To the east, the London Avenue Canal also broke -- in three places.

When Sins returned home, he found he'd been living closer to nature than he'd fully appreciated, because now it was in his swimming pool. He found fish, crabs and four turtles. "Three of them were gigantic, bigger than a big watermelon," he says, "way too big to get out with a fishing net."

Sins' house is now gutted, and he doesn't hesitate when you ask him whom he blames for all the misery. "The federal government," he says. "Absolutely."

'Hold the Corps Accountable!'

Across the street by a mangled garage door, someone else has hung a banner that reads "Hold the Corps Accountable!"

The Army Corps of Engineers oversaw the entire hurricane-protection system. The Corps is not terribly popular in town these days. One official says employees have been spit at in public meetings. And they don't always wear their bright red shirts with the castles on them when they go out anymore.

Sins believes the Army Corps botched the design for the canal walls. He talks about it as if anyone could have done a better job.

But the 17th Street Canal has a long, complex history. It has existed in one form or another since at least the mid-1800s.

If you climb up on the earthen levee and peer over the canal wall, you can get a pretty good view. The canal looks like a ruler-straight river. Workers dredged it so it could be used to carry rainwater out of the city and into Lake Pontchartrain. By 1915, huge pumps had been built to suck the city dry when the rain came. Rainwater would drain off the streets, and the pumps would lift the water up so it could flow out the canal into the lake by the force of gravity.

But during Katrina, everything went backwards. Water flowed into the city, instead of out. The storm drove huge quantities of water from the Gulf of Mexico into Lake Pontchartrain and then into the 17th Street Canal, which broke.

Building in 'Swamp Forest'

A military strategist looking at a map of New Orleans, thinking how to fight the water, would probably think the 17th Street Canal a little insane. The canal is a potential pipe for water to get into the city. It allows the enemy right into your backyard. And in fact, for more than 20 years, the U.S. Army Corps studied ways that would have tried to keep storm surge out of the canals.

The history of why so much depended on these slabs of concrete and this hill of earth begins 40 years ago.

Building hurricane-strength walls was not the Army Corps' first choice. The ground, after all, was notoriously difficult to build on. An 1849 map of the city labels Sins' neighborhood as "Swamp Forest," which is what it used to be before it was drained to accommodate houses as the city grew. The breach actually uncovered some old stumps of Cyprus trees.

A Promising Plan

Back in 1965, when the Army Corps was trying to figure out how to protect the city against a big storm, the initial plan was to build flood-surge gates at the eastern edge of Lake Pontchartrain, to keep storm surge out of the lake. It seemed simpler that way. Otherwise, the entire levee along the southern edge of the lake would have to be made higher, and the canal walls would have to be raised.

The floodgates would have blocked water at two deep channels, called Chef Menteur and the Rigolets. If a storm came, the gates could be lowered to block the two passes. The project was known as the "barrier plan;" some say it might have saved large sections of the city.

Environmentalists had concerns about the plan, though. They worried the gates would make it easier for developers to drain wetlands and build houses. And they worried it would reduce the flow of water into the lake.

"It was going to destroy the lake. It was going to make it a stagnant body of water," says Luke Fontana, executive attorney for Save Our Wetlands Inc. Fontana had spent years fishing the area by those channels, and even wrote a song opposing the plan. But more importantly, he filed a lawsuit.

Legal Roadblock

In 1977 a judge ruled in favor of Fontana and other critics, saying the Army Corps had not done enough to evaluate the impact the floodgate plan would have on the environment. Elvin Heiberg, who once ran the New Orleans district of the Army Corps, says that while the barriers would have had some effect on flow in and out of the lake, it would have been pretty minor. The gates would only be closed when a storm was coming.

But in 1985, a full 20 years after the barrier plan had been proposed, the Corps officially gave up. Heiberg by then had gotten a few promotions and was head of the entire Corps of Engineers.

"I think that's probably the biggest mistake I made, quitting instead of fighting," Heiberg says. "I think Katrina proved that."

There is some debate about whether the barrier plan would have neighborhoods flooded by the canals. Two scientists think it would have helped.

Missed Chances to Save the City?

Hassan Mashriqui at Louisiana State University and Joannes Westerink at the University of Notre Dame have both been running computer models of the storm. They say it looks like the barriers would have kept much of the surge out of the lake. Al Naomi, an engineer with the Corps, says no one there has had a chance to study those results.

There was also a second moment in history where a plan was put forward that might have saved the canal from Katrina. In 1990 the Army Corps considered putting floodgates (called "butterfly gates") right at the end of the canal, where it meets the lake. But again there was opposition, this time from city officials.

The New Orleans Sewerage and Water Board objected that if the gates closed even for a few hours, there would be no way to pump out the heavy rain that came with storms. And flooding from rainfall was a chronic problem.

So in the end, the Army Corps decided to leave the canal open to the lake. That meant a big storm surge would sweep in unimpeded, and the Corps would have to build the canal walls higher.

"In retrospect, that was a very unfortunate decision," says David Daniel, president of the University of Texas at Dallas. He says it would have made a lot more sense to build a small floodgate than to raise the height of miles of canal wall.

Misplaced Faith

The Corps went ahead and, working with contractors, drew up new plans for the new 17th Street Canal walls. The walls would be high enough for a large storm surge, with water 11 feet or more above sea level. The Corps designed the walls to have a 30 percent safety margin, meaning that on paper, they were 30 percent stronger than they had to be to minimally hold back the water.

Paul Kemp, an oceanographer with Louisiana State University, remembers that when authorities would "war game" hurricane scenarios, the question often asked was: "Where do we break the levees to let the water out?" It was assumed that the storm would push water over the top of the levees. There was an unspoken faith that the levees would hold.

When Katrina came and the 17th Street Canal broke, officials with the Army Corps assumed the storm had simply been bigger than the canal walls had been designed for. In the days after the hurricane, reporters repeatedly asked why the wall had broken.

The Army Corps said the leading theory was that water had poured in over the top and washed away some of the earth at the base of the wall, causing the wall to collapse. One official told NPR that an eyewitness had seen water coming over the top of the 17th Street Canal wall.

But this theory was wrong.

High-Water Marks

Shortly after the storm, Paul Kemp of LSU went looking for high-water marks to see how high the water had actually gotten. His group had run computer models that indicated the water had not gone over the top of the canal walls.

The models were right. In a beat-up yacht club by the lake, Kemp found a sheltered water mark slightly over his head. It was lower than the top of the canal wall.

"I had the first solid information that was incontrovertible," Kemp says. Investigators now think the wall failed when the water was several feet below the top of the wall.

Something had gone badly wrong.

Unearthing Evidence of a Flawed Design

Bob Bea is a tall, sinewy engineer with bright eyes who still uses a bulky calculator from the 1970s. He now teaches engineering at the University of California, Berkeley, but he knows New Orleans well; he lived there when he was chief offshore engineer for Shell Oil. And he knows hurricanes: During Betsy, he had to swim to his house.

After Katrina, Bea and some colleagues got funding from the National Science Foundation to try to figure out why the levees failed. They went to New Orleans, visited the broken levees and the canals, took pictures and gathered possible forensic evidence. They also began to gather the original design documents for the 17th Street Canal.

Bea went over them in his hotel room. His calculations turned up something he found disturbing.

Bea says the original soil tests used for the levee design show a weak layer underground. By his calculations, the layer was weak enough to explain why the wall broke. The wall would have been barely strong enough to hold back water at Katrina levels. Bea says the designers should have spotted the weak layer and built a better wall.

The Army Corps requires floodwalls to be designed 30 percent stronger than they need to be to hold back the water. And the design documents show the Corps calculated that would be true for the wall as a whole. But Bea says for this section of wall, the weak soil layer meant the margin would have been zero.

"I felt sick to my stomach," he says.

Did the Corps Break Its Own Rules?

There is some evidence the Army Corps did not follow its own guidelines. A levee manual available at the time says engineers should assume soils are weaker than the average measurement. It recommends picking a value not in the middle, but at the bottom third.

A design document for another canal in New Orleans contains a comment from an Army Corps reviewer instructing a contractor not to use the average, but instead to follow the one-third rule. In the margin someone has written, "Major change if we do this."

Not everyone agrees with Bea's conclusions that there was a weak layer, but he's not alone, either.

Two engineers -- Billy Prochaska and Louis Capozzoli -- with "Team Louisiana," a separate investigation funded by the state, agree and say the wall was destined to fail. Gordon Boutwell, an engineer with Soil Testing Engineers Inc., who works in the New Orleans area, also concurs.

How could the original designers have missed a weak layer? Bea and the others say that judging from the design documents, it looks as if the designers took an average of the soil strengths they measured over a large distance, possibly a mile or more along the length of the wall. This would have had the effect of masking any weak areas. "That's a long horizontal average, over a long swamp," says Bea.

'Smoking Gun' Disputed

NPR tried to get to the bottom of these allegations. The original soil testing was performed by a contractor, Eustis Engineering. Lloyd Held, who had signed off on the original soil report, would only comment that the matter would be "litigated in court."

NPR also sought out people who had worked on the original design of the 17th Street Canal at the New Orleans District of the Army Corps of Engineers. The District declined to set up an interview despite repeated requests.

NPR also sent Bea's calculations and the design documents to several geotechnical engineers at several of the most reputable university programs in the country. And while they found the averaging somewhat troubling, they did not see it as a smoking gun. Geotechnical engineers say their work involves judgment calls, and that not every weak measurement means the soil really is weak.

In fact, the official government investigation into what went wrong found no evidence of a weak layer.

Countering the Corps' Critics

After Katrina, the chief of the U.S. Army Corps of Engineers launched a large investigation into the levee failures. The effort, called the Interagency Performance Evaluation Task Force -- better known as IPET -- is expected to cost $20 million and has involved more than 100 people. As part of the work, the team took additional soil samples from the area of the breach and tested them. Their conclusions contradicted what critics have said.

IPET concluded that the strengths used by the original designers seem appropriate, at least under the levee itself. The analysis was performed by Virginia Tech's Tom Brandon and Mike Duncan, who is also the author of a textbook on soils.

New tests did show some of the soil areas to be weak, but Duncan is confident they don't really represent a weak layer.

"When you shove a sampler into the ground, you disturb the soil," Duncan says. "When you pull it out, you disturb it more. When you trim the sample you disturb it, and all of these things reduce the strength that you measure in the end."

Bea at Berkeley says he and his colleagues have also tested new soil samples. He still thinks there is a layer that is weaker than the original designers counted on.

A New Theory of Failure

The IPET investigation has a different explanation for why the wall failed. Team members say it is not in any textbook that they know of.

The explanation contends that when Katrina hit, the water rose in the canal and pushed the wall outward toward people's homes. The wall tilted, maybe just a few inches, but enough to open up a gap between base of the wall and the earthen mound that anchored it. The water would have rushed down into that gap and pushed against the full surface of the wall -- even the part that had been buried. That added pressure was enough to dislodge the wall, and a huge section of it and the supporting earth came loose.

There was a second component as well. IPET found a weak layer of clay in backyards around the area of the breach. When the wall gave way, it would have slid like a landslide on a huge section of earth. The weak clay would have helped allow that slide to happen.

The clay layer is everywhere underground in New Orleans. And the original designers measured its strength under the center of the levee. But the clay would have been weaker out in the backyards, where there was not so much weight on top of it. The government investigators say it doesn't appear that the original designers took this into account.

These two factors, the gap and the weak clay, together led to the collapse of the 17th Street Canal wall, according to the IPET report. Engineers with the team ran computer simulations that support the idea. Forensic evidence at the canal suggests the wall slid on the clay layer in the way the simulations predict.

Engineers also built precise dollhouse-sized miniatures of the canal wall and spun them in giant centrifuges to mimic the enormous pressure of the water. In those models, a gap opened at the base of the wall and the wall failed, just as the computer models had predicted.

Critical Blunders Cited

The report does not let the original designers off the hook, though. Army Corps officials acknowledge they should have caught the weak clay. Other engineers can't understand how they missed it.

"This is just sort of soil mechanics 101 principles that every educated person in this field knows," says David Daniel from the University of Texas.

Daniel was also disturbed by something else. He's leading a respected independent team brought in to monitor the government's own investigation. In a tough letter to Gen. Carl Strock, chief of the U.S. Army Corps of Engineers, the independent team wrote that the Corps' own research had hinted at the possibility of a gap opening up at the base of the wall. And it said that overall, the Corps had designed things too close to the edge.

Daniel gives the analogy of building a floor in an apartment building. A conservative engineer would try to think up a worst-case scenario, like the room being filled with 350- pound men. The designers of the 17th Street Canal didn't do that, Daniel says.

"We saw a rather disturbing pattern," he says. "When you added up all the pieces, it did not reflect a degree of conservatism you would expect for a critical life-safety structure."

Bitter Learning Lesson

Col. Murray Starkel is a big, gregarious man whose office looks out on the Mississippi River. Starkel is second-in-command for the New Orleans District of the Army Corps. During Katrina, water came up to the parking lot here.

A detailed map of the area hangs from his wall. It shows hundreds of miles of levees, many of which are undergoing repairs and inspections.

"Yes, I think we can say there were some problems with the initial design," Starkel told NPR, in reference to the levee gap that might have been caught and the weak backyard clay.

Starkel also said the Corps should have been more conservative when its engineers assessed the strength of the soil under the levee, the criticism raised by Bea at Berkeley and others.

Starkel said the Corps is learning from its mistakes.

"I would say [to the people of New Orleans that] we're sorry for what happened, for their loss to their property and those that lost loved ones," he said. "But we're doing everything within our power, our authority and our funding limits that Congress gives us to build them a much better system."

'Come On Home'

Many of the Army Corps' own employees live behind those same walls, he says. And he is urging people to come home.

Whole sections of New Orleans are still dark at night, and probably will be for some time. Some people are coming back. George Sins, the man who lived by the 17th Street Canal and found turtles in his swimming pool, says he will rebuild. He feels like he'll be protected by the new floodgate at the end of the canal.

But the city has hundreds of miles of levees. And like the 17th Street Canal, those levees didn't have one designer, but many: the Army Corps, Congress, local levee boards and various interest groups. Residents will have to decide if they trust that fractured system for the long haul.