The New Orleans levees failed during Hurricane Katrina in August 2005 because of a combination of flawed structural design, weak soil foundations, and a shipping channel that funneled storm surge directly into the city’s flood protection system. More than 50 breaches occurred across the levee system, flooding roughly 80% of the city and killing over 1,100 people in Louisiana alone. The failures were not simply a case of a storm being “too big” for the defenses. In most locations, the levees broke apart before the water even reached the top.
The Wall Design That Cracked Under Pressure
Much of New Orleans’ flood protection relied on a type of floodwall called an I-wall, named for its I-shaped cross section: essentially a thin sheet of steel piling driven into the ground with a concrete cap on top. I-walls are generally considered appropriate only when the exposed wall height is less than about 10 feet. In several critical locations around New Orleans, they were asked to do more than that.
The fundamental problem was how I-walls behaved when floodwater pressed against them. Detailed investigations after Katrina found that as water pushed on the wall, it deflected just enough to open a small crack between the base of the wall and the levee soil on the flood side. Once that crack formed, water filled it instantly, applying the full weight of the floodwater down the entire depth of the gap. The walls were never designed for this loading condition. The result was that the combined wall-and-earth structure failed through a mix of instability and water seeping under the wall to the dry side, often well before the surge reached the top of the wall.
A sturdier alternative, the T-wall, has an inverted T-shaped base that distributes force over a wider area. T-walls are typically anchored on deep pilings and include cutoff walls to block water from seeping underneath. They were used in some parts of the system, but the Army Corps of Engineers had opted for the cheaper I-wall design in many of the locations that ultimately failed.
Soft Soil and Buried Marsh Layers
Even a well-designed wall can fail if the ground beneath it gives way, and New Orleans sits on some of the weakest foundation soils in the country. The city was built on layers of soft clay, organic marsh deposits, and sand, all deposited over centuries by the Mississippi River delta. These layers behave very differently under stress.
I-walls that survived Katrina tended to be built on soils made up primarily of thick, firm clay, with no buried marsh layers underneath. The walls that failed sat on a very different foundation. Beneath much of the eastern part of the city, buried marsh layers extended to depths of more than 9 meters (about 30 feet). The steel sheet piling supporting the floodwalls had been driven only to depths above the bottom of these marsh layers. That meant storm surge water could travel through the soft, porous marsh beneath the wall and destabilize the entire structure from below.
Construction history made things worse. In some locations, previous excavation work had cut through those marsh layers and been backfilled with loose sand placed below the water table. Sand deposited underwater doesn’t compact well, creating an even easier path for water to move through. The floodwalls in these areas were essentially sitting on a foundation that could transmit water pressure directly to the protected side of the levee.
A Shipping Channel That Acted Like a Funnel
The Mississippi River Gulf Outlet, known locally as MRGO (pronounced “Mr. Go”), was a 76-mile shipping channel completed in 1968 to give cargo ships a shortcut from the Gulf of Mexico to the city’s Inner Harbor. It became one of the most consequential factors in the flooding of eastern New Orleans and the Lower Ninth Ward.
When MRGO was built, it was roughly 600 feet wide. But decades of erosion had widened some stretches to nearly 2,000 feet, destroying tens of thousands of acres of protective coastal wetlands in the process. During Katrina, the channel acted as a highway for storm surge, carrying Gulf water directly toward the city.
The geometry of the canal system created a particularly dangerous bottleneck. Where MRGO met the Gulf Intracoastal Waterway east of the city, the flow from both channels was squeezed into a single narrow passage, and the water from MRGO was forced to make a 90-degree turn. Researchers at the LSU Hurricane Center found that water velocity at this funnel point was nearly three times the speed of flow in Lake Borgne, the body of water feeding the surge. One LSU researcher called MRGO a “Trojan Horse” for the city.
Army Corps computer models later concluded that MRGO increased overall surge height by less than one foot, which sounds minor. But the more important finding was that the channel substantially increased the duration of overtopping, meaning levees that might have handled a brief surge were instead battered by fast-moving water for a longer period. That extended assault gave the water more time to scour the back side of levees and exploit structural weaknesses.
What Actually Broke the Lower Ninth Ward
One of the most iconic images of Katrina’s aftermath was a massive barge that ended up sitting in the middle of the Lower Ninth Ward after the Industrial Canal’s floodwalls gave way. For years, many residents believed the barge had physically rammed through the wall and caused the catastrophic flooding. A federal court examined the question in detail as part of the Katrina Canal Breaches Litigation. The court found that “the overwhelming physical and scientific evidence proved that the Barge did not cause the North and South Breaches to the eastbank floodwall” and did not cause the cataclysmic flooding of the Ninth Ward. The breaches were caused by the same structural and soil failures seen elsewhere in the system: I-walls on weak foundations giving way under sustained storm surge pressure.
Systemic Failures, Not a Single Cause
The levee failures were the product of decisions made over decades. The Army Corps of Engineers designed flood protection piecemeal, with different sections built to different standards at different times. Soil investigations were sometimes inadequate, meaning engineers didn’t always know what was beneath the structures they were building. The MRGO channel was maintained for shipping even as it visibly eroded and destroyed the wetland buffer that had historically slowed storm surge before it reached the city.
Cost-cutting played a role at nearly every stage. I-walls were chosen over T-walls to save money. Sheet piling was driven to shallower depths than would have been needed to get below the weak marsh layers. The system as a whole was designed to protect against a fast-moving Category 3 hurricane, but Katrina’s slow approach and massive size generated a storm surge that overwhelmed those assumptions in ways engineers had warned about for years.
How the System Has Changed
After Katrina, Congress authorized a $14.5 billion overhaul of the region’s flood defenses, resulting in the Hurricane and Storm Damage Risk Reduction System (HSDRRS). The new system is designed to reduce risk from a storm surge that has a 1% chance of occurring or being exceeded in any given year, commonly called “100-year” protection. The design accounts for expected sea level rise, land subsidence, settlement of the structures themselves, and possible increases in storm severity and frequency.
MRGO was permanently closed to navigation in 2009, and a massive surge barrier was built near the channel’s convergence point with the Intracoastal Waterway. The system performed well during Hurricane Ida in 2021, which made landfall as a Category 4 storm. Still, 100-year protection means there is a 26% chance over any 30-year mortgage that a storm will exceed the system’s design capacity. New Orleans is safer than it was in 2005, but the geography that made the city vulnerable, a bowl of soft soil sitting below sea level between a river and a lake, hasn’t changed.