Hurricane-induced seafloorfailures in the GoM

“The delta front has a bad reputation for seafloor problems,” says Hooper, who has more than 40 years of experiences with pipeline and platform damage and failures caused by seafloor movements. The problems are a consequence of the unique composition of the seafloor, and the way it reacts to hurricane waves when they roll into the delta.

Delta consequences

The Mississippi River carries anywhere from 160 to 500 million tons of sediment per year into the GoM, depending on the volumes of rain, runoff, and erosion. Summarizing the work of marine geologists who have studied the Mississippi River Delta, Hooper says that maps dating back more than 300 years show that the major outlets of the river (Southwest Pass, South Pass, Pass A’Loutre) have lengthened seaward during that period, at rates of 100 to 200 ft/yr.

Below the water’s surface, the submerged portion of the delta front has also been growing seaward, due to sediment deposition rates of 1-2 ft/yr in front of the major river outlets. These deposited sediments are now 200 to 300 ft thick above their pre-delta base, and they are composed primarily of low permeability clay and silt.

“The very weak strength profiles that have developed here are susceptible to failure in the form of massive regional landslides that have created a surprising seafloor geology,” Hooper says.

Delta geologists refer to this seafloor morphology as “mudflow gullies and lobes.” The mudflow gullies (essentially wide channels in the seafloor) are often more than 6 mi long and are filled with semi-fluid sediments that are mobile to depths of at least 80 ft below the seafloor. This thick mud will periodically move down the sloping seafloor from shallow areas near the mouths of the river distributaries, to deeper areas, where it stops.

Over time, the gully sediment discharge creates stacks of successive mudflow lobes that are more than 100 ft thick. Water depths where these mudflow lobes accumulate range from 250 ft to 600 ft along the west and central delta front, and about 100 ft to 200 ft along the eastern delta front. New mudflows often cross the tops of the earlier deposits, and move thousands of feet further seaward in a single event.

The net result, explains Hooper, is that the mudflows move the delta front seaward. While many deltas grow by slow deposition from their sediment-laden river plume, the deepwater Mississippi River Delta advances by erratic fits of seafloor failures and mudflows. There are other mudflow-dominated deltas around the world, but the Mississippi Delta is unique in the concentration of oil and gas production facilities that are exposed to their potentially destructive effects.

Waves as triggers

The great thickness of very weak clay layers beneath the seafloor makes it susceptible to attack by waves, especially waves of the size and strength churned out by a major hurricane. When these waves move into shallower water, the seafloor beneath the crest of the wave experiences increasing pressures, since the water surface rises higher, and the crest portion of the wave contains a greater weight of water than would be present in a period of calm seas.

Conversely, the region of seafloor beneath the trough of a wave experiences lower pressure for the opposite reason: the trough sinks lower and contains a lesser weight of water.

The result is a dynamic, shifting cycle of increasing and decreasing pressure changes that cause upward and downward motion of the seafloor and the weak, compressible sediments. Lateral sediment movements also occur at and below the seafloor where upward and downward motions converge. When the storm intensifies, the waves become higher, the distance between the crest and trough of the waves increases, and the upward, downward and lateral motions of the sediments all become greater.

The effects of the waves on the seafloor also depend on the length of time the storm waves pound on the delta front. Hooper notes that each large, heavy wave that loads and unloads the seafloor causes a small decrease in the shear strength of the underlying sediments. These effects are cumulative, and when a hurricane sends waves into the delta for many hours and days, the sediment strength can decrease to where the slope fails, and slides downhill.

“The effect is similar to when you first bend a thin metal rod, and it simply takes on the new shape.” he says, “but if you continue to bend the rod back and forth, the metal degrades and it breaks.”

In shallower waters, where mudflow gullies are common, the seafloor may start sliding early in a storm. The weak, watery muds that are contained in the gullies are generally prone to fail anyway, even in the absence of storm waves. Further down slope, the mudflow lobes are often stacked up in water depths that are too deep to be greatly affected by the smaller hurricanes that have brushed the delta during the last few decades. These deepwater mudflow lobes are susceptible to Katrina and Ivan-size waves, however, when they persistently load and unload these otherwise stable sediments. The resulting failures can occur a hundred feet or more below the seafloor.

Hooper also reports that while the analytical tools available to geotechnical engineers have become increasingly powerful and more subtle in the way they model the delta front, there are still problems in their application.

For example, since waves are a major factor in the analyses, the analytical programs require information about the number of large hurricane waves that will cyclically stress the seafloor. This requires input from oceanographers, including the probability of Katrina, Ivan, or Camille-size storms, compared to the probability of lesser storms. Also, geological analyses are crucial to unraveling the depositional history of the site and the nature of past failures, since they determine where field studies must be performed.

Geologic studies are also an important part of the “reasonableness check” of the answers that are uncritically presented by the computer program. The geotechnical engineer must resolve complexities such as natural variations in engineering properties uncovered by soil borings, as well as philosophical imponderables such as the likely degradation of sediment strength during a hurricane.

Sometimes, Hooper says, the various parts of the problem fit snugly together, and the answers can be comfortably applied to the economical design of a new delta front production facility designed to withstand seafloor failure. In other cases, the uncertainties of waves, geology, and engineering properties of the sediments, are too great, and must be compensated by greater caution for the facilities design, with all the accompanying costs. If the risks and costs are too great, the oil and gas are left where they lie.

A stormy history

Between 1900 and 2005, at least 10 major hurricanes of Category 3 or greater passed close to the Mississippi Delta. Hooper points to three that graphically demonstrated the challenges that the delta front poses. Hurricanes Camille, Ivan and Katrina were all considered either Category 4 or 5 at various stages of travel towards the delta, and they all had significant impact on platforms and pipelines in the area.

Hurricane Camille passed directly over the eastern portion of the delta in 1969 where Shell had recently installed a new platform in Block 70, South Pass Area. The structure sat in approximately 300 ft of water, supported by 16 piles that were driven nearly 400 ft into the seafloor. The platform was found lying on its side after the storm had passed.

According to post-failure studies reported in industry journals by Shell’s engineers, the hurricane wave-induced stresses imposed on the seafloor produced a 100-ft thick sliding sediment layer. The platform stood little chance against the loads created by the moving sediments, and the piles apparently snapped due to bending at the base of the jacket.

Thirty-five and 36 years later, Hurricanes Ivan and Katrina assaulted the delta, bringing unexpectedly large wave heights and lengths. Post-Camille, many platforms had been installed across the delta front in water depths ranging from 50 to 250 ft, located between mudflow gullies, to water depths near 500 ft, located on top of stacked mudflow lobe sediments. There were also new pipelines crisscrossing the seafloor between the shoreline and seaward onto the continental slope.

Hooper explains that the platforms had been designed to withstand various thicknesses of seafloor failures based on the level of geologic understanding and analytical methods that had been developed up to that time. Pipeline routes in shallow water regions crossed the most stable-looking seafloor, avoiding mudflow gullies whenever possible. In the deepwater delta, pipeline designers sometimes tried to run the pipelines down slope and beyond the seaward limits of the mudflow lobes, before turning to run along the face of the delta front to reach a region of seafloor that might offer a safe corridor to landfall.

Ivan and Katrina waves triggered massive flows of sediments within many of the mudflow gullies across the delta front, bending and breaking pipelines that had operated successfully for many years. These gullies discharged huge volumes of weak clay and silt on top of the older mudflow lobes. The discharged material then continued in front of the old lobes out to distances greater than 2,000 ft beyond the previous mudflows, snapping pipelines in the process.

But, as Hooper continues, each storm had an impact deeper into the seafloor, as very large waves persisted for many hours, causing stresses that reached deep beneath the seafloor. This degraded the strength of the sediment layers, causing some of the old mudflow lobes to fail a hundred feet and more below the seafloor.

The results of years of mudflow-resistant platform design advances, and pipeline routing to avoid mudflows, appear to be mixed. Some platforms had been sited in semi-stable regions between mudflow gullies where strength measurements had indicated the seafloor might not move during a major hurricane. They all survived Ivan and Katrina. “Other platforms had been designed to withstand the loads caused by predicted mudflow overruns of the site and from (variously) shallow or deep-seated failures of the existing seafloor,” Hooper says.

“Anecdotal information available to date indicates that at least several of these platforms experienced major seafloor failures that were within the design bounds, without serious consequence for the platform structures. However, one platform failed during Ivan from what may have been an unexpectedly deep-seated landslide.”

Pipelines did not fare as well. Some were damaged where they crossed previously stable mudflow gullies, when the mudflow activity blossomed and spread across the delta, and some were damaged when seafloor movements occurred around the base of mudslide-resistant platforms. Other pipelines broke in areas where they had been routed down slope along predicted mudflow paths where the sediment-imposed loading should be theoretically minimized, and some were damaged in deep water regions when they were overrun by mudflows that passed beyond their anticipated run out distance.

Consequences for production

“Large-scale seafloor failures are the primary geologic process for seaward growth of the delta,” Hooper says. “Past rates of seaward growth of the delta front will likely be maintained, and seafloor failures will continue to occur.” Also, “the number and the size of delta seafloor failures will vary with hurricane activity, such that future Ivan and Katrina-size storms will create similar scale seafloor failures.”

And for those that hope there may be at least some level of reduced risk once a mudflow has occurred at a site, Hooper notes that for some water depths, sediments are weak enough to fail and move about under big storm waves even if the seafloor is perfectly horizontal for miles in every direction.

With an assessment like that, and with history as a guide, producers would do well to carefully consider their options for where and how to place future production infrastructure in this hurricane-prone region.