Studies of the area where a sinkhole formed last summer in Assumption Parish indicate scientists suspected as early as 1991 that the Napoleonville Dome was not shaped as state regulators believed when they permitted industry to create caverns in the massive underground formation.
The shape of a salt dome is important because it dictates where caverns, which are created by pumping water underground to dissolve the salt and create a void, can safely be located.
Such caverns — often filled with oil, natural gas or brine — can become structurally unsound if too close to the edge of the salt formation in which they are carved. Scientists now think that’s what caused a Texas Brine Co. cavern to collapse in the Napoleonville Dome more than 5,000 feet beneath the surface.
Scientists hired by the state to study the collapse believe the failure of the salt wall separating the cavern from the surrounding natural formations allowed millions of cubic yards of rock and other material to flow into the cavern. The collapse deep underground caused the sinkhole to emerge near Bayou Corne in August, according to their scenario.
About 350 people remain evacuated from their homes because of safety concerns. The sinkhole continues to grow, and natural gas is still bubbling up in swamps and bayous.
The collapse of the Texas Brine cavern has changed regulators’ view of the dangers of locating caverns too close to the edge of salt domes, said Patrick Courreges, a spokesman for the Louisiana Department of Natural Resources.
Last week, the Office of Conservation — which is under DNR — proposed new rules governing salt dome operators. And last month, the Office of Conservation directed operators to show how close their caverns are to the edge of salt domes and, in some cases, whether they are structurally sound for the long term.
Courreges said operators with caverns close to the edge of salt dome who cannot show their caverns are structurally sound will have to shut down those operations.
“The game has changed here,” Courreges said.
Lisa Jordan, supervising attorney at the Tulane Environmental Law Clinic, helps represent people who are opposed to the proposed expansion of natural gas storage caverns in the Jefferson Island salt dome in Iberia Parish. She says scientists can’t always accurately predict what will happen deep underground.
“The bigger picture is just what we found out in Bayou Corne. We really don’t know,” Jordan said.
Long before the Texas Brine cavern failed, hints were emerging that the structure of the Napoleonville Dome was not as well understood as industry officials and state regulators had long believed.
Sandia National Laboratories studies salt caverns for the Strategic Petroleum Reserve, where the nation’s backup supply of oil is held, and has scouted the Napoleonville Dome for possible storage sites.
Two reports by Sandia, in 1991 and 1995, and two Texas Brine seismic surveys summarized in a 2010 report suggested that the dome had “overhangs,” which would give it more of a mushroom shape than once thought.
Independently done maps of the dome from the early 1960s — and in 1991 by a private company — suggested smoother sides that slope outward to the base of the formation.
The possibility of overhangs is significant to the location of the caverns. The salt rock in which the caverns are made provides the structure — roof, walls and floor — for the underground cavities, which can be as deep as the world’s highest skyscrapers.
Salt formations are considered prime spots for man-made caverns because the salt forms an almost impenetrable container and is structurally more stable than many other kinds of rock.
Overhangs can lead operators to believe a salt formation is thicker than it actually is where a cavern is to be created. The danger is that caverns will be created too close to the boundary of the salt dome.
While the 1991 Sandia report briefly mentioned a possible overhang on the southwest side of the dome, the 1995 Sandia report raised the prospect of an overhang on the northwest side.
Several Texas Brine caverns are near that northwest corner, including the one that failed in August.
The 1995 study also claimed the overhang was a factor in the mining of another Texas Brine cavern through the edge of the Napoleonville Dome and into shale. The shale sheath enveloping the edge of the dome prevented failure of the cavern, the study speculated.
“Because of the fortuitous presence of shale sheath, this cavern was not lost but easily might have been,” says the caption under an image of the dome in a 1997 publication of the study.
Mark Cartwright, a Texas Brine vice president, recently said that cavern is not in danger of collapsing. He said the cavern, created in 1976, was used without any problems until being shut down in 2011 because it was no longer commercially viable.
Cartwright added that Texas Brine officials initially believed they had drilled out of the salt dome and took steps to safely create the cavern above the shale. However, later testing proved Texas Brine never went through the edge of the salt formation, he said.
Courreges said that cavern, which is northeast of the one that failed, is being monitored.
The irregular shape of the dome was also suggested in a 2010 seismic report done for Texas Brine after it detected pressure problems in the access well for the cavern that would later fail.
Some experts say the study indicated an overhang in the salt dome there and suggested the cavern was much closer to the edge of the formation than once thought.
“To me, it looked like a complete piece of work, and certainly indicating some pretty dire consequences with respect to that cavern,” said John Ferguson, associate head of the University of Texas at Dallas’ Department of Geosciences.
He said he suspects little could have been done to stop the cavern failure by the time the 2010 study was conducted.
Cartwright said that while the 2010 study confirmed the salt dome was shaped differently than previously thought, the company had no reason to believe the cavern was at risk. The cavern, which was permitted in 1982 based on 1960s-era maps, was operating normally. Sonar images of its internal structure showed no problems, he said.
The data focused only on the top part of the salt dome, where the cavern’s access well was, because of the technical limitations of the test. The results didn’t cover the depth where the cavern itself is located.
David Borns, manager of the Geotechnology and Engineering Department at Sandia, said the study should have, at the least, prompted a deeper look into what was happening.
“It would have raised a red flag for me. That’s all I am saying,” Borns said.
As recently as late September, Office of Conservations officials contended the older concept of an outward-sloping dome was still accurate despite having been provided Texas Brine’s 2010 study in early 2011.
The 2010 study was limited in scope, and regulators didn’t have access to the proprietary 3-D seismic data used, said Courreges, the DNR spokesman.
As for the earlier Sandia studies, Courreges said the Office of Conservation was unaware of them until last fall, after the cavern failed.
After receiving the 2010 study, DNR officials felt that plugging and abandoning the cavern was the best course of action and didn’t expect any further problems, Courreges said. The cavern was plugged by June 2011.
Leon Thomsen, a geophysics professor at the University of Houston, said the 2010 study appears to show at least 100 feet of salt between the cavern’s access well and the edge of the salt dome at the point where the seismic survey stopped.
He said it was hard to know, based on the study, what the salt dome’s shape was like below that point.
“I am not sure there is any evidence here for any course of action other than plug and abandonment,” Thomsen said.
Courreges said the kind of failure suspected of happening at Bayou Corne is “unprecedented.”
Courreges said that in the past, there have been instances where a cavern’s side got close to the salt dome edge and resulted in a leak but not the kind of side wall collapse suspected at Bayou Corne. He said the wealth of research worldwide focuses on the failure of cavern roofs, not walls.
Such a roof failure happened in 1954 at the Bayou Choctaw salt dome in Iberville Parish. According to Sandia reports, the failure caused a flooded surface sinkhole now known as Cavern Lake.
Pierre Bérest, director of research at the Laboratoire de Mécanique des Solides at École Polytechnique near Paris, said he knows of several salt caverns that collapsed from the top in Texas, Kansas, France and Algeria.
“However, I know of no cavern on the flank of a dome which collapsed, except of course the Bayou Corne cavern,” Bérest said in an email.
Texas Brine’s Cartwright said that even after the 2010 study, his company had no reason “whatsoever” to believe the side wall of the cavern would collapse.
And conventional wisdom held that at those depths, about 5,000 feet, any rupture would be contained by the surrounding rock, he said.
“In fact, I think at the end of this process you’re going to see some really interesting papers written by some very smart people that will revolutionize our belief about what can happen deep,” Cartwright said.
DNR’s proposed new regulations on salt caverns were made public Wednesday and would bring the rules for brine caverns like Texas Brine’s up to what is required by rules on hydrocarbon storage caverns. While both types of caverns are created by injecting water into dissolvable salt formations, brine caverns are strictly used for the saltwater that is created and pumped out for industrial uses.
The new requirements include a minimum 300-foot separation between the walls of new brine caverns and a salt dome’s outer edge. The current rules for brine caverns have no minimum separation. Joe Ball, director of the Office of Conservation’s Injection and Mining Division, has said that a standing rule of thumb of 200 feet has developed through the permitting process.
New technology is helping lead to a better understanding of how salt domes are shaped and the implications for safely operating caverns.
A 2009 Sandia study underscored this evolution by comparing images of the Bayou Choctaw salt dome. Computer models of the dome, created at three different times since 1980, show that the smooth features initially envisioned become, under 3-D seismic imaging, a variable surface of overhangs, flutes and curves.
Borns, of Sandia National Laboratories, said the idea of a salt dome with smooth outward sloping edges permeated industry for 40 years even as seismic studies have shown domes are far more irregular in shape.
“I don’t know if everyone has changed the conceptual model in their minds,” Borns said.