On a July morning in 2008, the ground below southeastern New Mexico began to shift and crack, shooting a huge plume of dust into the air. Within minutes, a massive sinkhole emerged, which eventually grew to roughly 120 feet deep and 400 feet in diameter.
“At the time, it was an unfortunate situation, but most people considered it to be a one-off,” says Jim Griswold, a special project manager with New Mexico’s Energy, Minerals, and Natural Resources Department. But a few months later, in November, dust once again streamed toward the sky as another similarly sized sinkhole opened, cracking a nearby roadway.
Both holes—and later, a third in Texas—emerged at the site of brine wells, industrial wells through which freshwater is pumped into a subterranean layer of salt. The freshwater mixes with the salt, creating brine, which is brought to the surface for industrial purposes; in this case, oil drilling. After the second sinkhole emerged, Griswold’s department head gave him a new task: Characterize the stability of the state’s 30 other brine wells and report back on where the next crisis might arise.
What he found has been a source of near-constant worry for the past decade. While the first two sinkholes opened in remote areas, the next one, Griswold discovered, could hit the southern edge of Carlsbad: a city of 30,000 people.
At this location, the well’s role in dissolving the salt and drawing it to the surface had created a cavity wider than it was tall, a situation that destabilizes the ground above. And this particular well was situated near two highways, a rail line, an irrigation canal, a mobile home park, a church, a feed store, and several gas stations. If a similarly massive sinkhole were to open, it could endanger lives, stop road and rail traffic, and pollute groundwater with gasoline from ruptured pipes and storage tanks or saltwater from the well.
State officials have been fighting to hold off that fate, but it has taken a decade to research the problem and find funding to fix it. (The well’s owner, I&W Inc., filed for bankruptcy in 2010, and officials are not pursuing its potential liabilities.) Last year, the pieces finally came together, and an unprecedented remediation effort began. But the work unveiled subterranean secrets: The size and shape of the cavity isn’t what previous studies suggested, and this means it will be more costly to remediate. No one knows how long it will be before the ground gives way.
New Mexico’s statewide survey of brine wells gave scientists a rare chance to identify an area in danger of imminent collapse. Most other parts of the country have not been studied as closely, though risk exists there, too. Roughly 35 percent of the United States is underlain with karst, a landscape characterized by a network of sinkholes and caverns created when groundwater—or the chemicals it carries—dissolves the subsurface geological layers. States and the federal government don’t track the damage, making it tricky to pinpoint the frequency and cost of sinkholes. Still, the National Cave and Karst Research Institute, a government-supported nonprofit headquartered in Carlsbad, conservatively estimated the damage due to sinkholes at more than $300 million each year.
Carlsbad offers an extreme and completely human-caused example of the potential danger of sinkholes and how difficult—and costly—they are to prevent or mitigate.
Any number of human activities can disrupt underground karst, explains Jim Goodbar, who worked for the U.S. Bureau of Land Management (BLM) as a karst expert for 38 years. Rain gutters and highway runoff trenches can route water to one spot repeatedly, eventually wearing through the Earth’s surface open to cavities below. And pumping large amounts of groundwater can disrupt the water table, which in turn, destabilizes karst.
Florida is considered the hardest hit U.S. state, but other sinkhole-prone states include Texas, Alabama, Missouri, Kentucky, Tennessee, and Pennsylvania. In 2019, scientists used aerial surveys to find a cluster of 19 giant sinkholes, most of them more than 35 million cubic feet, in southern China, likely caused by underground rivers in the karst system in which they formed.
The big events draw headlines, but minor subsidence events, where the ground slumps slowly into a void, are a more common, constant and overlooked problem. When Goodbar was with the BLM, he recalls getting one or two calls a year about sinkholes in roads or near pipelines where water was channeled until the ground gave way.
“It’s a quiet disaster that we just don’t see hitting us all at one time in one place,” says George Veni, the National Cave and Karst Research Institute’s executive director and a member of the state-led group overseeing the brine well.
Carlsbad sits on the edge of the Permian Basin, an underground geological formation that stretches from southeastern New Mexico to West Texas and accounted for more than 35 percent of the U.S.’s domestic oil production in 2019. The surrounding desert is lined with rows of pumpjacks and the occasional white tower of a drilling rig. Carlsbad’s streets, like those in Lubbock, Midland, and Odessa, Texas, are crowded with the auxiliary businesses that supply that industry with water, sand, chemicals, and equipment.
Brine is one of those businesses. Drilling a well requires fluid—typically composed of water and chemical additives—to lubricate the bit. But in the Permian, a layer of salt, called the Salado formation, rests between the Earth’s surface and the desired fossil fuel reserves. Instead of staying in line with the drill bit, water will soak into that salt—unless the water is already saline. Brine wells in the region, therefore, help meet the needs of the local oil and gas industry.
People have mined salt through wells for thousands of years—one source notes that the first recorded brine well was sunk in China more than 2,200 years ago—and still do at thousands of sites around the world. In the U.S., the Environmental Protection Agency lists 165 permitted “solution mining” sites, with about 18,500 wells in operation, the majority of which mine uranium.
“Whenever you’re doing a brine well operation like that, you’re dissolving the salt out, you’re creating an artificial cave in the subsurface which could become unstable,” says Lewis Land, a hydrogeologist with the National Cave and Karst Research Institute who conducted some of the early geophysical surveys on the brine well near Carlsbad. “Everywhere you find salt in the subsurface, you’ll almost inevitably find brine wells.”
This is the case beyond the Permian Basin. In 2012, a salt dome in Louisiana being mined for brine punctured through to the bayou above, sucking down water and cypress trees and quickly spreading to impact 36 acres and force 350 people to evacuate their homes for years.
Operators often add a layer of oil or diesel fuel that floats on the saltwater and buffers the ground, preventing it from dissolving, Land says. That way, he says, “it’s less likely to collapse. That methodology is not practiced in New Mexico.”
In March 2009, regulators, technical experts, and industry representatives gathered in Santa Fe and concurred with Griswold’s conclusion: The I&W well was likely to collapse. In June of that year, the state installed an early warning system. On the highway, drivers pass a yellow and black sign cautioning that the road ahead is “subject to sinkhole.” On-site sensors monitoring the site are sensitive enough to register distant earthquakes, local rush-hour traffic, and, the hope is, rocks falling inside the cavity. That should give hours of notice to stop traffic on the highways and evacuate the area.
“I know people who have told me they won’t drive along that stretch of highway for fear that suddenly they’re in space, falling into a big hole in the ground,” Veni says. “I think the odds of that are incredibly slim.”
Were it to collapse, the costs to refill the hole and repair the roads, rail line, and businesses, plus the lost business, freight traffic, crops supplied by the irrigation canal, tourism to the nearby Carlsbad Caverns and Guadalupe Mountains national parks, and tax revenue could total $1 billion. Before remediation work began, geophysical data suggested the ground could fail as soon as 2021.
Over the last 12 years, research has worked to illustrate what’s impossible to see: the size and shape of the cavity itself. A formula correlates how much water has been pumped through a brine well with the volume of the cavity underground, but says little about whether that volume will be distributed in a wide, shallow space that might only cause the ground to sag if it were to collapse, or in a taller cavity that would swallow the highway and irrigation canal.
The state has run sonar, seismic, and electrical resistivity studies in search of answers. Likely, there’s a rubble-strewn chamber, much of it the cream color of the Salado formation, the dried footprint of an ancient inland sea. Some of the rocks may have piled up to create pillars and a tangle of corners. Remote sensing studies estimated that the main cavity was roughly 350 feet wide and 700 feet long, and 150 feet tall at its peak.
Shortly after Griswold recognized its instability, the well was pumped full of brine and capped with cement and cast iron. That brine has held up the cavern roof. But the surface measurably sank when a small percentage of the brine flowed out, and shifts each summer as water drops in the adjacent irrigation canal. Even in natural karst systems, changes in surrounding groundwater can alter the pressure that helps stabilize the system. The ground can fluctuate for a while, but for how long, no one is quite sure.
“The clock is ticking, but you don’t know what the clock is set on,” Veni says.
Still, some locals are taking things in stride. “Maybe it would be different if it were a permanent house,” says Charlie Kiser, who moved from Oklahoma with his wife to work in the Permian oil fields and lived for a while in a fifth-wheel parked within sight of the remediation work. “But if this thing falls in,” he says, gesturing to the RV, “I’ve got insurance. I’ll just take it as a sign from God to retire.”
When I&W, which operated the well for 30 years, was pressed to pay for fixing it, the company filed for bankruptcy. Liquidated assets produced $3 million, a small slice of a bill now estimated at above $54 million and growing. But the idea of fixing an invisible, indeterminate, potential disaster hit amid the very real pressures of the 2008 economic downturn, when lawmakers scrambling to close a budget gap raided the state’s reclamation fund. It took a decade for them to find money to tackle the task.
Griswold, whose full-time job now focuses on the brine well remediation, was already searching for someone who would take the work on. A global conference of “solution miners” that specialize in mining materials by dissolving them happened to meet in Albuquerque in 2017, and he presented the problem and invited solutions. The state group overseeing the remediation work settled on what they believed was the most promising plan: They would drill into the cavity and pump out the brine, while at the same time filling the cavity with grout, a mixture of cement, clay, and water. Workers would have to take care not to destabilize the system. Otherwise, the whole structure would collapse.
“I imagine it would be sort of like trying to tie your shoelaces while jogging,” says Land, with the National Cave and Karst Research Institute. “It’s obviously something that can be done, and that’s what the engineering firm that got the bid on this is doing, but it’s a delicate task.”
Activity has buzzed since last September. Initially, the goal was to have finished this past summer, but opening up the well to work on it offered a chance to collect new data on its dimensions. The southern portion was more stable than expected and filled quickly. That secured the area near the mobile home park and irrigation canal. But the northern end, some of which lies under Highway 285, proved larger than expected.
The cost of additional cement mixture would break the state’s budget, Veni says. “So what they started to do was to inject sand.”
Then, the cavity swallowed the quantity of sand expected to fill it to 70 percent, but remained only 20 percent filled. Sonar suggests two-thirds of the sand drifted into rubble from a previous internal collapse, instead of filling the void. “It’s going to take a hell of a lot of sand to do this,” says Veni—more than the state’s current budget can cover.
That’s typical, says Randall Orndorff, a karst specialist with the U.S. Geological Survey, even in the many cases when smaller sinkholes open and the work fills a visible hole.
“Without being able to see what’s underground, in a lot of cases they’ll come into it saying, ‘OK, we can grout this,’ and they’ll dig all the soil out around where the sinkhole is, down to the bedrock, then they start throwing the concrete grout into the hole,” he says. “In some cases, these are huge, huge caves, and it finally gets too expensive and they can’t do it.”
Work paused just before Memorial Day and sonar was run to see how much more sand might be needed. The work resumed, and then was suspended again at the end of July while the state looks for funding.
During injection, the effort costs more than $3 million a month, Griswold said during a September 2019 meeting, and, just as in 2009, the state is dealing with tight finances. Taxes and royalties from the oil and gas industry provide up to 40 percent of New Mexico’s general fund revenue, and lawmakers have had to cut spending after oil prices hit unprecedented lows this spring.
Sarah Cottrell Propst, who heads the state’s Energy, Minerals, and Natural Resources Department, said during a virtual meeting of the Carlsbad Brine Well Remediation Authority in July that one option is to seek pending congressional funding for reclaiming some of the nation’s estimated 2.1 million abandoned oil and gas wells. This is, as she put it, “the mother of all reclamation projects.”
If the ground gives way any time between now and when the work is done, it could still open into a sinkhole, though likely smaller than it would have been a year ago, now that the southern portion is filled, Veni says. That work may also have pushed out the timeline of collapse from next year.
But “there’s no guarantee,” Griswold says. “We need to finish this thing off.”