The cleanup after the meltdown at the Fukushima Daiichi nuclear power plant has been anything but smooth. First there was the denial that things were really so bad, then crews discovered leaky tanks and holes in radiation barriers, and finally Tepco, the utility company which owns the plant, admitted that tens of thousands of gallons of groundwater were flowing through areas contaminated with radiation.
But lately, there’s reason for hope—the ground surrounding the plant will be frozen, which if it works, will isolate the most contaminated areas from the steady seep of groundwater flowing toward the ocean. The ice wall will cost over $300 million and might take as long as two years to construct. It’s a scheme that sounds like it was ripped from the pages of a comic book, a half-cocked idea cooked up by an evil genius. But ground freezing isn’t as radical as it sounds. Engineers have been using the technique for over a century.
The basic premise behind ground freezing is that soil—made up of bits of minerals and organic matter, water, and air—becomes stronger and less penetrable when its water freezes and expands. To freeze the ground, steel pipes are drilled down about one meter apart along the perimeter of the site. An interior pipe is then filled with chilled brine that circulates through a refrigeration unit. As the brine circulates, it removes heat from the ground, and cylinders of ice begin to form around the steel pipes. In eight weeks or less, the ice cylinders merge to form a wall of frozen earth.
Once in place, an ice wall is typically maintained for months or years, but some have been kept in place for decades. Once a project is completed, the pipes usually remain in the ground. Thawing can take months.
Today, engineers are using ice walls to complete ever more daring projects. Some, like the proposed installation at Fukushima, keep harmful contaminants from escaping, while others prevent deep mines from collapsing. But some of the most breathtaking uses of ground freezing may have happened right in your backyard. They allowed workers to tunnel new underground routes for cars and trains while you went about your daily life, oblivious to the engineering magic that was going on below.
Civil engineers don’t call it an ice wall, though. Rather, they refer to it by less dramatic names—artificial ground freezing or the Poetsch process, after its inventor, German engineer F. H. Poetsch, who developed the process to combat water in Belgian coal mines. Praised for its novelty in the late 1800s, the process remains mostly unchanged today. “This method of shaft-sinking…by means of the artificial production of low temperatures,” wrote Charles E. Greene in a letter to Science in 1885, “is an illustration of the new and unexpected directions in which chemical and physical processes become of use.”
Greene predicted that this new method would enable startling feats of engineering, and he was right. More than 100 years later, Poetsch’s artificial ground freezing is used in civil engineering projects around the world, often with little fanfare. Now, after the news from Fukushima Daiichi, ground freezing has become something of a minor public obsession. In fact, so many people have been asking Masaru Mizoguchi, a soil scientist at the University of Tokyo, about ice walls that he now directs them to a video he posted to YouTube showing what happens when water comes into contact with frozen soil.
The video illustrates the concept fantastically well. Mizoguchi has four plastic cups filled with sand. The sand in two of the cups is wet and frozen. In the others it’s dry. When he dumps water into the cups, it passes through the dry sand quickly, almost as if there was nothing in the cup. But when he dumps water on the frozen sand, it barely trickles out the bottom.
“Normal soil is composed of soil particles and water and air,” Mizoguchi tells me. The amount space between soil particles, known as its porosity, determines whether or not groundwater is allowed to pass through, he says. But when the water part of the soil is frozen, it expands to fill the open spaces, locking the soil bits in place, and the frozen ground becomes an impenetrable barrier to moving water. A groundwater barrier, like the kind to be used in Fukushima, is usually two or three meters thick, says Joseph Sopko, director of ground freezing for New Jersey-based engineering firm Moretrench. For mines and other sites where the frozen ground must act as a structural support, the wall can be more than ten meters thick.
Though the basic theory remains largely unchanged, computers have allowed engineers to construct more elaborate systems. This allows engineers like Sopko, who helped bring ground freezing into the computer age as a graduate student at Michigan State in the 1980s, to plan ground freezing projects in three dimensions.
Such advances would later help one of Sopko’s colleagues at Moretrench, David Mueller, during ground freezing operations for Boston’s Big Dig, an enormous infrastructure project that buried Interstate 93, an elevated eyesore which cut the city in half. The Big Dig was conceived in the 1970s and planning began in the early 1980s. Ground was broken in 1991. The final cost, including interest on borrowed funds, exceeded $24 billion.
While the Big Dig is infamous for its long timeline and inflated cost, there were a few bright spots, including the ground freezing operations. Engineers froze heaping portions of Boston’s soil solid. Mueller says Big Dig contractors chose artificial ground freezing because it offered the best way to excavate without risking subsidence, or a gradual shifting of the ground. It also minimized disruptions. Even as a 3.5 mile tunnel was carved through the heart of Boston, cars, buses, and trains continued to move along the roads and rails overhead.
“The design had to be absolutely right.”
“The real challenge on that particular project was working in and around all the Amtrak trains and their infrastructure. We had to install the system from within the tracks,” Mueller says. The complexity and scale of the Big Dig made it one of the—if not the —largest artificial ground freezing projects ever attempted in the U.S., he adds. “The design had to be right. The top of the tunnel was something like 10 feet below where the trains were.” That gave them enough ground to support the tracks above, but there wasn’t much room for error. “The design had to be absolutely right.”
A few hundred miles southwest, in New York City, engineers employed ground freezing on a smaller but similarly audacious project. The East Side Access, which is expected to be completed in 2018, will finally bring the Long Island Railroad to Grand Central Station in Manhattan. After zipping along Long Island, the trains will submerge to travel beneath the country’s busiest rail intersection, below the East River, and under some of the most expensive real estate in the world.
Every part of the project presented engineers with a different challenge, but perhaps the most vexing was the tunnel beneath Northern Boulevard in Queens. It was only 100 feet long, but it had to be built beneath a five-track subway, the four-lane Northern Boulevard, and a three-track elevated railway. The importance of the infrastructure above and the instability of the soil made ground freezing a no-brainer. “This is really soft ground,” says Michael Horodniceanu, president of MTA Capital Construction.
The top of the tunnel would be just 10 feet below the bottom of the subway tunnel. Ground freezing pipes were laid horizontally and uncomfortably close to the subway’s support piles. The proximity meant that engineers had to be aware not only of where they froze, but how the frozen soil would behave. Because the density of ice is lower than the density of liquid water, as soil freezes, it expands or swells. That expansion can cause cracks and create large, air-filled spaces. If this happens, the functionality of the frozen soil could be lost.
“During the freeze periods, we were very careful not to have to much heaving. When water freezes, it expands, and there’s a rail above,” Horodniceanu says. Avoiding complications meant testing and modeling the characteristics of the soil long before freezing began. “We knew the limits, and we were constantly making sure that we stayed below the limits,” Horodniceanu says. “It sounds exotic, but this is done every day.”
In the Long Run
At Fukushima, experts agree that the actual system won’t be all that exotic, either. The biggest challenge will be ensuring the safety of the workers who will install the expansive system of pipes and coolant needed to create a frozen barrier. “We don’t see anything from a technical standpoint that would preclude ground freezing from being used as a barrier. It seems like an appropriate use,” Mueller says. Rather, it’s the site’s dangerous radiation levels that could complicate matters, he adds.
For Fukushima, ground freezing isn’t a long-term solution, experts caution. Maintaining the barrier will require constant refrigeration, which requires a continuous source of power. Any interruption could compromise the barrier. Furthermore, powering the refrigeration units over a long period of time can be expensive. And in the case of Fukushima, it only contains the contamination—it doesn’t eliminate it. “I think it’s an appropriate thing to do. It is a smart thing to do,” Horodniceanu says. “But at the same time, the polluted water has to be extracted from the clean water.”
The Japanese government is currently studying the frozen ground barrier proposal, and they expect to have a ground freezing system in place by April 2014, according to a document released in September by Japan’s Ministry of Economy, Trade and Industry. The proposed installation would encircle the four damaged reactors with an ice wall nearly one mile long and up to 100 feet deep, statistics that sound impressive but are quite achievable, Sopko points out. “It’s not a very challenging job, as far as depth goes,” he says. “The challenge is the men that have to drill the holes—keeping them safe, keeping them protected.”
Today, even this project seems normal to Sopko, but while in graduate school, he wouldn’t have predicted that ground freezing would be used to contain nuclear contamination. But after years of designing and creating ice walls, he now thinks that it could be used for “just about anything.” The current challenge, he says, isn’t about getting bigger. “I think we can handle the large projects, honestly.” Instead, he and others are hoping to shrink the process. “We’re looking at making smaller plants right now, and using it on smaller applications. How do we do a smaller job with less energy and more efficient freeze plants?” When that happens, ground freezing could really heat up.