Sri Lanka's Samanalawewa dam on the country's Walawe River has been leaking since the day it was completed in 1992. In the interim, the country has spent more than $65 million to plug the leaks in its second-largest dam, built to power the 120-million-watt Samanalawewa Hydroelectric Project. A 2005 study found that the reservoir—located near the town of Balangoda about 100 miles (160 kilometers) southeast of the capital Colombo—was leaking continuously at a rate of 475 gallons (1,800 liters) per second. And shotgun-type methods to solve the Samanalawewa dam problem—including the use of 13,640 tons of cement to reinforce the dam and the dumping of 1.8 million cubic feet (50,000 cubic meters) of clay to plug the holes—have failed.

The problem is that geologists and engineers do not know where all of the leaks are. So they turned to U.K. engineering consultant firm Atkins Global. Atkins performed a preliminary inspection of the dam and surrounding area for three weeks in February using AquaTrack technology developed by Draper, Utah–based Willowstick Technologies. The roughly $3-million project calls for Atkins Global to do additional survey work using AquaTrack this summer to pinpoint the sources of the leakage and spend the subsequent wet season planning precisely where to inject grout to plug those holes, work that Andy Hughes, the company's director of dams and reservoirs, anticipates will begin early next year.

Here's how AquaTrack works: Two electrodes—each three feet (one meter) long—are lowered down, one into the reservoir and the other someplace on the opposite side of the dam (typically in a sinkhole or other standing water downstream of the dam). The top of each electrode is connected with a wire. Once they switch on the electricity, "We've basically created a large circuit," says Paul Rollins, Willowstick's vice president of business development. Because groundwater is a conductor, the electrical current follows it between the electrodes, creating a magnetic field that can be detected on the surface using a sensitive magnetic receiver.

View images of how AquaTrack works

Once the magnetic field is generated, Willowstick's scientists walk the ground between the probes in a gridlike pattern with an instrument that collects data about the frequencies it detects underground. (The researchers are most interested specifically in the 380 hertz signals that AquaTrack's electrodes emit). The instrument is contained in a box that is three feet (one meter) tall and six inches (15 centimeters) square and held upright by a tripod and can collect thousands of readings in just five minutes, according to Rollins. (The technology has already been used successfully at a number of dams, including River Reservoir Dam No. 3 on the Little Colorado River in Arizona and Wolf Creek Dam on the Cumberland River in southern Kentucky.)

The circuit emits a magnetic field at 380 hertz that follows any groundwater it finds, Rollins says, "because water's really the best conductive [material] under the ground." The greater the amount of saturation, the greater the magnetic field, which emanates upward where it is recorded by Willowstick's surface sensing instruments. The gathered information is uploaded to computers at Willowstick's facilities, where researchers follow the thread of any 380 hertz readings to map the flow of underground water sources.

This will help determine the source of the leak, even if the leak is under the dam, Hughes says. "All dams leak to some extent," he adds, "but we don't want them to get out of hand."

U.S. companies have used AquaTrack to map dam seepage as well as determine the extent and location of groundwater those companies may have contaminated. Once a company that owns a plant or mine, for example, discovers it has polluted the local groundwater (or has been ordered by the U.S. Environmental Protection Agency to investigate the possibility that it has), the only real way to understand the problem to this point has been to dig a series of wells—generally six inches in diameter—to sample soil and underground water for contaminants, Rollins says.

Companies generally pay up to $120,000 to drill each well, so "they're not going to want to put 100 holes in the ground," Rollins says. "By creating theoretical flows in a modeling environment, the scientists can create theoretical magnetic fields," he adds. "They will then model these flows until the theoretical fields match [the data] collected in the field. Once they get the shape of the theoretical anomaly to match the actual data, then they can accurately determine depth of the dam seepage or groundwater." The goal here, as when the technology is used to find dam leaks, is to inform engineers as precisely as possible where they should drill to either pour concrete (in the case of a leaky dam) or take water samples to find the route of the contaminated water.

AquaTrack is designed to function much the way an MRI or X-ray is used locate a health problems within the body prior to surgery. "You wouldn't walk into a doctor's office and tell them to cut you open to find out what's wrong," Rollins says. "You'd first want to get an X-ray or MRI."

Of course, AquaTrack is not the only technology that allows scientists and business prospectors to better understand what lies beneath. Oil and gas companies for years have used the techniques of blasting or pounding into the ground and measuring the resulting shock waves to determine a site's crustal composition and, more importantly, where they might want to drill. "The acoustic signal travels through the Earth, and at each rock layer interface some of the signal bounces back up to the surface to be recorded by the sensor array," says Alex Krueger, vice president of research, development and marketing for Headwave, Inc., a Houston-based maker of software that can make maps out of raw data. "Thus, an image of the subsurface layers can be created."