Ever since the boom in hydraulic fracturing of shale gas formations picked up pace about five years ago, foes of "fracking" have argued that the cracking of shale formations was causing the migration of methane, or natural gas, into drinking water wells and aquifers.

But a new study has cast doubt on that explanation and points to different culprits: faulty drilling and well completion techniques.

That's both good news -- and bad news -- for the shale gas business. On the one hand, the study suggests the problem can be fixed and that methane contamination of water wells isn't an inherent part of the fracking process. Contamination can be minimized with proper drilling techniques, including higher standards for cement well linings and steel casings. "This is relatively good news because it means that most of the issues we have identified can potentially be avoided by future improvements in well integrity," said Thomas Darrah, the study leader and assistant professor of earth sciences at Ohio State.

On the other hand, with thousands of wells being drilled every year, even a very small rate of failure would mean the steady contamination of a significant number of water wells and aquifers. So far this year, the state of Pennsylvania -- home to the prolific Marcellus shale -- has issued drilling permits for more than two thousand wells. If just one percent of those had a cement or casing failure, that would mean the contamination of 20 drinking water sources in the state every nine and a half months.

At the moment, it is difficult to say with any certainty how often fracked wells suffer cement or casing flaws and state regulators are hard pressed to accurately monitor all the activity taking place.

Indeed, one area the new study examines is Parker County, Texas, the site of a high-profile lawsuit by an individual, Steve Lipsky, who alleged that fracking wells had spoiled his drinking water. He was also shown in a movie about fracking with flames coming out of his garden hose. The company that drilled that well denied any wrongdoing. The Texas Railroad Commission, which regulates oil and gas drilling, has blamed natural seeps for the contamination of local drinking water.

The study was done by a team led by Darrah and composed of others at Duke, Stanford, Dartmouth, and the University of Rochester. They devised a method of geochemical forensics to trace how methane migrates under the earth by tracking noble gases that are associated with methane.

The study also examined a well known naturally occurring instance of methane seeping into drinking water, but the study found that the seep had a distinctive ratio of methane to salt not found in instances of contamination from fracking wells.

"There is no question that in many instances elevated levels of natural gas are naturally occurring, but in a subset of cases, there is also clear evidence that there were human causes for the contamination," said Darrah. "However our data suggests that where contamination occurs, it was caused by poor casing and cementing in the wells," Darrah said.

In the hydraulic fracturing process, water and small amounts of chemicals are pumped underground to break up shale at a depth far below the water aquifers. That unlocks natural gas, which flows back up the well. The long vertical pipes that carry the gas upward are encircled in cement to seal the sides of the well and prevent the natural gas from leaking. The study suggests that natural gas that has leaked into aquifers is the result of failures in the cement used in the well.

The study identified seven clusters of contaminated drinking-water wells in Pennsylvania and one cluster in Texas. Then it looked at 113 drinking water wells near fracking sites in Pennsylvania's Marcellus shale and 20 wells in the Barnett shale in Texas. Their findings were published this week in the Proceedings of the National Academy of Sciences.

By tracking noble gas isotopes, the study linked four contamination clusters to gas leakage from failures of cement that should have sealed the space, or annulus, around the pipe; three to leaks that seemed to implicate faulty production casings; and one to an underground gas well failure.

"Many of the leaks probably occur when natural gas travels up the outside of the borehole, potentially even thousands of feet, and is released directly into drinking-water aquifers," said Robert Poreda, professor of geochemistry at the University of Rochester.

The study cast doubt on the Texas Railroad Commission report, which said that in Parker County, TX the landowners complaining of water contamination had drilled too deep and that methane had migrated from the relatively shallow Strawn shale formation. Darrah said that noble gases are "particularly good" for tracking the migration of gas and that they study did not find any evidence of migration from the Strawn formation.

"Well integrity failures are exceedingly rare, a fraction of one percent," said Katie Brown, a spokesman for Energy in Depth, a public affairs Web site launched by the Independent Petroleum Association of America. She cited statistics, however, that pre-dated the fracking boom. She also pointed to a study by the U.S. Geological Survey that said better information is needed about baseline conditions of water wells and aquifers.