PITTSBURGH -- A top Pennsylvania oil and gas regulator says the state's data "completely refute" a Duke University study saying that drinking water near Marcellus Shale gas wells is more likely to be contaminated with methane.

"We've collected enough data to completely refute the Duke study," Scott Perry, director of the Bureau of Oil and Gas Management, said today at a shale drilling conference here.

Rob Jackson, the Duke professor who authored the study, said he was not surprised by Perry's assertion. He said Perry and the administration of Gov. Tom Corbett (R) rejected his study when it was released -- before gathering any data.

"When Scott Perry says they have enough data to refute our study, what does that mean?" Jackson said in emailed comments. "Does it mean that the high methane concentrations we found were imaginary?"

Jackson's study, released in May, suggested the potential for widespread contamination of rural drinking water from drilling in the Marcellus Shale under Pennsylvania, New York and other states. It also backed up the claims of drilling opponents and drill-site neighbors who blame drilling for fouled drinking water (Greenwire, May 9).

The study did not find evidence that hydraulic fracturing fluid or flowback waste is getting into drinking water. The contamination was methane, the principal component in natural gas, which can build up inside houses and cause them to explode. The study found average methane was 17 times higher within 3,000 feet of drilling than water farther away.

The study won significantly more acceptance than a Cornell University study that said that the full cycle of shale gas drilling makes the fuel nearly as dirty as coal. A Department of Energy panel on drilling and fracturing recently deemed the Duke study "credible" and recommended follow-up research (Greenwire, Aug. 11).

But Pennsylvania Department of Environmental Protection Secretary Michael Krancer called it "biased science by biased researchers."

Perry said his assertion is derived from the review of baseline drinking water samples done by gas companies before drilling. The testing is not mandatory, but if companies do such tests, they must share them with the people who use the water well and the state.

Contaminated water within 1,000 feet of a gas well is presumed to have been polluted by drilling unless a company can prove otherwise with such baseline testing data.

"We have a large sample set," Perry said in an interview. "It really shows there's no correlation."

Industry groups criticized the study because Jackson did not use baseline data, gathered before drilling.

Jackson said that information was in the hands of the drilling companies, which declined to share it. He said today he would like to examine the state's data and said he would be more open-minded than Pennsylvania regulators have been about his study.

"I promise not to say it was 'biased science by biased researchers,'" Jackson said. "I'll look hard at it and try and learn as much as I can from it."