An analysis of drinking water sampled from three homes in Bradford County, Pa., revealed traces of a compound commonly found in Marcellus Shale drilling fluids, according to a study published on Monday.

The paper, published in the Proceedings of the National Academy of Sciences, addresses a longstanding question about potential risks to underground drinking water from the drilling technique known as hydraulic fracturing, or fracking. The authors suggested a chain of events by which the drilling chemical ended up in a homeowner’s water supply.

“This is the first case published with a complete story showing organic compounds attributed to shale gas development found in a homeowner’s well,” said Susan Brantley, one of the study’s authors and a geoscientist from Pennsylvania State University.

The industry has long maintained that because fracking occurs thousands of feet below drinking-water aquifers, the drilling chemicals that are injected to break up rocks and release the gas trapped there pose no risk. In this study, the researchers note that the contamination may have stemmed from a lack of integrity in the drill wells and not from the actual fracking process far below. The industry criticized the new study, saying that it provided no proof that the chemical came from a nearby well.

In 2012, a team of environmental scientists collected drinking water samples from the households’ outdoor spigots. An analysis showed that the water in one household contained 2-Butoxyethanol or 2BE, a common drilling chemical. The chemical, which is also commonly used in paint and cosmetics, is known to have caused tumors in rodents, though scientists have not determined if those carcinogenic properties translate to humans. The authors said the amount found, which was measured in parts per trillion, was within safety regulations and did not pose a health risk.

Dr. Brantley said her team believed that the well contaminants came from either a documented surface tank leak in 2009 or, more likely, as a result of poor drilling well integrity.

The nearby gas wells, which were established in 2009, were constructed with a protective intermediate casing of steel and cement from the surface down to almost 1,000 feet. But the wells below that depth lacked the protective casing, and were potentially at greater risk of leaking their contents into the surrounding rock layers, according to Dr. Brantley.

In April 2011 the three homeowners in Bradford County sued the drilling company, Chesapeake Energy Corporation, over reports of finding natural gas and sediment in their drinking well water. In May of that year, the Pennsylvania Department of Environmental Protection cited the oil and gas company for violating the Pennsylvania Oil and Gas Act and Clean Streams Law by letting natural gas enter the drinking wells, though the company admitted no fault. In 2012, the homeowners settled the lawsuit and the company bought the three households.

As a result of that suit, the state environmental protection agency recommended that the drilling company require that their wells extend what are known as intermediate casings beyond 1,000 feet.

Dr. Brantley described the geology in northern Pennsylvania as being similar to a layer cake with numerous layers that extend down thousands of feet to the Marcellus Shale. The vertical fractures are like knife cuts through the layers. They can extend deep underground, and can act like superhighways for escaped gas and liquids from drill wells to travel along, for distances greater than a mile away, she said.

Katie Brown, an energy consultant with Energy in Depth, an advocacy group for the Independent Petroleum Association of America, said the authors had no evidence that the small traces they found of 2BE, which is also used in many household items, came from a drilling site.

“The entire case is based around the detection of an exceedingly small amount of a compound that’s commonly used in hundreds of household products,” Ms. Brown wrote in an email. “The researchers suggest the compound is also found in a specific drilling fluid, but then tell us they have no evidence that this fluid was used at the well site.”

Garth T. Llewellyn, a hydrogeologist with Appalachia Hydrogeologic and Environmental Consulting and the lead author of the report, said that when his team sampled water wells that were farther away from the drilling sites, they did not find any of the compounds found in the three households. “When you include all of the lines of evidence, it concludes that that’s the most probable source,” he said.

Victor Heilweil, a hydrogeologist from the University of Utah who was not involved with the study but reviewed its details, said it was noteworthy for showing “the detailed geologic fabric explaining how these contaminants can move relatively long distances from the depth to the drinking well.”

An environmental scientist from Stanford University, Rob Jackson, who also reviewed the paper, said it “clearly shows an impact of oil and gas drilling on water quality.” But he emphasized that this instance was an exception.

The dates of the incident were not surprising to Scott Anderson, a senior policy analyst with the environmental advocacy group Environmental Defense Fund, who said that well integrity was generally poor around 2008 and 2009. He said that using casings of steel and cement at depths below 1,000 feet was a good idea in this region. But he also noted that the industry has strengthened its practices since then, including increased use of intermediate casings.

“Industry knows how to construct wells properly, but the fact is that they don’t always do so,” Mr. Anderson said. “My hope would be that papers like this will encourage industry and its regulators to do a better job of doing what they already know they are supposed to do.”