Whichoil and gas production sites and pieces of equipment are most likely to emit significantamounts of methane is difficult to predict, according to a study released April5 by the Environmental Defense Fund.

The studyinvolved using helicopter-based infrared cameras to survey more than 8,000 oil andgas well pads across seven U.S. basins and attempting to pinpoint common characteristicsof high-emitting sources.

But relativelyfew patterns emerged, according to the EDF. Statistical analysis looking at basinsand well pad characteristics found commonalities for only 14% of the variance inobserved emissions patterns.

"[S]uper-emittingsources are nearly impossible to predict. They can happen anywhere, anytime as aresult of malfunctioning equipment that goes unattended and sloppy mistakes in thefield," EDF said in an April 5 blog post.

The researchersfound a total of 494 unique high-emissions sources at 327 well pads out of the 8,220surveyed pads. Some characteristics — such as the amount of oil a particular wellsite produces — tended to have a slight correlation with high methane emissions,and the Bakken had more emissions than other basins. Further, over 90% of the roughly500 detected methane sources were from tank vents and hatches, which the researcherssaid might be happening on a routine basis when crude or condensate undergoes temperatureincreases or pressure drops.

But thestudy, which was accepted by the journal EnvironmentalScience and Technology, underscored that the presence or absence of high emissionswas mostly statistically random and that "statistical models have limited utilityfor predicting the occurrence of individual high emission sources."

MattWatson, associate vice president of EDF's climate and energy program, and DavidLyon, an EDF scientist, wrote in the April 5 blog post that the researchers' findingshighlight the need for consistent inspections at well sites.

"Ratherthan trying to guess where these super polluters will occur, it is clear from thestudy that regularly checking oil and gas facilities for leaky equipment is a moreeffective way to identify both high polluting sources, which this study sought toexamine, as well as the other low-polluting sources, which may be individually smallerbut are collectively significant and represent a substantial share of industry emissions,"the post said.