More studies on groundwater methane in Pennsylvania - no correlation with gas wells

Warning: High concentrations of methane in water wells, well enclosures and other confined spaces can cause explosions!

Here's a fact sheet from the Pennsylvania Dept of Environmental Protection providing information about how to deal with methane in your water well (it needs to be vented so that dangerous accumulations do not build up).

Have the DEP been forced to release this emergency information in response to increases in methane contamination as shale gas drilling spreads across the land?

No, in fact if you look closely in the bottom right corner, you can see that this information sheet was published in January 2004: long before shale gas came to Pennsylvania. This provides further demonstration of elevated methane in groundwater was common prior to drilling, as has already been indicated in baseline studies.

Why does this matter? Well, in a previous post I discussed the recent Duke findings of elevated methane in water near to gas wells in Pennsylvania, and I suggested that the very non-random way in which wells were chosen for sampling may well affect some of their conclusions. I suggested that to test their conclusions, more uniform and comprehensive sampling would be required.

Well, in a recent paper published in Groundwater, we have some new data. Molofsky et al tested 1701 samples (as opposed to only 141 tested by the Duke team). The two pictures below show the sampling from Molofsky (above) and the Duke paper (below), I leave it to you to judge which provides the more comprehensive sampling:

 

Of the 1701 samples tested by Molofsky, 322 were within 1km of a gas well, while 1379 are characterised as being 'pre-drill' - that is no gas well within 1km at the time of sampling, taken as part of a baseline surveys conducted by the DEP.

Molofsky et al found that 78% of sampled wells had detectable methane concentrations (hence the need for the DEP's fact sheet above), and 3.4% had levels exceeding the DEP's minimum level of 7mg/L.

The size of the circles in the Molofsky figure represent the amount of methane found in groundwater. They've helpfully plotted topography in their figure - even without the help of statistics you can see a correlation with being in a valley and having elevated methane (although the stats bear this correlation out), and upland areas with low methane. Why would being in a valley lead to elevated concentrations of naturally occurring methane? Well, a picture (from a Molofsky presentation I found online) tells a thousand words:

What about correlations between methane and natural gas wells, as found by the Duke study? Well, with 10 times as many data points, Molofsky et al find zero correlation between methane and natural gas wells. As their subsection title puts it: 'No Regional Association of Methane with Gas Production'.