Mike Stephenson, British Geological Survey

September 2012

Hydraulic fracturing or ‘fracking’ has had a bad press in the US; the chief worries there and elsewhere are methane contamination of groundwater, and earthquakes. The film ‘Gasland’ illustrated gas being ignited in someone’s kitchen sink. Some of the worries are no doubt justified: badly managed hydraulic fracturing and well design have recently been shown by the US Environmental Protection Agency to have contaminated water wells in Wyoming¹, though this was hydraulic fracturing of a sandstone gas reservoir, rather than shale.

But with so many vested interests involved on the ‘resource’ or the ‘environment’ side, getting reliable independent information is difficult. The internet is full of what looks like information, but most of it is from lobby groups from both sides, with barely hidden agendas. This is confusing for policy makers, regulators and the public alike. So, independent science has a role in deciding the real risks of shale gas extraction.

Most geologists think that methane or hydraulic fracturing fluid contamination of aquifers is unlikely because of the great difference between the depths at which hydraulic fracturing activities are usually carried out (usually 2-3 km down) and the very shallow aquifers from which water is abstracted (usually from the top few hundred meters). Hydraulic fracturing of shale has been done successfully and generally safely in a wide range of conditions since the late 1990s, and geologists and engineers who work for the companies that pioneered these techniques reinforce the message of safety. However to safeguard their intellectual property – their market advantage – they rarely publish their work. This is understandable: companies spend a lot of money developing their knowhow in a competitive commercial environment. However this perceived lack of clarity and independence has an impact on how the techniques are seen, and how much company research is believed by members of the public.

There are relatively few independent (as opposed to commercially-funded) technical studies of methane contamination related to shale gas hydraulic fracturing. A now well-known article by Osborn et al. (2011)² documented high levels of methane in water wells close to fractured wells, concluding that hydraulic fracturing of the deep Marcellus shale had contaminated the wells. The paper appeared in the Proceedings of the National Academy of Sciences and was peer-reviewed. The study has subsequently been criticised mainly because of the lack of data on background natural methane in groundwater³. Many aquifers naturally contain methane – though rarely enough to make tap water flammable⁴. So the presence of methane in tap water is not proof of contamination.

What scientific methods can be used to determine if hydraulic fracturing has contaminated aquifers? 

The simplest way to establish leakage from a deep hydraulic fracturing operation is to measure background or baseline amounts of methane in groundwater before hydraulic fracturing begins on a large scale. If higher amounts are measured than would be expected, then this would be evidence of leakage and contamination. In the early 2000s, UK groundwater CH₄ concentrations were measured in water-supply boreholes and other miscellaneous water sources, mainly to assess the possible climate-changing potential of methane emissions from groundwater. Concentrations in abstracted groundwaters ranged from <0.05–42.9 µg/l for Chalk, <0.05–22 µg/l for the Lower Greensand, 0.05–21.2 µg/l for the Lincolnshire Limestone and from <0.05–465 µg/l for the Triassic sandstone⁵. But baseline assessment is not as easy as it sounds because natural methane will likely vary with seasons and perhaps with other temporal cycles. The British Geological Survey is presently measuring baseline levels of groundwater methane systematically across Britain before any possible large-scale shale gas exploitation takes place⁶.

Can methane produced by hydraulic fracturing be distinguished from any other methane types in groundwater? There are two types of methane in groundwater, biogenic and thermogenic. Biogenic methane is generated by modern bacteria at the surface, or close to the surface in shallow rocks or in the soil. Shale gas is generally ‘thermogenic’, generated by heat acting on organic matter. Certain indicators help to distinguish the two types, for example: the δ¹³C and ¹⁴C of the carbon in the CH₄. The first indicator is perhaps the most useful. The ratio of the 12 and 13 isotopes of carbon in the methane (the δ¹³C of the C in CH₄) tends to be around -50 ‰ or less in biogenic methane and more than -50 ‰ in thermogenic methane (i.e. less negative). ¹⁴C has a short half life, so does not occur in thermogenic methane. In ideal conditions therefore it should be possible to distinguish thermogenic and biogenic methane. But this does not mean that we can distinguish ‘fracking methane’ and natural methane if the natural methane in the rocks is thermogenic. It seems that many areas (for example Pennsylvania) have natural thermogenic methane in groundwater that cannot have come from hydraulic fracturing⁷. So the technique is not infallible.

It is undeniable that hydraulic fracturing causes earthquakes because they are used by geologists to track the progress of hydraulic fracturing operations. The quakes are usually infinitesimally small, but recent events following hydraulic fracturing near Blackpool in the UK show that they may be larger, and felt by large numbers of people. Numerous claims were made about superficial damage, for example cracks in roads and bridges⁸. Two earthquakes were detected: on 1 April 2011 (magnitude 2.3), and a few weeks later (a 1.4 magnitude quake). Work by the British Geological Survey using its array of seismometers showed that the two earthquakes were clearly the result of hydraulic fracturing⁹. High pressure water probably found its way into small pre-stressed faults so that they moved slightly. This again shows the value of science in determining the cause of the earthquakes. The analysis was also able to show that the earthquakes could not have caused surface damage and so the claims made for earthquake damage were exaggerated.

As with methane in groundwater it is important to understand background levels of natural earthquakes, or earthquakes caused by old mine workings, because this allows us to distinguish anything unusual. The British Geological Survey maintain records of earthquakes¹⁰ but more local records of natural activity might be needed in areas of dense shale gas exploitation.

The examples of methane in groundwater and induced earthquakes show that science has an important role in controversial areas such as shale gas, where risk needs to be evaluated. Peer-reviewed science is often seen as esoteric and of little use to policy makers and regulators. It is unlikely that members of the public will read peer-reviewed articles on purported methane contamination or hydraulic fracturing -induced earthquakes, but the value of these studies lies in increasing the quality of the debate and helping to focus minds on the risks that do matter, rather than those that are overblown or false.

The British Geological Service Shale Gas Project website provides more information on shale gas in the UK.¹¹