Another Duke Rebuke?

One of the most under-reported aspects of the widely reported methane paper issued by Duke Univ. researchers in May 2011 was the fact that, try as they might, the authors could find no evidence of fluids from the fracturing process in, near or anywhere close to shallow sources of drinking water underground.

Of course, that non-discovery wasn’t all that significant in a geological sense: scientists, engineers and even EPA administrators have long known that the thousands of feet and billions of tons of impermeable rock that separate deep oil and gas formations from shallow water formations make it virtually impossible for such migration to occur. But it was significant in a political sense, because the Duke research was underwritten by groups that oppose the development of oil and gas – groups that couldn’t have been happy to once again find themselves denied of the one talking point they covet above all the rest.

Against that backdrop, some may view the release this week of the second installment of the Duke paper as an attempt at atonement – with researchers finally providing opponents the evidence they need to launch a thousand new ships against hydraulic fracturing. Unfortunately for them, though, what’s most notable about this paper – similar to the first — is what the authors did not find: no fracturing fluids in water wells, and no correlation between the phenomena they report and activities associated with natural gas development. According to the paper: “The occurrences of saline water do not correlate with the location of shale-gas wells and are consistent with reported data before rapid shale-gas development in the region.”

Still, though, while the paper’s findings are benign, and the authors’ insistence that development activities had nothing to do with the detection of salt in water abundantly clear, we’ve seen this saga play out before. Already, activists are pointing to the report as evidence that fracturing fluids may someday migrate up to drinking water sources, denying the facts of science, a history of experience and even the views of the researchers themselves. And reporters, having spotted the words “hydraulic fracturing” and “contamination” in the abstract, are now deciding whether they even need to read the rest of the paper before filing on it.

Below, we attempt to lay out in intelligible terms what the second Duke paper actually says, while also running through a couple issues that Penn State geologist Terry Engelder and others identified during the paper’s peer-review process. Of course, the fact that a paper is “peer reviewed” does not imply consensus among the reviewers, nor does it mean that suggestions for clarification are actually incorporated by the authors, as indicated by the following discussion between the Duke researchers and Engelder:

Issue #1: No discussion of time-scale

Did these brine samples migrate up from depth over 100 million years, 10 million years, 10,000 years, or 10 years? Arguing that brine and other fluids can mobilize underground over time isn’t new or controversial. What’s controversial is attempting to assert that these migrations are occurring on a time scale that actually matters to humanity. To their credit, the Duke researchers don’t do that. But without even a ballpark sense of the possible time horizons associated with these phenomena, it’s impossible to assess how important the “discovery” actually is.

  • Duke researchers: “In our paper we did not provide any indications for the time scale of the apparent brine migration. Entrapped brines in deep and confined geological units could be diluted during any time or stage of brine formation.” (Duke response to Engelder)
  • More from Duke: “The paper does not claim that the brines are actively moving on a modern-time scale … The time scale is not known but given the modern stable isotope composition, the potential risk of modern flow cannot be excluded.” (Duke response)
  • Duke researcher Rob Jackson: “There is a real time uncertainty. We don’t know if this happens over a couple of years, or over millennia.” (as quoted by ProPublica, July 9, 2012)
  • Engelder review: “As a given, everyone agrees that deep basin circulation can occur over time scales of millions of years … So, the factor that makes your paper original is time, short time.   My analysis of your paper assumes that significant circulation can occur during post-glacial time which means less than, say, 10,000 years (call this modern time). Otherwise, everyone agrees that there is connectivity over the long time frame.” (Engelder comments to Duke, p. 3)

Issue #2: No discussion of exposure pathways

If Marcellus brine is migrating up from formations thousands of feet below, why haven’t those same cross-formational fractures caused the natural gas in the Marcellus to leak-off and disappear over that same time?

  • Duke researchers: “While brine migration could have resulted from tectonic activity, other mechanisms such as changes in the hydrostatic pressure due to deglaciation or categenisis could trigger the flow of pressurized brines.” (Duke response)
  • Researchers say that “future studies” should look for exposure pathways: “We agree that future studies should test our hypothesis of possible saline groundwater flow through the shale formation overlying the Marcellus Shale and evaluate the hydraulic conditions that will allow such flow.” (Duke response)
  • Engelder suggests other mechanisms – unacknowledged in the Duke paper – that would lead to freshwater mixing with brine: “[H]ydrodynamic flow moves freshwater into rocks with more saline brine.  It seems to me that freshwater infiltration into elevated areas will mix with brine as it advances, but also will push the brine toward outlet areas in river valleys, and cause mixing with fresh water in that area.  Your samples demonstrate this phenomenon and this has nothing to do with ‘vertical’ migration from some deeper source.” (Engelder comments, p. 6)
  • Stanford geophysicist Mark Zoback: “Frankly, I think some degree of vertical hydraulic conductivity in the crust over geologic time is reasonable, but why dense brines would rise and mix with near surface aquifers is not clear. [Duke’s] supposition ‘therefore it implies a greater tendency for leakage from hydraulically fracturing in the shale’ is illogical.  Production from the Marcellus would lower the pressure and cause flow into the Marcellus, not out of it.” (Zoback email to Engelder)

Issue #3: No discussion of whether Marcellus even contains enough brinewater to leak.

  • As Engelder explains, the Marcellus is largely a dry formation: “This porosity [in the Marcellus] has virtually no free water.  In fact, high quality electric logs from the Marcellus show gas saturation greater than 95 percent.  …  Based on the free-circulation criterion established above, I hope we can agree that, in fact, there is no such thing as large volumes of Marcellus brine as you have implied.  … Any natural water that is produced might have come from other formations like the Onondaga or Oriskany which is known to contain brine.” (Engelder comments, p. 3)
  • Duke response cites “large volumes of brine” produced “directly from the Marcellus” as evidence that brine exists in abundance in the formation: “We agree with the reviewer that it is most likely to expect high brine volume in porous rocks relative to impermeable shale rocks, especially like the Marcellus Formation. Yet thousands of shale gas wells are producing large volumes of brine directly from the Marcellus shale, not from the underlying porous formations.” (Duke response)
  • But as Engelder explains, the water being produced at the surface isn’t brine from the Marcellus – it’s flowback from the production process itself: “Flowback water does not qualify as natural, deep-basin brine and I fear that you have mistaken flowback from the Marcellus as natural, deep-basin brine.  The same is true for produced water from the Marcellus, which most likely also originates partly as [fracturing] fluid.” (Engelder comments, p. 5)

 Issue #4: No discussion of transport or drive mechanism

What forces conspired to propel brine from a largely brine-less formation up through thousands of feel of solid rock, on an undetermined time scale?

  • Duke has no idea: “This paper does not address the hydrodynamics of brine flow and the specific hydrological conditions that would trigger such a flow.” (Duke response)
  • Assuming the drive mechanism is gravity, Engelder says the time-scale could be millions of years: “I think we can agree that the energy source for hydrodynamic flow is gravity.  [Previous research] makes it very clear that gravity-driven flow path that is as deep as the Marcellus requires millions of years and is thus not modern.  … Matrix permeability is insufficient to permit Darcy-flow on time scales that matter.” (Engelder comments, p. 7)
  • And even Duke acknowledges it’s no easy thing for fluids to migrate through Marcellus: “We do not dispute that the apparent permeability through the Marcellus matrix is negligible.” (Duke response)

So, to recap: Duke researchers say that small volumes of brinewater were detected in a few shallow water wells in northeast Pennsylvania. They believe this brinewater originated from the Marcellus, a conclusion they reached even without identifying a pathway for it to travel, a mechanism to propel it, or a time-scale that would at least narrow the possibilities of transport speed down to a couple million years. Notified of the fact that the Marcellus doesn’t actually contain much “free” brinewater, the researchers double-down and insist that it does, apparently confusing the collection of flowback and produced water on the surface for brine that’s native to the formation.

Of course, and very much to their credit, the Duke researchers make it clear that the phenomena they observed are in no way connected to development activities in the region. And although that is an important finding, and one that serious observers will note as they take time to actually read the paper, one can’t help but wonder whether that’s the audience for which this paper was actually written – and not this one instead.

Follow us on Facebook and Twitter!


  1. Scott Cline says:

    Good grief. Here we go again. The group from Duke has fallen on its sword again.

    To point 3 above:

    The Marcellus is a very dessicated formation with as Dr. Engelder says, up to 95% hydrocarbon saturation and almost no free or mobile water. The water that does come back (only about 20% of what is used in the frac) through production is predominantly the frac water that over time becomes more and more saline simply because of its prolonged contact with the formation at the fracture face. Most of that water comes back through production in the first month or two of production. It is predominantly not formation water but rather predominantly frac water in contact with the formation making it saline.

    Even up to 80% of frac water does not ever come back because it is imbibed in the first few inches of the extremely large fracture face surface area (12-50 milion square feet created per well) and then trapped in the Marcellus by capillary forces forever. Marcellus formation water is not mobile and any fluid that does move is directed down to the wellbore pressure sink and up the wellbore during production.

    As any first year petroleum engineer knows, fluid flow requires availability of fluid, a pressure gradient and a permeability path including relative permeability. It requires all three. None of these are present from the Marcellus to the surface through the overburden. End of story.

    I could go on and on.

    • VICTOR FURMAN says:

      Scott…. I really appreciate your educational background at times like these Dr.

  2. Concerned Scientist says:

    If there is any upward fluid flow at all that is naturally occurring from the Marcellus to the surface (which I doubt), production of the Marcellus will actually stop it and cause all fluids to migrate to the well bore. So if this imaginary problem actually existed, Marcellus development would improve it, not make it worse.

    There is brine in other formations above the Marcellus that are the source of the saline water. Watch the Brian Oram video. A decent percentage of wells there produce saline water because they drilled below the base of the fresh water. The formations are full of brine just below the aquifer.

    Big names and big pedigree and they have no idea what they are talking about. They don’t do their homework. Another C- effort.

    • Dr Scott Cline says:

      Great post Concerned Scientist! I agree on all of your points.

      It is hard to imagine that these poseurs from Duke would not understand these basic concepts. As with their first flawed paper it simply shows an extreme bias against natural gas development and a unsupported presumption that everything in groundwater is related to a single mile plus deep underground formation..the Marcellus.

      • Concerned Scientist says:

        Thanks Scott- your posts are always right on the money and much appreciated. The Duke people are doing a great disservice to science. I’d rank the Myers paper as the absolute worst, the Howarth et al in second place and these Duke papers a close third. At least they have some data. The irony is that their data always shows that hydraulic fracturing does not contaminate groundwater, yet they still find a way to make headlines with a fracking contaminates groundwater story.

        With the Duke people it looks like a classic case of knowing just enough to be dangerous. They need to do their homework – get up to PA and chat with the DEP and local geologists who understand the subsurface there. They are making really silly mistakes that are mainly a product of their enormous egos. Their arrogance is their greatest weakness.

Speak Your Mind