We’ve noticed a “study” by Dr. Marvin Resnikoff that came out in January of this year seems to be getting some attention recently. It’s entitled “Radon in Natural Gas from Marcellus Shale” and pitches the idea natural gas from Marcellus Shale is particularly loaded with radon. It further suggests this radon will not sufficiently degrade by the time it reaches metro area consumers because of the short distance involved from wellheads to consumers and, therefore, it represents a health threat.
We’re not radiologists and the subject is one almost guaranteed to produce confusion, given all the strange terminology involved. Notwithstanding this, Resnikoff’s work isn’t designed for academic use, as much as it is for political consumption. There’s more rhetoric than research. That much is clear from the manner in which it is written and his use of some pretty wild assumptions already exposed by professionals in that field. Resnikoff, in fact, repeats serious errors he has made previously, rendering his entire thesis unreliable to all but the true believers among the anti-natural gas community. You might even say it’s radioactive to any reasonable person.
Resnikoff’s entire premise rests on a few critical assumptions. Interestingly, some these assumptions are ones he used before when he issued a report entitled “Radioactivity in Marcellus Shale” for a NIMBY group known as “Residents for the Preservation of Lowman and Chemung.” This 2010 report included his analysis of certain wellhead data and information from a 1981 USGS study entitled “Geochemistry of Trace Elements and Uranium in Devonian Shales of the Appalachian Basin.” He included a similar analysis of that data in his most recent epistle to the anti-gas community. The earlier analysis was evaluated in a report by Ralph H. Johnson of Blank Rome LLP appropriately titled “A Critique of the Radioactivity Waste Management Associates’ ‘Radioactivity in Marcellus Shale’ Report.” Perhaps, “devastating” should have preceded “critique” but you can assess that for yourself.
Johnson pointed out so many flaws in Resnikoff’s analysis as to render it void, just as both Federal and state courts have done in previous cases. See section 1, pages 1-3 of the Johnson critique for the truth about Resnikoff’s non-existent qualifications as a health physicist. The key elements of the critique, however, are found in section 2, pages 3-7 where Johnson explains why Resnikoff’s interpretation of USGS and wellhead data, the same interpretation he makes in his latest piece of political physics, is completely wrong. Let’s examine what Resnikoff says and compare it to the facts laid out by Johnson:
Resnikoff claims Marcellus Shale radium concentrations are typically “8 to 32 times background.” He also says “this compares to an average radium-226 in surface soil in New York State of 0.81 picoCuries per gram.” This is similar to what he said in his 2010 report where he asserted, based on USGS data, “radioactivity in the Marcellus is more than 20 times higher than background.” He also, in 2010, used the same gamma-ray log he employs later in the 2012 report to say that data suggested radioactivity “25 times higher than background.”
Johnson notes the USGS data used by Resnikoff relies upon only four core samples (two from New York, one from Pennsylvania and one from Ohio). Moreover, these samples are all from different depths and some aren’t even Marcellus Shale. The average uranium concentration in those cores (properly excluding Ohio because it was not identified as Marcellus Shale) was only 34.4 parts per million (ppm). This equates to approximately 11.27 picoCuries per gram (pCi/g) of radium-226 (radon is a decay product of this) and the median concentration was only 25 ppm or about 8.19 pCi/g of radium-226.
Resnikoff assumes, based on a state by state examination of radioactivity levels conducted in 1975-1979, the background level of radium-226 is 0.81 pCi/g. However, Johnson notes the New York State Department of Conservation (DEC) has conducted its own studies of background levels. This agency found, based on 11 samples, that background concentrations average around 5 pCi/g total radium. Total radium is defined as the sum of the radioactivity of radium-226 and radium-228 but the former accounts for as much as 99% of that found in nature. Another report (see pages 12-13) indicates it’s typically 3:1, so the DEC data tells us Resnikoff’s baseline is way off the mark and the actual relationship of Marcellus Shale radioactivity to background, using these three valid USGS core samples, is, at worst, more like two or three times background than the 8 to 32 times he postulates. But, this isn’t all.
Both of Resnikoff’s reports also rely upon gamma-ray logs from wells in New York to obtain what he asserts is data that confirms his erroneous conclusions from the USGS report. His latest report cites three wells located in the towns of Dix, Pulteney and Reading (Schuyler and Steuben Counties). His earlier report only referred to the Reading well (a/k/a “Shiavone 2″). Even though, as Johnson notes, Resnikoff had elsewhere acknowledged “it is not possible to give the specific radioactivity measurement due to log quality,” he proceeded to do just that and claim the data demonstrated a typical spike in radioactivity to 24 pCi/g in the Marcellus Shale formation, which meant it was 25 times higher than background.
Johnson points out, however, Resnikoff had left out an important part of the formula for converting the gamma-ray readings into radium-226 readings by failing to recognize it measured other things as well (e.g. thorium and potassium-40). Johnson explains how the gamma-ray reading for the Reading well actually produces a radium-226 reading of approximately 1.4 to 2.9 pCi/g. Even adding in the radium-228 would yield total radium of just 3.7 to 7.3 pCi/g, not the 12 to 24 pCi/g suggested by Resnikoff and this is, obviously, much closer to background.
Resnikoff repeats his error in his latest report, doing so precisely. He provides more explanation of the gamma-ray analysis, even noting the difficulties with log quality, but never addresses the fundamental problem with his data, even to counter Johnson’s point. He just ignores it, tells us his approach is common (without addressing the point) and then offers another qualifier to the effect “it is unclear how the shifting of scale is recorded, but a certain depth the gamma ray line indicates measurements beyond the 0-200 GAPI range.” So, if you question his judgment, then it’s all “unclear,” but if you like his conclusion the number is 12-24 pCi/g. This is science?
He also throws the two additional wells into his 2012 report, as the above chart demonstrates. He replicates his error two more times, in other words. Curiously, Resnikoff never mentions DEC wellhead data in this report, even though he had included some in his earlier “study.” The reason for this may be what Johnson observed in his critique. Resnikoff, he points out, looked at DEC data from the 2009 version SGEIS regarding high-volume hydraulic fracturing and confused total radioactivity levels with radium concentrations, leading to the completely false conclusion radium readings were 25.4 and 29.2 pCi/g when they were really only 1.8 and 0.9 pCi/g.
No wonder Resnikoff thought DEC data confirmed his findings. When your fudge factor allows you to be off by a factor of 15 to one or more, it’s not hard to find confirmation. Moreover, as Johnson explains, the DEC details for two wells actually illustrated very low radium levels of 0.9 to 1.8 pCi/g that fall between Resnikoff’s and the state’s own estimates of background levels, as does the gamma-ray analysis he examined when it is properly interpreted.
Resnikoff ignores all this because it no longer fits his narrative. He mentions DEC only to denigrate it, stating the following:
None of this analysis appears in the Generic Environmental Impact Statement prepared by the New York Department of Environmental Conservation. In the entire 1400 page Environmental Impact Statement, one sentence appears. “Radon gas, which under most circumstances is the main human health concern from NORM, is produced by the decay of radium-226, which occurs in the uranium-238 decay chain.” (NORM refers to Naturally Occurring Radioactive Material.)
This one sentence is the full extent of the Department of Environmental Conservation’s analysis of the environmental impact of radon.
This is, to be blunt, utterly irresponsible. Like all great myths it includes a grain of truth in that the SGEIS only employs the word “radon” once, but that is irrelevant to the larger point that radon is a product of radium-226 and the SGEIS addresses that squarely. One only has to read to page 13 of the SGEIS to find this statement:
Gamma ray logs from deep wells drilled in New York over the past several decades show the Marcellus Shale to be higher in radioactivity than other bedrock formations including other potential reservoirs that could be developed by high-volume hydraulic fracturing. However, based on the analytical results from field-screening and gamma ray spectroscopy performed on samples of Marcellus Shale NORM levels in cuttings are not significant because the levels are similar to those naturally encountered in the surrounding environment. (emphasis added)
Apparently, this didn’t fit the narrative either. Resnikoff ignores Section 4.6 “Naturally Occurring Radioactive Materials (NORM) in Marcellus Shale,” and several others dealing with radioactivity, all of which undercut his basic arguments with respect to radon. He also leaves out an important part of the paragraph he quotes, the part that says the following:
Uranium and thorium, which are naturally occurring parent materials for radium, are contained in mineral phases in the reservoir rock cuttings, but have very low solubility. The very low concentrations and poor water solubility are such that uranium and thorium pose little potential health threat. (emphasis added)
Resnikoff also ignores the specific references, in the SGEIS, to this DEC website dedicated to NORM impacts and this Investigation of Naturally Occurring Radioactive Materials (NORM) in Oil and Gas Wells in New York State, which just happens to include 86 pages of detailed information on the subject and, after comparing a number of oil and gas wells with background tests concluded:
Given the NORM concentrations identified in this report, there are no plausible exposure scenarios that will yield 10 mrem/yr dose rates at New York State oil and gas wells (see following section). In fact, 91 percent of sample concentrations did not appear elevated above and/or were indistinguishable from background. The low survey instrument readings (within twice background) are consistent with the sample concentrations taken from the sites. Hence, NORM contamination at oil and gas mining sites poses no threat to the public or the environment. (emphasis added)
This same report also states the following (“this gas” refers to radon):
This gas tends to diffuse out of the sampled material (22% is lost from sludges; 5% is lost from scales; and an unknown amount is lost from brines), and thus radon and its progeny are reduced relative to the radium concentrations in the materials.
So, if radon is a product of radium and naturally occurring radium is not a threat and is not significantly above background levels in or outside wells, where is Resinkoff going? That, really, is the question, isn’t it? It is fairly obvious why he no longer uses DEC data to make his case. It simply doesn’t support his view, nor does his other data. The very foundation of his study and what he says is his “starting point, his contention radium-226 exists in Marcellus Shale at levels not just above but far above background, doesn’t hold up upon examination. The rest of what he argues is of no bearing because his basic premise is just plain wrong and wrong from the start. It’s recycled rhetoric even a laymen can decipher. Indeed, I’d call it radioactive.