Tony Ingraffea has spent a lot of time alleging gas wells routinely leak but his research has distorted and misused the figures to deliver the conclusion he desired.
Fractivist misinformation has a long half-life. We’re still dealing with otherwise intelligent individuals who think there’s really a Halliburton Loophole, for example – people who simply repeat the mantra without checking for themselves. Fresh evidence of this sad aspect of human nature is offered by a recent presentation in Michigan (29:00) where fractivist Chris Groebbel cites Tony Ingraffea’s report from 2012. He alleges Ingraffea found “7.2% of new gas wells were leaking” in Pennsylvania due to cement job failures. A great deal of license was used to arrive at that claim, which turns out not to be so.
When Tony Ingraffea came out with that Park Foundation funded 10-page report from Physicians, Scientists and Engineers for Healthy Energy (PSE) a couple of years ago indicating 6-7% of all new gas wells in Pennsylvania leaked, it was fairly obvious his study had manipulated the data. This is because it didn’t match with the far more extensive and much different experience with wells in Ohio and Texas.
An August, 2011 report issued by the Ground Water Protection Council (GWPC) indicates more than 220,000 oil and natural gas wells were developed in these two states over more than two decades and 16,000 of them were horizontal wells targeting deep shale formations. Ohio had but 12 incidents of leaking related to casing or cement failures, a failure rate of only 0.03 percent. Texas only had 21, a failure rate of only 0.01%. Yet, Tony Ingraffea claims numbers several orders of magnitude higher. How can this be?
The answer is to be found in the way Tony counted, of course. He examined the Pennsylvania Department of Environmental Protection (DEP) database of violations available on-line here to get his numbers. Table 1 of his report, offered below) indicated he used this list of codes to make his initial counts:
Tony explained the list as follows (emphasis added), clearly indicating these were codes that indicated leaking:
That data was obtained by searching the violations database for all violations indicating that a well was leaking outside its production casing. Table 1 shows all the violation codes used by PA DEP to indicate that a well is leaking outside its production casing, why it might have occurred, and the consequences of such failure. These were the codes used to filter the entire violations database to identify wells with compromised structural integrity presented in Figure 7.
The list of violation codes are related to specific DEP regulations found here and in subsequent sub-chapters following (use the “next” button to access them). The first problem with the list is that the descriptions from Tony’s list don’t match up with the regulations. There is no Section 78.73A, for example. There is a Sub-section 78.73(a) which says this:
(a) The operator shall construct and operate the well in accordance with this chapter and ensure that the integrity of the well is maintained and health, safety, environment and property are protected.
but, that is not how Tony describes it, as you can see here:
Operator shall prevent gas and other fluids from lower formations from entering fresh groundwater.
The section sub-section cited is a generic catch-all that says nothing about leaking. Tony’s description appears to be a condensed version of Sub-section 78.73(b), which says this:
(b) The operator shall prevent gas, oil, brine, completion and servicing fluids, and any other fluids or materials from below the casing seat from entering fresh groundwater, and shall otherwise prevent pollution or diminution of fresh groundwater.
but, Tony calls it “78.73B” and describes it as follows:
Excessive casing seat pressure
(a) The operator shall prepare and maintain a casing and cementing plan showing how the well will be drilled and completed. The plan must demonstrate compliance with this subchapter and include the following information:
the latter says this:
(a) For wells drilled, altered, reconditioned or recompleted after February 5, 2011, surface casing or any casing functioning as a water protection casing may not be utilized as production casing unless one of the following applies:
and, Tony says this:
Diameter of bore hole not 1 inch greater than casing/casing collar diameter
Finally, as you can see, Section 207B, listed by Tony as:
Failure to case and cement to prevent migrations into fresh groundwater
has nothing whatsoever to do with gas wells.
You can’t even follow what Tony has done, let alone conclude there was leaking involved. The fact a plan wasn’t perfect or there was some other failure to comply with these regulations doesn’t prove any leaking. Moreover, the purpose of citing the violation is to get something corrected before there is any leaking. Making the leap from the supposed violation of a planning requirement to the assertion there was leaking of methane is like jumping the Freedom Tower.
Tony said he was talking about gas wells “leaking outside its production casing” but, what he has produced as a list could indicate almost anything. If we are talking about failures that result in venting of methane or leaking gas wells, then the focus should have been on Section 78.86, which plainly says:
In a well that has defective, insufficient or improperly cemented casing, the operator shall report the defect to the Department within 24 hours of discovery by the operator and shall correct the defect. The operator shall correct the defect or submit a plan to correct the defect for approval by the Department within 30 days. If the defect cannot be corrected or an alternate method is not approved by the Department, the well shall be plugged under § § 78.91—78.98 (relating to plugging).
Even here, there is no real proof of leaking, but at least the distinct likelihood of it occurring is evident. Some of these counting problems, of course, are probably due to improper citation of code sections in violations reporting but this just further demonstrates what a flimsy foundation Tony’s assertions rest upon. There’s more, though.
Tony summed up his findings, such as they were, in Table 7 of the PSE (read Park Foundation”) report:
He thens proceeds to tell us this:
However, a recent re-review of this database revealed that the data shown in Figure 7 are inaccurate…However, recently it has come to our attention that this filtering process results in a lower-bound on the number of wells with compromised structural integrity. That is, more wells have failed cement jobs than have been reported through the violations shown in Figure 7. All inspection reports for the more than 6000 wells drilled to-date in the Marcellus in PA were reviewed; this is a more complete and revealing search than just filtering on certain violations. The inspection reports indicate that many failed wells were not issued violations. Rather, they received “Violation Pending” comments; or comments indicating that “squeezing”, a cement repair procedure which would only be done if a well was leaking outside its production casing, had been done or was to be done; or comments that repairs were underway for a perforated casing; or comments that gas was detected at the wellhead at or above the LEL (lower explosive limit).
Note the report says “re-review of this database revealed that the data shown in Figure 7 are inaccurate” but doesn’t really say why. It implies the reason is that additional evidence of leaking was picked up in comments. That sounds very reasonable as an explanation until you look at Table 2 and Figure 8 from the report, which are supposed to summarize the combined results:
Notice, in Table 2, how the new numbers in Table 8 were derived. Table 8 indicates there were 111 gas wells that failed (leaked) in 2010, this number being a combination of 64 gas wells with violations and “47 additional wells with loss of integrity noted in Inspection Comments.” But, Figure 7 indicated there 90 well failures based on violations data, not 64. Why did the number drop? It appears Tony recognized the problem with his data after it was initially published, needed to correct the mistakes of whatever student or volunteer he assigned to this project made and didn’t want to revise the numbers down, so he dug up some more from the comments to make the difference and then some.
The whole report is a sham, in other words. None of the violation codes, by themselves, prove leaking gas wells and if we’re going to rely upon comments and pending violations then we ought to see them. There is not one solid bit of evidence in Tony’s report to support the conclusion that 6-7% of new gas wells leak. That is not suggest some haven’t and those have either been corrected or capped but that’s the point of issuing violations, to correct problems. They don’t simply continue and a paperwork violation, which some of these are, doesn’t mean a leaking well at any time.
Yet, here is the misleading slide from Chris Groebbel’s presentation:
You can find the debunking of the other stats here, but the junk science generated by Tony Ingraffea for his Park Foundation funded PSE lives on in the minds of fractivists who hear what they want to hear and offer it up as gospel and verse. Take a couple hours off, Chris, and go on-line to check things out before you make more baseless allegations based on one very leaky report.