Coastal Bend Home Page
URANIUM NEWS UPDATE--DECEMBER 20, 2010
The U.S. Environmental Protection Agency (EPA) is currently reviewing its regulations for uranium and thorium milling to determine if the Agency should update them. EPA geologist Loren Setlow, during a meeting hosted by the EPA in Corpus Christi on November 4, presented an overview of current EPA standards and received comments from the public.
Coastal Bend Sierra Club (CBSC) members Venice Scheurich, Mina Williams, Mark Walsh, an officer in South Texas Opposes Pollution (STOP), and Sister Elizabeth Riebschlaeger participated by making presentations and asking questions. Also present was Goliad County Commissioner Ted Long who spoke in the interest of groundwater protection.
CBSC turned in a detailed report summarizing our concerns that statistically biased samples are being used to obtain baseline wells which provide data for estimating pre-mining groundwater quality values. These values are then also used as post-mining restoration standards.
Mr. Setlow informed us that there is a group at EPA headquarters in the Office of Water which is taking a look at the aquifer exemption activities by the regions, and that he is scanning our document and sending it to the people doing that review.
Following is the document (without insertions):
EPA MEETING COMMENTS RE URANIUM: CORPUS CHRISTI NOV. 4, 2010
My name is Venice Scheurich.
I am the Conservation Chair of the Coastal Bend Sierra Club.
My remarks will pertain to statistical methodology used in deriving estimates for the Restoration Table standards for post-mining groundwater restoration.
I have some expertise in both applied and mathematical statistics, and have included a copy of my CV with these comments.
Four years ago, when uranium companies intensified their interest in mining in several South Texas counties within our region, the Coastal Bend Sierra Club began studying the in situ mining process.
I believe the discoveries we made on statistical matters have an important connection to whether EPA is properly administering its mandate in the Safe Drinking Water Act.
I believe this because the statistical documentation attached to the following comments indicates that pre-mining baseline groundwater quality has been and is being incorrectly assessed by present State regulations.
What surfaced immediately in our study of uranium mining in Texas was the disturbing fact that post-mining efforts by companies to restore groundwater to pre-mining quality almost always failed. This was especially true for uranium. Therefore, my comments here are specific to uranium in groundwater.
In trying to understand why mining companies were not able to clean uranium out of the groundwater to levels they had agreed to when the Texas Commission on Environmental Quality (TCEQ) granted their permits, we focused on how the estimates for restoration values were obtained.
We learned that compiling a TCEQ Permit Applications Restoration Table which contains values for uranium and other components is a multi-step process.
To our astonishment, we found an extremely serious sampling design flaw in one of the very first steps of this TCEQ regulatory process.
The error is that TCEQs regulations allow companies to choose locations for all baseline wells which provide samples for estimating pre-mining groundwater quality. These selections are made after the company has test results from hundreds of boreholes drilled during the exploration phase. Data from this biased, statistically invalid sample is then used in subsequent steps which are involved in establishing the Restoration Table.
[SEE INSERT 1 FOR DISCUSSION.]
In addition to allowing companies to use statistically biased data sets to derive restoration standards, TCEQs regulations also allow companies to calculate these estimated values by questionable interpretation and manipulation of the data.
To see if companies were taking advantage of this opportunity, I examined the actual results of the baseline data analyses in the three counties mentioned above. Predictably, in all three cases, the companies chose analysis and interpretation of data which tilted baseline restoration values in their favor.
[SEE INSERT 2 FOR DISCUSSION.]
And further, in reading TCEQs responses to several sets of recent public comments, I found numerous examples of TCEQs having recommended or defended use of erroneous statistical procedures in their uranium mining regulations found in 30 TAC Chapter 331 and in the March 6, 2009, edition of the Texas Register which discussed recent revisions in uranium mining regulations.
[SEE INSERT 3 FOR DISCUSSION.]
Having spent over two decades of my professional life teaching college mathematics and statistics, and having also done some statistical consulting, I found these fundamental statistical errors in the regulations perplexing.
It is disturbing that TCEQs regulations contain no protocols to assure that samples are independent and representative, even though TCEQs statements (March 6, 2009, Texas Register) repeatedly stress the importance of samples having these properties. (See INSERT 3.)
The implications of this absence of protocols are profound because statistically biased baseline samples are being used to estimate pre-mining groundwater quality. The resulting flawed process of estimating groundwater quality has a direct impact on whether EPA will grant an aquifer exemption and therefore whether the spirit and intent of the Safe Drinking Water Act have been violated.
Perhaps a main reason there are numerous serious statistical flaws in TCEQs regulations and many of their official statements is, as we have learned, that the Agency employs no credentialed statisticians.
Given that EPA is responsible for granting aquifer exemptions prior to ISL mining and for enforcing the Safe Drinking Water Act, and given the ease with which a statistically unbiased sample of locations for baseline wells could be obtained in the production zone of the production area, the Coastal Bend Sierra Club asks the following questions:
1. What is the justification for EPAs continuing to accept estimates of pre-mining groundwater quality based on selected, biased samples when EPA makes decisions on whether or not to grant aquifer exemptions?
2. Does EPAs acceptance of flawed estimates of groundwater quality from a mining companys application for an aquifer exemption result in one or more violations of the Safe Drinking Water Act?
I am attaching three inserts and several additional pages of documentation to these comments to support and clarify the statement I have just made.
The Coastal Bend Sierra Club is grateful to the EPA for the opportunity to comment on this important matter.
CBSC has also participated in the Sunset Review of the Texas Railroad Commission and the Texas Commission on Environmental Quality. On November 26, our 53-page response to the Sunset Staffs Recommendations was submitted. The cover letter to this document, which is included in the official written comments of the 2010 Sunset Review, follows:
COASTAL BEND GROUP
P.O. BOX 3512
CORPUS CHRISTI, TX 78404
Sunset Commission November 26, 2010
P.O. Box 13066
Austin, TX 78711-3066
RE Comments on TCEQ Sunset Advisory Commission Staff Report
Attention: Ken Levine, Executive Director of the Sunset Advisory Commission
Dear Director Levine:
As Conservation Chair of the Coastal Bend Sierra Club (CBSC), I e-mailed (July 16, 2010) to the Staff of the Sunset Advisory Commission a request that two recommendationsincluding one on statistical review of permitting proceduresbe included in the Staff report to the Sunset Advisory Commission.
Perhaps because we did not submit supporting documentation, our recommendations were not included in the TCEQ Sunset Advisory Commission Staff Report which was issued November 18. Since TCEQ employs no credentialed statisticians, it is especially unfortunate that the Staff failed to address this glaring deficiency in the report.
The following comments and enclosures document the need for implementing a recommendation on statistical review which would improve TCEQs permitting procedures in a significant, even critical, way by providing necessary statistical support for the Agencys scientists and engineers.
Below is CBSCs recommendation:
A thorough statistical evaluation by independent, credentialed statisticians should
become an integral part of all evaluation of TCEQs regulations, permit applications, and
summary reports involving collection, manipulation, analysis, or interpretation of data.
(Note: Data includes assumed or hypothetical values used in mathematical modeling for
permit applications as well as actual measured values.)
The necessity for such a recommendation evolved as our CBSC Executive Committee studied TCEQ regulations in 30 TAC Chapter 331 for in situ uranium mining which was occurring in several counties within our geographic area. (Our studys report is enclosed as Attachment 1.)
This report, entitled EPA Meeting Comments RE Uranium: Corpus Christi Nov. 4, 2010 (see Attachment 1), clearly demonstrates the dire need for credentialed statistical consultation within TCEQs Office of Permitting & Registration. Our report has now been forwarded by the EPA geologist who presided at that meeting to a group at EPA headquarters in the Office of Water which is examining actions being taken within EPA regional offices re aquifer exemption activities.
Additional enclosures (see Attachment 2) make it clear that for years at least one division within TCEQs Office of Water has issued reports containing serious statistical errors to citizens in the Houston area. These reports contained erroneous estimates of some constituents in that areas drinking water which led residents to believe their water met safe drinking water standards when, in fact, there was a high likelihood that it did not. Finally, in 2009, the EPA intervened to stop this misleading and statistically unsound TCEQ practice.
In the area of air quality regulation, there has been wide media coverage of the on-going dispute between TCEQ and EPA over the issue of whether the Texas Flexible Air-Permitting Program is incompliant with the federal Clean Air Act. This is another situation which illustrates how TCEQs practices violate sound statistical principles at the most basic level.
For example, early in their training, statisticians learn to divide and conquer when they examine a data set and use its information to obtain good estimates of parameters. TCEQs flex permitting does the opposite: It lumps together the data from several sources which obscures the patterns in air emissions and makes it highly unlikely that accurate estimates are obtained.
If TCEQ had consulted independent statisticians prior to drafting their air-permit regulations for refineries, perhaps flex permitting would never have become one of this States regulatory practices and Texas would have avoided an expensive legal battle with EPA.
Also in the area of air-quality regulation, there are on-going debates with EPA over how to obtain estimates for various emissions which will result if permits for proposed power plants are granted by TCEQ. Since these plants do not yet exist, mathematical modeling is used to derive these estimates. Judging which mathematical model would be appropriate in a given situation often requires a deep understanding of mathematical statistics as well as a thorough working knowledge of applied statistics.
The current dispute between TCEQ and EPA in the purposed White Stallion Energy Center case is a good illustration of this (see Attachment 3). Note that EPAs correspondence with TCEQ expresses EPAs concern that TCEQ has obtained no appropriate modeling protocol from the applicant for assessing ozone impacts if the proposed plant is approved. In particular, note that EPA states that the Scheffe Point Source Screening Tables must not be used in this case. To emphasize this point, EPA attached a 2006 letter from Dr. Richard D. Scheffe himself (the scientist who derived the Tables in 1988).
Anyone who reads EPAs letters in Attachment 3 will note that an understanding of the theoretical concepts which were used in deriving Scheffes Tables and other models requires a rather sophisticated level of mathematical functioning in addition to a thorough grasp of both mathematical and applied statistics.
Given the reality that many judgments made at TCEQ require this level of expertise, it is profoundly unfortunate that TCEQ employs no statisticians at any degree level (see Attachment 4) to supply this sorely needed support for their scientists, engineers, or other technical staff.
However, it is fortunate that there are many highly qualified, independent statistical consultants available who could provide this critical support. If the Sunset Review Commission would agree to include CBSCs recommendation on statistical reviews within TCEQ, its adoption and enactment would benefit not only TCEQs staff, but would better protect the air and water resources of all Texas citizens.
Having spent more than two decades of my professional career (see CV in Attachment 1) teaching college mathematics and statistics, and having also done statistical consulting, I can say with confidence that no credentialed, independent statisticians who value their reputations would have approved the statistical manipulations allowed by TCEQ which I have just cited in this letter.
Members of the Coastal Bend Sierra Club appeal to the Sunset Review Commission to find a strategy which will translate into a policy that will provide strong statistical support for TCEQs technical decision makers.
Venice Scheurich, Conservation Chair
On December 15, CBSC members Hal Suter, Mark Walsh, Mina Williams, Sr. Elizabeth Riebschlaeger and Venice Scheurich testified in Austin at the Sunset Review Commission Hearings. See Coastal Bend Sierra Club Meeting Minutes, December 2010, for details.
Submitted by Mina Williams and Venice Scheurich, Dec. 21, 2010