COASTAL BEND GROUP SIERRA CLUB
P.O. Box 3512, Corpus Christi, TX 78404
Coastalbend current news and home page see below
JOIN ARCHIVES NEWS ITEMS KEMPS RIDLEY UPDATE SOUTH TEXAS URANIUM NEWS HOME
South Texas Uranium News
COASTAL BEND SIERRA CLUB'S COMMENTS ON TCEQ'S DRAFT RULES FOR
URANIUM MINING IN TEXAS
For more than a year, the Texas Commission on Environmental Quality (TCEQ) has been working on new regulations for uranium mining in Texas. Several members of Coastal Bend Sierra Club (CBSC) have participated in TCEQ's stakeholder meetings and CBSC has made detailed suggestions on some sections of the draft proposals.
After the draft proposals were posted in the Texas Register on September 5, 2008, stakeholders had until October 6 to submit comments and suggestions. To read CBSC's suggestions, see the letter below.
On January 9, 2009, TCEQ posted their responses to stakeholder comments on the TCEQ website and notified stakeholders of the posting by e-mail. On January 28, TCEQ e-mailed stakeholders to announce that another meeting to receive input on a few topics which were listed in that e-mail would be held in Austin on February 4, 2009.
One of the topics which TCEQ listed for additional comment was "clarification of statistical hypothesis test." Since the CBSC letter of October 6, 2008, had briefly addressed that topic, the CBSC officers decided to make a response at the February 4 meeting.
CBSC's response filed at the stakeholder meeting on February 4, 2009, appears below, following the letter dated October 6, 2008.
COASTAL BEND GROUP
P.O. BOX 3512
CORPUS CHRISTI, TX 78404
Texas Register TeamMC 205 October 6, 2008
General Law Division
Office of Legal Services
TCEQ Rule Project Number 2007-029-336-PR
P.O. Box 13087 SB 1604
Austin, TX 78711-3087 HB 3838
Dear Patricia Duran,
Coastal Bend Sierra Club (CBSC) represents approximately 400 members throughout South Texas counties whose residents are aware of uranium companies interest in leasing land for the purpose of seeking permits for in situ uranium mining. Consequently, many citizens and their county officials are following closely TCEQs work on revising the rules which will regulate permitting, in situ mining, and the post-mining restoration processes of uranium companies operating in Texas.
Also, because of past experiences with lack of groundwater restoration by the companies, many people in the affected counties have become aware of the weaknesses and inadequacies in the current TCEQ regulationsmany of which date back 20 or more years.
For several months, CBSCs Executive Committee has followed the Commissions work on the regulations to implement SB 1604 and HB 3838. We are aware of and appreciate the complexities and the hundreds of hours of staff work which have gone into these efforts toward regulatory improvements.
We also appreciate the opportunity to comment on the revisions and hope that you will agree that our suggestions on some items in 30 TAC 331 offer significant improvements which you will incorporate into your final version. We believe that the sections addressing determination of pre-mining water quality and enforceable requirements for post-mining restoration to pre-mining water quality are of prime importance.
Therefore, those sections must define protocols which are specific, unambiguous, complete, and sufficiently detailed to assure that owners and operators are required to use the best available scientific and statistical methodology.
For example, in establishment of baseline groundwater values, 331.104(a), the Commissions proposal to hereafter require that groundwater samples from baseline wells be independent and representative so that valid statistical analyses can follow, is a significant improvement in the wording over the present rule. But, 30 TAC 330 is far superior in its statistical approach for monitoring groundwater under landfills; similar requirements in 331.104 would be good.
Additionally, a valid statistical analysis of sample data requires that the samples be obtained from wells that are located on a systemic grid across the entire mining area surrounded by the monitor-well ring or randomly selected with an appropriate statistical procedure. The proposed rule does not require the systematic or random step for locating baseline wells. Without this step in obtaining the sample of baseline wells, the resulting data can not be representative in a statistical sense. Data flawed in this way can not then produce a valid statistical analysis.
Because data obtained from the sampling of baseline wells are all-important in laying the foundation for protecting groundwater by establishing values in the restoration table, CBSC strongly recommends that the Commission use some of its resources to allow staff scientists to seek out and consult the most highly qualified statisticians specializing in applied sampling design. These consultants would then work with Commission scientists to design an appropriate protocol (see 30 TAC 330, for example) for obtaining a systematic or random sample of baseline wells from which the resulting data would then achieve the legitimacy required to conduct valid statistical analyses.
This protocol for systematic or random sampling to locate baseline wells would become a valuable addition to 331.104. It would provide specific instructions for owners and operators at this very important step in data acquisition. And, it would assure that data used in subsequent analyses during groundwater restoration efforts described in 331.107(a) are statistically sound.
For example, the Commission proposes to amend 331.107(a) to allow that restoration may be demonstrated by either of two methods: the first uses the mean determined from a small sample of baseline wells to estimate the value of a parameter, and the second allows the owner or operator to propose a statistical test which will be used if approved by the Executive Director.
There are serious problems with both methods. In the first method, it is doubtful that the sample mean, which is very sensitive to even a single extreme data point, should be used as a measure of central tendency to represent the parameters value because it could be highly biased. In such a case, the sample median (middle of the ranked data) would be a much better estimate of the value desired, but the Commissions proposal does not offer the option of using the median.
To cite a specific example of how crucial it is to use the appropriate estimate for a parameter, consider the following actual sample data set which has been presented to the Commission by operators of the Vasquez mine in Duval County. The numbers are values for uranium (micrograms per liter) found in water from the five baseline wells comprising the sample to be used in establishing an amended restoration table for PAA2 of the mine.
The data points are: 2, 7, 8, 15, 131
The sample mean is 163 / 5 = 32.6. The sample median (middle of the ranked data) = 8.
This dramatic numerical difference between the mean and the median is highly meaningful because the mean is not valid in a statistical sense unless the data are shown to follow a normal or lognormal distribution. The 131 value in the example is clearly an outlier, and a normal or lognormal distribution would not be possible for the five data points. Therefore, the option is to use the median value or to collect more data points to retest the distribution assumption.
When one considers the fact that whether or not the pre-mining water quality would be judged to meet the EPA drinking water standard (no more than 30 micrograms of uranium per liter) depends upon the mean or the median, it is imperative that a proper statistical procedure be used in making the judgment. This judgment becomes important when a company requests that it be granted an Exempt Aquifer in a specific location, as well as when setting the value on a parameter for water restoration.
The second method given in proposed 331.107(a) as an option to demonstrate that restoration is complete calls for the Commissions Executive Director to approve using a statistical test proposed by the owner or operator. However, since the company is allowed to use their small selected sample, as in the Vasquez example, rather than required to obtain it through systematic or random sampling, there is no valid statistical test available.
Unfortunately, the language in some parts of 331.104 and 331.107, as we have shown, tilt the rules in favor of the industry to the detriment of all Texas citizens who are trusting TCEQ to protect their groundwater.
For the first time in years, since the rules for uranium mining are being revisited by the dedicated TCEQ staff, we in Coastal Bend Sierra have reason to hope that progress may be possible toward remedying some problems related to in situ uranium mining in Texas.
Patricia Suter, Chair Venice Scheurich, Conservation Chair
Coastal Bend Sierra Club Coastal Bend Sierra Club
Ph. 361-852-7938 Ph. 361-241-4289
E-mail PhSuter@stx.rr.com E-mail firstname.lastname@example.org
Comments on Clarification of Statistical Hypothesis Test for TCEQ
Stakeholder Meeting February 4, 2009
Good morning. My name is Venice Scheurich.
I am here from Corpus Christi today to explain in person my concerns regarding one of the issues which appears in your list of bullets to be discussed at this meeting: Clarification of Statistical Hypothesis Test.
Thank you for your all-inclusive definition of "stakeholder," which makes us all feel welcome and for the careful attention you have obviously given in responding to stakeholders comments.
In my personal letter to TCEQ, and also in the letter I wrote for Coastal Bend Group Sierra Club (CBGSC), which you received on October 6, 2008, I raised too briefly, at the expense of clarity, my concern about using a statistical test. I appreciate this opportunity to explain my concerns in detail.
I wrote the personal letter because my husband and I are natives of South Texas and property owners who are concerned that our groundwater resources be adequately protected and used wisely. I wrote and co-signed the CBGSC letter because I am secretary and conservation chair of that group, and because I have some expertise in statistics. (I will leave a copy of my CV with you today.)
When we became aware of controversies over efforts to restore groundwater after in situ uranium mining, it was only natural to search for how restoration values were established, how often restoration had been attempted, and what the restoration success rate was.
The shocking discovery was that in only one case (COGEMAS OHearn mine) in more than twenty years of mining had groundwater been restored to the baseline values agreed to when permits were issued. This led to a careful reading of sections 331.104 and 331.107 of the current rules and the proposed rules to learn how baseline wells had previously been selected and how they will be selected under the new rules, as well as how the data on pre-mining water quality obtained from these wells will be used in attempts at future restoration.
As you know from the letters, I made suggestions on sampling methodology for obtaining baseline wells and also on how to use the data in describing pre-mining groundwater quality. When we received your responses to our suggestions (pp. 30-31, pp. 57-59, and pp. 120-122 of the on-line link to Chapter 331 in the Executive Summary sent in B. Thatchers e-mail of January 9, 2009), we were pleased to find your acknowledgement of some of our main concerns and note revisions you made in the latest draft rules. (For example, see TCEQs response on page 58.)
We are especially pleased with the emphasis the Commission now places in 331.104 on the necessity that samples used to establish baseline be representative. Statements that these samples be representative and independent are made many times in your responses to comments TCEQ received from several groups and individuals by Oct. 6.
However, coming now to the matter of "clarification of statistical hypothesis test," it must be noted that the mathematical assumptions upon which such tests (and other inferential statistical techniques as well) are built requires that the sampling must include a random component. This randomness is essential so that the laws of mathematical probability used in constructing the inferential techniques and formulas can properly be used to analyze the sample data from which parameter estimates are calculated and inferences are made about the population from which the sample was taken.
Unfortunately, even though your thoughtfully written responses include many references to how the sample data may, could, or will be used in a statistical method approved by the Executive Director, I find no references to require a random step in obtaining the sample. Note that "random" in the context of sampling design and statistical practice has a technical definition which is not at all the same as "representative."
I do not doubt the Commissions good intentions. In both responses on p. 110, you state: "The Commission intends that any statistical test used to make an inference about a population should be valid." The problem, however, is that without requiring a random step in obtaining the sample data, a valid statistical hypothesis test for making an inference about a population does not exist.
To further clarify this assertion, I call attention to TCEQs response on pp. 119-121. In particular, the last paragraph in that response clearly states that a statistical test to compare two data sets to determine if they were from the same population is considered appropriate. The response contends that such a methodology could be proposed by an applicant under new 331.107(a)(1)(2). One of the sample data sets which would be used is the baseline well data from the production zone within the production area. This, however, would be incorrect because that sample would have been chosen without a statistically valid random step. Thus, its use here would be an abuse of statistical methodology. It must not be allowed.
To cite another example of TCEQs suggesting a statistical procedure which would not be valid because it, too, calls for data from the non-random sample of baseline wells to be used, refer to the response on pp. 121-122. The comments made in this response relative to increasing the sample size so the power of a statistical hypothesis test used to determine the distributional characteristic of the population from which the sample is drawn would also be invalid and must not be used. (Another TCEQ document, Response 96, p. 62, of Nov. 6, 2008, RE: Uranium Energy Corp., Permit No. UR 03075, indicates that the Executive Directors preference for assessing whether data are from a normal distribution is the Shapiro-Wilk Test for samples of size 50 or less. But, again, the assumption in using this test is that the sampling involved a random step. Thus, it would be incorrect to use on the non-random baseline well data from the production zone.)
These examples demonstrate that the proposed rules allow, and even sometimes recommend, incorrect use of inferential statistical methodology on data from a non-random sample of baseline wells.
Happily, there is a relatively simple solution for this problem: Include a section in 331.104 detailing a protocol which would require a random component in determining the locations of the baseline wells.
Fortunately, as you pointed out in your response to CBGSC on pp. 58-59, you have staff people with expertise in applied sampling design and your staff has access to a variety of reference manuals on that subject. We urge you to make use of these resources and include this vitally necessary random step in obtaining the baseline wells so that all of the sound inferential statistical practices which you so often and effectively advocate (for one example, see your response to CBGSC on pp. 30-31) will no longer be in conflict with the sampling method used to obtain baseline well locations.
Ever since TCEQ was created in 2002, you have been "stuck" with using rules which in some cases were written more than two decades ago and which have in some ways perhaps made it impossible for you to avoid violating the most fundamental statistical principles.
All of us who have done statistical consulting realize that frequently we are put in a position of working with data which did not involve the random sampling component. However, we are careful to advise our clients who brought us this data for analysis that necessary caveats must be included prominently in their reports to caution readers that summary numbers are being used in a descriptive, rather than inferential manner.
Now that you are revising the rules, there is no reason to continue using a selected rather than random sampling design. In this particular matter, we are not in a philosophical debate. The subject at hand is not subjective. There is a practical way to obtain unbiased, statistically sound data, and TCEQ must insist on that. Otherwise, Chapter 331 will contain internal inconsistencies and erroneous recommendations re applications of inferential statistical methods.
Also, a vitally important reason for insisting that the sampling technique used to obtain locations of baseline wells is statistically unbiased is the fact that data from these wells is the foundation on which the restoration table is built and on which a judgment is made regarding whether or not groundwater has been restored. Stated another way, it can easily be argued that a random sample is the "linchpin" of this whole process.
Today, perhaps more than ever before, people are questioning whether state and federal regulatory agencies have the poweror even the willto require and enforce rules which adequately protect them and their resources. Fortunately, at this time you are in the unique position to rewrite the rules to make a simple, but profoundly significant change in uranium mining regulations which would not only get the statistics right, but which would be one reassuring example to Texans that TCEQ is working for the good of all.
P.O. Box 10101
Corpus Christi, TX 78460
Submitted by V. Scheurich & M. Williams, 2/7/09
NEW USGS STUDY EXAMINES TCEQ RECORDS ON IN SITU URANIUM MINING IN TEXAS
For three years there has been a resurgence of interest in in situ uranium mining in South Texas counties. Since some Coastal Bend Sierra Club (CBSC) members live in these counties, and because the mining process contaminates with uranium and other harmful substances portions of aquifers supplying water to area residents, CBSC has studied post-mining efforts to restore groundwater to its pre-mining baseline quality.
A shocking discovery was that TCEQs own records reveal that mining companies have failed to restore groundwater uranium levels to pre-mining baseline values agreed to when the companies obtained permits to mine. For documentation, go to USGSs recently posted study (July 14, 2009) by geologist Dr. Susan Hall titled "Groundwater Restoration at Uranium In-Situ Recovery Mines, South Texas Coastal Plain."
This study examines and summarizes groundwater restoration data on file at TCEQs offices. To read the study, go to:http://pubs.usgs.gov/of/2009/1143/pdf/OF09-1143.pdf
Currently, Uranium Energy Corporation (UEC) has applied for the first production area authorization (PAA-1) to begin in situ mining in Goliad County. Officials and residents there have requested and were granted a public meeting and a contested case hearing on UECs application. The executive committee of CBSC, after studying how baseline groundwater quality was determined in the proposed production area, submitted comments to TCEQ.
As the following letters explain, our main concern is that UECs method for choosing locations of baseline wells and their method of summarizing data to derive values for the restoration table are statistically flawed in a most serious way.
Submitted by Venice Scheurich and Mina Williams, October 26, 2009
COASTAL BEND GROUP
P.O. BOX 3512
CORPUS CHRISTI, TX 78404
Office of the Chief Clerk, MC 105 July 20, 2009
Texas Commission on Environmental Quality
P.O. Box 13087 Re: UEC Permit Application UR030375 PAA1
Austin, TX 78711-3087 Request for Public Meeting
Dear Chief Clerk:
Given the historical lack of success by industry in their post-mining efforts to restore groundwater to pre-mining quality, the enormous amounts of groundwater these efforts require, and the difficulties and controversies in establishing accurate pre-mining baseline groundwater quality, Coastal Bend Sierra Club (CBSC) requests, on behalf of our Goliad County members, a public meeting on Permit Application UR03075 PAA1.
Some of these difficulties in establishing baseline groundwater quality have already surfaced in this case when our CBSC efforts to verify groundwater restoration values given in UECs revised Proposed Restoration Table of March 27, 2009, revealed significant inaccuracies in values for uranium and radium-226.
After these errors were brought to the attention of TCEQ, officials there contacted UEC for an explanation. UEC acknowledged the errors, gave their explanation, and Table 6.2 will need to be corrected.
Looking carefully at the entire process by which the Proposed Restoration Table for UR03075 PAA1 was derived necessitates asking other questionssome of which are already being discussed by many Goliad County and other South Texas citizens who are closely following issues involving in situ uranium mining. One of the things they are questioning is what provision, if any, TCEQ is making to prevent the issuing of future draft permits which might contain errors similar to those UEC made in this permit application.
This and other questions that are being asked are important, and we believe that citizens rightly have the expectation that UEC and TCEQ will give them their close attention and provide thoughtful and thorough answers.
One of the main concerns and obvious questions is the following: Why is there such a dramatic difference in values for uranium as well as for radium-226 from the September 2008 Proposed Restoration Table to the March 2009 Proposed Restoration Table?
That is, why did the eight additional baseline wells (PTWs 7 through 14) which were tested in the production zone between July 2, 2008, and September 20, 2008, differ so strikingly on values for uranium and radium-226 from the original ten baseline wells (RBLBs 1, 3, 4, 5 and PTWs 1 through 6) which were tested between July 11, 2007, and May 12, 2008? For example, the original ten baseline wells had an average uranium value of 33 micrograms per liter, but the additional eight baseline wells, which were tested several months later, had an average uranium value of 218 micrograms per liter.
And, therefore, UEC stated an initial uranium restoration value of 33 micrograms per liter (September 2008), but stated the revised restoration value for uranium, using all 18 baseline wells (March 2009), as 151 micrograms per litera value more than 4.5 times as high as the initial 33. Note that correcting one of UECs significant calculation errors in their revised Permit Application brought the value of 151 down to 115, but the 115 value is also dramatically higher3.5 times as high as the initial value of 33 micrograms per liter.
Moreover, it is also striking to note that the lowest value of uranium (86 micrograms per liter) in the eight additional wells which were tested at a later date is higher than all ten uranium values (highest = 80 micrograms per liter) for wells in the initial ten which were tested earlier and which provided values for the initial Proposed Restoration Table.
Many questions arise as to what the possible, as well as the most likely, explanations are for these striking differences in uranium values between the initial sample of ten baseline wells and the later sample of eight additional baseline wells. Note that both samples were from locations within the 36.1 acre proposed production zone.
Just a few of the more obvious questions begging to be considered follow:
How did UEC guard against selection bias when they chose locations for the samples of wells?
Did something occur during the time between drilling/testing the first set of ten wells and drilling/testing the second set of eight wells that influenced the amount of uranium in the groundwater?
What assurance is there that UECs restoration efforts in Goliad Countys PAA1 will be more successful than the almost total failure of restoration efforts in other Texas counties in the past?
CBSC believes that these and additional relevant questions from other groups and citizens should be addressed in a public meeting so that all concerned have the opportunity to gain insight from industry and from Texass regulatory agency, TCEQ.
Therefore, CBSC appeals to TCEQ to grant the people of Goliad County, those who will be most directly affected by UECs mining operations, an opportunity to express their concerns in a public meeting prior to the granting of Permit Application UR03075 PAA1.
Patricia Suter, Chairperson
PHsuter@stx.rr.com Ph. 361-852-7938 (home)
Corpus Christi, TX 78404
Venice Scheurich, Conservation Chairperson
email@example.com Ph. 361-241-4289 (home) 361-960-4289 (cell)
P.O. Box 10101
Corpus Christi, TX 78460
CC Representative Yvonne Gonzalez-Toureilles
721 E. 2nd St.
Alice, TX 78332-4714
` COASTAL BEND GROUP
P.O. BOX 3512
CORPUS CHRISTI, TX 78404
Office of the Chief Clerk, TCEQ Oct. 5, 2009
Mail Code MC-105, P.O. Box 13087
Austin, TX 78711-3087
Re: Proposed Production Area Authorization No. UR03075PAA1
Dear Chief Clerk LaDonna Castanuela,
Coastal Bend Group Sierra Club (CBGSC) represents over 400 members in South Texas. Since we have members in virtually every county involved in the recent resurgence of interest in in situ uranium mining, we have been following related activities closely in Goliad County. One of our main concerns has been trying to understand why, in more than two decades of mining in South Texas, groundwater has only once been restored to the baseline values agreed to by industry and state regulators when permits were first issued. This concern has brought us to focus on how restoration values are determined.
In the particular case of UR03075 PAA1, we have many questions about UECs process used in deriving their initial Restoration Table of August 2008, as well as their revised proposed Restoration Table dated March 27, 2009. (Uranium values are the main concern.)
Following is the background uranium data information on which our questions are based.
In the initial UEC application, 10 baseline wells were used, yielding uranium values (micrograms per liter) of:
5, 6, 9, 9, 10, 32, 59, 60, 62, 80 mean=33, median=21
Therefore, 33 was the proposed restoration value for uranium in the initial Proposed Restoration Table.
A statement was made in the application that additional wells had been drilled from which seven would be tested at a later date. UECs revised application shows that eight additional wells were sampled a few months later, yielding uranium values (micrograms per liter) of:
86, 99, 134, 135, 156, 163, 166, 804 mean=218, median =146
Therefore, in UECs revised Restoration Table (March 27, 2009), when these 8 values were combined with the initial 10 values, the restoration value for uraniumbased on the sample mean of 18 valueswas dramatically higher than the initial mean value of 33 in the proposed Restoration Table of August 2008. (UEC listed it as 0.151mg/l [151 micrograms per liter], which was acknowledged to be in error; the correct mean value is 115 micrograms per liter.)
This striking difference between the two sample uranium distributions, which caused a dramatic increase in the revised proposed Restoration Table value for uranium, led to the following questions:
NOTE: In the initial August 2008 application, UEC stated on p. 1-9 that 17 baseline wells had been completed and that analyses from 10 of them were included in that submission. However, when revisions to the proposed Restoration Table (March 27, 2009) were submitted, the table contained analyses from 18 wells. Thus, apparently at some time between August 2008 and March 27, 2009, UEC decided to include data from 8, not 7 wells in the second group. Some of the following questions are independent of this matter. Others identify and request explanation and clarification of confusion resulting from when and why UEC decided to use values from a total of 18, not 17 baseline wells in their revised proposed Restoration Table of March 27, 2009.
1. a) What was the reason for UECs plan, expressed in their Application for PAA-1
(August 2008, p.1-9) to have baseline wells sampled in two different groups, with
several months between sampling the first group of 10 and the second group of 7?
b) Could it have been anticipated, or even expected, that uranium levels in wells
sampled and tested in the second group several months later would be higher than
uranium levels in the first group of 10 wells?
c) Could knowledge of uranium levels at locations for the first 10 wells have been
used to help select locations for the second group of 8 wells so that chances of
finding higher levels of uranium in the second group would be increased?
d) Given that 239 exploration boreholes had been drilled in the Production Area prior
to any drilling of baseline wells, could knowledge of borehole test results have
assisted in choosing locations for baseline wells which would be likely to yield
high uranium values?
2. a) Did UEC use either a systematic grid or some type of probability sampling design
to determine the baseline well locations?
b) If the answer to 2.a) is N0, did UEC use some method relying on personal judgment
or choice (sometimes referred to as "search sampling" or "expert opinion"), which
introduces, even if unintended, a type of selection bias that cannot be quantified?
3. a) Does TCEQ acknowledge that increasing the sample size of baseline wells from 10
to18 did not assure more accurate estimates of groundwater quality unless UEC used
a statistically valid sampling design which is not subject to bias as described in 2.b)?
b) What criteria was used by TCEQ in judging whether or not UECs choice of
locations for the 18 baseline wells resulted in a representative sample?
c) What, if anything, in TCEQs criteria given in the answer to 3.b) would have
disqualified the sample of the first 10 baseline wells (RBLBs 1,3,4,5, PTWs 1- 6)
as representative and sufficient to provide values for a valid Restoration Table?
Given that TCEQ regulations in 30 TAC 331 require a specific, detailed protocol for determining locations of baseline monitor wells, why is a specific and valid protocol not required for determining locations of representative baseline wells in the Production Area?
If 30 TAC 331 also specified a required, detailed protocol for obtaining locations of Production Area baseline wells, as it does for baseline monitor wells, would it not be beneficial for the following reasons?
a) All stakeholders would understand why data used in constructing the Restoration
Table was taken from the specified locations and would be reassured that selection
bias was not involved;
b) Industry could no longer be justifiably accused of "cherry-picking" the locations for
c) TCEQ could no longer be suspected of favoring industry in this matter.
a) Why did UEC decide to sample 8 additional wells (PTWs 7 thru 14) instead of 7,
as they had initially indicated (August 2008, PAA-1 Application, p. 1-9) to
combine with the first sample of 10 wells (RBLBs 1,3,4,5, PTWs 1 thru 6)?
b) One of the additional wells, PTW-14, was renamed from its initial name of CBP-1
and it was sampled more than two months before PTWs 7 thru 13;
1) Why was this renaming and earlier sampling done?
2) Were there more CBP wells or other wells (in addition to OMWs 1 thru 9)
drilled in the Production Area which have not been designated as baseline wells,
and if so, how many and what are their locations in the Production Area?
3) What was the initial purpose of CBP-1 (before it was renamed and chosen as a
baseline well), and when was the first lab report for that well, whatever its name,
(prior to April 2, 2009), issued?
7. On page 1-9 of UECs initial PAA-1 Application, submitted in August 2008, they
state: "In developing Production Area baseline quality water, UEC exceeded the
minimum requirement by completing 17 wells. Sample analyses from 10 of these
wells are included in this submission."
Why were sample analyses submitted for only 10 wells at this time?
Are lab sheets available for the 7 wells which were completed but not included in the August 2008 submission?
What names or labels were given to these 7 wells?
Is a Production Area Map available indicating the locations of these 7 wells?
What was the purpose of these 7 wells?
8. On page 1-9 of UECs revisions to the initial PAA-1 Application, dated March 27,
2009, they state: "Seven additional wells are scheduled to be sampled in early
September. TCEQ is planning to collect samples from some of the baseline wells
during the September sampling period. UEC plans to supplement the production
zone water quality baseline data with results from the upcoming sampling."
Has this sampling been done by now (Oct. 5, 2009)? If not, when will it occur?
Has this new data supplemented the production zone water quality baseline data by now (Oct. 5, 2009)?
Has UEC submitted revisions containing this new data by now (Oct. 5, 2009)? If not, when will such revisions be issued?
9. a) Since the sample of 18 baseline wells contains an extreme value (outlier) of 804
micrograms per liter for uranium from PTW-7, what would be the justification
for using the sample mean instead of the sample median for the value of uranium
in the Restoration Table?
Comment: TCEQ has acknowledged that, under 331.107(b), use of the sample
median is allowed. Also, (see 34 TexReg 1668 March 6, 2009 Texas Register):
"The commission notes that in the case of a small data set that has an extreme
value, which can significantly affect the sample mean, use of the sample median
is an example of accommodation of an outlier."
Note: In UECs PAA-1 data for uranium (micrograms per liter), using all 18
baseline well values (the outlier of 804 included), gives a mean of 115 and a
median of 71. When the outlier value of 804 is deleted, the sample mean for the
remaining 17 values is 75. Clearly, including the outlier of 804 causes a dramatic
upward shift of the value of the sample mean.
b) In this a case, would TCEQ agree that using the sample median value for uranium
of 71 in the Restoration Table is better than using the sample mean value of 115?
10. a) Has UECs March 27, 2009, revised proposed Restoration Table, which
contained erroneous values (because of UECs acknowledged data processing
mistakes that were also included in the June 18, 2009, revised Final Draft Permit,
issued for PAA-1 by TCEQ) been corrected?
b) When will UEC and TCEQ make available the corrected copy of the Restoration
In the November 6, 2008, Executive Directors Response to Comment Permit No. UR03075,
concerned citizens were encouraged upon reading in response 36 on page 26: "Protection of groundwater quality is the most significant concern in regulating in situ uranium mining." The first critical step in such protection is establishing the Restoration Table, which provides the pre-mining values to which post-mining groundwater must be restored.
These pre-mining values for the Restoration Table must be determined by using valid estimation methodology which eliminates, or at least minimizes, the chance of introducing unquantifiable selection bias in the very first step of choosing baseline well locations.
Unless such a bias is carefully guarded against and controlled, the entire restoration process is so flawed as to become misleading and meaningless.
In 34 TexReg 1652 March 6, 2009 Texas Register, TCEQs position is stated as follows:
". . . the commission takes into consideration whether the samples used to establish baseline are representative. . . . Obtaining representative samples would certainly involve evaluation of the locations of baseline wells, and any evaluation by the commission regarding whether samples are representative would include consideration of how the baseline wells were located." [Emphasis added.]
Also, in 34 TexReg 1668 March 6, 2009 Texas Register, TCEQs position is stated as follows:
". . . the commission can determine that a sample data set is not representative, as required under revised 331.104 (a), and require additional samples from existing baseline wells or the completion of additional baseline wells."[Emphasis added.]
CBGSC and other citizens concerned about groundwater are depending on TCEQ to exercise the prerogatives cited above in the matter of locating baseline wells in UECs PAA-1.
CBGSC anticipates with appreciation TCEQs careful consideration of and responses to all of the questions, concerns, and requests contained in this letter.
Patricia Suter, Chairperson Venice Scheurich, Conservation Chairperson
1002 Chamberlain St. P.O. Box 10101
Corpus Christi, TX 78404-2607 Corpus Christi, TX 78460-0101
COASTAL BEND HOME PAGE ARCHIVES IMAGES
Website Design: Cheryl McGrath