DRC-2015-001201 - 0901a068804dfd4d

it-;
State of Utah
GARY R HERBERT
Governor
SPENCERJ COX
Lieutenant Governor
Department of
Environmental Quality
Amanda Smith
Executive Director
DIVISION OF RADIATION CONTROL
Rusty Lundberg
Director
DRC-2015-001201
January 22, 2015
Kathy Weinel, Quality Assurance Manager
Energy Fuels Resources (USA) Inc.
225 Union Boulevard, Suite 600
Lakewood, Colorado 80228
RE: Transmittal of Geotechnical Review of Energy Fuels Resources (USA) Inc., White Mesa
Mill, Tailings Data Analysis Report dated October 2014, and Probabilistic Seismic Hazard
Analysis Report dated July 2014, RML# UT 1900479, San Juan County, Utah
Dear Ms. Weinel:
Please find enclosed the Utah Division of Radiation Control (DRC) Technical Memorandum on
the geotechnical review of the MWH Americas, Inc. (MWH) Tailings Data Analysis Report dated
October 2014. Additionally included herein is the Technical Memorandum document regarding
AECOM's (formerly URS) review comments [URS Project No. UT11.1102.004] on the same
Tailings Data Analysis Report, and the Probabilistic Seismic Hazard Analysis Report dated July
2014 also prepared by MWH for Energy Fuels Resources (USA) Inc. (EFRI). The DRC has no
additional comments on the Probabilistic Seismic Hazard Analysis (PSHA) report.
The Tailings Data Analysis Report was prepared pursuant to the MWH July 2013, White Mesa
Mill Tailings Characterization and Analysis Work Plan. The Probabilistic Seismic Hazard
Analysis report was prepared pursuant to DRC's February 2013 review comments on EFRI's
August 2012 responses to DRC's Round 1 Interrogatories for the White Mesa Reclamation Plan
Rev. 5.0.
We anticipate that the License will be able to respond to these review comments within 45 days of
receipt. It is our expectation that the responses will be in the form of revised reports. Afterward
the DRC will review the Licensees response to support EFRI's efforts to subsequently provide
responses to the Round 1 Interrogatories dated February 7, and 13, 2013, by incorporating the
results of the Tailing Data Analysis and PSHA reports.
195 North 1950 West • Salt Lake City, UT
Mailing Address P O Box 144850 • Salt Lake City, UT 84114-4850
Telephone [phone] • Fax [phone] • T D D (801) 536^1414
www deq utah gov
Printed on 100% recycled paper
Page 2
If you have any questions or require clarification concerning the technical memos, please contact
Mr. Eric Boone or me at [phone].
Sincerely,
'John Hultquist, Licensing Program Mgr.
Division of Radiation Control
JH/EAB/,
Enclosures
cc: David C. Frydenlund, Energy Fuels Resources (USA) Inc.
Jon Luellen, AECOM (Formerly URS)
Department of
Environmental Quality
Amanda Smith
Executive Director
State of Utah DIVISION OF RADIATION CONTROL
Rusty Lundberg
Director
GARY R HERBERT
Governor
SPENCERJ COX
Lieutenant Governor
TECHNICAL MEMORANDUM
Geotechnical Review of Energy Fuels Resources (USA) Inc., White
Mesa Mill Tailings Data Analysis Report dated October 2014, Prepared
by MWH Americas, Inc., RML# UT1900479, San Juan County, Utah.
January 22,2015
Introduction
This technical memorandum prepared by Division of Radiation Control (DRC) staff presents
geotechnical review comments on the subject Tailings Data Analysis Report prepared by MWH
Americas, Inc. (MWH) for Energy Fuels Resources (USA) Inc., (EFRI). The subject report
presents the results of a work plan dated July 2013 (Revision transmittal dated August 1, 2013) to
collect site-specific tailings data on tailings Cells 2 and 3 at the White Mesa facility in San Juan
County, Utah. The objectives of the referenced work plan have been restated in Section 1.2 of the
report.
Findings
1. Section 2.1 - CPT Soundings
The widely dispersed cone penetration testing (CPT) soundings have provided a significant
improvement in the available data to model the geotechnical properties of the tailings soil
profile within Cells 2 and 3. MWH refers to the 1986 paper by Larson and Mitchell (L&M)
for the U.S. Department of Energy Uranium Mills Tailings Remedial Action (UMTRA)
Project which provides early experience interpreting CPT data to characterize uranium tailings
piles. Notwithstanding the variation of tailing soils over small distances causing a soil sample
taken at a given interval to potentially be quite different from the soil penetrated by an
adjacent CPT sounding, the paper is quick to point out that predicted Unified Soil
Classification System (USCS) material classifications within a typical CPT classification zone
may vary greatly with site specific classification testing results. This paper highlights the
importance of developing site specific correlations between the CPT record and site specific
laboratory classification and proposes a classification scheme unique to uranium mill tailings.
The L&M scheme utilizes three traditional brackets to capture and categorize mine tailings:
(1) Sand which is material with 0% to 30% passing the #200 sieve; (2) Sand Slime which is a
mixture that has 30% to 70% passing the #200 sieve; and (3) Slime which is a material with
70% to 100% passing the #200 sieve. MWH has recommended adjustments to the L&M
195 North 1950 West • Salt Lake City, UT
Mailing Address P O Box 144850 • Salt Lake City. UT 84114-4850
Telephone [phone] • Fax [phone] -TDD [phone]
www Jeq Utah gov
Printed on 100% recycled paper
Page 2
Geotechnical Review of Tailings Data Analysis Report
January 22, 2015
scheme which is to be discussed later.
The work plan anticipated 7 CPT soundings in each cell for a total of 14 soundings. More
than 14 CPT soundings were completed. Each CPT sounding was to extend into the tailings
profile to at least within 5 feet of the predicted depth to the cell liner. The CPT soundings
within Cells 2 and 3 typically reached to within 2 feet and 5 feet of the predicted liner depth,
respectively. Several noted exceptions were CPT soundings CPT-3-8S, CPT-3-4N, and CPT-
3-3S which were each terminated at depths slightly more than 5 feet from the predicted liner at
approximately 7.5 feet, 8.3 feet, and 9 feet, respectively. In general, the DRC acknowledges
that the CPT soundings collected field data as it was intended to undertake. However, the
work plan indicated the CPT soundings would be used to develop profiles that characterize the
tailings stratigraphy and thereby allow for interpretation and modeling of the various tailing
materials both vertically and laterally. Cross-sections (profiles) through the tailings
impoundments are absent from the Tailings Data Analysis Report. Without cross-sections
depicting the stratigraphy of the tailings materials at each CPT sounding it is unclear how the
tailings material types are distributed and therefore uncertain how the tailings should be
geotechnically modeled. Please provide profiles that depict the stratigraphy within each
tailings cell both vertically and laterally.
The Location Map identified as Figure 2-1 does not consistently call-out the depth penetrated
by each CPT sounding. Please review and update the map with the missing information.
2. Section 2.2 - Direct Push Sampling
The DRC acknowledges that additional direct push explorations were added over the work
plan amount of 2 explorations per cell. Sample data collected from the field direct push
sampling program will be invaluable to understanding the degree of variability of geotechnical
physical properties within the material placed in Cells 2 and 3. This data goes to the primary
goal of calibrating the abundant CPT soundings. However, before acknowledging the Direct
Push field program achieved the objectives of the work plan the following review comments
need to be addressed.
The eight representative logs in Appendix B need to be internally consistent with respect to
grammatical technique and symbol usage. Please indicate in a suitable place the standard
ASTM practice (2487 or 2488; or both) used to classify the soil encountered. Symbols
within the "Run" column appear to indicate on a few logs that there are sample runs over 24
inches in length up to 36 inches in length, however most sample runs were 24 inches. Review
this representation of 36 inch sample runs and confirm that it is correct. The work plan
described the sampler as being 12 to 18 inches in length and Section 2.2 of the report indicates
the sampler was 24 inches in length with an internal diameter of 1.5 inches. If it was possible
to achieve sampling runs over 24 inches in length please describe in further detail the
longitudinal dimensions of available sampling jars, as well as the length of the sample
sleeves or rings. Provide details of the alternate sampler set-up to assure that the
Page 3
Geotechnical Review of Tailings Data Analysis Report
January 22, 2015
sampler could accommodate accidental over driving without disturbing (compressing)
the sample. Please clarify/revise instances where the push sample symbol is absent from
the "Push Samples" column on the following logs: CPT-2W3; CPT-2W4-C; CPT-2W6-
S(3); and CPT-2E1.
Please describe the rationale that was used to determine what portion of the tailings
profile is represented by a typical 24 inch sample run that recovered less than 50% of the
penetration length attempted, especially for the longer sample attempts. For example
explain how 2 to 6 inches of recovered material from a 24 inch sample run was accurately
positioned on the log. On initial review there seems to be a bias to placing the recovered
sample as representative of the bottom of a 24 inch sample run and then scheduling and
developing lab results to establish a correlation to the CPT data from this designated "bottom
depth". This procedure could incorrectly place material that was captured at the initial
penetration to the bottom of the sample interval.
Difficulties with achieving decent sample recovery are a factor with every successful
exploration program. It is noted that based on current information on the Direct Push logs,
sample recovery achieved an overall success rate of approximately 40% recovered of the
sample run attempted. Furthermore of the nearly 160 lineal feet of Direct Push explorations
the total sample length recovered represents less than 20% of the lineal feet explored by the
Direct Push explorations. Ideal recovery rates would minimize introducing error and
uncertainty, below 50% recovery might be considered too uncertain given the narrative on
sampling procedures discussed in the preceding paragraph. MWH needs to clearly indicate
what recovery criteria would be appropriate for correlation and why.
These aspects of the sampling procedures as discussed in the preceding two paragraphs are
especially important to understand based on the adjacent CPT soundings the tailings profile
frequently changes classification vertically within several inches and certainly within a 24
inch sample run. Given the inherent frequent profile changes, the tailings
characterization report needs to explain clearly how any proposed correlation scheme
accounted for 1) an apparent overall low sample recovery; 2) an often limited amount of
material being recovered for testing; and 3) the apparent uncertainty of sample location
within the 24 inch interval, along with the associated biased to assign samples to the
bottom of a sample run.
It is noted that sampling within the upper sand section (interim cover / platform fill) of each
tailings cell is nearly absent, there are 2 possible representative samples collected at the
interface with the tailings soil, please indicate if this omission was intentional and describe
how this absence of data will be filled. To be complete the Tailings Data Analysis Report
should needs to include interpretation, past or present, on the geotechnical properties of this
sequence of material. Another material identified by the CPT soundings that was not sampled
and tested consists of sequences of Sensitive Fine Grained soil (this item was also identified in
Page 4
Geotechnical Review of Tailings Data Analysis Report
January 22, 2015
URS, 9/24/14). See CPT plots for SP2W3; SP3-3S; and SP3-6N and other plots which depict
sequences of Sensitive Fine Grained soil. This material falls within Zone 1 of typical soil
behavior classification charts. The L&M plot of data does not appear to have to account for
this zone as they didn't have data to plot within this zone. Please review the CPT data
within this zone and clearly justify within the report how the geotechnical properties of
the Sensitive Fine Grained soil are to be modeled.
The following are several editorial review comments. The elevation information is absent
from each log, please revise each log to include this information. The moisture content and
dry density for the sample from CPT-2W4-C @ 8.9 feet have been incorrectly posted to the
log of CPT-2W3. The moisture content and dry density for the sample identified as CPT-3-
6N @ 10.5 feet have been omitted from the log. The columns for % Gravel - % Sand - %
Passing No. 200 sieve would be expected to add up to 100%. While minor there are a few
instances where the % Sand is off by 0.1 % and appears to be associated with a rounding error.
A bigger deviation from the lab sheet result to the data placed on the log is noted for sample
CPT-3-6N @ 6.5 feet with the % Sand entered on the log. Please review these comments
and revise the logs and report as appropriate. The photo logs are very helpful and
appreciated. Please consider adding a running head and/or page numbering to the pages of
photos in Appendix C.
3. Section 3.0 - Laboratory Investigation
The following four paragraphs describe procedural aspects of the laboratory program that
were identified during the DRC review but were not thoroughly acknowledged within the
body of the Tailing Data Analysis Report. With the intent to develop a site specific
correlation to CPT soundings, please review these items and expand the narrative of the
characterization report to account for them and how they might or might not affect the
correlation. The subsequent review comments are based on technical or editorial items.
Delayed Testing
The DRC notes that with the delay in testing of often over 2 months, ordinary expectations for
timely geotechnical testing conditions were not observed. With an exception of one
consolidation test completed within approximately 1.5 months, the remaining four
consolidation tests where started more than 3 months after they were recovered from the
tailings. Ideally geotechnical laboratory testing for consolidation parameters would
commence directly upon returning from the field with the samples.
Shipping Disturbance
While understood that it was not originally anticipated, there is limited mention in the report
how the specimens were physically handled during the 1300 mile journey between Colorado
and Tennessee. Please indicate whether the samples were shipped commercially or not.
The DRC believes there would be considerable opportunity for sample disturbance caused by
Page 5
Geotechnical Review of Tailings Data Analysis Report
January 22, 2015
the shipping of the samples to Tennessee in lieu of the proposed laboratory situated roughly 70
miles south of MWH's Fort Collins office in Lakewood, Colorado (understood to be
subsequently disqualified).
Sample Tool Disturbance
Please research and interpret published studies on the potential disturbance of Direct Push
samples with inner diameters equal to or less than 1.5 inches that are used for geotechnical
testing. Section 6.2.2 of ASTM D2435 (Consolidation test method) states that the minimum
specimen diameter or inside diameter of the specimen ring shall be 2 inches. The samples
obtained are 1.4 inches in diameter or approximately 70% of the specified minimum diameter.
To further understand the impact of a smaller sample consider if the outer 1/8-inch perimeter
ofthe 1.4-inch diameter specimen is disturbed by internal wall friction, this results in 33% of
the specimen area being disturbed.
Gypsum Presence
There is the concern of the influence of gypsum (CaS04-2H20) being present in the tailings
samples and thus affecting the accuracy of several laboratory test methods. The 2nd paragraph
of report Section 3.0 acknowledges the potential for high moisture contents and high fines
contents. The method to determine moisture content of soil, ASTM D2216, specifically points
out that standard lab procedure may dehydrate the crystalline water contained in gypsum and
suggests that a lower drying temperature of 60° C be utilized in lieu of the standard 110° C.
As acknowledged the higher drying temperature burns off the hydrated water resulting in
erroneous higher moisture contents and the creation of anhydrite particles not normally
present in the natural tailings material. The potential error enters in the results of ASTM
Dl 140 (#200 Sieve wash) with potentially higher fines contents; the results of ASTM D4318
(Atterberg limits) which are entirely based on moisture contents; and the results of ASTM
D422 (gradation) which would be affected similarly to ASTM Dl 140. It is unclear what is
causing the abrupt curvature behavior of the hydrometer gradation curves.
MWH states in the second paragraph of Section 3.0 "The measured laboratory data used in
Larson and Mitchell (1986) study did not account for gypsum in the tailings. " This
conclusion may not be correct in as much as the L&M paper is silent on whether their test data
accounted for gypsum. The reviewer concurs that this concern will affect certain input
parameters for liquefaction hazard analysis which benefit from fines content. Possibly the
correction for fines content will need to be conservatively reduced.
General Laboratory Review Comments
The following review comments are based on technical or editorial items noted during DRC's
review of the Laboratory Investigation section of the report.
Page 6
Geotechnical Review of Tailings Data Analysis Report
January 22, 2015
Figure E.4-1 Summary of Atterberg Limits Tests Results has incorrectly plotted division
lines at the lower left corner of the standard plasticity chart. The "A"-line has been extended
diagonally to the X-axis instead horizontally at PI = 4 from an LL = 0 to 25.5. The "U"-line
has also been extended diagonally to the X-axis instead of vertically at LL = 16 to a PI = 7.
This is clearly depicted in Figure 4 of ASTM D2487. Please review the details of the
standard figure and make corrections as appropriate.
The consolidation test identified as CPT-2W6-S(2)@13 feet has been classified to be
representative of tailings slimes, however the total weight of the specimen used in the
consolidation test set-up is indicative of a sand - slime specimen. Similarly, the consolidation
test identified as CPT-2W6-S(3)@15 feet has been classified to be representative of tailings
sand - slimes, however the total weight of the specimen used in the consolidation test set-up is
indicative of a slime specimen. Please research and review the laboratory testing data as well
as the groupings and graphs that included these results to be sure it is being included with the
appropriate soil grouping. These are examples of the variability of the tailings profile within a
short distance. If appropriate please review and revise any other report component (such
as Table 3-2 or Figure E.l-1) that relied on this data or interpretation.
The eleven ASTM D422 lab test sheets report an increase in the percent passing the #200
sieve from the result of the ASTM Dl 140 test to the subsequent D422 test result. The amount
of increase ranges from 2.2% to 11.8% with an average increase of 4.9%. While an increase
in the % passing the #200 sieve from the initial wash (Dl 140) to the after dry sieve wash is
common, it is typically small. ASTM D6913 indirectly indicates that an increase greater than
2% could be indicative of a problem such as degradation during mechanical shaking; loss of
sample during testing, or other issues such as the influence of the dehydrating the gypsum
crystals and thus appearing to pass the crystalline water as wash water. Please research and
review the laboratory testing data and procedures for the eleven gradations with S&ME
to be sure the tests were performed correctly. If needed please review and revise any
other report component that relied on this data or interpretation.
4. Section 4.1 - Tailing Classification - Correlation
As indicated earlier a characterization scheme developed by L&M has been adopted by MWH
to capture site specific field and lab data with adjacent CPT sounding data and thereby making
it possible to classify material catalogued in the remaining CPT soundings. MWH has
interpreted their data and concluded an adjustment to the L&M brackets is necessary. MWH
has recommended a uniform lateral shift in the curve between the sand-slime and slimes; a
revision in the criteria for % fines content between the sand-slime and slimes from 70% to
60%; and finally the removal of the curve dividing sand from sand-slime material, resulting in
two material types sand-slime and slime. As discussed in the following paragraphs the
adjustments appear to be without merit, based on laboratory and field data uncertainty or
deficiencies.
Page 7
Geotechnical Review of Tailings Data Analysis Report
January 22, 2015
The classification curves by Larson and Mitchell are reported to be based on continuous data
which is neither the case for data presented in the report nor anticipated with the work plan.
The interpretation to adjust the L&M curves is based on 20 specimens from approximately
160 lineal feet of exploration, that were selected for correlation purposes and subjects of
gradation testing. Of the 20 specimens, 8 specimens were from sample runs with recovery
rates less than 50%. Therefore nearly half of the specimens are subject to the uncertainty
discussed previously with regards to sample location within a 24 inch sample run.
There also appears to be several plotting errors in the main interpretation graph, Figure E.l-1
Friction Ratio vs. Cone Resistance Tailings Classification. The graph appears to have
incorrectly plotted or transposed gradation and Cc data for the sandier sample from CPT-2W3
@ 7.0 feet with the more fine grained sample from CPT 3-6N @ 5 feet. Please review and
revise this figure and any other report component that relied on this data or
interpretation. While the plot of data from the sample at CPT-3-4N @ 9' was excluded it
emphasizes the complex nature of the tailings. The specimen consisted of 9 inches of soil
from a 30 inch sample run. The gradation result of 19.6% fines content classifies the
specimen as sand. The adjacent CPT log SP-34N appears to interpret the following 4 soil
behavior transitions between the 9 to 11.5 feet interval: Silt / Sensitive Fines / Clay / Sandy
Silt.
The DRC has the following observations with regards to removing the curve defining the
transition from sand to sand-slime. The combination plots of CPT Data from Cells 2 and 3
(Figure E.l-3 and Figure E.l-4, respectively) clearly indicate that there are sands in the
tailings profile. The field program recovered tailings that classified as sand as indicated with
4 of the 20 gradation tests. The number of samples appears to be justification to not remove
the published division line. Furthermore, a conclusion that there are no sands and that the
tailings are predominantly made up of sand slime tailings may be an unsupported conclusion.
Without cross-sections depicting the stratigraphy of the tailings this may be an unconservative
simplification ofthe tailings profile.
With the examples above as well as the numerous comments presented earlier in this review
memo with regards to uncertainties with the Direct Push exploration program and the
laboratory data it is not clear that the adjustments to the L&M classification scheme are
adequately justified.
Interim Cover Material and Sensitive Fines Grained Material
Additionally, the CPT soundings revealed two soil behavior types that have not been
adequately characterized in the Tailings Data Analysis Report. The first being the surface
sequence of sandy soil with debris in Cells 2 and 3. The second being the sequences of
Sensitive Fine Grained soil (See SP2W3; SP3-3S; and SP3-6N). The work plan indicated the
CPT soundings would be used to develop profiles that characterize the tailings stratigraphy
and thereby allow for interpretation and modeling of the various tailing materials both
Page 8
Geotechnical Review of Tailings Data Analysis Report
January 22,2015
vertically and laterally. Without cross-sections depicting the distribution of these tailings
materials at each CPT sounding it is unclear how the tailings should be geotechnically
modeled for these two soil behavior types. Please provide profiles that depict the
stratigraphy within each tailings cell both vertically and laterally.
References:
ASTM Designation: D2435-11, Standard Test Methods for One-Dimensional Consolidation
Properties of Soils Using Incremental Loading, American Society for Testing and Materials,
Annual Book of ASTM Standards, Section Four, Construction, Vol. 04.08, West
Conshohocken, Pennsylvania. 2013. www.astm.org.
ASTM Designation: D2487-11, Standard Practice for Classification of Soils for Engineering
Purposes (Unified Soil Classification System), American Society for Testing and Materials,
Annual Book of ASTM Standards, Section Four, Construction, Vol. 04.08, West
Conshohocken, Pennsylvania. 2013. www.astm.org.
ASTM Designation: D6913-04, Standard Test Method for Particle-Size Distribution (Gradation)
of Soils Using Sieve Analysis, American Society for Testing and Materials, Annual Book of
ASTM Standards, Section Four, Construction, Vol. 04.09, West Conshohocken, Pennsylvania.
2013. www.astm.org.
Larson, N. B., and Mitchell, B. (1986). Cone Penetrometer Use on Uranium Mill Tailings. In
Samuel P. Clemence, Editor, Use of In Situ Tests in Geotechnical Engineering: Proceedings
on In Situ '86, a Specialty Conference sponsored by the Geotechnical Engineering Division of
the American Society of Civil Engineers, Geotechnical Special Publication No. 6, pgs. 700-
713.
MWH Americas, Inc. (2014). Energy Fuels Resources (USA) Inc., White Mesa Mill, Tailings
Data Analysis Report, October.
MWH Americas, Inc. (2013). Energy Fuels Resources (USA) Inc., White Mesa Mill Tailings
Characterization and Analysis Work Plan, July.
Utah Department of Environmental Quality, Division of Radiation Control, Review of Energy
Fuels Resources (USA) Inc. White Mesa Mill Tailings Characterization and Analysis Data
[URS Project UDRC 1102.004] UT1900479. September 24, 2014.
URS Technical Memorandum
Date: January 21, 2015
To: John Hultquist, Utah Division of Radiation Control
From: jon Luellen, Christina Winckler, and Ivan Wong, URS Corporation
Subject: Review of Review of Energy Fuels Resources (USA) Inc., White Mesa Mill, Tailings Data
Analysis Report (October 2014) and Probabilistic Seismic Hazard Analysis Report (July
2014) - Reports prepared by MWH Americas, Inc. - URS Project No. UT11.1102.004
This Technical Memorandum documents URS's review comments on the following documents
prepared by MWH Americas, Inc. (MWH) for the Energy Fuels Resources (USA) Inc. (EFRI)
White Mesa Mill Site near Blanding, Utah:
1. White Mesa Mill, Tailings Data Analysis Report (October 2014); and
2. White Mesa Mill, Probabilistic Seismic Hazard Analysis Report (July 2014).
/. Review Comments on White Mesa Mill, Tailings Data Analysis Report (October
2014)
This report presents the CPT investigation and laboratory results for the tailings investigation
of Cells 2 and 3 at the White Mesa site.
MAJOR FINDINGS
1. Section 1.2, Objectives, and Sections 4.0 and 5.0: Overall, there are no conclusions or
recommendations on how the CPT and laboratory data will be used, or how these data
compare to previous work. Additionally, it is unclear in Section 1.2 whether additional
objectives of this investigation include, for example: (1) To acquire CPT and laboratory
data to be used to assist in development of cross-sections through the existing tailings
impoundment providing interpretation between various tailings types (i.e., sands, slimes
and transitional tailings) and their distributions for use in final cover design; (2) To
interpret the over-consolidation ratio and sensitivity of the tailings using CPT data to
assist in evaluation of material behavior; etc.? Please provide cross sections showing
inferred distributions of sand, sand-slime and slime tailings types in the two cells
investigated and indicating how the CPT characterization is used on sections. Please also
provide a summary of tailings data collected to date and their intended use(s), including
how the current data compare to previous information/data provided on tailings properties
and indicate whether data acquired to date are considered adequate for fulfilling the
intended use(s). (See also additional specific comments below).
2. Figure 2-2: The depths shown of Figure 2-2 are unclear. Were the CPTs pre-drilled and
hence the larger number shown on the figure? For example, sounding SCPT-2W2 shows a
depth of 20.34 ft on CPT log and on Figure 2-1. However, on Figure 2-2 it is shown as
1/22/15
URS Page 2 of 11
January 21, 2015
21.53 ft, which would suggest the sounding was pre-drilled to a depth of 1.19 feet. Please
clarify.
3. All CPT soundings appear to show a very clear upper layer with higher tip resistance and
higher skin friction (could be the interim cover/working platform fill) than the underlying
tailings with lower tip resistance and lower skin friction. Please confirm the distinction
between cover and tailings. All laboratory tests were performed on samples with lower tip
resistance and lower skin friction, i.e. tailings. Please provide information on test data that
are currently available for the interim cover/platform fill, and indicate whether such data
are considered adequate for final cover design.
4. Section 3.0 (all): Recommend an explanation be added to the discussion as to how the
Specific Gravity values determined for the different tailings samples tested might have
been affected by gypsum concentrations in the tailings (owing to the low specific gravity
of gypsum) and how this might impact any analyses completed for the Reclamation Plan
or the Infiltration and Contaminant Transport Modeling Report that incorporate Specific
Gravity values. Approximately what ranges of gypsum contents are expected to be present
in the tailings, according to tailings fraction?
5. Section 3.0: Tables 3-1 and 3-2: A total of 5 tailings samples were tested for hydraulic
conductivity, compared to the 6 hydraulic conductivity tests specified in the Tailings
Characterization Work Plan. Table 3-2 also indicates that no sand tailings were tested for
hydraulic conductivity. Please provide the following information with respect to the
characterization of hydraulic conductivities in the tailings:
i. A comparative analysis of the current hydraulic conductivity testing results (for
sand-slimes and slimes tailings only) relative to (higher) estimates of overall
hydraulic conductivity for the tailings previously developed based on White Mesa
tailings testing data collected in 1987 and 1999 and a comparison to a different off-
site tailings pile. In particular, the previous estimates suggested that the White Mesa
tailings consist of approximately 55 to 57 % sand (e.g., MWH 2010; MWH 2011;
Geosyntec 2007); and
ii. An assessment of the representativeness of the current tailings hydraulic conductivity
testing results with respect to the distribution of sand, sand-slime, and slime tailings
types in the various cells, with respect to the previously estimated tailings hydraulic
conductivity values, and with respect to dewatering and final cover design needs (see
also Comment No. 1 above).
6. The top of saturated tailings (listed in Table 4-1) has been estimated at the depth where
continuous elevated dynamic pore pressures have been encountered. This is not consistent
with the static pore pressure measurements at some locations (e.g. SCPT-2W3, see Figure
1 below) and also the degree of saturation measured in consolidation tests (e.g. SCPT-
2W2 at depth of 7 feet). It is not clear how these data would be used, but it is
recommended to establish a phreatic surface or zero pore pressure line for analysis. Please
provide a discussion on how the data will be used in future analysis.
URS Page 3 of 11
January 21, 2015
7. Section 4.2: The report states that "Equilibrium pore pressures measured during the pore
pressure dissipation tests often yield values that are higher than actual steady-state pore
pressure at the specific location and depth". The reviewer does not necessarily agree with
this statement. Typically, if sufficient time is provided to achieve equilibrium in tailings,
there is a good correlation between static dissipation tests and pore pressures measured by
vibrating wire piezometers, e.g., see Winckler et. al (2014). If no active piezometers are
available, then vibrating wire piezometers should be installed using the CPT rig to
evaluate pore pressures within the tailings. Piezometers would assist in evaluating the pore
pressure with time and provide guidance if drainage is occurring as predicted in analysis.
8. Section 4.2: The report also states "Dynamic pore pressures typically represent the upper
bound to the actual equilibrium pore pressures since they are the sum of the equilibrium
pore pressure and excess pore pressures due to shearing." This is not always true. Lower
or negative dynamic pore pressures could also be generated if the material is dilative i.e.
generates negative pore pressures upon shearing. The next sentence in the report states
"The pore pressures due to shearing are usually positive unless the tailings are heavily
overconsolidated." This does not agree with the laboratory data that showed over-
consolidated behavior in soundings SCPT-2W2 at a depth of 7 ft and SCPT-2W3 at depth
of 7.5 ft, and showed positive dynamic pore pressure at both locations. The dynamic pore
pressures may not reflect hydrostatic pore pressures nor the degree of saturation within the
tailings.
9. Section 4.2, p. 12: The statement is made that "there are also some lenses of elevated pore
pressures at shallow depths, but these are considered perched zones in the interim cover
and/or tailings due to seasonal fluctuations". Are there additional data available that
would confirm that such perched zones are seasonal vs. 'perennial' in nature?
10. URS has found that performing at least three static dissipation tests per sounding is helpful
in evaluating the in situ pore pressure profile. This has been performed at two sounding
locations within Cell 2. Three soundings had two dissipation tests and the remaining eight
locations have one dissipation test. Fortunately, it appears that the static dissipation tests
were run out long enough to reach equilibrium in all cases except for at sounding SCPT-3-
6N. If piezometer data is available the pore pressure profile should be confirmed with this
information. URS's interpretation of the CPT static dissipation tests indicates that there is
near hydrostatic pressure below the ground surface, see Figure 1.
URS Page 4 of 11
January 21, 2015
Pore Pressure (feet)
10 20 30 40
4^ Interim cover
Tailings
10
cu 15 cu
20
•
25
30
• SCPT-2W3 Dissipation
Test Results
^—Estimated Saturated
Tailings
100% Hydrostatic
35
Figure 1: Static dissipation test result for sounding SCPT-2W3 showing near hydrostatic
conditions below ground surface.
It also appear that there is a drainage toward the sump in Cell 2 based on dissipation test
results obtained in SCPT-2W6-S, -S(2), and -S(3), shown on Figure 2, which shows a
phreatic surface approximately 10 feet below the ground surface. These dissipation tests
were plotted together due to the proximity of the soundings.
URS Page 5 of 11
January 21, 2015
Pore Pressure (feet)
0 10 20 30 40
0 H 1 1 ' 1
Interim cover 1 10
Tailings cu 15 cu "
-20
4
25
30
• SCPT-2W6-S, -S(2), and
S(3) Dissipation Tests
Estimated Saturated
Tailings
100% Hydrostatic
35
Figure 2: Static dissipation test result for sounding SCPT-2W6-S, -S(2), and -S(3)
showing near hydrostatic conditions below a depth of approximately 10 feet.
11. It is not clear how the delineation of sand, sand-slime, and slime tailings will be used for future
calculations/analyses/models. Five consolidation tests were performed; based on fines content, two
were performed on slime tailings (67.4 and 97 percent fines) and three were performed on sand-
slime tailings (percent fines between 46.3 and 58.1). The reviewer recommends also looking at
plasticity indices and densities to evaluate material behavior. The interval at 15 ft for SCPT-2W6-
S(3) might be more slime-like than sand-slime like due to plasticity and density.
12. Section 4.3, p. 14: Please clarify/verify what figure or figures (e.g., Figures E.3-1 and E.3-2?) are
relevant to the derivation of the recommended density values listed in Table 4-3. Briefly describe
the basis for selection of the listed average values.
13. Section 4.4: Based on the tested hydraulic conductivity values, there appear to be minor differences
between slime and sand-slime tailings. Please provide a discussion on how these test data will be
used in future analysis (see also Comment Nos. 1 and 5 above and Comment No. 14). Also please
comment on the tested confining pressures that the hydraulic conductivity tests were performed at
and how the tested confining pressures were selected.
14. Section 4.4, pp. 15-16 and Table 4-6: Please provide a discussion comparing the estimated
hydraulic conductivity values listed in Table 4-6 for sand tailings to previous estimates of tailings
hydraulic conductivity described under Comment No. 5 above. Describe how the different
estimates were developed and provide a discussion of the reliability and representativeness of these
estimates of in-situ conditions in the tailings as they relate to characterization of areas/locations
within tailings cells that may consist of more sandy material based on the current investigation and
URS Page 6 of 11
January 21, 2015
previous tailings testing results. Provide additional information regarding how the current and
previous tailings testing data will be used to represent the potential variability in hydraulic
conductivity values across the tailings management cells, especially with regard to sand tailings.
15. The report shows estimated values of the horizontal coefficient of consolidation. Please provide a
discussion on how these data will be used in future analysis. Also discuss how the vertical
coefficient of consolidation (cv) values listed in Table 4-7 compare with estimates of cv that may be
derived from horizontal coefficient of consolidation (cn) estimates/values (e.g., estimated cn values
in Table 4-8) using published empirical methods (e.g., Robertson et al. 1992) and discuss
implications, if any, of apparent differences.
16. The direct push sampling consists of "piston-type" sampler deployed from the CPT rig. The sampler
have 1.5-inch inner diameter and are 24 inches in length. A total of 35 sampling intervals were
target with 24 of these locations with sampling recovery less than 16 inches. Three of the 35
locations had sample recovery greater than the sampler length. Please provide comment as to the
reason for the poor recovery and also indicate where (at what interval) the recovery was obtained.
17. According to ASTM D2435 the minimum specimen diameter or inside diameter shall be 2 inches.
This was not the case for the tested consolidation samples with a reported diameter of 1.4 inches.
Please comment on why larger samples were not obtained using a drill rig to collect samples for
engineering property testing. Typically push-in samples obtained using the CPT rig are for visual
confirmation only and in some cases index property testing if enough sample quantity can be
obtained.
18. Cell 3 was characterized by sampling at two locations, while Cell 2 was characterized by sampling
6 locations. At Cell 3, one location show fines content of less than 13% while the other show fines
content greater than 67%. Please provide comment on the adequacy of the sampling distribution
within each cell and spatial variation within each cell.
EDITORIAL COMMENTS
19. On drilling logs, the permeability should be raised to a negative number. There appear to be other
typos on the exploration logs with respect to reported cc and cv values.
20. On Figures E.2-1 through E.2-8 it would be helpful to show the pore pressure in ft. associated with
the static dissipation test and not only the elevation where the test was performed.
21. Discrepancies between values obtained from the laboratory testing and values reported in Table 4-7
have been noted. Please review and correct discrepancies.
22. The sample recovery noted on boring logs is different interval than noted on the laboratory testing
results. Please check for and correct discrepancies.
URS Page 7 of 11
January 21, 2015
//. Review Comments on White Mesa Mill, Probabilistic Seismic Hazard Analysis Report
July (2014)
This report presents results of a site-specific probabilistic seismic hazard analysis for the reclamation of
the White Mesa Mill site.
MAJOR FINDINGS
1. Overall, the report appears to be lacking sufficiently detailed discussion of the key inputs into the
Probabilistic Seismic Hazard Analysis (PSHA) and their technical bases. In a few cases, errors
were made in the estimation of input parameters. The major contributor to the hazard at the site is
the background seismic source zone in which the site is located. The calculation of the background
earthquake recurrence is thus important and so the compilation of an up-to-date seismicity catalog
that is properly treated must be done correctly.
As detailed below, URS's review indicates that the recurrence calculations do not appear to have
been performed correctly. In current state-of-the-art recurrence calculations, the recurrence needs to
be adjusted for magnitude bias. The approach that has been developed to make such a correction is
relatively new and based on URS's understanding such calculations have been done by very few
consulting professionals. This was not done by MWH in the current study. The basic process of
making the catalog uniform in terms of magnitudes, declustering, removing fault-related events if
there are any or induced seismicity needs to be done correctly. MWH has done most of this process
correctly. However in the computation of the Gutenberg-Richter relationship for the seismic source
zone, the process has not been done correctly as detailed below.
Also the discussion of the PSHA results needs to be expanded so that the potential impact of
earthquake ground shaking on the White Mesa Mill can be evaluated. As described below, only
PGA hazard results are briefly described but none of the other spectral periods. What are the
controlling earthquakes as a function of spectral period? Uniform hazard spectra (UHS) appear to
have been calculated although they are not discussed in the report nor are they tabulated so they
could be used. See detailed comments below.
2. The most significant deficiency of the study was the non-use of the most up-to-date ground motion
prediction models. The study uses the Pacific Earthquake Engineering Research Center's (PEER)
Next Generation of Attenuation (NGA)-West 1 models which were published in 2008. No mention
of the NGA-West 2 models is made in the report even though the models were released in mid-2013
and are currently being used by the earthquake hazards community. The authors of the NGA-West
1 models who are also authors of the NGA-West2 models have stated that the latter should be used.
The models have just been published in the journal Earthquake Spectra but they were readily
available through PEER publications in 2013 or through a spreadsheet that was pubhcally available
through PEER. The USGS used the NGA-West2 models in 2013 for the 2014 National Seismic
Hazard Maps which were released in mid-2014 (Petersen et al, 2014). It is puzzling why the
outdated NGA-West 1 models were used in the White Mesa Mill PSHA. Please provide rationale for
using the NGA-1 West models instead of the NGA-2 West models used in current practice.
URS Page 8 of 11
January 21, 2015
3. The PSHA results are only shown for PGA. UHS were calculated and are shown on Figure 10 but
not discussed. The results for at least 1.0 sec SA should be shown to indicate what seismic sources
are contributing to the long-period hazard. What is the period of engineering relevance to the site?
4. Page 5, Section 3.2.3, last para: It is stated that earthquakes within 5 km of faults were removed
from the earthquake catalog. Which faults? It is well known that in the Intermountain West
particularly within the Intermountain Seismic Belt (ISB), that very few earthquakes can be
associated with known faults. The vast majority of earthquakes in the ISB are background events
probably occurring on buried faults. For example, only one or possibly two small earthquakes can
be associated with the Wasatch fault zone (e.g., Smith and Arabasz, 1991). Therefore by removing
events within 5 km, background earthquakes are being removed that should be included in the
recurrence calculations.
5. Page 6, Section 3.2, Artificially induced earthquakes: It is agreed that induced earthquakes should
be removed from the historical earthquake catalog prior to calculating recurrence for a seismic
source zone. MWH removed earthquake associated with coal mining activities in the Glenwood
Springs and Paonia areas. However there are two areas of well-recognized induced seismicity
including the Book Cliffs-Eastern Wasatch Plateau near Price, Utah where coal-mining produces
earthquakes and the Paradox Valley where earthquakes are being induced by saltwater injection
(Ake et al., 2005). Earthquakes up to M 4.3 have been induced in the Paradox Valley area. Please
state that the Book Cliffs-Eastern Wasatch Plateau induced seismicity has been removed from the
parent catalogs. Please confirm that the induced seismicity in the Paradox Valley has already been
removed in the parent catalogs.
6. Page 9: The term "capable fault" has been abandoned by the Nuclear Regulatory Commission
(NRC). Although the NRC has not updated 10 CFR Appendix A Part 100, the term should
probably not be used. MWH should confirm this.
7. Page 11: It is stated that "the boundaries of the areal source zones were developed based on similar
patterns of historical seismicity". Although seismicity should certainly be considered, using
seismicity alone to define seismic source zones is a departure from state-of-the-practice. Why
wasn't regional geology and tectonics used as a criterion? The site is located within the Colorado
Plateau which can be distinguished from the neighboring provinces based on a number of geologic
and tectonic factors (Wong and Humphrey, 1989). In this same regard, the division of the two
source zones in Figure 5 is both simplistic and unrealistic. Isn't the Dispersed Earthquake Zone
(DEZ) just the Colorado Plateau? Using standard geologic and tectonic nomenclature would be
more appropriate.
8. Page 11: What is the technical basis for assigning the maximum earthquake in the two seismic
source zones by adding 0.5 magnitude unit? Such a practice has no technical basis and is not state-
of-the-practice. The maximum magnitude should be based on the minimum threshold for surface
faulting and the thickness of the seismogenic crust (e.g., dePolo, 1994). For example, MWH
adopted a moment magnitude (M) 7.0 for the ISB. There have been numerous published seismic
hazard studies that state a maximum earthquake for the ISB and provide a technical basis. The
typical value is M 6.5 to 6.75 ± 0.25. Note on page 16, the maximum magnitude for the ISB is
stated as M 6.7 which would be a reasonable value. The maximum background earthquake of M
6.2 for the DEZ is too low. See other published studies which suggest M 6.5 to 6.75 ± 0.25.
URS Page 9 of 11
January 21, 2015
9. Page 11: It is stated that the "largest event recorded in the ISB" was a M 6.5. Actually, the largest
historical earthquake in the ISB was the 1959 M 7.3 Hebgen Lake, Montana earthquake. URS
assumes that MWH is referring to the Utah portion of the ISB in which case, the 1901 earthquake is
correct. Was the 1901 event a background earthquake?
10. Page 11: The calculation of the recurrence for the two seismic source zones does not look correct.
The report states the maximum likelihood approach of Weichert was used. That is standard
practice. However, the truncated form ofthe exponential model should be used and Figures 6 and 7
indicate simply an exponential model was used. The fit of the model shown on Figure 6 does not
look correct using maximum likelihood. The number of events in each bin should be shown so one
can evaluate the sample size and the robustness of the recurrence estimates. Figure 6 shows the
recurrence relationship to extend out to M 7.0 when the report states the maximum M 6.2. Also the
relationship should be truncated.
11. Throughout the report, reference is made to "magnitude" without designation of what magnitude
scale is being referred to. Although this might seem redundant, the magnitude scale should be
specified consistently throughout the report.
12. Page 12: It would have been useful to see the results of the 13 seismic refraction surveys to see the
variability of Vp across the site; in particular, plots of Vp versus geology. Without such plots, it is
difficult to judge the validity of the Vp estimate of 6,500 ft/sec used for the top 30 m and the lower
and upper bounds.
13. Page 13: There are a number of sites where both Vp and Vs have been measured in situ in
sandstone thus avoiding the use of laboratory measurements of Poisson's ratio which are known to
be suspect. Measurements have been made in California (see Wills and Clahan, 2006) for example.
Examination of such data will show that a range of Poisson's ratio of 0.25 to 0.30 is probably too
low. A Poisson's ratio of 0.25 is assumed to be appropriate for crystalline rock at crustal depths. It
is the value used in most crustal velocity models for seismic networks. Hence a Vp/Vs of 1.5 to 1.9
is biased low and should not be used.
14. Page 15: Even though the NGA-Westl models should not have been used in the PSHA (see
Comment #2), the parameters used in the NGA models should have been stated so they could be
evaluated for their accuracy.
15. Page 17: There is no mention of Figure 10 in the text. Figure 10 is also out of order. It should
follow Figure 13.
16. Page 18: The 2014 USGS National Seismic Hazard Maps have been published (Petersen et al,
2014). It would have been more appropriate to compare the results of the study with the most
recent set of maps.
17. Attachment 2: Some of the dip values and slip rates can rounded so as to not indicate a degree of
accuracy that is not present. For example, a slip rate of 0.231 mm/yr can be rounded to 0.23 mm/yr
without a loss of information. Dips such as 66 degrees indicate a level of accuracy that is not there.
URS Page 10 of 11
January 21, 2015
EDITORIAL COMMENTS
18. Page 1, first line: PSHA stands for "probabilistic seismic hazard analysis" as stated in the title of
the report not "assessment".
19. Page 1, Section 1.1, 2ndpara: its "seismically induced ground motions". Delete "amplification".
20. Page 2, Section 1.3, 2nd para: the National Seismic Hazard Maps do not provide the "seismic
event". The maps depict ground motion hazard.
21. Page 3, Section 2.1, 2nd para: "epicenter depths" is an improper term. Epicenters do not have a
depth. They are defined at the earth's surface. "Hypocentral depth" is the correct term.
22. Page 3, Section 2.1, 5th para: "Quaternary" should be capitalized.
23. Page 5, Section 3.2.2, 2nd para: Its stated that the Gardner and Knopoff approach was not used
because it had not provided "good results". On the previous page for the Petersen catalog, the
Gardner and Knopoff algorithm was used. Is this an issue?
24. Page 7, Section 3.4.1, 3rd para: The Paradox Basin network was installed in 1979 not 1962. The
sentence is confusing.
25. Page 7, 2ndpara: "Stepp" is misspelled.
26. Page 10, Section 4.1, last para: The probability of activity is the probability that a fault is
seismogenic, not that it will rupture. A fault could rupture aseismically such as the Paradox Basin
faults that exhibit fault displacement due to salt movement.
27. Page 12, Section 4.3: Vs30 is the "time-averaged shear-wave velocity in the top 30 m".
28. Page 15, Section 5.0. It should be noted that for the Colorado Plateau, the USGS uses ground
motion models appropriate for the central and eastern U.S. in the National Seismic Hazard Maps,
not the NGA models. I agree that the NGA models are most appropriate and should be used in the
analysis. However see my comment #2.
29. Page 17, Section 6.2, 1st line: Please correct: Fault "recurrence" is modeled as both
characteristic...
30. Page 17: "Probabilistic seismic hazard analysis"
31. The choice of a return period of 9,900 years has no regulatory precedent or basis. Why wasn't a
return period of 10,000 years used, which appears in many regulatory guidance documents for a
range of critical structures use? Although the difference between 9,900 and 10,000 years is trivial,
it just seems odd why 9,900 years was selected.
URS Page 11 of 11
January 21, 2015
REFERENCES
Ake, J., Mahrer, K., O'Connell, D., and Block, L., 2005, Deep-injection and closely monitored
induced seismicity at Paradox Valley, Colorado: Bulletin of the Seismological Society
of America, v. 95, p. 664-683.
dePolo, C. M., 1994, The maximum background earthquake for the Basin and Range Province,
western North America: Bulletin of the Seismological Society of America, v. 84, p. 466-
472.
Geosyntec Consultants 2007. "Analysis of Slimes Drain", White Mesa Mill - Cell 4B. Signed
December 3, 2007.
MWH (MWH Americas, Inc.) 2010. "Tailings Hydraulic Conductivity Evaluation", Appendix I of
Revised Infiltration and Contaminant Transport Modeling Report, White Mesa Mill Site.
March 2010.
MWH (MWH Americas) 2011."Tailings Cell Dewatering Modeling", Appendix J of Updated
Tailings Cover Design Report. September 2011.
Petersen, M.D., Frankel, A.D., Harmsen, S.C., Mueller, C.S., Haller, K.M., Wheeler, R.L.,
Wesson, R.L., Zeng, Y., Boyd, O.S., Perkins, D.M., Luco, N., Field, E.H., Wills, C.J.,
and Rukstales, K.S., 2014, Documentation for the 2014 update of the United States
National Seismic Hazard Maps, U.S. Geological Survey Open-File Report 2014-1091.
Robertson, P. K., Sully, J.P., Woeller, D.J., Lunne, T., Powell, J.J.M. and Gillespie, D.G. 1992.
"Estimating coefficient of consolidation from piezocone tests," Can. Geotech. J., 29:
pp. 539-550.
Smith, R.B. and Arabasz, W.J., 1991, Seismicity of the Intermountain seismic belt in D.B.
Slemmons, E.R. Engdahl, M.D. Zoback, M.L. Zoback, and D. Blackwell (eds.),
Neotectonics of North America: Geological Society of North America, SMV V-l,
p. 185-228.
Winckler, C, Davidson, R., Yenne, L., and Gallegos, M., 2014, Pore Pressure Characterization
of Impounded Tailings, USSD 2014, San Francisco conference, p. 1437-1451.
Wills, C.J. and Clahan, K.B., 2006, Developing a map of geologically defined site-condition
categories for California: Bulletin of the Seismological Society of America, v. 96, p.
1483-1501.
Wong, I.G. and Humphrey, J.R., 1989, Contemporary seismicity, faulting, and the state of
stress in the Colorado Plateau: Geological Society of America Bulletin, v. 101, p. 1127-
1146.
CLOSING
We hope that these comments are helpful and would be pleased to discuss these with you and/or MWH
or EFRI at your convenience. Please contact Jon Luellen (303-796-4738) if you have any questions on
the above review comments on the subject reports or if you require any additional information.