ML20210D902
| ML20210D902 | |
| Person / Time | |
|---|---|
| Issue date: | 06/15/1992 |
| From: | Surmeier J NRC OFFICE OF NUCLEAR MATERIAL SAFETY & SAFEGUARDS (NMSS) |
| To: | Chernoff A ENERGY, DEPT. OF |
| References | |
| REF-WM-74, REF-WM-75 NUDOCS 9206180169 | |
| Download: ML20210D902 (11) | |
Text
F m.
1
,)ljN 151992
' Albert-R. Chernoff, Project Manager.
Uranium. Mill Tailings Remedial Action
' Project Office U.S. Department of Energy Albuquerque Operations'0ffice P.O. Box 5400-Albuquerque, New Mexico 87115
Dear Mr. Chernoff:
-The U.S. Nuclear Regulatory Commission (NRC) staff and its contractor, Pacific Northwest Laboratories, have completed their review of the Belfield and Bowman, North Dakota final Remedial Action Plan (RAP) dated December,1990. As a result of this review, the NRC staff concluded that the information provided in the final' RAP.did not resolve the open issues identified in the draft Technical Evaluation Report (dTER).
Therefore, an updated set of dTER open issues is provided as Encloeure 1 to this-letter.. The review also resulted in a set of general comments w eich are included as Enclosure 2. _The NRC staff discussed the updated dTER open issues with DOE during a telephone conference call on June 9, 1992.
As you are aware, until these open issues are resolved, the NRC is not prepared to concur in the RAP'or the Remedial Action Inspection Plan (RAIP).
In order to make.the NRC review more timely, a."roadmap" should be submitted with the revised RAP. This."roadmap" should discuss where and how each updated dTER open issue (Enclosure 1), dTER confirmatory item, dTER Appendix A comment, and general comment of Enclosure 2 is addressed.
.If you have any questions regarding the information in the enclosure, please contact me at-(301) 504-3439 or the NRC Project Manager, Sandra Wastler, at (301) 504 2582.
Sincerely, MGERS$Nagy John J. Surmeier, Chief Uranium Recovery Branch-Division.of low-level Waste Management and Decommissioning, NMSS
Enclosures:
As stated cc:.
S..Arp, 00E/AL Distribution:
U"JSifrmeier"tentrklTFile P F NMSS r/f LLUR r/f Plohaus RBangart JAustin DGillen DRom
~-SWastler Mlayton LJCallan,RIV RHall,RIV,URF0 WBrach=
hj PDR-YES X
ACNW YES X
\\ -( p~
SUBJECT ABSTRACT: BELFIELD/ BOWMAN - ADDITIONAL INFORMATION ON OPEN ISSUES OFC :LLUR-
- LL
- L z.....
NAME:S ol r
- DGillen
- JSurmu er i
bkTb06//[/92
- 06/b[9h b6/$h/92
{'
920A19 W e 970115
'FFICIAL RECORD COPY PDR HASTE WM-74
~ ppa
~.
ENCLOSURE 1 UPDATED dTER OPEN ISSUES The NRC staff considers the level of hydrogeologic characterization at the l
Belfield/Bowmm site (s) inadequate to demonstrate compliance with EPA's pronosed ground-water..cotection standards in 40 CFR 192.
In addition, dTER issue 'o 3.
was revised to address a recently issued staff policy on hydraulic conductiv1;y.
Therefore, additional information related to the open issues identified in the j
draft Technical Evaluation Report is needed as follows:
DTER OPEN ISSUE (la)
The potentiometric surface maps presented in the RAP for the Bowman site show two unusual features: (1) a recharge mound in the upper fine grained sediment, and (2) a discharge " sink" in the middle lignite zone. These features are anomalous considering the vertical hydraulic interconnection between upper and middle units of the uppermost aquifer.
The RAP does not provide any explanation of these features, nor their implicatiots in meeting the EPA ground-water protection standards. The NRC staff believes that some of the ambiguity presented in the potentiometric maps is the result of inadequate (number and distribution) spacial coverage of monitoring wells.
None of the potentiometric maps use water-level _ data within the planned disposal cell.
It should be noted that monitoring wells 501, 502, and 503 are located within the planned disposal cell, but data from these wells were not used to develop the potentiometric information. DOE should collect and provide additional water-level data to address the ambiguities and inadequacies identified in the RAP. If necessary, new monitoring wells should be installed to meet these data needs.
(Ib)
The hydraulic properties of the uppermost aquifer at the Bowman site have not been adequately characterized.
In particular, the horizontal hydraulic conductivity, that was estimated from the aquifer tests, may not be representative of the conditions in the uppermost aquifer because the aquifer analyses presented in the RAP are inappropriate for the hydraulic conditions exhibited at the site.
The ar,alyses performed do not account for leaky-aquifer conditions that are exhibited in the raw field data collected during the pumping tests.
Additionally, the dominant flow path in the uppermost aquifer appears to be the middle lignite unit.
The hydraulic characteristics of this unit were measured with only one pumping test, which leaves the majority of the performanca assessment demonstration (volumetric mixing calculation) based on a single test.
1 I
ENCLOSyRE1 D0E should reanalyze the previously performed aquifer tests, using the proper analysis
- methods, to provide a
representative characterization of the uppermost aquifer. Additional aquifer tests should be-performed if the existing data do not support a technically-defensible analysis.
DOE should also perform an additional aquifer test in a different pumping well in the middle lignite zone to adequately characterize the hydraulic properties of this primary flow path.
(Ic)
DOE should' provide a technically-defensible rationale for the selection of the effective porosity values used in the ground-water velocity-calculations.
These values should be compared to any available information on the measured void ratio or grain-size distribution within the aquifer.
The appropriateness of the estimated values should be based on this comparison.
(Id)
DOE has not a#quately delineated the lateral extent of ground-water
-contamination within the uppermost aquifer at the Bowman site. DOE addressed this dTER _ issue by adding eight monitoring wells in the upper fine-grained unit at the Bowman site, and collected additional samples for analysis.
Even with these additional wells, the contaminant plume is only reasonably delineated near the eastern end
-of the site, and only in the upper fine-grained unit.
The NRC staff cannot concur on DOE's proposed deferral of ground-water restoration to a later project phase, as long as the existing contaminant plume has not been adequately delineated.
In addition, a cell performance demonstration through ground-water monitoring may not be possible without an adequate characterization of the existing ground-water contamination.
(2)
DOE must provide a more rigorous, technically-defensible methodology to demonstrate that geochemically reducing conditions exist at the Bowman _ site.
-00E previously collected core samples from the surficial soils and the. lignite zone and performed analyses for organic content. The results of these analyses were not presented nor discussed in the RAP.
DOE concluded that reducing conditions occurred because of the presence of organic matter.
DOE has not measured nor demonstrated through analysis that reducing conditions exist within any of the monitored units.
Also, the lignite zone has been characterized as being from 4 to 7 feet in thickness. Any preferential flow through vertical fractures
- within-the lignite would result in a shortened residence time for completion of the attenuation. DOE must adequately demonstrate that reducing conditions exist along the flow paths beneath the disposal cell and that the ground-water travel time through the lignite zone provides an adequate residence time and mass to attenuate the hazardous constituents throughout the design-life of the facility.
2
ENCLOSURE 1 (3).
Recently, staff policy on the evaluation procedure for hydraulic
-conductivity of radon / infiltration barriers for Title I and Title II mill tailings sites was issued by this office.
The rationale for the_ guidance is restated herein.
~ Since it. is necessary to limit radon emissions from, and water infiltration into stabilized mill tailings, the radon / infiltration barrier component of the layered cover must have a minimal hydraulic conductivity. The hydraulic conductivity is typically minimized by compacting fine-grained soils for a suffic' ent depth above the stabilized tailings.
.%:n natural borrow soils have insufficient fines content to effectively reduce the barrier's hydraulic conductivity, bentonite amendments are often specified.
In response to the EPA groundwater standards, it is evident that designers of radon / infiltration barriers for both Title I and Title II mill tailings sites are proposing increasingly limited design hydraulic conductivity values.
It is not unusual for laboratory permeability test values to yield results of 10 to 10" cm/sec.
Such tests are performed on compacted soil samples considered by the design engineer to be representative of the soil to be used for the radon / infiltration barrier.
Considering both the experience of our technical staff and a review of current tachnical papers, we have concluded that field testing to verify the design is warranted for those cases where a radon barrier k-value less than 10" cm/sec ' is specified.-
Such testing is necessary. since construction operations and soil material variability can create preferred pathways, f sints, seams, holes, and flaws which effectively increase a barrier's hydraulic conductivity.
According to researcher David Daniel, the hydraulic conductivity may be increased by "several" orders of magnitude.
Considering the above rationale, and Belfield/ Bowman design criteria.
which state that a hydraulic conductivity value of 2x10 is to be used, dTER_ open issue No. 3 is revised to reflect the NRC staff position as follows:
(a)
The rational basis for the design hydraulic conductivity (k) value for the radon / infiltration barrier shall be provided by DOE.
In-all cases where k< 10",
a test fill shall be constructed and the hydraulic conductivity value verified by in-place testing with double-ring infiltrometers and/or other approved methods.
The test fill construction plan and verification program should be provided in the RAP.
(b)
For all cases where k<10" and the test fill program has been provided in the RAP, specifications and related documents (RAIP, etc.) shall include a strict quality control program.
The prog am shall provide a mechanism for the improvement of 3
I s 4 w =
4
ENCLOSURE 1 construction and testing techniques on the cell barrier.
Assurance that-uniform and high-quality construction of the cell barrier will be accomplished shall be the ot>jective of the program.
(4)
DOE should add nitrate, thallium, and gross alpha activity to the list of hazardous constituents and provide the appropriate concentration limits for those constituents.
These constituents have been identified in ground-water samples collected at the Bowman tite.
DOE shoulc also identify, through analysis, the organic constituents that contributed to the Total Organic Carbon (TOC) concentrations reported in Table 0.5.18 of the RAP. The appropriate hazardous organic constituents should then be added to the list of hazardous constituents.
(5)
The POC should be extended to the northeast border of the disposal cell to accommodate th? uncertainty expressed in the ground-water flow directions at the Bowman site.
If additional hydraulic chr.recterization reveals a substantially different flow direction in the aprermost aquifer, then the P0C should be adjusted to accom C te the new information.
(6/7)
DOE's performance assessment to demonstrate compliance with the EPA ground-water protection standards is based on a calculation to show that tailings seepage from the disposal cell will mix with the ground-water in the uppermost aquifer and result in hazardous constituent concentrations that comply with the EPA standards. DOE has made the following assumptions in the volumetric mixing calculation:
The steady-state infiltration flux through the radon barrier is equal to the design permeability of I x 10" cm/s; The hazardous constituent concentrations in the seepage is equal to the concentrations measured in lysimeters beneath the contaminated materials; and Contaminated scepage from the disoosal cell will mix evenly with ground-water in the uppermost aquifer.
DOE must perform a technically-defensible demonstration that the infiltration flux through all components of the cover will meet the rate required for the mixing calculation to demonstrate compliance with the ground-water standard. If the stringently low permeability n
of the radon barrier is the limiting factor in demonstrating compliance, then DOE must perform a
technically-defensible demonstration that the stringently low permeability design specifications can be constructed -(see comments in Geotechnical Section).
4
I
_ENCLOSpRE1 00E. must show that compliance with the - ground-water protection standards will be achieved in all dominant flow paths within the uppermost aquifer. The calculation presented in the RAP relies on averaging the various hydraulic gradients measured in the uppermost aquifer. NRC staff considers the averaging of the flow gradielts as being'non-conservative, because approximately 80 percent of me flow is through the -lignite zone.
DOE must perform the mixing calculation for each designated, major flow-path in the uppermost aquifer and demonstrate compliance for each zone.
(8)
CLOSED (9)
Performance monitoring of the disposal cell at the Bowman site will be severely hampered without an ' adequate characterization of the existing ground-water contamination. The NRC staff cannot evaluated the adequacy of the proposed performance monitoring system without an adequate characterization of the axisting ground-water contamination. -DOE must adequately characterize the extent of the existing contamination at the Bowman site.
This Open Issue is related to the iscue identified in item ld above.
(10)
DOE has provided a list of credible failure scenarios and conceptual corrective actions to address the potential failures. The NRC staff notes that two critical scenarios were not included: (1) side slope failure, and (2) failure due to differential settlement of the radon barrier. : 00E should-revise the list of credible failure scenarios to include these-two scenarios and provide conceptual corrective actions that will address these potential failures.
-(11)
CLOSED (12)
Open, pending resolution of updated dTER open issue 1(d).
5
i.
h ENCLOSURE 2
GENERAL COMMENT
S GROUNDWATER (GW) 32.
The DOE considers the upper fine-grained and middle lignite zones to be separate hydrogeologic units, but also as the uppermost aquifer system to be monitored at the Point of Compliance.
The NRC agrees that both hydrogeologic units should be monitored, but for future wells to be constructed for the purpose of delineating more completely the extent of ground-water contamination and/or for monitoring purposes, it is not clear whether DOE plans to complete wells separately or have a combined well completion for these two separate hydrogeologic units, it is NRC's opinion that future characterization and monitoring wells be completed separately in these individual hydrogeologic units - as was done by DOE in the initial characterization investigations.
33.
Page 45, Vol. I and Page D-155, Vol. 2:
The average transmissivity and hydraulic conductivity for the upper fine-grained unit is reported to be 6.5 f t2/d and 0.23 ft/d, respectively.
Pumping test results reported in Table D.5.11 indicate an average transmissivity of 10.2 ft 2/d and an average hydraulic conductivity of 0.39 ft/d.
34.
Page D-9, Vol.2:. The calculation document for " Contaminated Materials Excavation - Bowman, Excavation Plan and Quantities" should be listed as B/B-915-01-01.
35.
The reference of Fisher et al.,1985 (Ref. 22) should be removed from the text (e.g., one occurrence is on page D-152, Vol.2). The U.S. Geological Survey has issued a general disclaimer on all works by their former employee Robert L.
Houghton (an apparent co-author in the notcJ t
reference), due to evidence of scientific fraud (see Graund Water 1990; Ref. 25).
36.
The potentiometric maps p,esented for several of the hydrogeologic units characterized indicate very low hydraulic gradient conditions. Since the determination of ground-water flow velocity and direction is required for the overall performance assessment cf the disposal cell, it is important that the uncertainty of factors that can affect the determination of hydraulic gradients and flow directions be also stated.
One obvious controlling factor would be the level of accuraty ascribed to the well elevation surveys. An uncertainty of 0.1 ft in the well casing elevation, for instance, would have a pronounced influence on calculating gradients and flow directions for the middle lignite zone at Bowman.
37.
The reference of Table 0.5.12 on page D-160 is incorrect.
38.
Table 0.5.10 on page D-284 has a
column title that states "Hydrostratigraphic Unit Screened".
Change to "Hydrogeologic Unit Screened" to be consistent with table designation on page D-285.
1 i
l
ENCLOSURE 2 39.
Page 32 (Ground Water Monitoring Program, Sec. 5.3):
The more the hydrogeochemical system approximates steady-state leaching and transport, i
the more difficult it becomes to deduce the rate of transport from chemical concentration data alone.
If construction of the disposal cell is sufficient over time to result in meeting performance goals at the P00, the actual rate of transport is immaterial.
If, on the other hand, the cell alone does not accomplish the desired goals, the rates of contaminant and ground water movement become important from the standpoints of not plume movement and of the ultimate volume of ground water that may require remediation.
NRC recommends that best estimates of seepage velocity be used to predict the amount of time following waste emplacement that is reasonably required to determine if the cell can reduce leaching of the wastes sufficiently to meet contaminant concentration goals at the POC.
That is, an evaluation of the extent of ground water remediation that may he required, if any, should be deferred until performance of this waste disposal emplacement can be assessed by direct measurement.
(Similar argum"*
apply to the Belfield site; that is, removal of contaminated surfece material will obviously remove the source of leachable wastes, but the effect on contaminant concentrations i ground-water will not be immediate.)
40.
Page 0-384 Typographical error:
One sample temperature listed as 1010 degrees.
41.
Page 82 (Ground Water Protection St'ndards, Sec 5.2):
In the RAP, DOE is proposing concentration limits for the POC that are equal to either EPA MCLs or statistical maximum background concentrations, the latter to be applied where background concentrations are greater than the EPA ground-water protection standards.
00E needs to define what a " statistical maximum" is, what sampling points will be used to determine such concentrations, and what will be done if upgradient concentiations change over time.
The RAP should be mere carefully worded so that the performance goals at the POC reflect the actual changes in ground water wality between upgradient and downgradient monitoring wells.
Assuming that DOE adequately defines background concentrations, it must also be recognized that choosing these background
.oncentrations as P0C performance goals requires either that enginet. ring control of wastes (i.e., the disposal cell) is virtually.100% effectly_q, or that ground-water iemediation will ultimately be required.
42.
Page 82:
In the ground-weter monitoring 9rogram discussion, the RAP stater that monitoring will occur during %d after the remediti action period.
However, monitoring should also bc performed just prior to remedial action.
NRC recommends four quarterly sampling events in the year preceeding construction / excavation at the sites, using a monitoring netu rk that represents both the PCC and upgradient ground water.
This vili be necessary to provide baseline data.
43.
Page 0-159:
The RAP argues that nitrate concentrations in the ground-water at the Bowman site are not due to processing activities at that 2
4 EllCLOSURE 2
~ ite, and are instead caused by nearby agricultural activity.
This is s
possible, and perhaps even likely, but NRC considers that DOE has not sufficiently established its case (especially in light of the fact that
- the potentiometric relationship presented in Figure D 5.19 does not supportsuchahypothesis). Much of the difficulty in assigning a source to the apparent nitrate contamination lies in the high spatial variability of n?trate concentrations.
If DOE vishes to pursue this matter, NRC suggests the _use of Piper diagrams, ternary plots, etc., which are condructed using concentration
- ratios, rather than absolute concentrations. That is, it is nften possible to separate the effects of comingled waste plumet by considering the characteristic constituent ratios of the centaminant sources, independent of the dilution effects of seasonal or local variations in precipitation or recharge, or of dispersion.
44.
Page D-144:
One approach to estimating effective porosity and average rour.d water velocity that may be useful is described by Hall et al.,
- Ref. 26).
45.
Page D 160 ar.d 161:
The RAP suggests that reducing conditions in the lignite zone at the_ Bowman will aid in attenuating uranium. The NRC does not accept that the mere presence of lignite shows that conditions are actively reducing. -Further, the presence of oxidized anionic chemical species 'such as nitrate or soluble uranium (presumably hexavalent and complexed with carbonate), ugests that reduction substantially useful in attenuating _ contaminants is iwt occurring.
If the DOE wishes to pursue this matter, NRC suggests a two-stage approach.
First, ground water samples should be analyzed for evidence of active reduction: absence of dissolved oxygen, presence of sulfide (sulfate reduction), presence of ammonium (nitrate reduction).
Then, if these kinds of indicators show that the ground water is reducing, injection-withdrawl tests using elevated concentrations of nitrate, hexavalent uranium, and perhaps hexavalent chromium, could be used to broadly and semi-quantitatively assess the casacity and rate of reduction. This level of investigation is not warrantec at this time, and should only be considered in the event that engineering control proves to be inadequate. Nevertheless, without these kinds of evidence, DOE should not continue to cite reduction as a probable attenuation mechanism.
Similarly. 00E has stated that laboratory examination of soil samples indicates favorable capacity for the attenuation of cationic trace metals, but field data to support this-conclusion are lacking.
A useful approach to search for evidence of contaminant attenuation,
- regardless of mechanism, and which circumvents the nead for injection-withdrawal tests, etc., is to compare on site and downgradient constituent ratios using-existing chemical monitoring data.
This is especially helpful-if contaminants can be compared to a conservative chemetal species that can be related to the source of contaminants.
3 s +9n,gt---y
- Pew +Nv-W'*1PT+
ENCLOSURE 2 In any case, NRC considers that the only real significance of contaminant attenuation mechanisms, as applied to either the Belfield or the Bowman
-site, lies in evaluating the ultimate mobility of contaminants that have passed or will pass beyond the site bounda-v (i.e., P0C) before remedial action is taken.
Nevertheless, attenuation mechanisms, if established, may be useful in determining the practical extent of ground-water remediation that 15. necessary to clean up relict contamination or to correct any future excursions.
GE0 TECHNICAL 46.
Pape E-23 Table E.3.4.:
Table E.3.4 summarizes the f ailure :.c enarios considered.
Two critical scenarios were not included,1) a ;.de slope failure and 2 a failure uue to differential settlement of the barrier.
Ali. hough conse)rvative designs with safety factors were used, these two scenarios should not be omitted. A side slepe failure could expose much, if not all, the barrier cross-section to creaion and provide easy access for infiltration. Differential settlement tuy not be as catastrophic as a side slope failure but could greatly enhance infiltration through surface cracking and surface depression ponding.
47.
RAP vol. 3 (Erosion Protection); The specifications discussed could not be found in the text for the Specification 02278 and page number listed.
48.
RAP vol.1 (Compaction requirements - Section 4.3.9 pg. 69): The percent compaction has n,gi been corrected to 95Y. of Standard Proctor Density.
49.
Section D.3.4:
The discussion of the Bowman site geology needs clarification.
This discussion could be followed only when read -in
-conjunction with the stratigraphy section (Figure D.3.2) provided.
The NRC recommends that the geologic section be discussed in order from the lowest identifiable strata to the surface.
The discussion in section D.3.4_does not approach the stratigraphy in sequence. - This section should also site supporting data, preferably from the well borings performed during the study.
It is noted, however, that this section meets the guidelines specified by the NRC and EPA.
-5C.
Ceneral Comment:
It is recommended that the field notes from site i
inspections and from ground truth surveys of the areas identified by aerial photography be included in the appendices, it would be also helpful to include some'of the marked up airphetos identifying the linear features and geomorphic features that are discuned in the text.
It is also recommended that the lithologic descriptions in the well logs be more detailed-to include physical characteristics (eg. color, sorting and percentages, texture, and mineralogy).
liany of the field logs did not contain these-items identified in the Standard-Review Plan for VMTRCA Title I Mill Tailinas Remedial-Action Plans section 2.2.2.1.1.
4
= _ _ - _ - ___-__--__ _- _ _-
ENCLOSURE 2 51.
General Question: Are the monitoring wells currently locked or otherwise secured to ensure representative water samples from the monitoring network?
The As luilt diagrams included in Appendix D do not indicate these wells are capped. if they are capped, the As-Built diagrams need to be more detaild 'bowir.g %s surf ace protection.
52.
General Comment: Since the final calculations for the required thickness of the radon barrier assumed 0.0 pCi/l ambient radon concentration, the actual measurements of these concentrations in the field (which DOE hs committed to) are probably unnecessary, at least for radon barrier des...
53.
Page 15 Vol. 1:
There are other ihan lived radionuclides in the uranium decay series, but all of the long lived ones have been named here.
54.
Page 15, Vol.1:
- 1) Alpha radiation is also emitted by V 234, Pb-210, Bi-210, and Po 210, all members of the uranium decay series, for consistency with paragraphs 1 and 5 of this secticn they should be listed also. There are actually 10 alpha emitting daughters, but At-218 can be omitted because of it's very low branching ratio.
- 2) Alpha emitters are also hazardous if they enter the body through open wounds.
55.
Volume 1 Table 3.1 (page 1) and Table 3.2 (page 22):
The soil radioactivity level of Ra 226 reported in these tables (1986 reference) is 4
significantly higher than that determined in calculation B/8 950 0100 (reference calculation completed 1988-89; in Final DesiSD for Construction Volume 11:
Embankment Desian).
This discrepancy should be explained.
56.
Volume 1, page 18. Table 3.1:
"Ra-266" should be Ra-226; 2 separate occurrences.
5 1
_ _ _ _ _ - - - _ - _ - - - _ - _ _ - - - - - - - - - - - - - - - - - - - - - - - - = - - - - - - - - - - - - - - - - - - - - - - - - -