ML20196B427
| ML20196B427 | |
| Person / Time | |
|---|---|
| Issue date: | 01/26/1988 |
| From: | Starmer R, Tokar M NRC OFFICE OF NUCLEAR MATERIAL SAFETY & SAFEGUARDS (NMSS) |
| To: | Fliegel M NRC OFFICE OF NUCLEAR MATERIAL SAFETY & SAFEGUARDS (NMSS) |
| References | |
| REF-WM-68 NUDOCS 8802110253 | |
| Download: ML20196B427 (12) | |
Text
[-
lb7n48 MEMORANDUM FOR: Myron Fliegel, Section Leader Operations Branch Division of Low-level Waste Management 21 Rxcrf Fge and Decomissioning, NMSS
{7o FROM:
R. John Starmer, Section Leader
- - @~'~
Technical Branch i.;
p, g, s.,i
~
Division of Low-Level Waste Management'
~
and Decomissioning, HMSS Michael Tokar, Section Leader Ed
~
Technical Branch
~ ~~~~ -
Division of Low. Level Waste Management and Decomissioning, HMSS
SUBJECT:
NRC TECHNICAL STAFF COMMENTS ON THE GREEN RIVER, UTAH UMTRA PROJECT SITE, FINAL DESIGN This memorandum transmt ts the Technical Branch's (LLTB) conclusions regarding the Green River Final Design. Enclosed are a st.mmary of information needs to support staff review and the overall and specific coments detailing our Concerns.
This action has been coordinated with Susan Bilhorn of your staff.
M ininal Sipvd b R. John Starmer, Section Leader Technical Branch Division of Low-Level Waste Management and Decomissioning, NMSS Chriginal Sipua %
8802110253 900126 Michael Tokar, Section Leader PDR WASTE PDR Technical Branch WM-68 Division of Low-Level Waste Management and Decomissioning, NMSS
Enclosures:
As stated cc:
P. Lchaus, LLOB DISTRIBUTION:
'LLWM/SF.
MTokar MRKnapp NMSS'r/f RJStarmer JGreeves LLTB r/f BJagannath KWestbrook MHYoung TLJohnson JGrim DFC :LLTB
- LL
- LLT p
....t..:............:___________:.......__...:..._____..
NAME :MHYoung Jr.H[:RJStartner
- MTokar DATE:01/p/88
- 01/Zf/88
- 01/g/88 OFFICIAL RECORD COPY
ENCLOSURE 1
, 6 BACKGROUND INFORMATION AND INFORMATION NEEDS TO SUPPORT NRC STAFF REVIEW GREEN RIVER, UTAH UMTRA PROJECT SITE Division of Low-Level Waste Management and Decomissioning, Technical Branch (LLTB) staff reviewed the DOE contractor-submitted final design for the Green l
River, Utah remedial action disposal unit and found that the submitted material l
does not contain sufficient information for the staff to make substantive findings regarding the acceptability of the design. Without adequate information and analyses, the staff cannot determine if the proposed design makes a reasonable contribution to meeting the EPA standards for stability and ground-water protection, l
i Critical information and data on cover design, ground-water hydrology, ground-water quality and design performance were absent. The design documents did not provide an acceptable specification of rock durability, or incorporate the proposed thickened bedding layer of lower quality material to protect the radon attenuation layer from frost. Assertions made about expected ground-water quality have inadequate data support. No analysis of the performance of the buried riprap is presented in the document, particularly regarding potential migration of the overlying earthen cover in to the voids of the riprap and the potential for increased infiltration into the tailings and leachate production. No analysis of the performance of the proposed thickened bedding was provided either from the standpoint of stability or the potential for increased infiltration.
NRC technical staff cannot effectively continue their review of the proposed DOE action until the following information and analyses are provided in writing:
1.
Hydraulic conductivity of the Upper Cedar Mountain Formation measured by a comonly acceptable technique.
2.
Hydraulic conductivity of the alluvium measured by a comonly acceptable technique. This data is required only if the planned action requires drainage through the alluvium to convey surface runoff from the buried riprap toe illustrated by the design orawings.
3.
Degree of hydraulic connection between the Upper and Lower Cedar Mountain Fonnations using observed site environmental conditions.
4.
Specification for riprap based on quality and reasonable availability.
5.
If a thick bedding layer is ultimately proposed for th9 design, an analysis of the affect of any increase of the currently proposed thickness on infiltration and subsequent leachate production.
6.
Data on the redox conditions of the aquifer.
Staff would consider it wise for DOE to demonstrate that some other attenuation factors could be relied on to improve estimated groundwater quality based on predicted geochemical behavior of the contaminants.
7.
Assessments of the potential for migration of sediment from the overlying backfill into the riprap toe, and potential for increased erosion initiated by the resulting subsidence of the backfill in the toe area.
8.
A performance assessment of the proposed design inter-relating the factors above, with particular emphasis on infiltration and potential for leachate production.
Staff does not encourage DOE to proceed with plens to implement the proposed design until they have upgraded the design criteria for the riprap rock to a level acceptable, and analyzed the proposed buried riprap and thickened bedding. DOE should also assess hydrologic properties of the site and design for expected performance with respect to probable rainfalls, surface runoff and infiltration of groundwater into tailings and leachate production.
Finally, an analysis should be presented of the expected interaction of leachate with the site groundwater based on field data. These analyses should be related to the applicable EPA standards.
NRC staff emphasizes the need for analysis by DOE, and the staff position that analysis without data seldom provide adequate support for concurrence or licensing decisions. LLTB staff is willing to meet for discussion of the enclosed comments at the request of 00E.
Such a meeting could be best accomodated at our offices or via a telephone conference.
NRC/LLTB STAFF COMMENTS ON THE FINAL DESIGN GREEN RIVER, UTAH UMTRA PROJECT SITE OVERALL COMMENTS GROUND-WATER RESOURCES 1.
Characterization and Performance of the Upper Cedar Mountain Formation In our review of the draft design (DOE,1987), NRC staff comented on the use of the upper Cedar Mountain Fomation (Kemu) for conveying surface runoff away from the tailings embankment, and stated that water could pond up in the buried riprap channel and increase flux into the embankment.
The staff concludes that the success of the Green River design is heavily dependent upon the ability of the Kcmu to convey surface runoff away from the tailings. DOE responded that the hydraulic conductivities of the Kcmu and lower Cedar Mountain Fonnation (Kcml), as determined by pumping and slug tests, are the same, and almost four-orders of magnitude greater than that of the radon barrier.
Staff has reviewed DOE's response and the DOE summary report, "Geochemical Modeling and Dilution Estimates for the Proposed Disposal Area Green River, Utah, Tailings Site," and concludes that DOE has not demonstrated, with reasonable assurance, that the Kcmu is capable of conveying surface runoff away from the embankme9t.
The staff concludes that the field tests used to characterize the hydraulic properties of the bedrock units did not provide sufficient information on the Kcmu to make conclusions as to its hydraulic conductivity. The staff concludes (a-d below) that the hydraulic conductivity value of 1.6E-03 cm/s for the Kcml cannot be applied to the Kemu, as was suggested in the DOE's response to NRC coments. Thus, the ability of the Kcmu to adequately convey surface runoff away from the embankment is in question.
This conclusion is based on the following points:
DOE performed pump and slug tests on well GRN01-813, which is completed in a.
the Xcml, and concluded that the hydraulic conductivity of both the Kcmu and Kcml is 1.6E-03 cm/s. However, DOE did not indicate whether any observation wells completed in the Kcmu were used during the test.
Without an observation point in the upper unit, no conclusion can be i
reached on the hydraulic conductivity of the upper unit. 00E should either perform in-situ hydraulic conductivity tests on the Kcmu or utilize an observation well in the Kcmu when testing the Kcml.
b.
DOE should provide separate hydraulic conductivity results from the pump and slug aquifer tests, rather than an average value.
This will reduce error incurred when combining results from different test methods.
l J
c.
DOE assumed that the fracture system of the Kcmu and Kcml is uniform.
Other than pictures of cores retrieved from recent well drilling, DOE has not provided information or data that substantiates this claim. Without knowledge of the fracture geometry, unanticipated migration of leachate could occur, which will make future monitoring of water quality difficult.
DOE should characterize the aperture, direction and extent of the fractured system to determine its uniformity and potential hydrogeologic behavior.
i d.
DOE assumed that the hydraulic conductivity value of 1.6E-03 cm/s for the Kcmu and Kcml, "is representative of the horizontal and vertical hydraulic conductivity of the foundation bedrock." However, DOE did not differentiate between vertical and horizontal conductivity of the aquifers in their estimate of the overall rock conductivity. The value obtained from the aquifer tests could represent primarily the horizontal conductivity. Vertical conductivity may be very low relative to horizontal conductivity, since the vertical fractures are relatively few.
Ground water percolating into the Kcmu could flow laterally into the tailings embankment and not downward as DOE asserts. DOE should differentiate between vertical and horizontal hydraulic conductivity, or analyze the performance of the design under a conservative scenario assuming predominantly horizontal flow in the Kcmu.
Because of the importance of the Kcmu to convey surface runoff from the embankment, additional information on this unit is critical when assessing performance of this disposal unit.
U.S. 00E, 1987, Preliminary Design for Review, Vranium Mill Tailings Remedial Action Project, Green River, Utah, September.
2.
Possible Decrease of Hydraulic Conductivity of Kcmu Caused by Clogging of Fractures The success of the Green River design, from a ground water standpoint, relies upon the ability of the Kcmu to convey water from the stabilized tailings.
However, DOE has not considered possible failure scenarios of the design caused by clogging of the fractures in the Kcmu by migration of the sediment from the overlying fill material, or sediment transported by surface runoff through the buried riprap channel.
DOE did not assess the possibility that the Kcmu' will become clogged with i
transported sediment in the immediate vicinity of the contact with the riprap lens. DOE should expect sediment to be transported to the bedrock during larger intensity storms and to eventually clog the fractures that are asserted to be present during the 200 - 1000 year design life.
Fractures clogged with sediment will dramatically reduce the hydraulic conductivity of this unit and increase the likelihood that water will pond up in the buried riprap lens and increase infiltration in to the disposal cell. This possibility needs to be i
assessed before NRC staff concurs that the design will meet the EPA standards for 200 - 1000 years.
GE0 CHEMISTRY 1.
"Geochemical Modeling And Dilution Estimates For The Proposed Disposal Area, Green River, Utah, Tailings Site" The geochemical reaction path code PHREEQE was used by DOE to calculate the equilibration of leachate-groundwater mixtures with secondary minerals present in the Cedar Mountain Formation.
In the first case presented, the leachate groundwater mixture is equilibrated with calcite, gypsum, and Fe(OH)3.
The speciation of this water resulted in calculated uranium calculations of 22.62 mg/1. DOE stated that uranium concentrations beneath the present tailings pile do not exceed 1.5 mg/l "indicating there are probably attenuating mechanisms (such as adsorption) not accounted for in these calculations" (page 20, last sentence). A second computation was performed by equilibration of the leachate-groundwater mixture with calcite, gypsum and pyrite to calculate predicted uranium concentrations on the order of what is found beneath the current tailings pile. DOE stated that pyrite occurs in the shallow saturated zone. Uranium concentrations were calculated to be 7.5x10-6 mg/l and oversaturated with respect to uraninite when in the reduced environment determined by pyrite equilibrium.
From the results of this report it is apparent that predicted uranium concentrations are strongly dependent on the redox conditions of the aquifer.
The NRC staff concludes Gat there is insufficient evidence that uranium species will be dominant.'y in the reduced state thus invalidating DOE conclusions on expected uranium concentrations. The existence of pyrite in the Cedar Mountain Formation and the low concentration of uranium beneath the existing tailings pile are the only evidence provided by DOE which suggests reducing conditions are present in the aquifer. The existence of pyrite is not sufficient evidence to determine the rednx conditions of the groundwater because pyrite may not be in equilibrium with the Cedar Mountain groundwater.
No field data c1 the redox conditions of the aquifer are presented.
Measurement of the dissolved oxygen content, redox couples or a field Eh would
)
support conclusions on the redox conditions of the aquifer.
The staff notes, however, even if these measurements did indicate reducing conditions, it is not i
certain that uranium species would be reduced. Lindberg and Runnells (1984) i have computed Nernstian Eh values from multiple redox couples. The computed Eh values do not agree with each other nor do they agree with the field Eh measurement.
Due to this disequilibrium, the use of any measured Eh may result in misleading conclusions.
There are also concerns with regard to using measured uranium concentrations beneath the current tailings pile as an indicator of expected geochemical conditions beneath the proposeo site.
The relatively low concentrations of uranium beneath the current tailings pile as compared to the first leachate-groundwater mixture (in equilibrium with calcite, gypsum, and Fe(OH)3) was used by TCE to support the conclusion that the groundwater is in equilibrium with pyrite.
If the groundwater beneath the current tailings pile
i is in equilibrium with pyrite (and the subsequently calculated Eh is in agreement with the uranium redox couple) then uraninite should be precipitating out.
Field evidence of uraninite precipitation beneath the current tailings pile would support the existence of reducing conditions in the aquifer.
As stated earlier, the staff conclude that there is insufficient evidence that uranium species will be dominantiy in reduced state and therefore does not accept DOE conclusions on expected uranium concentrations.
LLTB staff suggests that DOE provide field evidence of the redox conditions of groundwater beneath the proposed disposal site if reducing conditions are to be assumed when predicting uranium concentrations. Due to the uncertainty with regard to redox conditions in groundwater which were discussed earlier, the staff would not accept predicted uranium concentrations based solely on a reduced environment.
staff suggests that reoucing conditions, if present, be part of an evaluation of all dominant hydrological and geochemical retardation mechanisms.
REFERENCE:
Lindberg, R. D., and Runnells, D. D. (1984)
Ground water redox reactions: an analysis of equilibrium state applied to Eh rneasurements and geochemical modeling. Science, 225, 925-927.
GE0 TECHNICAL ENGINEERING 1.
Typical Riprap Toe Protection Detail in Drawing No. GRN-PS-10-0517, Rev. B.
There is no filter layer between the Type B Riprap material and the overlying backfill material. There is a potential for the backfill material to migrate in to the voids of the riprap and/or fracture openings in the bedrock with time resulting in clogging their voids. This particle migration will also result in surficial subsidence of the backfill which will either require maintenance with time or initiate surficial erosion as result of uneven surface at the toe of the tailings pile slope.
The effect of clogging on the drainage and other functional aspects of the riprap should be examined in detail. The design of the riprap toe protection should be examined in the light of the above comfrents.
EROSION PROTECTION 1.
Specifications for Riprap Material We note that the specifications for the riprap to be provided have not been modified to reflect placement of good-quality rock. Specifications for the riprap material indicate that the rock will have to meet only the following durability requirements:
Specific Gravity - Not less than 2.45 Absorption
- Not greater than 4.00 NASO 4
- Not greater than 30% loss (5 cycles) 1
l l
In general, staff considers that these requirements will not be acceptable to assure that the EPA long-term stability requirements will be met. Based on research performed for the staff (NUREG/CR-2642, Table 6.7), rock meeting only these minimum specifications is likely to weather severely and thus may not meet EPA longevity standards.
Based on meetings held on the subject, we recognize that a search for good-quality rock is currently underway and that the selected rock will likely exceed these requirements; however, it is nevertheless possible that rock approaching the minimum limits could be used if these specifications are followed. The minimum limits should be raised and/or an acceptable quarry should be specified.
We would like to emphasize that the riprap described in the design is not considered acceptable, even if oversized.
DOE should provide additional information regarding durability of the selected rock source, including any information and evaluations related to oversizing.
In addition, DOE should document the efforts that have been made to 'ocate sources of good-quality rock.
SPECIFIC COMMENTS GREEN RIVER FINAL DESIGN GROUND-WATER RESOURCES 1.
Hydraulic Connection Between the Kcmu and the Kcml, DOE Response to NRC Coments on Preliminary Design, Coment GW/6 DOE stated, in their response to NRC coment GW/6, that high hydraulic conductivity results derived from pump and slug tests indicate that the Kcmu and Kcml are interconnected. Thus, hydraulic conductivity results from pump tests in the Kcm1 are representative of hydraulic conductivity values for the Kcmu. LLTB staff concludes, however, that DOE has not demonstrated that the units are interconnected. When considering other site conditions, the two units may not be fully connected.
Thus, the assumption that surface runoff will percolate down to the water table may not be valid.
The following points support the NRC staff position:
a.
The hydraulic gradients and flow directions differ between the two units.
The gradient and flow direction for the Kcmu are.023 and N-NW, whereas the gradient and flow direction for the Kcml are.012 and NW. Thus, the different gradients indicate that the units are under different pressures and are not fully connected.
j b.
A significant head difference exists between the Kcmu and Kcml, with the hydraulic gradient upwards. This gradient increases in a northward direction from approximately six feet belcw the disposal site to up to 20
l
[
1 feet near the current location of the tailings. Artesian pressure in the lower hydrostratigraphic unit indicates that the Kcmu is acting as a confining unit.
If the Kcmu and Kcml are interconnected, as DOE asserts, then the degree of artesian pressure would be very little.
DOE needs to consider these hydrologic conditions when assessing the connnectivity of these two units. This will improve predictions for leachate migration direction, as required by the EPA ground-water protection standards.
2.
Characterization of the Alluvium, Drawing GRN-PS-10-0517 According to Drawing GRN-PS-10-0517, the buried riprap channel terminates into the alluvial material overlying the Cedar Mountain Fonnation. All routed surface runoff must percolate through the alluvium before reaching the fractured Kcmu.
Information supplied by DOE in the Final Design document indicates that the alluvial material has a much lower capacity for transmitting water than does the Kcmu.
For example, data on the site foundation soil (i.e.,
the alluvium), found on sheet 44/63 of calculation 10-536-01-00. indicates that the hydraulic conductivity may be as low as 5.2E-06 cm/s. This value is within one order of magnitude of the cover hydraulic conductivity.
Because the hydrogeologic system will be governed by material with the lowest hydraulic conductivity, it does not appear that the system is capable of conveying surface runoff away from the embankment at a rate high enough to prevent ponding in the buried riprap channel.
NRC staff recognizes that information on this remolded sample may not be representative to undisturbed foundation soil, but no other information is available on the alluvium.
DOE needs to provide data and information on the alluvial material that demonstrates its ability to remove the volumes of surface runoff expected to be routed through the buried channel during the design life of the facility. This information will improve DOE's assessment of the drainage characteristics of the alluvium.
GE0 CHEMISTRY 1.
Geochemical Modeling and Dilution Estimates for the Proposed Disposal Area, Green River, Utah, Tailings Site, Page 20, Section 4.2 00E used a uranium concentration value of 1.5 mg/l for their geochemical modeling analyses and stated that this was the maximum uranium value found beneath the present tailings pile.
Table B.2.21 (DOE,1987) lists uranium concentration values for well 701 as high as 3.11 mg/l for a sample collected on 6/6/86. Staff is unsure whether this discrepancy effects the geochemical results listed in the report.
DOE should clarify why this higher i
value was not used for the geochemical modeling analyses.
j
REFERENCE:
U.S. DOE,1987, Environmental Assessrrent of Remedial Action at the Green River Uranium Mill Tailings Site, Green River, Utah (Final), August.
GE0 TECHNICAL ENGINEERING 1.
Specifications, Page 02200-4, Item 1.5.A.1 ASTM 0 422, Sieve analysis, should be added to the list of ASTM tests.
2.
Specifications, Page 02200-7, Item 6.b.
The maximum particle size permitted in a fill is generally 75 percent of the compacted thickness of each layer. The specifications provide for a maximum particle size not greater than the compacted thickness of the layer. Particle size equal to the ccmpacted thickness of the lift would pose problems in obtaining uniform compaction, particitlarly in its imediate vicinity.
Justification that unifonn compaction will be obtained inspite of the maximum particle size being equal to the compacted lift thickness should be provided.
3.
Radon Barrier Material, Specification 02200, Section 2.1.8 Pages 02200-7 and 8.
The specifications for the radon barrier material should be revised to limit the maximum size of the particle permitted. The current specifications require only that a minimum of 50 percent of the material be finer than a No. 200 sieve size. The subcontractor could bring a material with 50 percent of maximum 9 in. size and 50 percent of passing No. 200 sieve size, and would be in compliance with the specification. This may result in a poorly graded material which will exhibit a higher coefficiqnt of permeability than a well graded material with a minimum of 50 perce. passing No. 200 sieve. The specifications should be revised to reflect the gradation of the material that was tested and recomended for use as a radon barrier material.
4.
Specifications, Page 02200-17, Item B. 9 Specifications on page 02051-4 permits contaminated wood, concrete or masonry to be cut up to a maximum size of 3 feet in any dimension and this is expected to be disposed in the tailings embankment. Since the maximum loose thickness of a lift is 12 inches it will be difficult to compact each layer to the specifications if there are 3 feet size concrete blocks in the compaction area.
The specifications should provide specific guidance for compacting around such large size particles to ensure that there will be no under-compacted zones with in the tailings embankment.
5.
Specifications, Page 02200-17, Item C.2 The moisture content control for the placement of the contaminated material has no upper limit and very high moisture content would result in lower density and strength, and the possibility of contaminated water oozing out of the tailings i
---g n---
i as compaction and construction progresses. The specification should be revised to make the moisture content upper limit a function of the optimum moisture content for the desired level of compaction.
6.
Specifications, Page 02200-18, Item 3.6 B.
a.
In order to ensure that the test locations are selected without any bias and are not lumped with in a limited area, the specifications should state that the test locations should be distributed over the compacted area of the particular layer so that the test results are a reasonable representation of the compacted mass, b.
The frequency of gradation testing for the radon barrier material should be increased to one test per 1,000 cyd of material placed. Since the radon barrier is only 2 feet thick (12,800 cyd), increased testing frequency is desirable to ensure compliance with the specifications for this critical item of the remedial action.
c.
The gradation testing for gravel fill should be a minimum of one test per day of operation or for a full shift of placing material in excess of 150 cyd per day.
7.
Specifications, Page 02200-19. Item 3.7 Specifications on preparation of subgrades for areas other than bedrock excavation should be provided.
8.
Specifications, Page 02200-19, Item 3.8.
Specifications on compaction should include guidance on moisture control during compaction for all items mentioned in Item 3.8 A.
9.
Specifications, Page 02278-7, Item 3.3 A.5 Specifications should include guidance on the frequency of testing for gradation of the bedding and filter material.
10.
Radon Barrier Design; Calculation No. 10-536-03-00 DOE should provide justification and backup data for the long-term moisture content used in the design, and field and laboratory test data that will be performed to validate the perneability parameter value proposed in the design.
s 9
- 11. Geotechnical Design Parameters-Radon Barrier Constructibility and Its Pent.eability Value; Calculation No. 10-536-01-00 The two-foot thick radon barrier layer will probably be compacted in three lifts. The first lift will be placed above sandy tailings and the the third lift will be beneath the filter / gravel-bedding layer. The top and bottom zones of the radon barrier, i.e. transition zones between the tailings and radon barrier and the filter / bedding layer and radon barrier, will not have the same permeability characteristics as in the center zone of the compacted radon barrier. This is because it is difficult to construct a two-foot thick low permeability layer between two high permeability layers and expect the low permeability layer to have the laboratory tested permeability properties for the entire thickness of the layer, particularly when it is only two feet thick and placed in three lifts. Justification to show that the radon barrier as constructed will have uniform permeability characteristics, assumed in the design, for its entire thickness should be provided.
GEOLOGY Design Calculation 10-536-08-00 DOE's response to staf f's coiments regarding rock units forming the disposal foundation is unclear and inconsistent.
A statement in the subject calculation that disposal would not take place on the Cedar Mountain Fomation has remained unchanged contrary to data presented to the staff at the on-board review 1
meeting held in Silver Spring, MD (September 15,1987).
In addition, R. Rager, TAC, acknowledged by telephone (October 23,1987) that the specification is I
unrealistic and would be revised.
DOE's response (November 12,1987) states "excavation into the Cedar Mountain Formation for the tailings embankment would be pemissible" and concludes the calculation sheet would be revised to that effect. No revision has been made in the final design document.
In addition, design and remedial action documents have not provided adequate evidence that disposal on the Cedar Mountain is indeed permissible with respect to EPA standards.
Please refer to staff coments related to characteristics of the site's bedrock and protection of ground-water resources.
--