ML20206H542
| ML20206H542 | |
| Person / Time | |
|---|---|
| Issue date: | 02/05/1986 |
| From: | Weber M NRC OFFICE OF NUCLEAR MATERIAL SAFETY & SAFEGUARDS (NMSS) |
| To: | |
| Shared Package | |
| ML20206H509 | List: |
| References | |
| NUDOCS 8606260179 | |
| Download: ML20206H542 (20) | |
Text
.
-o Estimating contaminant discharge rates from stabilized uranium tailings embankments Michael F. Weber, Division of Waste Management, U. S. Nuclear Regulatory Connission 8th Annual Symposium on Geotechnical and Geohydrological Aspects of Waste, 2/5-7/86, Fort Collins, Colorado 1 ASSTRACT Estimates of contaminant discharge rates from stabilized uranium tailings embankments are essential in evaluating long-term impacts of tailings disposal on groundwater resources. Contaminant discharge rates are a function of water flux through tailings covers, the mass and distribution of tailings, and the concentrations of contaminants in percolating pore fluids. Simple calculations, laboratory and field testing, and analytical and numerical modeling may be used to estimate water flux through variably-saturated tailings under steady-state conditions, which develop after consolidation and dewatering have essentially ceased. Contaminant concentrations in water discharging from the tailings depend on tailings composition, leachability and solubility of contaminants, geochemical conditions within the
, embankment, tailings-water interactions, and flux of water through the embankment. These concentrations may be estimated based on maximum reported concentrations, pore water concentrations, extrapolations of column leaching data, or geochemical equilibria and reaction pathway modeling. Attempts to estimate contaminant discharge rates should begin with simple, conservative calculations and progress to more-complicated approaches, as necessary.
2 INTRODUCTION ,
Since the late 1970's, disposal of uranium mill tailings in the United States has been regulated to reduce potential detrimental effects of the tailings on the environment and humans. Uranium tailings embankments are typically stabilized at the time of closure beneath compacted earthen covers to isolate the tailings and reduce discharges of hazardous contaminants. Long-term estimates of contaminant discharge rates are needed to assure that tailings stabilization does not adversely impact the quality of groundwater resources. Practical methods for ~ estimating discharges of contaminants from stabilized tailings have not been identified.
, This paper identifies and evaluates alternative approaches to estimate contaminant discharge rates from stabilized uranium tailings embankments. After a brief description of water flow and contaminant leaching in variably-saturated tailings, the paper describes alternative approaches to estimate contaminant discharge rates from stabilized 8606260179 860619 PDR MISC 0606260147 PDR t - _ _
. m l
tailings. For each alternative approach, the paper describes the application of the approach and identifies uncertainties that may
{- detract from the defensibility of contaminant discharge estimates. The i paper emphasizes practical approaches (e.g., defensible estimates at a i reasonable cost) and does not consider embankments using low-j penneability synthetic membranes, or leaching of organic constituents.
3 CONTAMINANT LEACHING IN STABILIZED TAILINGS l
I At most uranium milling sites in the western United States, a fraction
} of the precipitation that falls on stabilized tailings infiltrates the
- surface and percolates downward until it recharges the saturated zone. ;
Previous assessments have shown that the most significant effects of l 1 uranium tailings on groundwater resources occur during the active life i of tailings impoundments, when ponded process waters increase hydraulic <
i gradients that accelerate transport of contaminants away from the
- impoundments (NRC,1980). During closure of the impoundment, tailings i dewatering caused by draining and consolidation may discharge large .
volumes of highly contaminated tailings fluids (Fayer and Conbere 1985). '
j In comparison with the operational and deconunissioning phases of tailings impoundment operation, the discharge of contaminants from i stabilized (post-closure) tailings has been assumed to be negligible.
- Long-term contaminant discharge rates increase in significance.
l however, when the duration of their occurrence is considered. The EPA i standards in 40 CFR Part 192 require that uranium tailings stabilization j be effective for 1000 years, to the extent reasonably achievable, and, in any case, at least 200 years for radiological hazards. For non- !
radiological hazards, the standards require tailings disposal areas to
- be closed in a manner that controls, minimizes or eliminates, to the extent necessary, post-closure escape of hazardous constituents and
- leachate to groundwater, surface water, and the atmosphere.
1 Water flowing through stabilized ' uranium tailings may leach high
, concentrations of uranium, arsenic, molybdenum, selenium, thorium-230,
! radium-226, iron, chloride, sulfate, and other constituents (Dreesen, et I
al. 1983). Martin, Optiz, and Serne (1985) detennined that the main processes controlling contaminant leaching are displacement of residual ,
2 pore fluids and dissolution of readily soluble evaporites and moderately I sol.uble gypsum. Contaminant concentrations vary significantly because i of differences in ore composition, milling processes, and geochemical
- environments. For example, compositional analyses of tailings samples i from 12 inactive uranium mill's indicated that the concentration of l arsenic varied from 6 to 1100 ppm, with a mean concentration of 155.3 i ppm (Dreesen, et al. 1983).
I Long-term discharge of contaminated water from tailings embankments l will be primar,ily controlled by the downward flux of water through the l tailings. Even at arid sites where potential evaporation greatly >
4 exceeds annual precipitation, net recharge through tailings embankments ;
to the saturated zone may be expected. The flow of water through !
- partially-saturated tailings, however, is complicated by non-linear ,
i relationships that govern flow under these conditions. i l Following stabilization and dewatering of the tailings, which may last !
15 years or more, uranium tailings drain to partial saturation in semi-arid to arid environments where embankments are located above the local water table. If draining continued without recharge or j evaporation, the downward flux of water would decrease to zero as time r
! approached infinity. Klute and Heerman (1978) demonstrated using l l
- . - - - . - - - - - - - , - - - - . - - - - - _ - - , . - . - , - _ - . - - ~- -
L e t
numerical models that draining columns of medium-sized tailings assymptotically approached hydraulic equilibrium (zero flux) in 1000 to i 2000 days, even though - downward flux may decrease several orders of l
magnitude in the first 10 days of draining. This reduction in flux is
- i. caused by the rapid, non-linear decrease in the hydraulic conductivity 1 of the tailings as the moisture content decreases.
Hydraulic equilibrium cannot be achieved, however, as long as recharge y
?
or evaporation occurs. Evaporation of near-surface soil moisture may -
- reverse the downward-directed head gradient causing upward flow of 1
moisture to the land surface, which creates a hydraulic divide within -
i.' the shallow flow system. Water above the divide flows to the surface, -
whereas water below the divide continues to flow downward toward the i water table.
Although evaporation causes water flow in the near-surface hydraulic I
system to be transient, surface evaporation does not typically influence '
moisture content profiles'significantly below depths of about 1 m (Klute ,
and Heerman 1978; Gee, Nielson, and Rogers 1984; Lewis and Stephens 1 1985; DOE 1985). Below hydraulic divides, water content profiles are relatively constant, indicating that downward percolation approximates a steady-state process. Even in semi-arid and arid environments, the net
- movement of water in the vadose zone is generally directed downward to
! recharge the saturated zone. As an example, Klute and Heerman (1978) 1 determined a cumulative evaporative loss 2.1 cm in contrast with a 37.3-i cm loss of water that drained downward during a 15-day drainage period
! with a potential evaporation rate of I cm/ day in medium-grained uranium j tailings.
{ Infiltration of incident precipitation also causes transient flow in i shallow hydraulic systems in tailings embankments. Water that
' infiltrates the surface propagates downward as a wetting front. For a-given amount of precipitation, the greater the water content in cover materials near the surface, the greater the depth of wetting tront l! penetration. Using numerical models of infiltration, Klute and Heennan
! (1978) determined that wetting fronts from most precipitation events would not penetrate deeper than 5 to 10 cm, and only rarely penetrate 1 deeper than 15 to 20 cm in medium to coarse uranium tailings at Salt i Lake City, Utah. As the wetting front propagates, it displaces
! antecedent precipitation downward which increases the downward flux of
! water through tailings (Lewis and Stephens 1985). Water displaced below l the hydraulic divide continues to percolate downward until it recharges the saturated zone.
Water flow in stabilized tailings embankments may be complicated by rock armoring or transpiration by plants. In semi-arid environments, l for example, transpiration is expected to consume most, if not all, of '
i the precipitation that infiltrates into the land surface. Available
} research indicates that rock armoring tends to increase soil moisture i i retention by reducing surface evaporation, while transpiration decreases i soil moisture to depths up to the base of plant root zones (Beedlow i 1984; DOE 1985). These and other factors contribute complexity and i uncertainty in the estimation of contaminant discharge rates from j stabilized tailings embankments.
l l 4 ESTIMATING CONTAMINANT DISCHARGE RATES l Contaminant discharge rates from stabilized tailings equal the product l y
, of the net downward water flux through the tailings and contaminant I j concentrations in the percolating water. Approaches for estimating i
l l I !
e long-term discharge of contaminants from stabilized uranium tailings range from highly conservative, simplified calculations to advanced numerical analyses that require detailed supporting infonnation.
Selection of approaches for estimating contaminant discharge rates on a site-specific basis should consider the objective, desired precision, and defensibility of the estimates. Each combination of flux and concentration estimates has its strengths and limitations, and each may be considered appropriate for application depending upon site characteristics and the intent of the application. Estimation of contaminant discharge rates should begin with the simplest, most conservative approaches and progress with more-complicated approaches, as necessary.
Alternative approaches for estimating the flux of water through stabilized embankments include (1) assuming that tiux equals annual precipitation, (2) assuming that flux equals the saturated hydraulic conductivity of the cover, (3) assuming that flux equals the difference between annual precipitation and surface runoff, (4) measuring flux in demonstration embankments representative of stabilized conditions, and (5) calculating flux using analytical or nuerical models of moisture migration through stabilized tailings embankments. For estimating contaminant concentrations in water discharging from the tailings, alternative approaches include (1) assuming conservative concentrations based on available analyses, (2) assuming concentrations based on column-leaching studies, (3) assuming concentrations based on analyses of pore water samples from tailings, and (4) determining ranges and trends of concentrations using geochemical equilibria and reaction pathway modeling.
4.1 Estimating water flux through tailings The simplest and most conservative approach for estimating water flux through stabilized tailings is to assume that all incident precipitation on tailings embankments percolates through the tailings. For most uranium processing sites in the western United States, annual precipitation ranges between 10 and 30 cm (Gee, Nielson, and Rogers 1984). The approach assumes steady-state percolation through tailings to the saturated zone. For example, the groundwater flux through a 30 hectare (75 acre) embankment with an annual precipitation of 30 cm (11.8 in) would be 9.0E4 m3/yr (45.2 gpm). This flux of contaminated water may not significantly degrade groundwater quality beneath the embankment depending upon ' site-specific hydrogeologic conditions. A high estimate of contaminated water discharge from the tailings, however, may indicate that the quality of groundwater resources would be significantly degraded. Thus, a less conservative approach for estimating water flux may be more appropriate.
Water flux may also be ' estimated by assuming that it equals the saturated vertical hydraulic conductivity of the cover. For a uranium tailings embankment cover with a saturated hydraulic conductivity of 1E-7 cm/s, the net water flux could be estimated as 3.15 cm/yr assuming I steady-state flow under a unit hydraulic gradient through the cover.
Assuming this low conductivity, the net flux of contaminated water is an order ' of magnitude less than that for the example discussed above.
As an alternative to the first two approaches, water flux could be assumed to equal the difference between annual precipitation and estimated runoff from tailings embankments. Runoff may be estimated by several methods, including -direct measurement, empirical formulas, and
)
graphical methods. Because of the need to construct demonstration covers to measure runoff, direct measurement is not considered a practical approach to estimate runoff prior to construction of tailings embankments. The Rational Equation is an empirical formula that may be used to estimate runoff when site-specific runoff infonnation is not available. In the Rational Equation, runoff (peak discharge) equals the product of the uniform rate of rainfall intensity, the area across which rainfall occurs, and a runoff coefficient, which is a function of surface characteristics, type and extent of vegetation, surface slope, moisture conditions of the soil, and other factors (Lu, Eichenberger, and Stearns 1984). For gently-sloped (less than 7%) covers of compacted silt and clay, runoff coefficients may vary between 0.2 and 0.8 based on comparisons of cover characteristics with native and disturbed soils (cf. ASCE 1960; Perry 1976). Runoff can also be estimated using graphical methods such as the Soil Conservation Service curve method (SCS 1972). Runoff estimates prepared using the SCS curve method or the Rational Equation may be uncertain because of the inability of the techniques to consider such factors as riprap armoring or surface compaction (Lu, Eichenberger, and Stearns 1984).
Direct measurement of net water flux may be accomplished by monitoring moisture migration within demonstration embankments. Estimates of downward water flux can be compared qualitatively with estimates from sites in similar hydraulic environments or with estimates of regional recharge. This approach may not be practical, however, because of the need to construct demonstration embankments for measurement purposes. In addition, analysis of moisture migration data may be delayed until after perturbations to hydraulic conditions caused by construction have ceased.
Developing analytical or numerical models of water balances or moisture migration is the most complicated approach for estimating water fluxes througn stabilized tailings embankments. Models developed in the past to estimate fluxes through tailings have seldom been supported by sufficient data and consideration of reasonable ranges of input parameters and alternate conceptual models. Without sufficient information about site characteristics and consideration of ranges of predicted infiltration rates, models cannot defensibly estimate water flux through stabilized tailings.
For example, the development of deterministic numerical models of moisture migration in variably-saturated tailings requires specification of moisture characteristic curves, which are relationships between moisture content and matric potential. Given the heterogeneity of pore size distributions in uranium tailings (Martin et al 1980; 00E 1985),
moisture characteristic curves may change significantly within several centimeters because the relationships are a function of pore size distribution. Thus, defensible characterization of moisture characteristic curves to support numerical modeling of flow through partially-saturated tailings may require analysis of hundreds of tailings samples in a large tailings embankment. Even if such characterizations were available, modelers would be faced with the formidable task of accounting for this heterogeneity in the simulations.
Although stochastic simulation approaches may account for hatcrogeneous properties more easily than deterministic simulations, such approaches have not been sufficiently developed to predict water infiltration on a site-specific basis.
Analytical and numerical models are useful in estimating water flux in generic simulations. These models can help investigators conceptualize water flow in tailings, evaluate sensitivities of flow estimates to
[ _
~
variations in parameter values, and identify infomation needs for
, detailed simulation of moisture migration. They can also be used to determine whether estimated fluxes derived from simpler approaches are conservative. Defensible application of models to simulate l site-specific water discharge rates, however, will continue to be '
limited because of insufficient site characterization and inadequate consideration of ranges of parameter values and alternative conceptual models.
4.2 Estimating contaminant concentrations The simplest approach for estimating contaminant concentrations in water discharging from stabilized uranium tailings is to assume conservatively j high concentraticns that remain constant with respect to flow volume.
These assumed concentrations could be based on existing analyses of mill process liquors, raffinates, tailings pond solutions, and groundwater samples immediately beneath tailings impoundments. Concentrations should be selected to overestimate actual contaminant concentrations to i
determine worse case impacts of contaminated water discharge on groundwater resources. Table 1 lists examples of contaminant concentrations. In comparison with other approaches for estimating discharge water quality, this approach is relatively simple, inexpensive, and efficient.
If contaminant concentrations are assumed to teach from tailings at constant concentrations, the duration of leaching should be identified i to ensure that the contaminant mass leached from the tailings does not exceed that originally present in the embankment. Simple parametric analysis shows that the time (T) required to leach all of a contaminant 1 equals the product of contaminant concentration (C) and mass of the tailings (M) divided by the constant leaching concentration (L) and the water flux (Q) from the tailings, or j T = (CM)/(LQ).
1 Assumption of constant concentrations may not be appropriate, however, to estimate the quality of water discharging from tailings of unusual composition because of unusual ore composition or milling processes.
Such assumptions require consideration of site-specific factors that determine the validity of the assumptions.
As an alternative approach, contaminant concentrations might be
' estimated based on column leaching tests of uranium tailings.
Estimation of contaminant concentrations based on batch leaching tests is difficult because of uncertainties about what batch concentrations represent. Concentrations detemined in column leaching experiments are considered more representative of the magnitude and temporal variation of contaminant concentrations leached from uranium tailings. Existing column leaching data can be supplemented with site-specific leaching data for representative tailings samples to estimate variation of contaminant concentrations as a function of flow volume. As shown in Figure 1, column leaching experiments show that most contaminant concentrations decrease exponentially as a function of pore volume of flow through the columns. Because the kinetics of processes that
- control contaminant leaching in tailings are relatively rapid, leached concentrations appear to be independent of retention time within the columns. Concentrations of most contaminants decrease to relatively low i
~
levels after 4 to 6 pore volumes of flow through the tailings for contaminants whose leaching is controlled primarily by tailings-water 3 interactions such as absorption and co-precipitation (Martin, Opitz, and )
Serne 1985). Leaching behavior of mobile contaminants such as calcium l or chloride may deviate from this general behavior because their l leaching is. controlled by gradual dissolution or their transport is dominated by advection rather than tailings-water interactions.
The leaching behavior of individual contaminants may vary significantly. For example, pH cenerally increases as a function of leached pore volume for acidic tailings, which may decrease leaching of arsenic while increasing molybdenum leaching. The mobility and transport behavior of tailings contaminants is discussed in greater detail in Weber and Dam (1984), Dreesen et al (1983), and Peterson et al (1983).
TABLE 1. Comparison of contaminant concentrations (mg/1) in uranium mill tailings solutions Assumed high g Column leach Pore water Contaminant concentration concentration 2 concentrations 3 u 250 116 0.02 - 23.3 As 42 30.2 ---
Mo 93.7 1.31 0.004 - 49.0 Se 36.7 2.11 0.004 - 0.098 Th-230(PCi/1) 250,000 144.687 ---
Ra-226(PCi/1) 35,000 5,992 ---
Fe 5,450 7,700 2.8 - 56,500 C1 2,400 2,900 14 - 250 50 4 100,000 53,000 1970 - 268.000 I Weber and Dam,1984 2
Sample collected at 0.23 pore volumes (adjusted) from column leach test using acidic uranium tailings; Opitz, Dodson, and Serne 1985.
Ranges of concentratio'is in pore water samples collected at an inactive uranium tailings mill processed uranium ore using both acid and alkaline leaching; DOE 1985.
Column leaching studies have demonstrated that the quality of water initially displaced from uranium tailings may be significantly worse than the quality of mill process water. Opitz, Dodson, and Serne (1985) suggest that increases in contaminant concentrations in water displaced from tailings columns may be caused by continued interaction between pore water and tailings and dissolution of evaporitic salts that form naturally or during preparation of the columns. Continued interaction between pore water and tailings within stabilized embankments appears likely because the geochemical environment within the tailings is similar to the highly oxidizing, high tonic strength, low pH environment used in most milling processes to leach uranium from the ore.
1 l
i Another approach for estimating contaminant concentrations in water discharging from stabilized tailings is to assume concentrations based on analyses of pore water samples collected within tailings. This approach could easily be integrated with estimation of water flux through tailings by measuring moisture migration wich tensiometers and lysimeters. Analysis of samples trom two inactive tailings sites has indicated that contaminants may exist at extremely high concentrations within tailings pore water. For example, concentrations of sulfate in samples from one site ranged from approximately 2000 to 268,000 mg/l (DOE 1985). Table 1 compares ranges of contaminant concentrations observed in pore water samples with conservative concentrations of contaminants in process waters and in a sample from a column-leach test.
Similar to the results of column leaching experiments, contaminant concentrations vary considerably because of variability in ore composition, mill processing, evaporation of process and meteoric water, and tailings-water interactions.
O w.i.
, . n ..~.
} . % !
5 O I 3 a % is 3 Nr 2 . .
g i.
30000 r y ,. O O" J D - =
o og -
O
, .. Figure 1. Sulfate and
, , arsenic concentrations as
. a function of adjusted pore volume through columns
. of acid-leach uranium mill tailings (Martin, Opitz,
'a~a'**'**
and Serne (in press)J.
Analytical and numerical models may also be developed to estimate contaminant concentrations in water discharging from uranium tailings. ,
Development of geochemical equilibria or reaction pathway models requires substantial information and consideration of reasonable ranges )
of input parameters and alternative conceptual models. There are substantial uncertainties associated with concentration estimates developed using geochemical modeling. They are caused by incomplete I understanding of the chemical and physical processes that control !
- . m . I
l these concentrations, inadequate characterization of solid phases and solid-water interactions, sampling and analytical errors, insufficient thermodynamic data for constituents of interest, and approximations and simplifications invoked in developing the models. For example, geochemical equilibrium models that assume relatively dilute solutions are unable to simulate equilibria in high tonic strength tailings solutions where ion shielding, complexing, and common ion effects may significantly affect contaminant solubilities and speciation. An electrically neutral solution of calcium and sulfate ions with a sulfate concentration of 200,000 mg/l has an ionic strength of approximately 8.3, far in excess of the range of applicability for the Davies activity coefficient equation. As with the application of analytical and numerical models to estimate water fluxes through tailings embankments, geochemical models can be used to help conceptualize geochemical systems and estimate ranges and trends in contaminant concentrations.
Defensible modeling applications to predict contaminant concentrations on a site-specific basis, however, require substantially more supporting information and technical analysis than more conservative approaches.
5
SUMMARY
Prediction of the impacts of uranium tailings stabilization on groundwater resources requires estimates of contaminant discharge rates from partially-saturated tailings embankments. Water flowing through the tailings leaches contaminants such as uranium, arsenic, radium-226, and sulfate that may contaminate groundwater resources. Contaminant discharge rates may be estimated by multiplying estimates of water flux througn stabilized tailings by estimates of contaminant concentrations in the water. Attempts to estimate contaminant discharge rates should begin with simple, conservative approaches and progress to more complicated approaches, as necessary, considering the significance of contaminant discharges and desired defensibility of estimates.
REFERENCES American Society of Civil Engineers 1960. Design and construction of sanitary and storm sewers: Manual of engineering practice number 37.
Beedlow, P.A. 1984. Revegetation and rock cover for stabilization of inactive uranium mill tailings disposal sites. U.S. Department of Energy, DOE /UMT-0217.
Dreesen, D.R., M.E. Bunker, E.J.Cokal, M.M.Denton, J.W.Starner, E.F.
Thode, L.E.Wangen & J.M. Williams 1983. Research on the characterization and conditioning of uranium mill tailings. U.S.
Department of Energy, D0E/UMT-0263.
Fayer, M.J. & W.Conbere 1985. The analysis of drainage and consolidation at typical uranium mill tailings sites. U.S. Nuclear Regulatory Commission, NUREG/CR-4192.
Gee, G.W., K.K.Nielson & V.C. Rogers 1984. Predicting long-term moisture contents of earthen covers at uranium mill tailings sites. U.S.
Department of Energy, DOE /UMT-0220.
Klute, A. & D.F.Heerman 1978. Water movement in uranium mill tailings profiles. U.S. Environmental Protection Agency, ORP-LV-78-8.
Lewis, G.J. & D.B.Stephens 1985. Analysis of infiltration through mill tailings using a bromide tracer. In proceedings of the Seventh symposium on management of uranium mill tailings, low-level waste, and hazardous waste, February 6-8, Colorado State University, p.347-359.
Lu, J.C.S., B.Eichenberger & R.J.Stearns 1984. Production and management of leachate from municipal landfills: sumary and assessment. U.S.
Environmental Protection Agency, EPA-600/2-84-092.
Martin, J.P. , G.E.Veyera, D.M.Nasiatka & L.F.0sorno 1980.
Characterization of the inactive tailings sites. In proceedings of the Symposium of uranium mill tailings management, November 24-25, Colorado State University, p.421-488.
Martin, W.J., B.E.0pitz & R.J.Serne 1985. The effects of column dimensions on uranium mill tailings leach curves. In proceedings of the Seventh symposium on management of uranium mill tailings, low-level waste, and hazardous waste, February _6-8, Colorado State University, p.361-370.
Martin, W.J., 8.E.0ptiz & R.J.Serne (in press). The effects of column dimensions on uranium mill tailings leach curves. Uranium Journal.
Opitz, 8.E., M.E.Dodson & R.J.Serne 1985. Uranium mill tailings neutralization: contaminant complexation and tailings leaching studies. U.S. Nuclear Regulatory Comission, NUREG/CR-3906.
Perry, R.H. 1976. Engineering manual. New York: McGraw-Hill Book Company.
Peterson, S.R., A.A.Felmy, R.J.Serne & G.W. Gee 1983. Predictive geochemical modeling of interactions between uranium mill tailings solutions and sediments in a flow-through system. U.S. Nuclear Regulatory Commission, NUREG/CR-3404.
Soil Conservation Service 1972. National engineering handbook, section 4, supplement A (hydrology). U.S. Department of Agriculture.
U.S. Department of Energy 1985. Environmental assessment of remedial actions at the Riverton uranium mill tailings site, Riverton, Wyoming (draft). UMTRA Project Office, DOE /EA-0254.
U.S. Nuclear Regulatory Commission 1980. Final generic environmental impact statement on uranium milling. Office of Nuclear Material Safety and Safeguards, NUREG-0706.
Weber, M.F. & W.L. Dam 1984. Effects of uranium mill tailings on groundwater quality: a historical perspective. In proceedings of the Seventh national ground water quality symposium, September 26-28, National Water Well Association, p.193-208. ,
4 i Estimating contaminant ' discharge rates from stabilized uranium tailings embankments
- Michael F. Weber,
- Division of Waste Management, U. S. Nuclear Regulatory Connission i
! 8th Annual Symposium on Geotechnical and l
Geohydrological Aspects of Waste, ,
7 2/5-7/86, Fort Collins, Colorado l
- 1 ABSTRACT [
! Estimates of contaminant discharge rates from stabilized uranium i
- tailings embankments are essential in evaluating long-term impacts of l tailings disposal on groundwater resources. Contaminant discharge rates !
1 are a function of water flux through tailings covers, the mass and l
, distribution of tailings, and the concentrations of contaminants in
{ percolating pore fluids. Simple calculations, laboratory and field ;
testing, and analytical and numerical modeling may be used to estimate
{ water flux through variably-saturated tailings under steady-state
- conditions, which develop after consolidation and dewatering have i essentially ceased. Contaminant concentrations in water discharging i from the tailings depend on tailings composition, leachability and j
- solubility of contaminants, geochemical conditions within the
2 embankment, tailings-water interactions, and flux of water through the ,
i embankment. These concentrations may be estimated based on maximum
- reported concentrations, pore water concentrations, extrapolations of ;
i column leaching data, or geochemical equilibria and reaction pathway .
- modeling. Attempts to estimate contaminant discharge rates should begin i
- with simple, conservative calculations and progress to more-complicated j approaches, as necessary.
2 INTRODUCTION ,
3' Since the late 1970's, disposal of uranium mill tailings is the United States has been regulated to reduce potential detrimental etfects of the
, tailings on the environment and humans. Uranium tailings embankments i t are typically stabilized at the time of closure beneath compacted
- earthen covers to isolate the tailings and reduce discharges of hazardous contaminants. Long-term estimates of contaminant discharge '
rates are needed to assure that tailings stabilization does not adversely impact the quality of groundwater resources. Practical
- . methods for estimating discharges of contaminants from stabilized l tailings have not been identified.
j This paper identifies and evaluates alternative approaches to estimate
- contaminant discharge rates from stabilized uranium tailings i embankments. After a brief description of water flow and contaminant l leaching in variably-saturated tailings, the paper describes alternative 1 approaches to estimate contaminant discharge rates from stabilized
- - - , - - - - - - - - . . - - - - - , - , -.---.~,----..m e-. .
- - - - . - -,---r- , . - - - - - , . - - - . - . . - - - - - . , - - - . , . , - - - - . , - - - - , . . + - . - - , - -
- _ . . -. .=. _ - . -- .- . - - - - - - _ __
tailings. For each alternative approach, the paper describes the
, application of the approach and identifies uncertainties that may detract from the defensibility of contaminant discharge estimates. The i paper emphasizes practical approaches (e.g., defensible estimates at a j reasonable cost) and does not consider embankments using low- i permeability synthetic membranes, or leaching of organic constituents.
i 3 CONTAMINANT LEACHING IN STABILIZED TAILINGS
! At most uranium milling sites in the western United States, a fraction i of the precipitation that falls on stabilized tailings infiltrates the j surface and percolates downward until it recharges the saturated zone.
Previous assessments have shown that the most significant effects of uranium tailings on groundwater resources occur during the active life
- I of tailings impoundments, when ponded process waters increase hydraulic
! gradients that accelerate transport of contaminants away from the
- impoundments (NRC 1980). During closure of the impoundment, tailings dewatering caused by draining and consolidation may discharge large i volumes of highly contaminated tailings fluids (Fayer and Conbere 1985).
In comparison with the operational and decommissioning phases of tailings impoundment operation, the discharge of contaminants from stabilized (post-closure) tailings has been assumed to be negligible.
t Long-term contaminant discharge rates increase in significance, however, when the duration of their occurrence is considered. The EPA standards in 40 CFR Part 192 require that uranium tailings stabilization ;
be effective for 1000 years, to the extent reasonably achievable, and, 1
in any case, at least 200 years for radiological hazards. For non-J radiological hazards, the standards require tailings disposal areas to i i be closed in a manner that controls, minimizes or eliminates, to the I
extent necessary, post-closure escape of hazardous constituents and
- leachate to groundwater, surface water, and the atmosphere.
- Water flowing through stabilized uranium tailings may leach high
! concentrations of uranium, arsenic, molybdenum, selenium, thorium-230, i radium-226, iron, chloride, sulfate, and other constituents (Dreesen, et j al. 1983). Martin, Optiz, and Serne (1985) determined that the main processes controlling contaminant leaching are displacement of residual pore fluids and dissolution of readily soluble evaporites and moderately soluble gypsum. Contaminant concentrations vary significantly because of ' differences in ore composition, milling processes, and geochemical environments. For example, compositional analyses of tailings samples 1
from 12 inactive uranium mill's indicated that the concentration of arsenic varied from 6 to 1100 ppm, with a mean concentration of 155.3
- ppm (Dreesen,etal.1983).
Long-term discharge of contaminated water from tailings embankments will be primar,ily controlled by the downward flux of water through the tailings. Even at arid sites where potential evaporation greatly exceeds annual precipitation, net recharge through tailings embankments to the saturated zone may be expected. The flow of water through
! partially-saturated tailings, however, is complicated by non-linear '
relationships that govern flow under these conditions. i
- Following stabilization and dewatering of the tailings, which may last 1 15 years or more, uranium tailings drain to partial saturation in 3
semi-arid to arid environments where embankments are located above the i local water table. If draining continued without recharge or i evaporation, the downward flux of water would decrease to zero as time l approached infinity. Klute and Heerman (1978) demonstrated using l
,,-m.- - - -- - - . - - . - , - - _ ,. .-- . _ _ _ - _ . - . _ _ , _ , . . _ _ . . #~,,,_.___-.,.__<__m_-....-__,_m- , - - - , . _ . , ~ _ . _ _ _ _ _ - - - -
numerical models that draining columns of medium-sized tailings assymptotically approached hydraulic equilibrium (zero flux) in 1000 to 2000 days, even though downward flux may decrease several orders of magnitude in the first 10 days of draining. This reduction in flux is caused by the rapid, non-linear decrease in the hydraulic conductivity of the tailings as the moisture content decreases.
Hydraulic equilibrium cannot be achieved, however, as long as recharge or evaporation occurs. Evaporation of near-surface soil moisture may reverse the downward-directed head gradient causing upward flow of moisture to the land surface, which creates a hydraulic divide within
. the shallow flow system. Water above the divide flows to the surface, I whereas water below the divide continues to flow downward toward the water table.
Although evaporation causes water flow in the near-surface hydraulic system to be transient, surface evaporation does not typically influence '
. moisture content profiles significantly below depths of about 1 m (Klute j and Heennan 1978; Gee, Nielson, and Rogers 1984; Lewis and Stephens ,
! 1985; DOE 1985). Below hydraulic divides, water content profiles are 1 l relatively constant, indicating that downward percolation approximates a i steady-state process. Even in semi-arid and arid environments, the net i movement of water in the vadose zone is generally directed downward to i recharge the saturated zone. As an example, Klute and Heerman (1978) )
{ determined a cumulative evaporative loss 2.1 cm in contrast with a 37.3 :
cm loss of water that drained downward during a 15-day drainage period I with a potential evaporation rate of I cm/ day in medium-grained uranium !
j tailings. !
Infiltration of incident precipitation also causes transient flow in ;
shallow hydraulic systems in tailings embankments. Water that infiltrates the surface propagates downward as a wetting front. For a
- given amount of precipitation, the greater the water content in cover materials near the surface, the greater the depth of wetting tront penetration. Using numerical models of infiltration, Klute and Heennan (1978) determined that wetting fronts from most precipitation events would not penetrate deeper than 5 to 10 cm, and only rarely penetrate deeper than 15 to 20 cm in medium to coarse uranium tailings at Salt Lake City, Utah. As the wetting front propagates, it displaces antecedent precipitation downward which increases the downward flux of
- water through tailings (Lewis and Stephens 1985). Water displaced below the hydraulic divide continues to percolate downward until it recharges the saturated zone.
i Water flow in stabilized tailings embankments may be complicated by j rock armoring or transpiration by plants. In semi-arid environments, for example, transpiration is expected to consume most, if not all, of 3
1
. the precipitation that infiltrates into the land surface. Available research indicates that rock armoring tends to increase soil moisture
) '
retention by reducing surface evaporation, while transpiration decreases i soil moisture to depths up to the base of plant root zones (Beedlow 4 1984; DOE 1985). These and other factors contribute complexity and i
, uncertainty in the estimation of contaminant discharge rates from !
stabilized tailings embankments.
4 ESTIMATING CONTAMINANT DISCHARGE RATES j
1 Contaminant discharge rates from stabilized tailings equal the product !
l of the net downward water flux through the tailings and contaminant )
l concentra tions in the percolating water. Approaches for estimating '
l
l l
. . t l
l long-term discharge of contaminants from stabilized uranium tailings
! range from highly conservative, simplified calculations to advanced !
i numerical analyses that require detailed supporting in formation.
! Selection of approaches for estimating contaminant discharge rates on a i site-specific basis should consider the objective, desired precision.
l and defensibility of the estimates. Each combination of flux and i concentration estimates has its strengths and limitations, and each may
- be considered appropriate for application depending upon site i characteristics and the intent of the application. Estimation of f i contaminant discharge rates should begin with the simplest, most conservative approaches and progress with more-complicated approaches,
( as necessary.
- Alternative approaches for estimating the flux of water through ,
j stabilized embankments include (1) assuming that flux equals annual j precipitation, (2) assuming that flux equals the saturated hydraulic l conductivity of the cover, (3) assuming that flux equals the difference i between annual precipitation and surface runoff, (4) measuring flux in l
demonstration embankments representative of stabilized conditions, and 1
(5) calculating flux using analytical or numerical models of moisture migration through stabilized tailings embankments. For estimating
! centaminant concentrations in water discharging from the tailings, alternative approaches include (1) assuming conservative concentrations
}; based on available analyses, (2) assuming concentrations based on column-leaching studies, (3) assuming concentrations based on analyses !
. of pore water samples from tailings, and (4) determining ranges and trends of concentrations using geochemical equilibria and reaction \
pathway modeling.
i
! 4.1 Estimating water flux through tailings
! The simplest and most conservative approach for estimating water flux
~
through stabilized tailings is to assume that all incident precipitation on tailings embankments percolates through the tailings. For ,most
! uranium processing sites in tha western United States, annual precipitation ranges between 10 and 30 cm (Gee, Nielson, and Rogers 1984). The approach assumes steady-state percolation through tailings f to the saturated zone. For example, the groundwater flux through a 30 hectare (75 acre) embankment with an annual precipitation of 30 cm (11.8
- in) would be 9.0E4 m3/yr (45.2 gpm). This flux of contaminated water may not significantly degrade groundwater quality beneath the embankment.
- depending upon ' site-specific hydrogeologic conditions. A high estimate l of contaminated water discharge from the tailings, however, may indicate
- that the quality of groundwater resources would be significantly
]
degraded. Thus, a less conservative approach for estimating water flux may be more appropriate.
] Water flux may also be estimated by assuming that it equals the saturated vertical hydraulic conductivity of the cover. For a uranium tailings embankment cover with a saturated hydraulic conductivity of
, IE-7 cm/s, the net water flux could be estimated as 3.15 cm/yr assuming i steady-state flow under a unit hydraulic gradient througn the cover.
j Assuming this low conductivity, the net flux of contaminated water is an
- order of magnitude less than that for the example discussed above.
a As an alternative to the first two approaches, water flux could be
- assumed to equal the difference between annual precipitation and estimated runoff from tailings embankments. Runoff may be estimated by .
several methods, including direct measurement, empirical formulas, and .
i
,e-.n,-,.- , - - - . - _
_.,.._.-,-,.--m,n p.,,,- _.. ,,,- .,._..,, -, ,,_.,. _,, _ -
_.,,-~y- ,-,,9._,y g, ,.- _e-m,-
3, . _ - . .- -3 -, m--.., .,
graphical methods. Because of the need to construct demonstration covers to measure runoff, direct measurement is not considered a practical approach to estimate runoff prior to construction of tailings i embankments. The Rational Equation is an empirical fomula that may be :
used to estimate runoff when site-specific runoff infomation is not l available. In the Rational Equation, runoff (peak discharge) equals the product of the unifom rate of rainfall intensity, the area across which rainfall occurs, and a runof.f coefficient, which is a function of surface characteristics, type and extent of vegetation, surface slope, moisture conditions of the soil, and other factors (Lu, Eichenberger, i and Stearns 1984). For gently-sloped (less than 7%) covers of compacted silt and clay, runoff coefficients may vary between 0.2 and 0.8 based on comparisons of cover characteristics with native and disturbed soils (cf. ASCE 1960; Perry 1976). Runoff can also be estimated using graphical methods such as the Soil Conservation Service curve method (SCS 1972). Runoff estimates prepared using the SCS curve method or the Rational Equation may be uncertain because of the inability of the techniques to consider such factors as riprap amoring or surface compaction (Lu, Eichenberger, and Stearns 1984). ,
Direct measurement of net water flux may be accomplished by monitoring !
moisture migration within demonstration embankments. Estimates of downward water flux can be compared qualitatively with estimates !
from sites in similar hydraulic environments or with estimates of regional recharge. This approach may not be practical, however, because of the need to construct demonstration embankments for measurement purposes. In addition, analysis of moisture migration data may be delayed until after perturbations to hydraulic conditions caused by construction have ceased.
Developing analytical or numerical models of water balances or moisture migration is the most complicated approach for estimating water fluxes through stabilized tailings embankments. Models developed in the past to estimate fluxes through tailings have seldom been supported by sufficient data and consideration of reasonable ranges of input parameters and alternate conceptual models. Without sufficient information about site characteristics and consideration of ranges of predicted infiltration rates, models cannot defensibly estimate water flux through stabilized tailings.
For example, the development of deterministic numerical models of moisture migration in variably-saturated tailings requires specification of moisture characteristic curves, which are relationships between ,
moisture content and matric potential. Given the heterogeneity of pore size distributions in uranium tailings (Martin et al 1980; DOE 1985), ,
moisture characteristic curves may change significantly within several :
centimeters because the relationships are a function of pore size distribution. Thus, defensible characterization of moisture characteristic curves to support numerical modeling of flow through partially-saturated tailings may require analysis of hundreds of tailings samples in a large tailings embankment. Even if such t characterizations were available, modelers would be faced with the famidable task of accounting for this heterogeneity in the simulations. :
Although stochastic simulation approaches may account for heterogeneous.
properties more easily than deterministic simulations, such approaches have not been sufficiently developed to predict water infiltration on a site-specific basis.
Analytical and numerical models are useful in estimating water flux in i generic simulations. These models can help investigators conceptualize water flow in tailings, evaluate sensitivities of flow estimates to
- . - -- - - - - - - _ _ - - .- = _ _ _ _ _ _ _ _ _ _ _ _ _ ___
i variations in parameter values,. and identify information needs for detailed simulation of moisture migration. They can also be used to detemine whether estimated fluxes derived from simpler approaches are conservative. Defensible application of models to simulate site-specific water discharge rates, however, will continue to be limited because of insufficient site characterization and inadequate consideration of ranges of parameter values and alternative conceptual models. -
4.2 Estimating contaminant concentrations The simplest approach for estimating contaminant concentrations in water discharging from stabilized uranium tailings is to assume conservatively .
- high concentrations that remain constant with respect to flow volume.
! These assumed concentrations could be based on existing analyses of mill process liquors, raffinates, tailings pond solutions, and groundwater samples imediately beneath tailings impoundments. Concentrations should be selected to overestimate actual contaminant concentrations to '
determine worse case impacts of contaminated water discharge on groundwater resources. Table 1 lists examples of contaminant concentrations. In comparison with other approaches for estimating '
discharge water quality, this approach is relatively simple, inexpensive, and efficient.
If contaminant concentrations are assumed to leach from tailings at constant concentrations, the duration of leaching should be identified to ensure that the contaminant mass leached from the tailings does not 4
exceed that originally present in the embankment. Simple parametric i
analysis shows that the time (T) required to leach all of a contaminant 1 equals the product of contaminant concentration (C) and mass of the
, tailings (M) divided by the constant leaching concentration (L) and the water flux (Q) from the tailings, or
, T = (CM)/(LQ).
Assumption of constant concentrations may not be appropriate, however,
' to estimate the quality of water discharging from tailings of unusual 1 composition because of unusual ore composition or milling processes. l Such assumptions require consideration of site-specific factors that '
determine the validity of the assumptions.
As an alternative approach, contaminant concentrations might be
' estimated based on column leaching tests of uranium tailings.
i Estimation of contaminant concentrations based on batch leaching tests is difficult because of uncertainties about what batch concentrations represent. Concentrations determined in column leaching experiments are considered more representative of the magnitude .and temporal variation of contaminant concentrations leached from uranium tailings. Existing
- column leaching data can be supplemented with site-specific leaching data for representative tailings samples to- estimate variation of contaminant concentrations as a function of' flow volume. As shown in l Figure 1, column leaching experiments show that most contaminant concentrations decrease exponentially as a function of pore volume of flow through the columns. Because the -kinetics of processes that control contaminant leaching in tailings are relatively rapid, leached concentrations appear. to be independent of retention time within the columns. Concentrations of most contaminants decrease to relatively low l
l
l 1
f levels af ter 4 to 6 pore volumes of flow through the tailings for l contaminants whose leaching is controlled primarily by tailings-water interactions such as absorption and co-precipitation (Martin, Opitz, and l
Serne 1985). Leaching behavior of mobile contaminants such as calcium or chloride may deviate from this general behavior because their leaching is. controlled by gradual dissolution or their transport is dominated by advection rather than tailings-water interactions.
The leaching behavior of individual contaminants may vary significantly. For example, pH generally increases as a function of leached pore volume for acidic tailings, which may decrease leaching of arsenic while increasing molybdenum leaching. The mobility and transport behavior of tailin is discussed in greater detail in Weber and Dam (1984)gs contaminants, Dreesen et al (1983), and Pe (1983).
TABLE 1. Comparison of ccntaminant concentrations (mg/1) in uranium mill tailings solutions Assumed high g Column teach Pore water Contaminant concentration :encentration 2 concentrations 3 U 250 116 0.02 - 23.3 As 42 30.2 ---
Mo 93.7 1.31 0.004 - 49.0 Se 36.7 2.11 0.004 - 0.098 Th-230 (PCf/1) 250,000 144,687 ---
Ra-226 (PC1/l) 35,000 5,992 ---
Fe 5,450 7,700 2.8 - 56,500 C1 2,400 2,900 14 - 250 50 100.000 53,000 1970 - 268,000 4
I Weber and Dam,1984 2
Sample collected at 0.23 pore volumes (adjusted) from column leach test using acidic uranium tailings; Opitz, Dodson, and serne 1985.
Ranges of concentrations in pore water samples collected at an inactive uranium tailings mill processed uranium ore using both acid and alkaline leaching; DOE 1985.
Column leaching studies have demonstrated that the quality of water initially than thedisplaced fromprocess quality of mill uranium tailings water. may Opitz, be significantly Dodson, and Serne(worse 1985) suggest that increases in contaminant concentrations in water displaced from tailings columns may be caused by continued interaction between pore water and tailings and dissolution of evaporitic salts that form naturally or during preparation of the columns. Continued interaction between pore water and tailings within stabilized embankments appears likely because the geochemical environment within the tailings is similar to the highly oxidizing, high ionic strength, low pH environment used in most milling processes to leach uranium from the ore.
I i
l
Another approach for estimating contaminant concentrations in water discharging from stabilized tailings is to assume concentrations based on analyses of pore water samples collected within tailings. This approach could easily be integrated with estimation of water flux thrnugh tailings by measuring moisture migration with tensiometers and lysimeters. Analysis of samples trom two inactive tailings sites has indicated that contaminants may exist at extremely high concentrations within tailings pore water. For example, concentrations of sulfate in samples from one site ranged from approximately 2000 to 268,000 mg/l (DOE 1985). Table I compares ranges of contaminant concentrations observed in pore water samples with conservative concentrations of contaminants in process waters and in a sample from a column-leach test.
Similar to the results of column leaching experiments, contaminant concentrations vary considerably because of variability in ore composition, mill processing, evaporation of process and meteoric water, and tailings-water interactions.
i l
Q sese.
. s,. .
} . % !
5 O i
- - Q % i 3 N s
! i.
00 0 0 l
o O O- 0 "
I 06 -
O
, .. Figure 1. Sulfate and
. arsenic concentrations as
, a function of adjusted pare volume through columns
= of acid-leach uranium mill tailings (Martin, Opitz,
^*""'*'** and Serne (in press)).
Analytical and numerical models may also be developed to estimate contaminant concentrations in water discharging from uranium tailings.
Development of geochemical equilibria or reaction pathway models requires substantial information and consideration of reasonable ranges of input parameters and alternative conceptual models. There are substantial uncertainties associated with concentration estimates developed using geochemical modeling. They are caused by incomplete understanding of the chemical and physical processes that control
i l
l these concentrations, inadequate characterization of solid phases and I solid-water interactions, sampling and analytical errors, insufficient )
thennodynamic data for constituents of interest, and approximations and cimplifications invoked in developing the models. For example, geochemical equilibrium models that assume relatively dilute solutions are unable to simulate equilibria in high ionic strength tailings solutions where ion shielding, complexing, and comon ion effects may significantly affect contaminant solubilities and speciation. An electrically neutral solution of calcium and sulfate ions with a sulfate concentration of 200,000 mg/l has an ionic strength of approximately 8.3, far in excess of the range of applicability for the Davies activity coefficient equation. As with the application of analytical and numerical models to estimate water fluxes through tailings embankments, geochemical models can be used to help conceptualize geochemical systems and estimate ranges and trends in contaminant concentrations.
Defensible modeling applications to predict contaminant concentrations on a site-specific basis, however, require substantially more supporting information and technical analysis than more conservative approaches.
5
SUMMARY
Prediction of the impacts of uranium tailings stabilization on groundwater resources requires estimates of contaminant discharge rates from partially-saturated tailings embankments. Water flowing through the tailings leaches contaminants such as uranium, arsenic, radium-226, and sulfate that may contaminate groundwater resources. Contaminant discharge rates may be estimated by multiplying estimates of water flux througn stabilized tailings by estimates of contaminant concentrations in the water. Attempts to estimate contaminant discharge rates should begin with simple, conservative approaches and progress to more complicated approaches, as necessary, considering the significance of contaminant discharges and desired defensibility of estimates.
REFERENCES American Society of Civil Engineers 1960. Design and construction of sanitary and storm sewers: Manual of engineering practice number 37.
Beedlow, P.A. 1984. Revegetation and rock cover for stabilization of inactive uranium mill tailings disposal sites. U.S. Department of Energy, DOE /UMT-0217.
Dreesen, D.R., M.E. Bunker, E.J.Cokal, M.M.Denton, J.W.Starner, E.F.
Thode, L.E.Wangen & J.M. Williams 1983. Research on the characterization and conditioning of uranium mill tailings. U.S.
Department of Energy, D0E/UMT-0263.
Fayer, M.J. & W.Conbere 1985. The analysis of drainage and consolidation at typical uranium mill tailings sites. U.S. Nuclear Regulatory Commission, NUREG/CR-4192.
Gee, G.W., K.K.Nielson & V.C. Rogers 1984. Predicting long-term moisture contents of earthen covers at uranium mill tailings sites. U.S.
Department of Energy, 00E/UMT-0220.
Klute, A. & D.F.Heerman 1978. Water movement in uranium mill tailings profiles. U.S. Environmental Protection Agency, ORP-LV-78-8.
.+.
Lewis, G.J. & D.B.Stephens 1985. Analysis of infiltration through mill tailings using a bromide tracer. In proceedings of the Seventh symposium on management of uranium mill tailings, low-level waste, and hazardous waste, February 6-8, Colorado State University, p.347-359.
Lu, J.C.S., B.Eichenberger & R.J.Stearns 1984. Production and management of leachate from municipal landfills: sumary and assessment. U.S.
Environmental Protection Agency, EPA-600/2-84-092.
Martin, J.P. , G.E.Veyera, D.M.Nasiatka & L.F.0sorno 1930.
Characterization of the inactive tailings sites. In proceedings of the Symposium of uranium mill tailings management, November 24-25, Colorado State University, p.421-488.
Martin, W.J., B.E.0pitz & R.J.Serne 1985. The effects of column dimensions on uranium mill tailings leach curves. In proceedings of the Seventh symposium on management of uranium mill tailings, low-level waste, and hazardous waste, February 6-8, Colorado State University, p.361-370.
Martin, W.J., 8.E.0ptiz & R.J.Serne (in press). The effects of column dimensions on uranium mill tailings leach curves. Uranium Journal.
Opitz, B.E., M.E.Dodson & R.J.Serne 1985. Uranium mill tailings neutralization: contaminant complexation and tailings leaching studies. U.S. Nuclear Regulatory Connission, NUREG/CR-3906.
Perry, R.H. 1976. Engineering manual. New York: McGraw-Hill Book Company.
Peterson, S.R., A.A.Felmy, R.J.Serne & G.W. Gee 1983. Predictive geochemical modeling of interactions between uranium mill tailings solutions and sediments in a flow-through system. U.S. Nuclear Regulatory Commission, NUREG/CR-3404.
Soil Conservation Service 1972. National engineering handbook, section 4, supplement A (hydrology). U.S. Department of Agriculture.
U.S. Department of Energy 1985. Environmental assessment of remedial actions at the Riverton uranium mill tailings site, Riverton, Wyoming (draft). UMTRA Project Office, DOE /EA-0254.
U.S. Nuclear Regulatory Commission 1980. Final generic environmental impact statement on uranium milling. Office of Nuclear Material Safety and Safeguards, NUREG-0706.
Weber, M.F. & W.L. Dam 1984. Effects of uranium mill tailings on groundwater quality: a historical perspective. In proceedings of the Seventh national ground water quality symposium, September 26-28, National Water Well Association, p.193-208.
e -