Search

Processing U ores in the JEB Mill of the McClean Lake Operation in northern Saskatchewan produces spent leaching solutions (raffinates) with pH ⩽ 1.5, and As and Ni concentrations up to 6800 and 5200 mg L−1, respectively. Bench-scale neutralization experiments (pH 2–8) were performed to help optimize the design of mill processes for reducing As and Ni concentrations in tailings and raffinates to ⩽1 mg L−1 prior to their disposal. Precipitate mineralogy determined by chemical analysis, XRD, SEM, EM, XM and EXAFS methods, included gypsum (the dominant precipitate), poorly crystalline scorodite (precipitated esp. from pH 2–4), annabergite, hydrobasaluminite, ferrihydrite, green rust II and theophrastite. The As was mostly in scorodite with smaller amounts in annabergite and trace As adsorbed and/or co-precipitated, probably by ferrihydrite. Geochemical modeling indicated that above pH 2, the ion activity product (IAP) of scorodite lies between the solubility products of amorphous and crystalline phases (log Ksp = −23.0 and −25.83, respectively). The IAP decreases with increasing pH, suggesting that the crystallinity of the scorodite increases with pH. Forward geochemical models support the assumption that during neutralization, particles of added base produce sharp local pH gradients and disequilibrium with bulk solutions, facilitating annabergite and theophrastite precipitation.

Published solubility data for amorphous ferric arsenate and scorodite have been reevaluated using the geochemical code PHREEQC with a modified thermodynamic database for the arsenic species. Solubility product calculations have emphasized measurements obtained under conditions of congruent dissolution of ferric arsenate (pH < 3), and have taken into account ion activity coefficients, and ferric hydroxide, ferric sulfate, and ferric arsenate complexes which have association constants of 10 4.04 (FeH2AsO4 2+ ), 10 9.86 (FeHAsO4 + ), and 10 18.9 (FeAsO4). Derived solubility products of amorphous ferric arsenate and crystalline scorodite (as log Ksp) are � 23.0 ± 0.3 and � 25.83 ± 0.07, respectively, at 25 � C and 1 bar pressure. In an application of the solubility results, acid raffinate solutions (molar Fe/As = 3.6) from the JEB uranium mill at McClean Lake in northern Saskatchewan were neutralized with lime to pH 2–8. Poorly crystalline scorodite precipitated below pH 3, removing perhaps 98% of the As(V) from solution, with ferric oxyhydroxide (FO) phases precipitated starting between pH 2 and 3. Between pH 2.18 and 7.37, the apparent log Ksp of ferric arsenate decreased from � 22.80 to � 24.67, while that of FO (as Fe(OH)3) increased from � 39.49 to � 33.5. Adsorption of As(V) by FO can also explain the decrease in the small amounts of As(V)(aq) that remain in solution above pH 2–3. The same general As(V) behavior is observed in the pore waters of neutralized tailings buried for 5 yr at depths of up to 32 m in the JEB tailings management facility (TMF), where arsenic in the pore water decreases to 1–2 mg/L with increasing age and depth. In the TMF, average apparent log Ksp values for ferric arsenate and ferric hydroxide are � 25.74 ± 0.88 and � 37.03 ± 0.58, respectively. In the laboratory tests and in the TMF, the increasing crystallinity of scorodite and the amorphous character of the coexisting FO phase increases the stability field of scorodite relative to that of the FO to near-neutral pH values. The kinetic inability of amorphous FO to crystallize probably results from the presence of high concentrations of sulfate and arsenate. � 2006 Elsevier Inc. All rights reserved.

At the McClean Lake Operation in the Athabasca Basin of northern Saskatchewan, the untreated acid raffinate solutions associated with U mill tailings contain up to 700 mg/L dissolved As. To reduce the concentration of As and other contaminants in acid tailing slurries at the JEB mill at McClean Lake, ferric sulfate may be added to the acid raffinates to assure that their molar Fe/As ratio equals or exceeds 3. Tailings slurries are then neutralized with lime to pH 4, and subsequently to pH 7–8. The neutralized tailings contain minerals from the original ore, which are chiefly quartz, illite, kaolinite and chlorite, and precipitated (secondary) minerals that include gypsum, scorodite, annabergite, hydrobasaluminite and ferrihydrite. Most of the As is associated with the secondary arsenate minerals, scorodite and annabergite. However, a few percent is adsorbed and/or co-precipitated, mainly by ferrihydrite. Of major concern to provincial and federal regulators is the risk that significant amounts of As might be released from the tailings to pore waters after their subaqueous disposal in the tailings management facility. A laboratory study was performed to address this issue, measuring readily desorbed As using a method known as equilibrium partitioning in closed systems (EPICS). The EPICS method was selected because it employs a leaching solution that, except for its As concentration, is identical in composition to the neutralized raffinate in contact with the tailings. Laboratory experiments and modeling results demonstrated that the As that could be readily released to pore waters is about 0.2% of the total As in the tailings. Long-term, such releases may contribute no more than a few mg/L of dissolved As to tailings pore waters.

The proposed JEB Tailings Management Facility (TMF) to be emplaced below the groundwater table in northern Saskatchewan, Canada, will contain uranium mill tailings from McClean Lake, Midwest and Cigar Lake ore bodies, which are high in arsenic (up to 10%) and nickel (up to 5%). A serious concern is the possibility that high arsenic and nickel concentrations may be released from the buried tailings, contaminating adjacent groundwaters and a nearby lake. Laboratory tests and geochemical modeling were performed to examine ways to reduce the arsenic and nickel concentrations in TMF porewaters so as to minimize such contamination from tailings buried for 50 years and longer. The tests were designed to mimic conditions in the mill neutralization circuit (3 hr tests at 25°C), and in the TMF after burial (5–49 day aging tests). The aging tests were run at, 50, 25 and 4°C (the temperature in the TMF). In order to optimize the removal of arsenic by adsorption and precipitation, ferric sulfate was added to tailings raffinates 1 having Fe/As ratios of less that 3–5. The acid raffinates were then neutralized by addition of slaked lime to nominal pH values of 7, 8, or 9. Analysis and modeling of the test results showed that with slaked lime addition to acid tailings raffinates, relatively amorphous scorodite (ferric arsenate) precipitates near pH 1, and is the dominant form of arsenate in slake limed tailings solids except those high in Ni and As and low in Fe, in which cabrerite-annabergite (Ni, Mg, Fe(II) arsenate) may also precipitate near pH 5–6. In addition to the arsenate precipitates, smaller amounts of arsenate are also adsorbed onto tailings solids. The aging tests showed that after burial of the tailings, arsenic concentrations may increase with time from the breakdown of the arsenate phases (chiefly scorodite). However, the tests indicate that the rate of change decreases and approaches zero after 72 hrs at 25°C, and may equal zero at all times in the TMF at 4°C. Consistent with a kinetic model that describes the rate of breakdown of scorodite to form hydrous ferric oxide, the rate of release of dissolved arsenate to tailings porewaters from slake limed tailings: (1) is proportional to pH above pH 6–7; (2) decreases exponentially as the total molar Fe/As ratio of tailings raffinates is increased from 1/1 to greater than 5/1; and (3) is proportional to temperature with an average Arrhenius activation energy of 13.4 ± 4.2 kcal/mol. Study results suggest that if ferric sulfate and slaked lime are added in the tailings neutralization circuit to give a raffinate Fe/As molar ratio of at least 3–5 and a nominal (initial) pH of 8 (final pH of 7–8), arsenic and nickel concentrations of 2 mg/L or less, are probable in porewaters of individual tailings in the TMF for 50 to 10,000 yrs after tailings disposal. However, the tailings will be mixed in the TMF, which will contain about 35% tailings with Fe/As = 3.0, and 65% tailings with Fe/As = 5.0–7.7. Thus, it seems likely that average arsenic pore water concentrations in the TMF may not exceed 1 mg/L.

Rodney C. Ewing and Allison Macfarlane (“Yucca Mountain,” Policy Forum, 26 April, p. [659][1]) suggest that the decision to site a high-level nuclear waste repository at Yucca Mountain, Nevada, should be delayed until all relevant scientific issues are addressed. Insisting on comprehensive

Uranium ores at the McClean Lake Operation in the Athabasca Basin of Northern Saskatchewan can produce elevated levels of arsenic (up to 700 mg/L) and nickel (up to 500 mg/L) in acidic (pH

A method to quantify the amount of readily released arsenic in uranium mill tailings was developed using a technique known as Equilibrium Partitioning In Closed Systems (EPICS). The method employs a gentle leaching solution that, except for its arsenic (As) concentration, is identical to the neutralized raffinate that contacts the tailings. Prior to implementation, the experimental design and mathematical approach were verified in geochemical models using PHREEQC. Laboratory experiments using tailings from the Athabasca Basin of Northern Saskatchewan demonstrated that As that could be readily released to pore waters is about 0.2 % of the total As in the tailings.

Radionuclide (RN) adsorption has long been recognized as important to assure the isolation of nuclear wastes in a geological repository [1]. Laboratory measured RN adsorption data have generally been expressed as distribution coefficient (Kd) values or adsorption isotherms. The proper application of these models is to site conditions nearly identical to those used in the laboratory adsorption experiments. This has required that multiple Kd's and isotherms be determined in a wide range of experiments designed to bracket expected repository conditions.The surface complexation (SC) adsorption models were introduced in the late 1970's. The best known of these models incorporate electrical double layer (EDL) theory [2]. Their use requires that the water chemistry and surface properties of adsorbing rocks and minerals be fully characterized. Adsorption is then studied as reactions involving specific aqueous RN species (often complexes) and specific surface sites. Because the SC models are relatively mechanistic, they may allow extrapolation of adsorption results to repository conditions that lie outside the limited experimental range used to parameterize a given model. Turner [3] has shown that the diffuse layer model (the simplest SC model) fits a wide range of RN adsorption data as well as the more complex models. Others have suggested ways to generalize and estimate SC model parameters for a variety of minerals, rocks and engineered materials (cf. [4,5,6,7,8,9,10,11,12]. Degueldre and Werlni [12] and Degueldre et al. [13] have proposed a simplified SC model for RN adsorption that avoids EDL theory, in which the adsorption of RN species is estimated from linear free energy relationships.It is appropriate to ask how accurately RN adsorption behavior must be known or understood for total system performance analysis (TSPA). In most geological settings now being considered for repository development globally, it may suffice to select bounding Kdvalues for the different rock types (cf. [14,15]). Use of the SC models to describe RN adsorption can provide us with increased confidence that minimum Kd's and the distribution of Kdvalues we might propose for TSPA are in fact conservative.

The clay backfill that will surround a buried high level nuclear waste package in most national repository programs, could be modified to play a greater role as a barrier to radionuclide (RN) releases. The RN steady state release (Mb) rate from a clay backfill to adjacent rock is directly proportional to backfill porosity (ε), RN diffusion rate In the backfill (Ds), and RN solubility at the waste form surface (C*), and Inversely proportional to RN half-life (λ) and RN retardation coefficient (R) in the backfill [1]. We propose ways to reduce ε, Ds and C* and Increase R for important radionuclides, mostly through the addition of reactive minerals to the backfill. Silica, calcite and anhydrite may be added to precipitate and clog porosity. Increased backfill compaction similarly reducesε, Ds and Mb for all the RN's. Strongly sorbent phases can be added to selectively adsorb both cationic and anionic RN's (e.g. 1–129). However, adsorption will not Importantly reduce peak release rates of most long-lived RN's. The backfill can be poised at reducing Eh's with mineral additives to lower Ds and so immobilize radioisotopes of NI, Np, Pa, Pu, Se, Tc and U. Minerals of stable or more stable isotopes of Cs, NI, Se, Sn and U can be added to lower Ds values of the RN's and to coprecipitate them in solid solution. Phosphorite-apatites, which are known to have high selectivities for rare earths and RN's, may be added to coprecipitate Am, Np, Pu, Sr, Th and U.

Adsorption processes are important in controlling U concentrations in ground water. Quantifying such processes is extremely difficult in that in situ conditions cannot be directly measured. One rock characteristic that must be known to quantify adsorption is the specific surface area of reactive minerals exposed to the ground water. We evaluate here three methods for estimating specific surface area in situ. The first is based on the dissolution kinetics of sodium feldspars, the second on emanation of radon-222 and the third on adsorption of naturally-occurring U. The radon-222 method yields estimates 5 to 8 orders of magnitude greater than those obtained via the other two methods; too large probably because of effects related to fracture geometry. Estimates of specific surface area based on modelling adsorption of natural U by aquifer materials are of comparable magnitude to those from the feldspar-dissolution kinetics approach. These conclusions are based on analyses of water from 145 wells in crystalline-rock aquifers from Pennsylvania, New Jersey, Maryland, and Colorado. Computer modelling of the chemical data using PHREEQE [1] showed that uraninite or coffinite approach saturation in reducing water, limiting total U to −9 m. Generally, U minerals are below saturation in oxidizing ground water, where uranyl-carbonate complexes are the dominant dissolved U species. Autoradioluxographs of thin sections show areas of concentration of radioactivity in the rocks and establish that U is concentrated along fracture boundaries and on ferric oxyhydroxide grain coatings. Because U minerals generally are undersaturated, U mobility is limited by adsorption onto ferric oxyhydroxides and other mineral surfaces. Calculations of uranyl adsorption from the ground water onto goethite using the program M1NTEQ [2] show that adsorption decreases with increased carbonate concentrations due to the formation of uranyl-carbonate complexes. Results of this paper improve our understanding of the mobility of U that might be released into oxidized ground water in crystalline rock from a breached radioactive-waste repository.

Based on potentiometric data and gypsum solubility in mixed salt solutions, respectively, the activity coefficients of MgCOo3 and CaSOo4 ion pairs decrease with ionic strength (I) at 25°C. Computed γ's for the ion pairs fit the empirical equation log γi = − BI. B coefficients of 0.63 ±0.10 for MgCOo3 and 0.45 ± 0.15 for CaSOo4 are obtained from linear regression of log γi values vs I between 0.04 and 0.6 molal. Assumptions that the activity coefficients of these neutral ion pairs equal unity or are approximated by the Setchenow expression (log γi = kI) are therefore invalid at moderate ionic strengths. Log γi = − BI is the same general form as the equation of Kirkwood (Chem. Rev.24, 233–251, 1939) for neutral dipoles.

The chemical and stable isotope compositions of unpolluted ground waters in carbonate terranes are a function of the pH, PCO2, 13C content of the ground water recharge, the 13C content of the carbonate rock, and the manner in which the rock is dissolved or precipitated. Physico-chemical models show that significantly different relationships exist between Ca2+ plus Mg2+, HCO3−, pH and 13C content of unpolluted ground waters when carbonate solution occurs in the presence or absence of a gas phase. A study was made of these relationships in waters from 21 springs and 13 wells in Nittany Valley, Pennsylvania. Assuming that CO2 in the recharge zone has a 13C contents below −21%. vs PDB, the data indicate that ground water flow to wells and springs, and carbonate rock solution probably occurs chiefly in the absence of a gas phase. This is in spite of the fact that most of such flow is under water table conditions. 13C contents averaged −12·3%. for the spring waters and −13·3%. for the well waters. Five well waters polluted by septic tank or sewage effluents had carbon isotopic compositions from −13·5 to −16·4%.,vs −11·3 to−12·7 %. for the eight well waters relatively free of organic wastes.

Uranyl adsorption was measured from aqueous electrolyte solutions onto well-characterized goethite, amorphous ferric oxyhydroxide, and hematite sols at 25°C. Adsorption was studied at a total uranyl concentration of 10−5 M, (dissolved uranyl 10−5 to 10−8 M) as a function of solution pH, ionic strength and electrolyte concentrations, and of competing cations and carbonate complexing. Solution pHs ranged from 3 to 10 in 0.1 M NaNO3 solutions containing up to 0.01 M NaHCO3. All the iron oxide materials strongly adsorbed dissolved uranyl species at pHs above 5 to 6 with adsorption greatest onto amorphous ferric oxyhydroxide and least onto well crystallized specular hematite. The presence of Ca or Mg at the 10−3 M level did not significantly affect uranyl adsorption. However, uranyl carbonate and hydroxy-carbonate complexing severely inhibited adsorption. The uranyl adsorption data measured in carbonate-free solutions was accurately modeled with the surface complexation-site binding model of Davis et al. (1978), assuming adsorption was chiefly of the UO2OH+ and (UO2)3(OH)+5, aqueous complexes. In modeling it was assumed that these complexes formed a monodentate UO2OH+ surface complex, and a monodentate, bidentate or tridentate (UO2)3(OH)+5surface complex. Of the latter, the bidentate surface complex is the most likely, based on crystallographic arguments. Modeling was less successful predicting uranyl adsorption in the presence of significant uranyl carbonate and hydroxy-carbonate complexing. It was necessary to slightly vary the intrinsic constants for adsorption of the di- and tricarbonate complexes in order to fit the uranyl adsorption data at total carbonate concentrations of 10−2 and 10−3 M.

Gibbs free energies, enthalpies and entropies of 42 dissolved uranium species and 30 uranium-bearing solid phases have been critically evaluated from the literature and estimated when necessary for 25°C. Application of the data shows that the uranium in natural waters is usually complexed. At typical concentrations of chloride, fluoride, phosphate and sulfate, uranous (U4+) fluoride complexes are important in anoxic waters below pH 3–4. An intermediate Ehs (between about +0.2 and −0.1 V) and pH values 1–7, UO2+ ion may predominate. In oxidized waters, uranyl (U22+) fluoride complexes and uranyl ion predominate below pH 5; from about pH 4 to 7.5, UO2(HPO4)22− is the principal species; while at higher pHs, UO2CO30 and the di- and tri-carbonate complexes predominate. Uraninite [UO2-UO2.25], α-U3O8 and schoepite are the stable uranium oxides and hydroxides in water at 25°C. Coffinite, USiO4 (c), is probably stable relative to UO2(c) when dissolved silica exceeds about 60 ppm (as SiO2). At low Ehs and pH 4–6, the solubilities of stoichiometric crystalline uraninite and coffinite are below roughly 10−4 ppb. But at intermediate Ehs and neutral to alkaline pHs in the presence of phosphate or carbonate, the formation of uranyl phosphate or carbonate complexes can increase the solubilities of these minerals by several orders of magnitude. The uranyl minerals carnotite, tyuyamunite, autunite, potassium autunite and uranophane are least soluble at pHs in the range 5–8.5 and, in the case of carnotite and tyuyamunite, have solubilities as low as 0.2 and 1 ppb uranium, respectively. The autunites and uranophane are usually several orders of magnitude more soluble than this, consistent with their natural occurrences. Sorption of uranyl on to natural materials is maximal in the same pH range of 5–8.5.

Ground water samples were collected from 41 wells penetrating basal Oakville Sandstone (Miocene) sedimentary rocks in southern Texas and chemically analyzed for major and trace elements to identify chemical changes related to nearby uranium ore bodies. The sample coverage included a 240 km2 area which contains several fault-related uranium deposits. Factors affecting the hydrochemical system of the study area include : (1) the relatively high permeabilities of buried fluvial channel sediments ; (2) upward leakage of brines along growth faults into the aquifer ; (3) the development of a redox interface (Eh=0 volts) within the aquifer ; and (4) the semi-arid climate. Variations in the saturation indices (SI) for chemically reduced minerals of U, As, Mo, Se, and for associated minerals such as pyrite outlined the position of known deposits. The SI increased towards zero (the equilibrium value) as the deposits were approached from updip distances of 3 to 4.5 km, then decreased again downdip. Ground water saturation data for calcite were strongly correlated with the position of the present-day redox boundary. SI values of minerals of U, As, Mo and Se were far more useful for locating the ore deposits than were individual element concentrations. The radiogenic pathfinders Ra and Rn exhibited very strong anomalies within ore, but diminished to background levels at short distances from ore. A strong helium anomaly is deflected in the direction of ground water flow away from the ore. The approach described in this paper constitutes an effective regional scale exploration guide which would lead to a more efficent test-drilling program., Des échantillons d'eaux de subsurface ont été récoltés dans 41 puits atteignant les grès de base Oakville d'âge Miocène dans le Sud du Texas et ont été analysés pour les éléments majeurs et les éléments en traces afin de mesurer les variations chimiques liées à la proximité de minéralisations uranifères. L'échantillonnage s'étend sur une surface de 240 km2 qui contient plusieurs gisements d'uranium liés à des failles. Les facteurs qui influent sur le système hydrochimique de la région étudiée sont : (1) les perméabilités relativement fortes des sédiments enfouis issus de chenaux fluviatiles, (2) l'écoulement vers le haut des saumures le long de failles vivantes dans l'aquifère, (3) la formation d'une interface redox (Eh=0 volt) dans l'aquifère et (4) le climat semi-aride. Les variations des indices de saturation (SI) pour les minéraux réduits d'U, As, Mo, Se et pour les minéraux associés tels que la pyrite déterminent la position des gisements connus. Les indices de saturation tendent vers zéro (la valeur d'équilibre) lorsque l'on approche les gisements à des distances de 3 à 4-5 km en amont et décroissent de nouveau en aval. Les données de saturation pour la calcite dans les eaux de subsurface sont fortement corrélées avec la position de la limite redox actuelle. Les valeurs des indices de saturation pour les minéraux d'U, As, Mo et Se se sont révélées beaucoup plus utiles pour localiser les gisements que les concentrations des éléments prises individuellement. Les traceurs radiogéniques Ra et Rn sont fortement anomaliques dans le minerai, mais descendent rapidement au niveau du fond à de faibles distances du minerai. Une intense anomalie en hélium sous forme de traînée est observée dans la direction de circulation de l'eau de subsurface. L'approche décrite dans cet article constitue un guide d'exploration efficace à l'échelle régionale qui devrait conduire à un programme de sondages d'évaluation plus sûr., Wanty Richard B., Chatham James R., Langmuir Donald. The solubilities of some major and minor element minerals in ground waters associated with a sandstone-hosted uranium deposit. In: Bulletin de Minéralogie, volume 110, 2-3, 1987. Les mécanismes de concentration de l'uranium dans les environnements géologiques.

The enthalpy, Gibbs free energy, and entropies of aqueous radium species and radium solids have been evaluated from empirical data, or estimated when necessary for 25°C and 1 bar. Estimates were based on such approaches as extrapolation of the thermodynamic properties of Ca, Sr, and Ba complexes and solids plotted against cationic radii and charge to radius functions, and the use of the Fuoss or electrostatic mathematical models of ion pair formation (Langmuir, 1979). Resultant log K (assoc) and ΔH0 (assoc) (kcal/mol) values are: for RaOH+ 0.5 and 1.1; RaCl+ −0.10 and 0.50; RaCO03 2.5 and 1.07; and RaSO04 2.75 and 1.3. Log Ksp and ΔH0 (dissoc) (kcal/mol) values for RaCO3(c) and RaSO4(c) are −8.3 and −2.8, and −10.26 and −9.4, respectively. Trace Ra solid solution in salts of Pb and of the lighter alkaline earths, has been appraised based on published distribution coefficient (D) data, where D ∼- (mM2+)(NRaX)/(mRa2+)(NMX) (m and N are the aqueous molality and mole fraction of Ra and cation M in salt X, respectively. The empirical solid solution data have been used to derive both enthalpies and Gibbs free energies of solid solution of trace Ra in sulfate and carbonate minerals up to 100°C. Results show that in every case D values decrease with increasing temperature. Among the sulfate and carbonate minerals, D values decrease for the following minerals in the order: anhydrite > celestite > anglesite > barite > aragonite > strontianite > witherite > cerussite.

The cumulative association constant (β 2 ) for the geochemically important aqueous complex UO 2 [HPO 4 ] 2− 2 has been determined by potentiometric titration in Na 2 HPO 4 -UO 2 (NO 3 ) 2 solutions in the pH range 3.9–4.7, at ionic strengths below 0.024 molal with the Newton-Raphson method used to compute β 2 from the chemical analytical data. Based on 25 measurements we obtain log β 2 = 18.3 ± 0.2 at 25°C. From the same experiments we compute that the association constant of UO 2 OH + is 8.9 ± 0.1, in disagreement with the value of 8.3 ± 0.3 for this constant given by Baes and Mesmer (1976).

Mineralogy and specific surface area are major controls on the stabilities of ferric oxyhydroxide microparticles in natural waters. The thermodynamic stabilities of ferric oxyhydroxides, as described by the activity product in solution pK = −log [Fe3+] [OH−]3 range from pK = 37.1 for freshly precipitated amorphous oxyhydroxide to pK = 44.2 for well crystallized goethite. The sizes of suspended oxyhydroxide particles in natural waters range from less than 0.01 μm to greater than 5 μm. Oxyhydroxides precipitated in the laboratory from solutions simulating high-iron natural waters are needlelike or lathlike in shape and have mean thicknesses as small as 60 Å. Large specific surface areas resulting from the small sizes of ferric oxyhydroxide particles cause increased solubilities and thus decreased pK values. Specific surface areas of 40–170 m2/g determined for laboratory precipitates gave computed decreases in pK of 0.4 to 1.6 units. ImagesFIGURE 1.FIGURE 3.FIGURE 3.FIGURE 4.

Thermodynamic properties of 32 dissolved thorium species and 9 thorium-bearing solid phases have been collected from the literature, critically evaluated and estimated where necessary for 25°C and 1 atm pressure. Although the data are incomplete, especially for thorium minerals and organic complexes, some tentative conclusions can be drawn. Dissolved thorium is almost invariably complexed in natural waters. For example, based on ligand concentrations typical of ground water (ΣCl = 10 ppm, ΣF = 0.3 ppm, ΣSO4 = 100 ppm, andΣPO4 = 0.1 ppm), the predominant thorium species are Th(SO4)02, ThF2+2, and Th(HPO4)20below pH ≈ 4.5; Th(HPO4)2−3 from about pH 4.5 to 7.5; and Th(OH)04 above pH 7.5. Based on stability constants for thorium citrate, oxalate and EDTA complexes, it seems likely that organic complexes predominate over inorganic complexes of thorium in organic-rich stream waters, swamp waters, soil horizons, and waterlogged recent sediments. The thorium dissolved in seawater is probably present in organic complexes and as Th(OH)04. The tendency for thorium to form strong complexes enhances its potential for transport in natural waters by many orders of magnitude below pH 7 in the case of inorganic complexing, and below about pH 8 when organic complexing is important. The existence of complexes in addition to those formed with hydroxyl, is apparent from the fact that measured dissolved thorium in fresh surface waters (pH values generally 5–8) usually ranges from about 0.01 to 1 ppb and in surface seawater (pH = 8.1) is about 0.00064 ppb. This may be contrasted with the computed solubility of thorianite in pure water which is only 0.00001 ppb Th as Th(OH)04 above pH 5. Although complexing increases the solubility of thorium-bearing heavy minerals below pH 8, maximum thorium concentrations in natural waters are probably limited in general by the paucity and slow solution rate of these minerals and by sorption processes, rather than by mineral-solution equilibria.

From conductance measurements, the negative logarithm of the dissociation constant of the CaHCO 3+ ion pair, p K ( CaHCO 3 + ), is 0.7, 1.0 and 1.35 within ±0.05 units at 0, 25 and 60°C, respectively. A revaluation of published and unpublished data yields p K ( CaCO 3 0 ) ≈ 3.2 at 25°C. Use of these p K 's to compute the dissociation constant of calcite ( Kc ) from published calcite solubility measurements in pure water gives p Kc values which increase markedly with ionic strength. However, if the ion pairs are ignored, computed p Kc values are nearly constant with ionic strength. All reasonable attempts to eliminate the trend in p Kc by adjusting ion activity coefficients, and/or values of K ( CaCO 3 0 ) failed, so the dilemma remains. Kc values computed from the most reliable published calcite solubility data are in good agreement with such values based on solubility data measured in this study at 5, 15, 35 and 50°C. Study results ignoring ion pairs are accurately represented by the equation log Kc = 13.870 — (3059/ T ) −0.04035 T , and correspond to −8.35, −8.42, and −8.635 at 0, 25 and 50°C, respectively. The logarithmic expression leads to ΔH r o = −2420 ± 300 cal/mol, ΔCp = −110 ± 2 cal/deg mol, and ΔS r o = −46.6 ± 1.0 cal/deg mol for the calcite dissociation reaction at 25°C. The dependence of Kc on temperature when CaCO 3 0 and CaHCO 3 + are assumed, is described by log Kc = 13.543 − (3000/ T ) − 0.0401 T which yields −8.39, −8.47, and -8.70 at 0, 25 and 50°C. This gives ΔH r o = −2585 ± 300 cal/mol, ΔCp = −109 ± 2 cal/deg mol, and ΔS r 0 = −47.4 ± 1.0 cal/deg mol at 25°C.

A synthetic δMnO 2 was characterized with respect to intrinsic equilibrium constants and surface charge density as a function of pH. Site binding-surface complexation theory was used to model adsorption of Cu, Pb, and Zn onto the δMnO 2 surface. The model response was tested with respect to changes in solution ionic strength, total trace metal concentration, δMnO 2 in suspension, substrate surface area, and the presence of competitive trace metals. Results indicate that both freshly precipitated and aged δMnO 2 act as heterogeneous surfaces with a spectrum of binding site energies. The distribution of site energies was examined by determination of apparent equilibrium binding constants. Highest energy sites were saturated once less than 0.1% of all surface sites had been occupied. Adsorption experiments were performed for solutions which simulate conditions in the channel of a mountainous stream. The natural system is actively precipitating δMnO 2 in a watershed containing base metal sulfide deposits.

The geochemistry of Ca, Sr, Ba and Ra sulfates in some deep brines from the Palo Duro Basin of north Texas, was studied to define geochemical controls on radionuclides such as 90Sr and 226Ra. Published solubility data for gypsum, anhydrite, celestite, barite and RaSO4 were first reevaluated, in most cases using the ion interaction approach of Pitzer, to determine solubility products of the sulfates as a function of temperature and pressure. Ionic strengths of the brines were from 2.9 to 4.8 m, their temperatures and pressures up to 40°C and 130 bars. Saturation indices of the sulfates were computed with the ion-interaction approach in one brine from the arkosic granite wash fades and four from the carbonate Wolfcamp Formation. All five brines are saturated with respect to gypsum, anhydrite and celestite, and three of the five with respect to barite. All are undersaturated by from 5 to 6 orders of magnitude with respect to pure RaSO4. 226Ra concentrations in the brines, which ranged from 10−11.3 to 10−12.7 m, are not controlled by RaSO4 solubility or adsorption, but possibly by the solubility of trace Ra solid solutions in sulfates including celestite and barite.

A study was made of the chemistry of soil moisture collected over a period of 12 months from sandy loam soils in central Pennsylvania. Samples were taken at depths of from 1 to 9 m by means of suction lysimeters. The prevalent chemical character of soil moisture was generally Ca 2+ > Na + > Mg 2+ ≈ K + , with HCO − 3 > Cl − > SO 2− 4 ≈ NO − 3 on both a mole and weight basis. Soil-moisture pH ranged from 5.20 to 6.74. Dissolved silica averaged 54 mg/l (as SiO 2 ) and was as high as 117 mg/l. The specific conductance ranged from 20 to 400 μS/cm. Total dissolved solids (TDS) in the water increased with decreasing atmospheric precipitation and increasing temperature. This was due to evapotranspiration, which increased TDS directly, and also indirectly by reducing soil relative permeabilities and increasing the contact time between soil minerals and water. A secondary cause was greater precipitation acidity during the warmer months and during periods of lower precipitation. Soil colloids are predominantly Fe- and Mn-oxyhydroxides, with usually smaller amounts of illite and kaolinite. Total cation exchange capacity is 2–3 meq./100 g. Regression analysis of dissolved-ion activities showed that the aquo-cations are strongly buffered by Donnan equilibrium and selective adsorption with selectivities in the order K + > H + > Mg 2+ > Ca 2+ . Calculation of the degree of saturation of the soil moisture with respect to soil minerals showed that the soil moisture is undersaturated with respect to calcite, dolomite, amorphous silica and illite, and super-saturated with respect to quartz and probably also kaolinite. Sorption equilibria controlled major-cation concentrations at all depths in the soil, whereas solution-mineral equilibria did not limit concentrations of these cations.

A study of the character and movement of landfill leachate through unsaturated soil was begun in 1967 at the State College (Pennsylvania) Regional Sanitary Landfill which has operated since 1962 employing the trench method of waste disposal. The landfill occupies a gently sloping underdrained valley with a water table more than 200 feet below land surface. Precipitation averages about 37 inches as rain per year. Residual sandy-clay to sandy-loam soils range from a few feet to greater than 70 feet in thickness on a sandy dolomite bedrock. Soil moisture samples were extracted at different depths from beneath two of the refuse cells using suction lysimeters. Water samples were also bailed from these cells and pumped from a water table well beneath the landfill. Monthly or less frequent analyses performed on water samples included Eh, pH, temperature, specific conductance, BOD, Cl, SO4, total alkalinity, NH3, NO2, NO3, PO4, Ca, Mg, Na, K, and total Fe. Soil samples from beneath the refuse were subjected to particle size and x-ray analysis, and chemical analysis of soil pH, soluble salt content, exchangeable Ca, Mg, Na, and K, cation exchange capacity, and extractable P content. The study showed that the quality and quantity of leachate beneath a landfill varies considerably with the topographic setting of landfill trenches or cells. Leachates 2 feet under an upslope cell which received only direct precipitation, had the following maximum values 3-13 months after refuse burial: specific conductance 8445 μmhos, Cl 1890 mg/1, BOD 3300 mg/1, NH3-N 540 mg/1, and total Fe 225 mg/1. Upon reaching a depth of 14.5 feet after about 21/2 years or more, maximum values of these species in the leachate had been reduced by 83%, 80%, >99%, >99%, and 98% respectively. In contrast, more water, including precontaminated surface and subsurface runoff from adjacent upslope cells; infiltrated a downslope cell, saturating the refuse. Even after moving downward in the soil to a depth of 36 feet in 7 years, the leachate beneath this cell had a conductance of 6600 μmhos, 600 mg/1 Cl, over 9000 mg/1 BOD, 40 mg/l NH3-N, and 100 mg/1 total Fe. A practically continuous depletion of inorganic species in the refuse as indicated by the quality of leachate from both cells has occurred with time. However, concentrations of BOD and redox sensitive species such as Fe and NH3 in the leachate have fluctuated in response to changes in the moisture content and temperature of the refuse. Leachate beneath instrumented cells is moving downward in the subsoil at the rate of 6-11 ft/yr. Observed mechanisms of leachate renovation during this downward percolation, along with supporting evidence for each listed parenthetically, include: dilution and dispersion (decrease in Cl with depth); oxidation (Eh and pH measurements, decrease in BOD and Fe with depth); chemical precipitation (decrease in soil extractable P after leachate percolation); cation exchange (increase in percent base saturation of clays affected by leachate, and depletion of NH3 under reducing conditions—bacterial growth may also retard or remove NH3); and anion exchange (decrease in SO4 with depth). Removal of bacteria, suspended ferric oxyhydroxides and other particulates by soil filtration undoubtedly also occurs. Although renovation takes place, it is incapable of preventing highly contaminated leachate from moving to depths of 50 feet or more in soils beneath downslope cells. Ground-water contamination has occurred, probably by leachate channelled down fractures in locally shallow bedrock, or by leachate contaminated runoff from heavy storms which has entered sinkholes or infiltrated along the valley bottom. The study shows that improper design of landfills emplaced above the water table in relatively permeable soils and bedrock such as are found in many carbonate-rock terranes, can result in serious ground-water pollution.

A review and revaluation of calcite solubility measurements in pure water indicates that the activity product of calcite at 25°C and 1 atm pressure is 10−8.40±0.02.ΔGf0 for Ca2+(aq) has been revaluated as −132,350 ± 250 cal gfw−1. Based on the above data and ΔGf0 = −126,170 ± 70 cal gfw−1 for CO32−(aq), ΔGf0 for calcite is −269,980 ± 260 cal gfw−1. The study includes a detailed appraisal of the various errors relating to the solubility measurements and their interpretation.

In carbonate-rock terranes in central Pennsylvania a spring's discharge characteristics and water chemistry can be predicted from its geographic and hydrogeologic setting. Springs located near non-carbonate ridges (conduit-type) receive much of their recharge as mountain runoff entering sinkholes, have waters low in solutes, and a large variability in water quality and discharge. Springs in mid-valley areas (diffuse-type) receive water chiefly by infiltration through soils, or from a deeper circulating ground water flow system, and discharge water with higher total dissolved solids (TDS) and less variation in solute amounts and discharge. A spring's flow is a much more important control on its water chemistry than is season, particularly for diffuse-type springs. In conduit spring waters TDS increase in the early part of the storm hydrograph, decrease during the middle of the storm peak because of recharge via sinkholes, and finally increase as the storm pulse dissipates. In diffuse spring waters TDS increase and decrease nearly simultaneously with discharge following a storm as infiltration through soils near the spring leaches salts to the ground-water table. Based on precipitation and discharge records it took 6 and 60 days respectively, for 50% of the ground water introduced following a storm to leave a conduit and a diffuse type spring system.

The field pH and content of Ca2+, Mg2+, HCO3−, dissolved oxygen and other aqueous species have been measured in 29 spring waters and 29 well waters in folded and faulted Paleozoic carbonate rocks near State College, Pennsylvania. Most of the springs issue from limestone; most of the well waters are pumped from dolomite. Average specific conductances of the spring and well waters were 347 and 499 μΩ−1, respectively, with polluted well waters having conductances as high as 945 μΩ−1. Molar Ca2+Mg2+ ratios as low as 0.6 in the well waters have resulted from incongruent solution of dolomite. Theoretical treatment of relationships between PCO2, pH and HCO3− content shows that solution of calcite or dolomite by undersaturated waters leads to increases in both pH and HCO3− content. This relation is approximately obeyed by the spring waters at an average CO2 pressure of 10−2.2 atm. In contrast, solution or precipitation of carbonates by waters near saturation with respect to them, results in inverse relationships of pH and HCO3−; a behavior closely obeyed by the well waters which have CO2 pressures up to 10−1.6 atm. Computer calculation of ground water saturation with calcite and dolomite at ground water temperatures (8–14°C) was made using remeasured solubility product data for calcite, and considering the effect of CaSO40, MgSO40 and MgHCO3+ ion pairs. Within the uncertainty of the chemical analyses and thermochemical data, 12 well waters and 3 spring waters were saturated with calcite, whereas 7 well waters and 3 spring waters were saturated with dolomite. None of the ground waters significantly exceeded saturation with respect to either carbonate.

Published free-energy data on carbonates in the system $$MgO-CO_{2}-H_{2}O$$ are largely either suspect or unavailable. This paper is a review and revaluation of information bearing on the stabilities of the carbonates nesquehonite, lansfordite, hydromagnesite, artinite, and magnesite. Whenever possible, several independent methods have been used to establish free-energy values. Free energies of formation from the elements at 25° C. and 1 atm. total pressure have been calculated from data on solubility, synthesis, thermal decomposition, heats of solution, and heats of combustion, as well as from geological occurrences. Selected free-energy values for magnesite and hydromagnesite differ substantially from published values. Stability diagrams based on preferred free-energy data are presented in summary, and appear to be essentially consistent with the low-temperature laboratory and geologic behavior of phases in this system.

Insights into the chemical and Hydrogeologie history of spring waters in carbonate rocks can be gained from a study of the chemical quality of such waters and of the waters which feed springs. The median Ca2+/Mg2+ ratio was 3.4 in 29 spring waters from folded and faulted carbonate rocks near State College, Pennsylvania, indicating ground-water flow predominantly in limestone. The mean dissolved oxygen content was 7.7 mg/1 or 71% of saturation consistent with subsurface flow chiefly under water table conditions. Most of the dissolved solids (as specific conductance) including pollutants such as Cl and NO3 were added to spring waters during ground-water flow. Specific conductances ranged from 180 to 476 micromhos, with a mean of 347 micromhos. CO2 pressures were from 10−1.9 to 10−2.6 atm with a mean of 10−2.2. atm. The CO2, which is chiefly introduced as gas by downward diffusion from the soil zone, gives spring waters the capacity to hold about 3 times more hardness and alkalinity than streams in equilibrium with the atmosphere. However, all spring waters were undersaturated with respect to calcite and dolomite because of their short residence times as ground water. Dye tracing of ground-water feeding springs in two drainage basins gave residence times of 2–6 days for subsurface flows of about 4,000 feet. The specific conductance and carbonate saturation data indicate that ground-water residence times of the 29 spring waters average somewhat longer than 2–6 days.

To give due credit to the barrier of the far field geologic environment in many host media it is necessary to account for sorption processes. The ultimate impact of sorption modelling will be in helping to define the degree of confidence that may be placed on geochemical retardation occurring in this barrier. For those involved in site characterisation, systems' design, design implementation and regulation, this confidence is best derived from a combination of a fundamental phenomenological understanding of the sorption process with empirical observations of sorption in natural environments. Neither alone is adequate. By performing a few additional measurements during classical Ko experiments, the data necessary for the more fundamental models, such as that of double-layer or surface ionization and complexation, may also be provided. The basis of these models and their integration into broader performance analysis are outlined in the context of how this maximises confidence in the geologic barrier of critically concerned groups.

The work reported here involved application of the ion-interaction approach of Pitzer [1] [2] to model the activities of some major and minor cations and major anions in deep brine systems at elevated temperatures and pressures.The solubilities of anhydrite (CaS04), gypsum (CaSO4·2H2O), celestite (SrS04), barite (BaS04), and radium sulfate (RaSO4) in brines from the Wolfcamp Formation and granite wash facies, Palo Duro Basin, Texas, were modeled using ion-interaction equations. Waters with ionic strengths ranging from 2.89 to 4.76 m, and temperatures and pressures up to 40°C and 130 bars were sampled from five horizons in three wells, at depths between 970 and 1,670 meters. Theoretical solubility products as a function of temperature and pressure were obtained for comparison with ion activity products computed using ion-interaction theory. The effect of temperature on model calculations was found to reside almost entirely in the Debye-Hückel AΦ parameter, which accordingly was corrected for brine temperatures [3].Modeling results indicated that all five brines are saturated with respect to anhydrite and celestite, and three of the five with respect to barite. Saturation may result from hydrologic connection with adjacent overlying evaporite units, or increased Ca, Sr and Ba concentrations caused by their desorption from clays. Radium concentrations, which range from 10-11.3 to 10-12.7 m, are not controlled by RaSO4 solubility or adsorption, but probably by solid solution in other sulfate minerals in these low pH (4.4 – 6.3) high calcium (0.18 – 0.52 m) brines [4] [5].

The inability of empirical single value elemental Rd values to model radionuclide retardation in natural groundwater systems has demonstrated the need for developing a comprehensive integrated phenomenological model of the physicochemical and hydrologic mechanisms involved in retardation. The model must account for: (1) radioactive decay, (2) chemical precipitation/dissolution of bulk phases, (3) chemical substitution reactions, (4) isotopic exchange reactions, (5) cation and anion exchange, (6) specific adsorption, (7) diffusion into adjacent flow paths, (8) diffusion into fluid not involved in flow, (9) diffusion into the solid matrix and (10) ultrafiltration. The formulation of such a model is proposed, based on a reaction-path-simulation code with modifications for adsorption processes (by a surface-complexation site-binding model) and diffusional processes. The coupling of the physicochemical portion of this model to a ground-water transport code can be achieved by alternately iterating through the chemical and physical portions of the code during each time step. Sensitivity analysis used to determine dominant species, dominant retardation mechanisms and temporal and spatial homogeneities in the flow system will allow the integrated model to be reduced to a manageable form.

The empirical data on adsorption of metal cations by naturally occurring soil materials can be systematized in terms of power exchange functions. For example, given the reaction PbX + Ca2+ = CaX + Pb2+, the exchange function is: Kex = ([Pb2+]/[Ca2+]) (CaX/PbX)n where Kex and n are constants, the brackets denote activities of the ions, and CaX and PbX are their mole fractions on sorbent X. This approach is mathematically equivalent to regular solution exchange when the mole fractions of two competing cations sorbed lie between 0.25 and 0.75, and to Freundlich isotherm-type behavior when the mole fraction of the minor cation is < 0.05. Combined literature review and laboratory study show that the exchange behavior of H+, Na+, K+, Ca2+, Mg2+, Cd2+, Co2+, Ni2+, Pb2+, UO 2 2+ and Zn2+ and their hydroxy-complexes on a variety of adsorbents (montmorillonites, beidellite, illite, ferric oxyhydroxides, zeolites, soils and humic materials) can be accurately described for a wide range of competing sorbate concentrations and ratios using from one to three power exchange expressions. The adsorption of the alkali metal and alkaline earth cations on pure clays at about 10−2 to 10−4 M often follows the power exchange function with n = 1, corresponding to simple ion exchange. Adsorption of heavy metals (between 10−3 and 10−7 M) is usually more complex, and fits power exchange functions with n = 0.8 to 2.0. Log-linearization of power exchange expressions yields lines with correlation coefficients usually in the range from 0.98 to 1.00. The power exchange model is a potentially useful predictor of heavy metal levels in the subsurface controlled by adsorption processes.

Tripathi argues that we could not determine the stability constant ( β 2 ) of UO 2 [HPO 4 ] 2 2− in our solutions because they were supersaturated with respect to solid uranyl triphosphate. This is irrelevant because no phosphate solids were present in solution during our experiments. He further argues that the conditions of our experiments were not ideal for us to determine the potentially most accurate value of β 2 . This is also irrelevant. The point is that our published value of log β 2 = 18.3 ± 0.2 is given with an uncertainty which clearly acknowledges the inaccuracy of its determination based on our experimental and computational approach.