This Bureau of Mines publication provides information to assist in selecting a lixiviant (leach solution) for in situ uranium leaching. The cost, advantages, and disadvantages of lixiviants currently used and proposed are presented. Laboratory and field tests are described, and applications of geochemical models are discussed.

Environmental, economic, and technical factors should all be considered. Satisfying environmental regulations on restoring groundwater quality is becoming an overriding factor, favoring sodium bicarbonate or dissolved carbon dioxide over ammonium carbonate. The cheapest lixiviant is dissolved carbon dioxide, but it is not effective in all deposits. Technical factors include clay swelling by sodium, acid consumption by calcite, and the low solubility of oxygen in shallow deposits.

Laboratory leaching tests can provide useful data. However, they can be misleading if, for example, the ore is allowed to oxidize before testing or if distilled water instead of formation water is used for making solutions for permeability tests. Geochemical models presently are more useful for indicating trends in solubility than in reliably predicting concentrations.

The Bureau of Mines, as part of its mission to help assure that the Nation has an adequate supply of minerals, conducts research on alternative recovery techniques such as in situ leaching. An important consideration in any in situ leaching operation is the choice of lixiviant (the leach solution). This publication is intended to assist companies planning an in situ uranium leaching operation to select a lixiviant. General knowledge of in situ leaching is assumed; readers wanting a basic introductory publication should consult other literature.

Selection of the lixiviant is of critical importance to the success of an in situ leaching operation. The lixiviant affects not only the recovery of uranium and the cost of chemicals, but also the difficulty of meeting environmental regulations concerning restoration of groundwater quality after leaching. Thus, the choice of lixiviant is not simple, as technical, economic, and regulatory factors should all be considered.

No manuscript specifically on lixiviant selection has previously been made available to the public. Much research has been done by companies, but the results have usually been considered proprietary. However, useful literature is available on topics that are important parts of the lixiviant selection process. The chemistry of conventional uranium milling is thoroughly discussed by Merritt. The similarities in chemistry between milling and in situ leaching make this a very useful reference. However, differences such as the generally higher concentrations and consumptions of chemicals and higher recovery of uranium in milling should be kept in mind when considering in situ leaching. (The section on in situ leaching does not represent the current state of the art because this reference was published in 1971.) Extensive column leaching studies were performed for the Bureau of Mines by Westinghouse Electric Corp. These reports compare results using various lixiviants on two sandstone ores and provide recommendations concerning methods of conducting column leaching tests in general. The lixiviants were usually made with distilled water instead of compatible formation water, so permeability results should be interpreted with caution. The influence of various lixiviants on the difficulty of restoring the groundwater quality after leaching is an important factor and is discussed in several publications. Well drilling and completion affect lixiviant selection indirectly, as good injection and production wells are vital to the success of field tests.

 

Uranium Leaching Agents

Lixiviants that have been used for in situ uranium leaching include solutions of ammonium carbonate-bicarbonate, sodium carbonate-bicarbonate, carbon dioxide, and sulfuric acid. Potassium carbonate-bicarbonate is technically attractive but has been considered too expensive. Hydrochloric and nitric acids ; have been proposed for leaching carbonaceous ore. The carbonate-bicarbonate and sulfuric acid lixiviants contain an anion that will form a soluble complex with uranium in its +6 charge state. The cation does not directly affect the solubility of the uranium but is important because of its effect on permeabi1itv, cost, and groundwater quality restoration.

An oxidizer is required to convert unoxidized uranium from its insoluble +4 charge state to its soluble +6 charge state. Oxidizers that have been used include oxygen, hydrogen peroxide, and sodium chlorate.

The costs, not including delivery, of chemicals used for making lixiviants are listed in table 1 in the forms commonly used in dollars per kilogram and per ton. Costs were obtained from discussions with suppliers and leaching companies and from published prices in late 1980. These units are not the most useful for comparing costs because they do not directly compare the cost of providing the significant component. For example, 1 kg of potassium carbonate provides less carbonate that 1 kg of ammonium carbonate.

To facilitate comparing the costs, tables 2, 3, and 4 present the cost of a kilogram-mole and a pound-mole of alkaline lixiviants, acid lixiviants, and oxidizers, respectively. Table 2 lists the costs of bicarbonate and carbonate lixiviants separately to permit calculating the cost of lixiviant containing any proportion of the two. Costs of lixiviants of typical concentrations are also given in terms of cents per liter and cents per gallon in the discussion of each lixiviant that follows the summary tables. These lixiviant costs were calculated using the costs of chemicals in table 1, so changes in those chemical prices will cause proportionate changes in the corresponding lixiviant costs. The cost of $80/ton of carbon dioxide used in formulating table 2 is typical, but it can vary a great deal.

When assessing the significance of the chemical costs, it is useful to express them in terms of the ppm U3O8 in solution that pays for the chemical costs of a typical strength lixiviant. Accordingly, table 2 includes the ppm U3O8 required to pay for 3 g/l carbonate or bicarbonate, a typical concentration. Table 3 lists the ppm U3O8 required to pay for acid molar and normal concentrations equivalent to 5 g/l sulfuric acid. The costs of equivalent normalities are included because they are comparisons of the costs of obtaining a selected pH. Table 4 lists the ppm U3O8 required to pay for 0.3 g/l oxygen, which is a typical concentration, provided by each of the oxidizers. A value of $66/kg is assumed for U3O8.

The ppm U3O8 listed in tables 2-4 were calculated assuming no recycling of the lixiviants, and so are upper limits. Recycling was not included because it depends on site-specific factors. Discussions with leaching company personnel suggest that 60 to 90 pct of the lixiviant can be recycled at most sites. The ppm values can be compared with the 17 to 200 ppm U3O8 in the pregnant solutions from successful operations. The comparisons can help avoid incorrect conclusions. For example, one might infer that sodium bicarbonate should not be used because it costs twice as much as an equivalent concentration of dissolved carbon dioxide. However, when recycling is considered, the cost difference is equivalent to only about 1 ppm U3O8 and so will have less impact than a very small difference in leaching efficiency. Tables 2, 3, and 4 also summarize the advantages and disadvantages of the various alkaline and acid lixiviants and the oxidizers, respectively. More detailed discussion of each follows.

Alkaline (Carbonate Bicarbonate) Lixiviants

All the currently used alkaline lixiviants contain carbonate-bicarbonate that combines with uranium oxide to form a soluble anionic complex. The ratio of carbonate to bicarbonate can be raised by adding hydroxide and can be lowered by adding carbon dioxide. Leaching companies have tried carbonate-bicarbonate lixiviants having pH values from 6.5 to 10. The current trend appears to be to use neutral or only moderately alkaline lixiviants, avoiding pH values over 9 except for special circumstances. The lower pH values can result in somewhat slower dissolution of uranium, but they minimize the geochemical disturbance to the formation, thereby facilitating restoration and usually reducing plugging.

Compared with acids, alkaline lixiviants offer more selectivity for uranium (that is, dissolve less of the unwanted substances) and can be used in deposits , having a higher carbonate (for example, calcite) content. In deposits amenable to both alkaline and acid lixiviants, the alkaline lixiviants often recover somewhat less uranium.

Ammonium Carbonate-Bicarbonate

Ammonium carbonate-bicarbonate has been the most widely used lixiviant. Its advantages include fairly low cost and little damage to permeability. However, restoration regulations are causing companies to avoid ammonium lixiviants, especially in Wyoming. A Wyoming regulation, promulgated in 1980, requires that the post-restoration concentration of ammonium ions be no greater than 0.5 ppm if the preleach water was suitable for domestic use. Such restoration cannot be achieved at a reasonable cost. In fact, restoration of ammonium ion levels to below 50 ppm is usually expensive and requires extended flushing of the ore body. Companies evaluating ammonium carbonate-bicarbonate should carefully consider the restoration requirements that they will have to meet.

Ammonium carbonate-bicarbonate is usually made onsite by combining ammonia, water, and carbon dioxide. The chemical reactions are as follows:

2NH3 + H2O + CO2 → (NH4)2 CO3

NH3 + H2O + CO2 → NH4HCO3.

The cost of ammonia is about $0.10/lb not including transportation to the site. The cost of carbon dioxide can vary considerably, as will be explained in the section on carbon dioxide, depending on the quantity purchased per year. For many operations purchasing 500 to 2,000 tons per year, it will cost about $0.040/lb, so this figure was used in this report when calculating lixiviant costs. Ammonium carbonate then costs $11.38/kg-mole, $5.16/lb-mole, $0.12/kg, or $0.054/lb. Ammonium bicarbonate costs $7.63/kg-mole, $3.46/lb-mole, $0.097/kg, or $0.044/lb.

The concentration of ammonium carbonate-bicarbonate is usually 1 to 5 g/l, although high-strength high-pH lixiviants containing up to 10 g/l at pH 10 have been employed when attempting to prevent calcite dissolution. The lower strength lixiviants may dissolve uranium more slowly, but they made restoration less difficult and reduce total lixiviant consumption. The significant concentration is that of the anion. Ammonium carbonate is 62.5 wt-pct carbonate and ammonium bicarbonate is 77.3 wt- pct bicarbonate, so providing 3 g/l carbonate or bicarbonate requires 4.8 g/l ammonium carbonate or 3.9 g/l ammonium bicarbonate, respectively. These solutions cost 0.057¢/l (0.22¢/gal) and 0.038¢/l (0.14¢/gal), respectively.

When estimating total cost of the lixiviants, it would be misleading to multiply the number of liters of lixiviant expected to pass through the formation by the cost per liter. The fact that most of the lixiviant can be recycled by adding a relatively small amount of ammonium carbonate-bicarbonate lowers total lixiviant consumption considerably. However, the loading of ion-exchange sites can consume significant amounts of lixiviant. At many sites, the lixiviant cation is exchanged for calcium, thereby consuming carbonate through calcium carbonate precipitation. Thus, the exchange capacity can be an important factor in projecting costs. Judging from discussions with leaching company personnel, it would be reasonable to assume that about 80 pct of the lixiviant can be recycled after the exchange sites have been satisfied.

The reactions with uranium can be described as follows:

UO3 + H2O + 3(NH4)2 CO3 → (NH4)4 UO2(CO3)3 + 2NH4OH

UO3 + 2NH4HCO3 → (NH4)2 UO2(CO3)2 + H2O

UO3 + 2NH4HCO3 + (NH4)2CO3 → (NH4)4 UO2(CO3)3 + H2O

Sodium Carbonate-Bicarbonate

The use of sodium carbonate-bicarbonate is increasing, as companies seek alternatives to ammonium carbonate-bicarbonate. Its major drawback has been that it tends to swell smectite (montmorillonite) and plug the formation. Sodium has a high hydration energy, so water tends to enter clays where sodium is absorbed. It is monovalent, so it does not hold the clay layers together as tightly as calcium and magnesium do. Ammonium and potassium ions cause less swelling because they fit better between the clay layers and have a lower energy of hydration.

Sodium carbonate-bicarbonate can be used where the natural sodium content of the groundwater is high, or where the concentration of swelling clays is very low. For example, clay swelling by sodium will not be a problem where nearly all the clay is a nonswelling type such as kaolinite. Companies have had some success in keeping clay swelling to tolerable levels, even where those conditions are not satisfied, by keeping the pH near neutral. The reason has not been established, but it may be that fewer cation exchange sites are occupied by sodium as more are occupied by hydrogen. This technique is not always successful.

The advantages of sodium carbonate-bicarbonate in formations where it does not cause plugging include low cost and ease of restoration. A good example of its successful use is the OPI-Western joint venture pilot operation in Wyoming in 1979. Acceptable groundwater restoration standards were met after circulation of only six pore volumes of surface-treated water through the mined aquifer. The lixiviant just before the start of restoration contained 1.0 g/l sodium and 0.9 g/l bicarbonate.

Sodium carbonate lixiviant is usually made by dissolving soda ash (anhydrous sodium carbonate) or by bubbling carbon dioxide through caustic soda (sodium hydroxide solution). In Wyoming, soda ash is cheaper. In Texas, using caustic soda may be cheaper because of freight charges on soda ash shipped from Wyoming. Sodium carbonate from soda ash costs about $10.52/kg-mole, $4.77/lb- mole, $0.099/kg, or $0.045/lb, not including transportation to the site. It costs considerably more when made from caustic soda and carbon dioxide, $47.98/kg-mole, $21.76/lb-mole, $0.45/kg, or $0.21/lb.

Sodium bicarbonate made from soda ash and carbon dioxide costs $7.19/kg- mole, $3.27/lb-mole, $0.086/kg, or $0.039/lb. When made from sodium hydroxide and carbon dioxide, it costs $25.93/kg-mole, $11.76/lb-mole, $0.31/kg, or $0.14/lb. Sodium bicarbonate flakes or powder can be purchased for $22.22/kg- mole, $10.08/lb-mole, $0.26/k g, or $0.12/lb.

Sodium carbonate-bicarbonate can also be made from purified trona, marketed as sodium sesquicarbonate (Na2CO3 • NaHCO2 • 2H2). Trona is mined primarily in Wyoming and is processed to make soda ash. Even though less processing is required to make sodium sesquicarbonate, its cost per ton in 1980 was similar to that of soda ash because of the smaller market. Using the prices in table 1, sodium carbonate made by combining sodium sesquicarbonate and caustic soda costs $17.60/kg-mole, $7.98/lb-mole, $0.17/kg, or $0.075/lb. Sodium bicarbonate made by bubbling carbon dioxide through sodium sesquicarbonate solution costs $8.67/kg-mole, $3.93/lb-mole, $0.10/kg, or $0.047/lb. If a sufficient market develops, sodium sesquicarbonate may provide a cheaper source of sodium carbonate-bicarbonate. It dissolves more readily than soda ash, thereby reducing mixing costs. It would be very convenient for operators who wish to leach with a lixiviant containing equal molar concentrations of carbonate and bicarbonate because no carbon dioxide would have to be added.

The concentration of sodium carbonate-bicarbonate in a lixiviant is usually 1-5 g/l, just as for ammonium carbonate-bicarbonate. A pH up to 9 or 10 has been used, but the trend now appears to be to keep the pH near neutral to minimize clay swelling. Sodium carbonate is 56.6 wt-pct carbonate and sodium bicarbonate is 72.6 wt-pct bicarbonate, so providing 3 g/l carbonate or bicarbonate requires 5.3 g/l sodium carbonate or 4.1 g/l sodium bicarbonate, respectively. If made using soda ash, these solutions cost 0.053¢/l (0.20¢/gal) and 0.035¢/l (0.13¢/gal), respectively. If made using caustic soda, they cost 0.24¢/l (0.91¢/gal) and 0.13¢/l (0.48¢/gal), respectively. If made by combining sodium sesquicarbonate and caustic soda or carbon dioxide they cost 0.090¢/l (0.34¢/gal) and 0.041¢/l (0.16¢/gal), respectively. Thus, sodium carbonate-bicarbonate made using soda ash may be slightly cheaper than ammonium carbonate-bicarbonate. The difference in total chemical costs will be enhanced by the fact that less sodium will be absorbed on clays, so the consumption of a sodium lixiviant will be smaller than that of an ammonium lixiviant.

The chemical reactions of oxidized uranium with sodium carbonate-bicarbonate are similar to those with ammonium carbonate-bicarbonate. For example, the chemical equation corresponding to the equation under ammonium carbonate-bicarbonate is as follows:

UO3 + Na2CO3 + 2NaHCO3 → Na4 [UO2(CO3)3]+ H2O.

Potassium Carbonate-Bicarbonate

Potassium carbonate-bicarbonate is very attractive technically and environmentally. It does not swell clays as sodium does. Meeting restoration requirements should be less difficult than with ammonium lixiviants. Potassium is normally found in the preleach groundwater, so its presence after leaching does not change the potential uses of the water. However, its high cost has prevented its use.

Potassium carbonate can be made by bubbling carbon dioxide through caustic potash (potassium hydroxide solution). Made in this way, it costs $63.26 kg- mole, $28.69/lb-mole, $0.46/kg, or $0.21/lb. A more convenient but somewhat more expensive source is a liquid containing 47 pct potassium carbonate. It then costs 376.18/kg-mole, $34.55/lb-mole, $0.55/kg, or $0.25/lb. (All costs are on a 100 pct basis.) Potassium bicarbonate made by bubbling additional carbon dioxide through caustic potash costs $33.56/kg-mole, $15.22/lb-mole, $0.34/kg, or $0.15/lb. Potassium bicarbonate granules can be purchased for $30.90/kg-mole, $14.00/lb-mole, $0.31/kg, or $0.14/lb. If an operator wished to avoid handling carbon dioxide, the potassium bicarbonate granules could be combined with the 47 pct potassium carbonate in the proportion to give any desired pH.

A new technique developed by the University of Texas at Austin under Bureau of Mines research contract H0282016 promises to greatly reduce the cost of using potassium carbonate-bicarbonate and also reduce permeability loss from calcium carbonate precipitation. The technique involves flushing the formation with potassium chloride before injecting potassium carbonate- bicarbonate. This satisfies the cation exchange sites with potassium from relatively cheap potassium chloride, which costs about $4.92/kg-mole, $2.24/lb-mole, $0.066/kg, or $0.030/lb. In a laboratory column leaching experiment, the consumption of potassium carbonate following chloride preflush was only 17 pct of that without the chloride preflush. The preflush reduces permeability loss because soluble calcium chloride is formed instead of insoluble calcium carbonate. The calcium can be removed aboveground before injecting potassium carbonate. The problem of what to do with calcium aboveground is not negligible, especially since the calcium solution will contain some radium. However, the problem of a plugged formation can have a much more severe impact on a leaching operation than the calcium disposal. The chloride preflush is more fully described in another publication.

The desired pH range and concentration of carbonate-bicarbonate for potassium carbonate-bicarbonate should be similar to those for ammonium or sodium carbonate-bicarbonate. An operator may have the option of using a higher pH and concentration because the problems of high-pH, high-strength ammonium lixiviant creating restoration difficulties, and high-pH, high-strength sodium lixiviant creating clay swelling difficulties are both avoided. Potassium carbonate is 43.4 wt-pct carbonate and potassium bicarbonate is 60.9 wt-pct bicarbonate, so providing 3 g/l carbonate or bicarbonate requires 6.9 g/l potassium carbonate or 4.9 g/l potassium bicarbonate, respectively. If made using caustic potash, these solutions cost 0.32¢/l (1.20¢/gal) and 0.16¢/l (0.62¢/gal), respectively. The carbonate solution made from 47 pct potassium carbonate liquid costs 0.38¢/l (1.43¢/gal). The bicarbonate solution made from granules costs 0.15¢/l (0.57¢/gal).

The chemical reactions of oxidized uranium with potassium carbonate- bicarbonate are similar to those with ammonium carbonate-bicarbonate.

Carbon Dioxide

Lixiviants made by bubbling carbon dioxide through groundwaters from the formation to be leached offers technical, economic, and environmental advantages. It is the cheapest lixiviant on a molar basis. It creates the least geochemical disturbance in the formation and should not cause clay swelling. It should allow easier restoration than any other lixiviant. Its use is increasing; it is being used in at least two commercial operations.

This lixiviant does not leach effectively in all formations, however. Discussions with leaching companies indicate that it is successful in deposits having clean sands, but does not leach effectively if much organic carbon is present. Unpublished laboratory data indicated that it leached the Wyoming ore being tested somewhat slower than ammonium carbonate-bicarbonate. One leaching company employee felt that it can leach faster than ammonium carbonate-bicarbonate in some deposits and usually leaches almost as fast in amenable deposits.

The lixiviant produced by bubbling carbon dioxide through the groundwater should not be thought of as carbonic acid. Dissolved carbon dioxide is most effective where there is enough carbonate in the groundwater to serve as a buffer and form bicarbonate. Then the reactions are primarily of the form

CO2 + Na2CO3 + H2O → 2NaHCO3

and

CO2 + K2CO3 + H2O → 2KHCO3

rather than

CO2 + H2O → H2CO3

The pH of this lixiviant is usually 6.5 to 7.5.

The cost of carbon dioxide varies a great deal, depending primarily on the quantity purchased per year, and also on the distance from the seller and on local supply-demand factors. The prices listed by one seller in Texas are as follows:

These prices include delivery within 50 miles. Transportation charges add $10/ton for each additional 50 miles. The prices quoted by all sellers similarly depended on quantity purchased per year. However, for a given quantity, the price varied more with location than it did for most other chemicals.

To determine a typical price to use when calculating costs of lixiviants it is necessary to choose a typical consumption per year. For most operations, the ratio of carbon dioxide consumed to U3O8 produced will be between 5 and 10 on a weight basis. Thus, a large commercial operation producing 500,000 to 1 million lb U3O8 per year might consume 2,000 tons of carbon dioxide per year. Most commercial operations would consume over 500 tons per year. Discussions with carbon dioxide sellers and in situ leaching companies indicated that $80/ton would be a typical price for consumption of 1,000 to 2,000 tons per year at most locations in Wyoming and Texas. Using $80/ton, the cost of carbon dioxide is $3.88/kg-mole, $1.76/lb-mole, $0. 088/kg, or $0.040/lb.

Ideally, the molar concentration of bicarbonate in this lixiviant should be similar to that in ammonium or sodium bicarbonate. In practice, dissolving that much carbon dioxide without excessive lowering of the pH can seldom be done, so the concentration is usually lower. However, useful cost comparisons can be made assuming equivalent concentrations. Each mole of carbon dioxide can provide a mole of bicarbonate, assuming that all the carbon dioxide dissolves. Thus, the cost of solution made from carbon dioxide providing 3 g/l bicarbonate (2.2 g/l carbon dioxide) is only 0.019¢/l (0.073¢/gal). about half that of the second cheapest lixiviant. The difference in total chemical costs will be even greater because this lixiviant should have smaller consumption by cation exchange than any other.

To obtain the potential cost savings from using carbon dioxide, it is vital that an efficient sparging system producing very small bubbles be employed for dissolving the carbon dioxide in the lixiviant. A system that is not carefully designed and operated can easily waste half the carbon dioxide by allowing it to pass through the solution as large bubbles that escape before dissolving. This is especially true as the pH is brought down to near neutral. An additional cost factor is that storing carbon dioxide in bulk as a liquefied gas requires a special tank and equipment because it must be cooled during storage.

The chemical reactions with oxidized uranium are similar to those with ammonium bicarbonate.

Acid Lixiviants

Acid lixiviants are used much less than alkaline lixiviants. Many (perhaps most) deposits are not amenable to acid leaching because they contain too high a concentration of acid consumers, usually calcium carbonate. In amenable deposits, acid lixiviants generally recover somewhat more uranium than alkaline lixiviants do. They also dissolve more of most of the other elements, especially metals, so the total dissolved solids are higher. Sulfuric acid has been used successfully. Hydrochloric and nitric acids have been proposed for special applications, but are not being used.

Sulfuric Acid

Sulfuric acid leaches very effectively in amenable deposits. Its technical advantages and disadvantages both stem from the fact that it dissolves more gangue than alkaline lixiviants do. Sulfuric acid often can recover more uranium than alkaline lixiviants, based on published and unpublished laboratory data. The higher recovery may be due to more of the uranium being exposed to the lixiviant, as more of the gangue is dissolved.

The acid dissolves higher concentrations of some pollutants such as thorium and toxic metals, but less of radium and selenium. Restoration after sulfuric acid leaching at the Rocky Mountain Energy Co.’s Nine Mile operation in Wyoming required extensive flushing but was successful. It appears that restortion after sulfuric acid is more difficult than after sodium carbonate- bicarbonate, but easier than after ammonium carbonate-bicarbonate.

There is disagreement as to whether harmful channeling occurs with acid leaching. Theoretically, acid could dissolve the gangue to form pathways for the lixiviant that bypass most of the ore.

Concentrated sulfuric acid (96-98 pct) costs about $8.65/kg-mole, $3.92/lb-mole, $0.088/kg, or $0.040/lb. A pH of 1.8 to 2.0 is preferred. Unlike carbonate-bicarbonate lixiviants, the pH cannot be controlled independently of the concentration. Theoretically, obtaining a pH of 2 in distilled water would require only 0.5 g/l H2SO4. At the Rocky Mountain Energy Co.’s operation, about 5 g/l were required to obtain a pH of 2. A 5 g/l solution costs about 0.044¢/l (0.17¢/gal).

When considering the amenability of a deposit to acid leaching, both the cost of the consumed acid and the possible blockage from calcium sulfate and from carbon dioxide gas should be considered. Calculation of the cost of the acid consumed by calcite is straightforward if the average calcite concentration is known. For example, assume that the deposit is 1.0 pct calcite. Because 98 g H2SO4 is required to consume 100 g CaCO3, the calcite in each kilogram of ore will consume (0.01) (98/100) – 0.0098 kg H2SO4, costing 0.086¢. If the ore contains 0.05 pct recoverable uranium worth $66/kg, then 1 kg of ore yields uranium worth 3.3¢. Thus, if using acid increases uranium recovery by 3 pct, the value of the extra uranium recovered exceeds the cost of the acid consumed by the calcite in this example. This example suggests that a calcite concentration of 1 or 2 pct does not by itself lead to prohibitively high costs for acid consumption. Some acid will be consumed by other components in the ore, of course.

Blockage by calcium sulfate or carbon dioxide can occur if their solubility is exceeded. The problem can be minimized by increasing the acid strength slowly, so that the products are formed slowly and lixiviant can transport them to a production well without the solubility being exceeded. Operators familiar with acid leaching have stated that calcium sulfate is a more serious problem than carbon dioxide. Carbon dioxide forms only when the acid dissolves calcite or other carbonates. It has a high solubility under typical formation pressures. Calcium sulfate can form when calcium is removed from minerals by ion- exchange, in addition to forming when calcite is dissolved.

Although ionic strength affects solubilities, the published solubilities in water provide a guide as to the approximate conditions under which solubilities in lixiviants might be exceeded. The solubility of calcium sulfate in water at 18° C is 2.0 g/l (28, p. 249). To calculate the concentration of calcite in the ore that could produce a saturated solution of calcium sulfate by reacting with a stoichiometric amount of sulfuric acid, assume that the porosity of the ore is 33 pct and that the density of the rock particles is 2.6 g/cm³. Then, to produce saturation, 100 cm³ of ore must contain sufficient calcite to produce (0.33 l) (2.0 g/l) = 0.66 g of calcium sulfate. The corresponding amount of calcite is (0.66)(100/136) = 0.49 g, which gives a calcite concentration of (0.49)/(667)(2.6) or 0.03 pct. A greater than stoichiometric concentration of sulfuric acid would allow the solubility product of calcium sulfate to be exceeded with an even lower concentration of calcite. These calculations show that in nearly all formations, rapid introduction of strong acid could produce calcium sulfate greatly in excess of its solubi1ity.

Similar calculations show that more calcite is required to exceed the solubility of carbon dioxide, but blockage could be a problem if strong acid is rapidly injected, especially into shallow deposits. Consideration of published data (28, pp. 365-368) indicates that in the range of 170 to 680 ft of water absolute pressure (5 to 20 atmospheres) and 15° to 25° C, the solubility (S) in grams per liter of carbon dioxide in water can be approximated by

S = 9.6 + 0.044(P-170) – 0.0012P(T-15)

where P is the absolute pressure in feet; of water and T is the Celsius temperature. (Extrapolating outside the stated range of pressure and temperature will produce large errors because the experimental solubility is a nonlinear function of pressure and temperature.) Within the stated range, the lowest solubility is at 170 ft of water and 25° C, where S = 7.6 g/l and the highest is at 680 ft of water and 15°, where S = 32 g/l. Calculations similar to those for calcium sulfate show that when S = 7.6 g/l, the solubility of the carbon dioxide from calcite could be exceeded if the ore is greater than 0.3 pct calcite. When S = 32 g/l, the solubility could be exceeded if the ore is greater than 1.4 pct calcite.

Sulfuric acid reacts with oxidized uranium to form soluble anionic complexes of sulfate and uranium dioxide, according to the following equations.

UO3 + H2SO4 → UO2SO4 + H2O

UO2SO4 + H2SO4 → H2 (UO2 (SO4)2)

H2 (UO2 (SO4)2) + H2SO4 → H4 (UO2 (SO4)3)

Nitric and Hydrochloric Acids

Nitric and hydrochloric acids are not being used for in situ leaching but have been proposed for leaching ore that cannot be leached by other lixiviants. Citric acid offers the advantage of being an oxidizer, and both acids are said to leach more effectively than sulfuric acid in ores containing much shale, clay, or carbonaceous material. A disadvantage is that they do not form anionic complexes with uranium. Thus, while uranium can be extracted from sulfuric acid with highly selective anionic exchange resins, uranium must be extracted from nitric or lydrochloric acids with less selective cationic exchange resins (19, p. 60).

They have environmental disadvantages. They share the nonselectivity of sulfuric acid and in addition will readily dissolve radium. Radium nitrate and chloride are much more soluble than radium sulfate or bicarbonate. In addition, nitric acid will almost certainly be subject to more restrictive restoration requirements than sulfuric or hydrochloric acid. The EPA drinking water standards for sulfate and chloride are both 250 ppm, but the standard for nitrate nitrogen is only 10 ppm. (The limit is expressed in terms of nitrogen attributable to nitrate because standard tests measure nitrogen; 10 ppm nitrogen from nitrate is equivalent to 44 ppm nitrate.)

Comparing the cost of nitric and hydrochloric acids with that of sulfuric acid is complicated by the difference in reaction mechanisms. Perhaps the most useful comparison is on the basis of normality, which is equivalent to comparing the costs for producing a given pH. Using that basis, nitric and hydrochloric acids cost; about twice as much as sulphuric acid. Concentrated nitric acid (70 pct) costs about $8.33/kg-mole, 3.78/lb-mole, $0.13/kg, or $0.060/lb on 100 pct HNO3 basis. Concentrated hydrochloric acid (37 pct) costs about 0.035/lb. and so costs $7.61 kg-mole, 3.45/lb-mole, $0.21/kg, or $0.095/lb on 100-pct basis. Solutions of 0.102 N the same normality as 5 g/l sulfuric acid nitric and hydrochloric acids cost 0.085¢/l (0.32¢/gal) and 0.078¢/l 0.29¢/gal), respectively. The total chemical cost when using nitric acid may be reduced by savings in oxidizer.

Oxidizers

The oxidizers currently used for in situ uranium leaching are oxygen and hydrogen peroxide. Chlorates have been used and are discussed for comparison.

Oxygen

Oxygen is used at most commercial scale operations because of its low cost. It is usually shipped and stored as a liquid in bulk, although pilot operations have used compressed oxygen in bottles. In bulk, it costs about $3.40/kg-mole, $1.55/lb-mole, $0.11/kg, or $0.048/lb in Texas and $5.10/kg-mole, $2.33/lb-mole, $0.16/kg, and $0.073/lb in Wyoming. Typical concentrations in lixiviant are 0.1 to 0.3 g/l. In Texas 0.3 g/l costs 0. 0033¢/l ( 0.012¢/gal), and in Wyoming costs 0. 0048¢/l (0.018¢/gal). Although this is a lower cost per liter than the other components in the lixiviant, the total cost of the oxidizer can be greater because so little of it is recycled.

The disadvantages of oxygen are caused by its limited solubility. Litz reported that the solubility (S) in grams per liter of oxygen in lixiviants could be approximated by

S = 0.064 p/33.5 + T (1.107 – 0.07 log P).

where P is the absolute pressure in feet of water and T is the Celsius temperature. The formula indicates that the solubility under atmospheric pressure (33.9 ft of water) at 20° C is only 0.04 g/l, and that 300 ft of water is required to produce a solubility of 0.3 g/l. Thus, oxygen must be injected under pressure to obtain sufficient solubility. It is usually injected downhole at the leaching depth, thereby requiring a separate oxygen line and sparger for each injection well.

An additional disadvantage is the risk of bubbles forming and causing gas phase blockage. Bubbles may form even when the solubility is not exceeded because thorough mixing of the oxygen and lixiviant is not instantaneous. Small bubbles can form in and close to an injection well. These bubbles do not pass between sand grains as lixiviant flows into the formation, so the bubbles can build up and cause some blockage even when the theoretical solubility is not exceeded. An efficient sparger that produces very small bubbles helps to minimize such blockage.

Hydrogen Peroxide

Hydrogen peroxide has been used for the majority of pilot scale operations. It is convenient because it can be added to the lixiviant at one central point under atmospheric pressure instead of requiring separate lines to all injection wells. A higher concentration of hydrogen peroxide than oxygen can be injected without causing gas phase blockage. This advantage is especially important for shallow deposits.

Studies have indicated that hydrogen peroxide oxidizes more effectively than oxygen. However, there is disagreement as to the extent to which this advantage assists in the field. Most investigators believe hydrogen peroxide decomposes to oxygen so rapidly that it is not superior to an equivalent amount of oxygen. A differing opinion is that even though most of the hydrogen peroxide decomposes, much of the ore is contacted by enough hydrogen peroxide to affect the leaching rate.

The decomposition is slower in acid than in alkaline lixiviants. Hydrogen peroxide is much more stable in acid than in alkaline solutions of the same purity. Researchers familiar with hydrogen peroxide stated that the effect of alkalinity would be greater than the catalyzing effect of the higher concentration of dissolved metals in acid lixiviants.

Hydrogen peroxide for lixiviants is purchased and transported in solutions of either 50 or 70 pct H2O2. Because of requirements of the National Fire Protection Association, it is usually stored onsite at less than 51 pct. Shipping it as a 70 pct solution can save some freight charges, but then facilities for diluting the solution with high quality water must be provided.

In August 1980, 50 pct hydrogen per oxide cost 25.75¢/lb and 70 pct cost 36.00¢/lb, f.o.b., 3,500 gal minimum. The freight cost to most locations would be 3¢/lb to 5¢/lb on an “as-is” basis. This report uses 58¢/lb on a 100-, pct basis as a representative price. Then the cost of hydrogen peroxide is $43.48/kg-mole, $19.72/lb-mole, $1.28/kg, or $0.58/lb on a 100-pct basis. Each mole of hydrogen peroxide provides half a mole of oxygen, and each kilogram of hydrogen peroxide provides (16/34) kg of oxygen, so the effective cost of oxygen purchased as hydrogen peroxide is $86.96/kg-mole, $39.44/lb-mole, $2.72/kg, or $1.23/lb. Typical concentrations of hydrogen peroxide in lixiviants are 0.3 to 1.0 g/l, which would provide the same oxygen as 0.14 to 0.47 g/l O2. The cost of 0.3 g/l oxygen supplied by hydrogen peroxide is 0.082¢/l (0.31¢/gal).

The high cost of hydrogen peroxide, coupled with the fact that little can be recycled, results in it being the major chemical cost at most operations using it. To illustrate, if the hydrogen peroxide costs 0.082¢/l of lixiviant and uranium is worth $66/kg, then 12 ppm U3O8 in the pregnant lixiviant is required to pay for the oxidizer.

Chlorates

Sodium chlorate has been used as an oxidizer, and potassium chlorate could be used. Their cost is between those of hydrogen peroxide and oxygen. Sodium chlorate costs about $46.97/kg- mole, $21.30/lb-mole, $0.44/k g, or $0.20/lb. Potassium chlorate costs only 4 pct more on a molar basis. Considering the fact that each molecule of sodium chlorate provides 3 atoms of oxygen shows that the cost of oxygen from this source is $31.31/kg-mole, $14.20/lb-mole, 0.98/kg, or $0.44/lb. Providing 0.3 g/l oxygen would cost 0.029¢/l (0.11¢/gal). The total cost of chlorate relative to the other oxidizers may be reduced somewhat by the ability to recycle unconsumed chlorate, whereas unconsumed oxygen goes out of solution in the surface plant and is lost.

Chlorates are effective oxidizers in acid solutions that contain iron (19, p. 64). Apparently, the chlorates oxidize iron and the iron oxidizes the uranium. However, neither sodium nor potassium chlorate oxidizes efficiently in carbonate-bicarbonate lixiviants (19, p. 104). One researcher, discussing unpublished experiments, stated that chlorates in carbonate-bicarbonate lixiviants oxidized uranium to some extent but were less effective than oxygen.

Other disadvantages result from the buildup of chloride as the lixiviant is recycled repeatedly. The chloride increases the corrosiveness of the lixiviant and decreases the uranium loading capacity of ion-exchange resin. Some companies believe that the chloride would make restoration more difficult. The sodium in sodium chlorate could reduce permeability. These problems do not completely rule out the use of chlorates. They could be advantageous in very shallow deposits, for example, especially with acid lixiviants.

Methods of Testing Lixiviants

The costs, advantages, and disadvantages previously presented provide only a general guide for lixiviant selection. To determine the suitability for a specific deposit, thorough laboratory and field testing is necessary.

Laboratory Tests

Both batch leach tests (sometimes called agitation leach tests) and column leach tests are used in selecting the lixiviant. Batch leach tests consist of placing the ore and lixiviant in a container, often a sealed flask, and gently agitating them. Column leach tests consist of passing lixiviant through a column packed with ore.

Batch Leaching

Although batch leaching tests do not simulate downhole conditions, they do provide useful relative data. They show the relative rate and amount of uranium extraction with tested lixiviants and can give an indication of lixiviant and oxidant consumption.

The ore sample must be large enough to allow solution samples to be withdrawn without unduly disturbing the geochemistry and to provide sufficient dissolved uranium for analysis. Typical ore sample sizes are 100 to 500 g, with a liquid-solid volume ratio of three.

The variables are usually type of lixiviant (for example, sodium or ammonium bicarbonate), pH (for example, 7, 8, and 9), carbonate or bicarbonate concentration (for example 1, 3, and 5 g/l), and oxidizer concentration (for example, equivalent to 0.1, 0.3, and 0.5 g/l). When performing tests to select a lixiviant for a particular site, the solutions should be made from groundwater from the formation to be leached, not from distilled water. (Simulated groundwater made by adding measured quantities of chemicals to distilled water may be preferred for general studies of reaction mechanisms.) The ore should be blended to ensure homogeneity, but not ground.

The measured observables are usually pH and concentrations of uranium, carbonate and/or bicarbonate, oxidizer, and site-specific elements such as vanadium that may be present in sufficient concentration to interfere. Some experimenters measure Eh, others feel that Eh measurements in this type of experiment are not meaningful. Typical sampling times are 0, 1, 4, 8, 12, and 24 hours after the start of the test and once per day until equilibrium is reached, usually in 3 days or less. The uranium in the ore should be measured before and after leaching and a material balance made to check the validity of the measurements. There is disagreement among experimenters as to whether fresh lixiviant should be added to replace lixiviant withdrawn for sampling. Either method can give good results. Replacing the lixiviant allows a smaller sample to be used but complicates the material balance.

Measuring the consumption of oxygen or hydrogen peroxide requires sealed experimental equipment and a method for withdrawing samples without losing oxygen. Some experimenters favor a chlorate oxidizer to avoid the need for pressurization. The consumption of chlorate can be determined by measuring the increase in chloride. Because the oxidizing effectiveness of chlorate may be different from that of oxygen, the consumption of chlorate will not necessarily be equal to that of oxygen. Nevertheless, oxidation tests with chlorates can be a useful and economical method of comparing the amounts of oxidizer consumed by different samples.

Obtaining meaningful results from oxidizer consumption tests requires care. The results of a careless experiment can range from much less to much more than the correct value. If the ore is inadvertently oxidized before being leached, then the results will be too low. Avoiding preoxidation of the ore requires that the ore be protected as soon as it is cored until it is placed in the leaching flask. The ore should be sealed in air-tight cylinders in nitrogen or frozen in dry ice. Sealing it in plastic is not adequate. Wrapping it in foil and sealing it in wax may be adequate if done carefully and the storage time is short. A plexiglass cylinder that can be flushed with nitrogen after inserting the core and then sealed airtight for transporting has worked well for the Bureau of Mines. If the ore is not preoxidized, then the results will provide an upper limit to the oxidizer consumption. The results are an upper limit because the oxidizing reactions will be more complete in an agitated flask with a liquid-solid volume ratio of three than in actual in situ leaching conditions. Some commercial service laboratories feel they have reliable methods of predicting what the field consumption of oxygen will be. Those methods are proprietary.

X-ray diffraction measurements of clay swelling can indicate which lixiviants cause the most swelling and hence allow qualitative predictions of permeability loss. Usually, however, permeability loss is studied with column leaching experiments.

Column Leaching

Column leaching tests simulate field conditions more closely than batch tests, but caution must still be used when extrapolating from laboratory to field. The contact between ore and lixiviant is more complete than in actual in situ leaching. Therefore, the measured consumption of lixiviant and oxidizer and the extraction of uranium should be viewed as upper limits of what might be expected in the field.

Column leach tests can indicate permeability losses, but to obtain meaningful results, water from the formation should be used and the ore should be disaggregated and blended. Making lixiviants by adding only the primary components (for example, ammonium carbonate) to distilled water will cause misleadingly large permeability losses during column leaching. Attempts to use intact cores in hopes of better simulating downhole conditions have not been satisfactory. Meaningful comparisons of lixiviants require similar cores, but cores vary considerably in permeability and uranium content. Further, the lixiviant flows through the core in the ore’s vertical direction but flows horizontally in situ. Therefore, a thin clay lens can drastically reduce the measured permeability of a core but have very little effect on the in situ horizontal permeability. The ore should be disaggregated gently to avoid breaking grains and thoroughly blended. Most sandstone uranium ore can be disaggregated easily.

To obtain meaningful data on the effect and the consumption of oxidizer, the ore must be protected from preoxidation, just as for batch tests. The type of oxidizer and the formation of bubbles can affect permeability, so ideally the column would be pressurized to simulate downhole conditions. Using sodium chlorate in these tests could reduce permeability.

There is disagreement as to the best column configuration. For nonpressurized systems, the authors prefer a vertical column with lixiviant entering at the bottom. Admitting the lixiviant at the bottom maximizes the speed at which any air or other gases in the core at the beginning of leaching will be swept out. Horizontal tubes present the risk that ore will slump slightly, thereby creating a high permeability channel between the top of the ore and the inside of the column. Some experimenters favor horizontal columns. They are more convenient, especially if long columns are used.

There is also disagreement as to the choice of diameter and length. Theoretically, the most reliable results will be obtained with the widest and longest columns. Greater length better simulates chromatographic effects (differing migration speeds of different ions). Greater diameter minimizes edge effects; in particular, the tendency for the flow resistance to be smaller between the ore and the column than in the ore. In practice, the amount of core is usually quite limited and the number of tests to be run quite large, so rather small columns must be used. A column 5 cm in diameter and 120 cm long appears to be a good compromise unless chromatographic effects are of special interest. Then a 2.5 cm diameter sectional tube with a total length of 300 to 500 cm would provide more information. In one unpublished experiment, 1.6-cm-ID horizontal columns were used with good results by employing careful packing techniques. There is some disagreement as to the need for simulating chromatographic effects when selecting a lixiviant.

For testing oxidizer consumption and response to oxygen under downhole conditions, a smaller pressure cell can be used. Hassler pressurized cells 1.75 cm in diameter and 10 cm long have given good results at the University of Texas. With such small cells, it is important that the sides of the cell press tightly against the ore to avoid large edge effects.

Typical variables, observables, and sampling times are similar to those for batch tests. In addition, permeability can be measured either by maintaining a constant pressure and measuring the flow rate or maintaining a constant flow with a positive displacement pump and measuring the pressure difference across the column. The second method is generally preferred. The speed of the lixiviant should be similar to that in the field, often about 3 m/day. Excess flow rate can lead to channeling, especially with horizontal columns. The pump should not pulse the flow, as excessive pulses can increase the movement of fines and reduce permeability.

Pilot Field Tests

A pilot-scale field test is essential before starting commercial operation. It is needed not only as an aid to making the final choice of lixiviant, but also for evaluating well construction and completion techniques and for demonstrating restoration procedures.

Types of Tests

Pilot-scale tests can be divided into two classifications. The first type is called push-pull, or huff-and-puff. The lixiviant is injected and recovered from the same well. The second type can be called flow-through. The lixiviant is injected, flows through the formation, and is recovered from other wells. Consultants disagree as to the value of push-pull tests. Some believe that these tests provide a reliable and relatively economical method of evaluating the lixiviant under field conditions. Professor Robert Schechter of the University of Texas at Austin determined that one well serving first for injection and later for production, together with two observation wells, would provide sufficient data for a field evaluation of the chloride preflush previously described. The well pattern is an L, with the injection-production well at the corner.

A push-pull test can be misleading, especially one with no observation wells, if adsorption is not considered. Adsorbed species do not move as far from the well as would be calculated if only fluid volume were considered. In one field test, most of the ammonium ions remained within one-fourth of the ore contacted by the fluid. In that case, the number of pore volumes of groundwater sweeping required for restoration of groundwater quality would have been underestimated by a factor of 4 if adsorption had been neglected.

Flow-through tests resemble most commercial operations more closely than do push-pull tests. A commercial operation may be forced to use a push-pull method of leaching if numerous clay lenses and pinchouts prevent good hydrologic communication between wells, but the flow through method is preferred and far more common.

Most operators believe that a flow-through pilot test before beginning a flow-through commercial operation is advisable even if a push-pull test has been used for a preliminary evaluation. A common configuration is a five-spot pattern, injecting at the corners and producing at the center, because usually there is more fluid resistance to injection than to production. The results of flow-through tests can be quite different from those of push-pull tests. In unpublished flow-through tests at one site, increases in uranium concentration and Eh lagged behind the changes in the other components. Over three pore volumes of lixiviant were injected before the uranium concentration and Eh rose. However, in a previous push-pull test at the same site, that lag was not observed. Therefore, the push-pull test resulted in an overly optimistic prediction about the ease of leaching.

Problems and Solutions

Because problems can occur that render a pilot field test useless as a guide in making the final choice of lixiviant, these problems are briefly described below. Problems that have occurred in past tests include the following:

1. Leaking casings.
2. Clogging of well screens or nearby formation.
3. Clogging of formation near a production well.
4. Reprecipitation of uranium.

The most serious leaks are caused by well completion tools gouging holes in PVC casing. For example, leaks have been caused by underreamers when the tool has been pushed down or pulled up with the blade not fully retracted. Sand can interfere with the retracting of the blade. Damage can also be caused when any tool that fits tightly is forced past casing that is somewhat curved because of hole deviation. Smaller leaks can occur where screws penetrate all the way through a joint, or where a joint is leaking because of improper gluing. Holding a glued joint immobile with screws may be advisable, but the screws should not penetrate all the way through the joint. The Bureau has published other information concerning casings.

Clogging of well screens or the nearby formation can occur before lixiviant is injected as a result of wall cake left from drilling. Techniques that produce acceptable exploration holes do not necessarily produce acceptable injection wells. Care should be taken to avoid drilling fluids that leave a thick wall cake. If underreaming or perforating is used for well completion, then the problems from the well cake are minimized. However, if conventional well screens are used, then polymer drilling fluids or bentonite with a polymer additive generally allow better injectivity than just bentonite. The Bureau has prepared more detailed publications on this topic.

Clogging of the formation near an injection well screen can occur upon the injection of lixiviant incompatible with the formation or groundwater. An example is clay swelling caused by injecting sodium carbonate-bicarbonate into a formation that has a high concentration of swelling clay and where the preleach groundwater is low in sodium. Usually, the incompatibility can be identified in laboratory column leaching tests.

Clogging of the formation near a production well can be caused by precipitation of calcium carbonate. The pressure near a production well is lower than in the rest of the formation, so some of the carbon dioxide can come out of saturated solution. Apart from the gas phase blockage, the lowering of dissolved carbon dioxide can contribute to clogging by raising the pH, thereby decreasing the solubility of calcium carbonate and perhaps causing its precipitation. The chloride preflush appears to be a promising method of minimizing this problem. An alternative approach is to begin injection with a high pH lixiviant, to precipitate calcium in place so it cannot become concentrated near the production well. This approach has the disadvantage that uranium will be coprecipitated and become less accessible to the lixiviant.

Clogging of production well screens or the nearby formation can also occur as a result of fines migration. Periodic vigorous well development such as air-lifting may be required.

Reprecipitation of uranium can occur if the strength of the lixiviant (pH, Eh, or concentration) changes excessively as the lixiviant moves from injection to production well. This problem can be minimized by increasing the strength of the lixiviant gradually as leaching begins and by selecting a lixiviant compatible with the formation. The problem can be reduced by reducing well spacing, but that is an expensive solution.

Geochemical Models

The geochemistry associated with in situ leaching of uranium is extremely complex and is difficult to characterize without the use of a computer. Geochemical computer models can be divided into two major categories. The first type of model, the equilibrium approach, is useful for describing numerous interactions of a complex system of aqueous species and solid phases. Equilibrium programs generally require a complete chemical analysis of a solution including field measurements of Eh, pH, and temperature. The output of this type of model can be used to determine the reactions that are likely to occur within a given system but gives no information concerning the rates of the reactions. Therefore, equilibrium models cannot adequately describe time dependent reactions, which are affected by fluid velocity and dispersion.

The second type of model, the kinetic model, simulates the progress of kinetic reactions as a function of time and location. Kinetic models are coupled with hydrology models and thus account for formation factors such as porosity, permeability, and dispersivity. A hydrology model adaptable to such coupling has been developed by Schmidt. The main disadvantage of kinetic models is that they are limited to relatively few chemical reactions.

In addition to hydrologic data for the aquifer, kinetic models require a large amount of chemical information pertinent to the reactions being modeled. Chemical data include reaction rate coefficients, which are determined by extensive laboratory experiments. Once the pertinent physical and chemical properties have been established, an operator can determine the effect of well pattern and spacing, pumping and injection rate, and injected oxygen concentration on uranium production. Because kinetic models cannot be used in selecting a suitable lixiviant and oxidant unless pertinent reaction rates are first determined through laboratory experiments, this report will concentrate on equilibrium modeling.

Numerous equilibrium models have been developed in recent years for the purpose of modeling geochemistry of natural water. A useful summary and comparison of these models has been published by Nordstrom. Although these models can be applied to a wide range of hydrogeochemical problems, most of them are unable to model reactions associated with in situ leaching of uranium roll-front deposits. As of 1980, an updated version of WATEQF is the most suitable program for describing the aqueous geo-chemistry of uranium.

The original version of this program (WATEQ) was written in PL/l and was published by Truesdell and Jones in 1973. In 1976, Plummer published a revised version, WATEQF, in the programming language FORTRAN IV. This revised version recently has been enlarged by Runnells to include many dissolved species and solid compounds of uranium, vanadium, and molybdenum.

Detailed explanations of theory and operation for WATEQ and WATEQF are provided by Truesdell and Jones; Plummer (21); and Lueck. Briefly, the purpose of WATEQF is to solve a large number of simultaneous equations that describe all of the known equilibrium reactions that may occur in a given solution. The program requires as input a relatively complete chemical analysis of the solution of interest. This analysis includes temperature, Eh, pH, concentrations of all major cations and anions, and concentrations of the minor and trace elements that are relevant to the investigation. These data are summarized on the first page of output (table A-1). In addition to the chemical analysis of the solution, a table of thermodynamic data for all reactions modeled by the program must also be read into the computer. This feature enables the user to update thermodynamic data without having to modify the program itself.

After the data are read, the program adjusts the equilibrium constants for temperature effects using the Van’t Hoff equation or a user defined analytical expression. It then uses an iterative procedure to determine the activities and concentrations of all aqueous species. At the start of each iteration, activity coefficients are calculated from the Davies equation, the extended Debye-Huckel equation, or the Guntelberg equation. Then a back substitution procedure is used to solve mass action and mass balance equations for dissolved species. When the computed concentrations of major anions agree with the respective analytical values within a specified tolerance, iteration stops and the final activities are retained for computing the stabilities of solid phases.

The second page of output (table A-2) lists the differences between computed and analytical values of major anions for each iteration. It also describes the solution in terms of important parameters such as ionic strength, total dissolved solids, pH, Eh, electron activity, and partial pressures of dissolved gases, (O2, CO2, and CH4). The concentrations, activity coefficients, and activities of aqueous species are tabulated on pages 3-6 of the output (table A-3). This distribution of species list yields valuable information on the speciation of dissolved constituents. For example, the sample output indicates that about 59 pct of the total molality of dissolved calcium is free ionic Ca++; about 40 pct is aqueous CaSO4; and about 1 pct is CaHCO3.

After the distribution of aqueous species is determined, WATEQF computes the state of saturation of the solution with respect to various minerals and amorphous solid compounds. The diagnostics that describe the stabilities of the solid phases constitute the final section of output (table A-4). When interpreting these diagnostics, it is important to remember that each phase in the first column actually represents a reaction involving that phase with the appropriate aqueous species. These reactions are included in several references describing WATEQ or WATEQF.

To determine the stability of a solid phase within a solution, it is necessary to compare the activity product of the ions involved in the appropriate reaction (ion activity product, or i.a.p.) to the thermodynamic equilibrium constant (K) for that reaction. This comparison can be expressed as a ratio (i.a.p./K) or as a log of that ratio (log 10 (i.a.p./K)). The log of the ratio is termed the saturation index (S.I.) and is a useful indicator of a mineral’s stability in a solution. Another useful indicator is the Gibb’s free energy of reaction (∆Gr), which is the amount of energy (expressed in kilo-calories) that must be supplied to allow the reaction to proceed. The following guidelines should be used in interpreting saturation indices and free energies of reaction:

1. S.I. and ∆Gr less than zero: The reaction will proceed spontaneously, although the rate may be extremely slow. Since the reactions modeled by this program describe the dissolution of solid phases, this would indicate that the solution is undersaturated with respect to the solid and that the solid (if present) would go into solution. A highly negative S.I. or ∆Gr may indicate that a particular mineral does not exist in the system.
2. S.I. and ∆Gr close to zero: The reaction is at equilibrium.
3. S.I. and ∆Gr greater than zero: The safest interpretation of this condition is that the solution is supersaturated with respect to a given mineral and that the mineral is stable in that particular aqueous environment. This does not indicate precipitation. The formal interpretation of a positive ∆Gr is that the reaction cannot proceed unless energy is supplied from an external source. In some cases, positive values suggest that a particular solid may precipitate from solution, but caution should be used in this type of interpretation. It is possible for solutions to remain supersaturated with respect to some minerals for long periods of time without precipitation. Some minerals are not known to precipitate from solution at all.

The modified version of WATEQF has had many applications related to the genesis of uranium and vanadium deposits, hydrogeochemical exploration for uranium, and in situ leaching uranium roll front deposits. In every application, it allowed interpretations that could not have been made from chemical data alone. Application of WATEQF to in situ leaching of uranium can provide operators with valuable information on the speciation of uranium and vanadium complexes in solution; the solubility of uranium and vanadium minerals; the formation of gaseous oxygen and carbon dioxide, which could reduce permeability, and the precipitation of solid phases, which could reduce permeability or remove uranium from solution.

WATEQF is more useful for predicting changes in solubility than for predicting the solubility itself. This is because supersaturations of 10 to 100 are often found, so the calculated solubilities are not reliable predictors of concentrations.

As of 1980, at least two companies are using WATEQF (as modified by Runnels) to assist in determining how the lixiviant composition should be changed to improve leaching. One company uses it to help select the most cost-effective lixiviant composition for dissolving the uranium minerals. The cost of a solution providing a given pH and Eh can be estimated, and the solubilities of the minerals can be predicted with WATEQF. Thus, for a given lixiviant cost, the program can help select the combination of pH and Eh maximizing solubility. Judgment is still required for balancing cost versus solubility, however.

Another company uses WATEQF to predict whether solubilities will increase or decrease with changes in carbonate concentration, pH, or Eh. It is especially helpful in determining the probable cause and suggesting a cure when pilot tests are yielding much less uranium than expected. This company also uses WATEQF to predict the relative amounts of uranium species. Uranium as a monocarbonate complex will not load on anionic exchange resins and so is undesirable. WATEQF predicts what fraction will be in monocarbonate, dicarbonate, and tricarbonate complexes. The program has also been used to predict fouling from minerals precipitating in pipes and to study restoration geochemistry.

An additional application of WATEQF can be demonstrated through the following example supplied by Professor Donald Runnels of the University of Colorado. At the 1979 Symposium on the Grants Mineral Belt, chemical evidence was presented in support of the existence of a calcium-uranium-phosphate association in uranium ores (personal communication with Dr. Runnels, Sept. 22, 1980). Results of geochemical modeling suggest a potentially important role of dissolved phosphate in controlling the solubility of this association in acid lixiviants and in affecting the recovery of uranium from in situ operations.

The data for this example consist of complete chemical analyses of solutions that were collected from observation wells at the Rocky Mountain Energy Co.’s Mine Mile Lake Site, Wyoming. Although the concentration of dissolved phosphate never exceeds 13 ppm, many of the solutions appear to be supersaturated with respect to the mineral ningyoite, (U,Ca)2(PO4)2 · 1 – 2H2O. Ningyoite and other solid compounds of uranium and phosphate appear to be insoluble under most conditions of the sulfuric acid lixiviant, except at very high Eh. This suggests that an operator might be able to increase the efficiency of the leaching process by reducing the concentration of phosphate in the lixiviant.

At the present time, WATEQF seems to be the most suitable equilibrium model for problems related to in situ leaching of uranium because it includes reactions for minerals associated with uranium deposits. However, it is limited by its inability to model mass transfer, adsorption, or solid-solution reactions. Other models having one or more of these capabilities include REDEQL2 and EQ3/6. These models, however, must be modified to include reactions associated with in situ leaching of uranium.

Summary

The selection of a lixiviant proceeds through three phases. First, general advantages and disadvantages of lixiviants are considered. These general considerations include technical, economic, and environmental factors. Currently, restoration of groundwater quality are causing a movement away from ammonium carbonate-bicarbonate toward sodium bicarbonate and dissolved carbon dioxide. The cost of the oxidizer should be carefully considered, because it can exceed the cost of all the other chemicals.

Second, lixiviants that seem promising are tested with ore (cores) from the site to be leached. Laboratory batch and column leaching experiments measure leaching efficiency, consumption, and effect on permeability. These tests can be misleading if not conducted and interpreted with care.

Third, a pilot-scale field test is conducted. Proper well construction is vital to the success of this test. The test can be either the push-pull or flow through type. The former is cheaper, but the latter simulates most commercial operations more closely.

Computer modelling of the geochemistry can aid in the selection process. Such models are being used by at least two leaching companies to predict changes in solubilities associated with possible changes in lixiviant composition.