Any liquid HLW that has been generated either in national programs or commercially is presently contained in tanks, some of them still with artificial cooling. In the United States, there is more than 200,000 m3 of HLW from defense programs. Most of it is not as highly radioactive as waste solutions from commercial reprocessing will be. But the volume cor­responds to more than 200,000 MT of commercial fuel or about 7 X 106 MWe-years of nuclear energy, and this is about twice as much as commercial reprocessing in the United States will generate up to the year 2010.

If the storage conditions are properly chosen (which has not always been so in the past), tank storage is a perfectly safe technique. This has been proved by almost three decades of experience in the United States and Great Britain. Proper conditions mean storage of acidic waste in stainless steel tanks to minimize corrosion and to avoid formation of sludges.

However, liquid storage cannot be the ultimate solution to HLW management. A surface storage system, which is relatively vulnerable, with a large and steadily increasing inventory of

liquid highly radioactive material, means a potential hazard that is not acceptable in the long run. Moreover, the high degree of maintenance and surveillance that is required would have to be provided for an extremely long period of time.

The general concept how to achieve permanent isolation of radioactive waste from the environment without human activity being required in the long run has been well established internationally. First, the waste solution will be immobilized by solidification after some time of intermediate storage as a liquid. The solid product should have a high degree of stability to ensure that the radionuclides contained in the waste remain immobile for a long time. With present knowledge, materials that meet this requirement are glasses and ceramics. The solidified waste contained in a steel canister will then be stored irretrievably in a stable geologic formation deep underground.

The most likely path for such radioactive material to find access to the environment is transport by groundwater. Three barriers are provided in the above concept to prevent this: (1) the inherent stability of the solidification product, particularly against attack of water, and the durability of the waste container; (2) the stability of the geologic containment, which prevents water from providing means of communication between waste and circulating groundwater; (3) the large sorption capacity provided by rock and soil on the long distance from the underground repository to the surface combined with a low groundwater velocity.

In various countries this concept may materialize in different ways. Alternative solidifica­tion processes and solidification products have been developed, and some of them are in or close to the prototype stage. As for final disposal, some countries have made a decision in favor of rock salt as the most suitable geologic formation. In West Germany all effort is focused on salt domes. In the United States much effort is devoted to bedded salt, although alternatives such as granite and shale are under consideration as well. Other countries are investigating those geologic formations to which they have easy access. Because of strong public opposition, most governments have decided not to take any foreign waste for final disposal.

Some more basic alternatives of waste disposal are studied in some countries. Seabed disposal is one that attracts some attention in the United States and in Great Britain. The idea is to drill holes into the bottom of the deep sea that will be self-sealing after they have been filled with solidified waste. The water covering the disposal site will act as an additional barrier between waste and human beings. An obvious advantage of the seabed option is the very remote location. An obvious disadvantage is the complex and not at ah conventional technology required.

Another approach to HLW management takes into account that after some 400 years the radiotoxicity of the waste begins to be determined by the actinides. Consequently, if the actinides can be removed from the waste to a sufficiently high degree, the integrity of the geologic containment is required only for hundreds of years rather than for thousands. However, it will be difficult to achieve the high degree of separation required, and it will even be difficult to deal effectively with the separated actinides. The very substantial additional effort to separate actinides can be justified only if the separated actinides can be eliminated. Presently no way seems feasible other than transmutation by neutron bombardment to yield shorter-lived fission-product nuclides. As yet, the technology for both, actinide separation and actinide transmutation, is not available. Moreover, it seems doubtful whether there is a sound incentive to undertake the extra effort required.

Occasionally some more exotic alternatives to get rid of radioactive waste are mentioned, such as disposal into the earth’s ice caps or into outer space. They may look promising at first sight. Studying them in greater depth, however, makes obvious that there may be insur­mountable problems. Both ice-cap disposal and disposal into outer space do not seem safe as yet and have political implications that will be hard to resolve.

For the first generation of the world’s commercial reprocessing plants, there is little doubt that the HLW will be solidified as a glass or as a ceramic and, after some interim storage, will be disposed of into a geologic formation deep underground.

1.7 Uranium Valence States

Uranium compounds have been prepared with positive valence states of 3, 4, 5, and 6. In addition, U03 is a weak acid, like M0O3 and W03, and forms uranates such as Na2U04 and diuranates such as Na2U207.

Trivalent uranium ion reduces water to hydrogen. Hence, stable aqueous solutions of trivalent uranium compounds cannot be prepared. Compounds of tetravalent uranium are generally similar to those of zirconium or thorium, except that some uranium compounds can be oxidized to the hexavalent form. Compounds of pentavalent uranium are of little importance because they disproportionate readily into tetravalent and hexavalent forms. The properties of hexavalent uranium are generally similar to those of hexavalent molybdenum or tungsten. In aqueous solution hexavalent uranium forms the uranyl ion U02 2+

1.8 Uranium Oxides

Data on the most important oxides of uranium are given in Table 5.7.

Uranium dioxide U02 is the form in which uranium is most commonly used as a reactor fuel for light-water, heavy-water, and fast-breeder reactors. It is a stable ceramic that can be heated almost to its melting point, around 2760°C, without serious mechanical deterioration. It does not react with water, so that it is not affected by leakage of cladding in water-cooled reactors. Its principal disadvantages compared with uranium metal are its lower uranium atom density and lower thermal conductivity. At 100°C, thermal conductivities are metal, 0.25; U02,

0. 09 W/(cm-°C).

U308 occurs naturally as the mineral pitchblende. It can be made by oxidizing U02 or heating U03.

Uranium trioxide U03 is made by igniting uranyl nitrate, U02(N03)2-6H20, or иОд^НзО, the two principal forms in which uranium purified in aqueous solution is prepared. It is an intermediate in the preparation of U02 or UF6.

Uranium peroxide U04’2H20 is prepared by precipitation of an aqueous solution of uranyl nitrate at 70 to 80°C and a pH of 3 to 4 with H202. Because few other cations form precipitates under these conditions, this is an effective way of purifying uranium.

232 Th and 228Th. Natural thorium consists almost entirely of 232Th, plus 1.35 X 10"8 percent of 228Th (radiothorium), and small, but variable, amounts of 234Th, 230Th, 231 Th, and 227Th.

The 238Th/232Th ratio in natural thorium equals the ratio of their half-lives, as 228Th, a decay product of 232Th, is in secular equilibrium with its parent. In irradiated thorium, however, the 228Th/232Th ratio may be much higher because of formation of additional 228Th by alpha decay of 232U, as explained in Chap. 8.

234Th and 230Th. The members of the 238U decay chain, 234Th (UX-1) and 230Th (ionium), Table 5.2, are found in natural thorium when uranium is present in thorium ores. Because of the short half-life of 234 Th, its concentration in thorium is inconsequential, and it soon decays from separated thorium. Eighty-thousand-year 230Th, on the other hand, is a significant constituent of thorium from ores containing uranium. The 230Th/232Th atom ratio is given by

Here 0.9927 is the atom fraction of 238U in natural uranium, rU Th is the atomic ratio of U to Th in the ore, and 4.51 X 109 years is the half-life of 238U. The ratio of the activity of 230Th to 232 Th is

where t is the half-life of the designated isotope.

Thorium has been recovered as a by-product of uranium production from ores of the Blind River district in Ontario in which the uranium:thorium ratio is 6:1 [С5]. In such thorium the 230Th activity is 3.1 Ж 6 = 18.6 times the activity of the 232Th.

231 Th and “’Th. The members of the 235U decay chain, 231 Th and 227Th, Table 5.3, are trace constituents of thorium from ores containing uranium. Because of their short half-lives and the small proportion of 23s U in natural uranium, they are of no significance in thorium technology.

1.1 Synthetic Thorium Isotopes

233Th. The isotope 233Th, produced when 232Th captures a neutron, is an important short-lived intermediate in production of 233 U, as explained in Sec. 1.

229Th. The isotope 229Th, with a half-life of 7340 years, is the longest-lived radioactive daughter of 233U (Table 5.4). It is present at very low concentration in aged irradiated thorium.

Aqueous solutions of zirconium or hafnium compounds may be obtained by dissolving the corresponding hydrous dioxide in the appropriate strong acid. Because zirconium and hafnium oxides are such weak bases, these aqueous solutions tend to hydrolyze, with formation of zirconyl salts such as ZrO(N03)i.

Crystals of zirconium nitrate, Zr(N03)4’5^0, can be obtained by evaporating a solution of Zr02 in strong nitric acid at a temperature not higher than 15°C. Evaporation at higher temperatures or low acid concentration yields zirconyl nitrate, Zr(XN03)2 ’2^0. Zirconyl nitrate is less readily extracted by tributyl phosphate than zirconium nitrate, a property made use of in separating thorium from fission-product zirconium, Chap. 10.

Evaporation of aqueous solutions of ZrCU or of solutions of hydrous zirconia in hydrochloric acid yields the oxychloride, ZrOCl2.

Sulfuric acid solutions of hydrous zirconia contain very little Zr4* because of complex

Source: National Bureau of Standards, JANAF Thermo chemical Tables, 2d ed., U. S. Government Printing Office, Washington, D. C., June 1971.

formation. Zirconium-containing species present include un-ionized Zr(S04)2 and a number of complex sulfatozirconic anions, of which the best known is the disulfatozirconic acid anion ZrO(S04)22 ~ ■ Evaporation of such solutions produces crystals of disulfatozirconic acid trihydrate, H2Zr0(S04)2’ЗНгО, formerly regarded as zirconium sulfate tetrahydrate, Zr(S04)2’4Н20. Dehydration of these crystals at 100°C produces the anhydrous acid H2Zr0(S04)2. Further heating at 380°C produces anhydrous Zr(S04)2. A number of salts of this acid have been prepared.

ZrF4 is partially hydrolyzed by water and is only slightly soluble in it. It is soluble in aqueous solutions of HF. From such solutions a fluozirconate, M2ZrF6) can be crystallized by

addition of an alkali chloride, carbonate, or hydroxide. Potassium fluozirconate can be made by reacting zircon with potassium fluosilicate at 1000°C.

ZrSi04 + K2SiF6 -+ KjZrF6 + 2Si02

Its solubility in water at 20°C is 0.055 g-mol/Uter. Fractional crystallization of a fluozirconate was one of the early methods used to separate hafnium from zirconium.

A much more complete description of the chemistry of zirconium compounds is given by Blumenthal [B3].

1 GENERAL CHEMICAL PROPERTIES OF THE ACTINIDES

1.1 Electronic Configurations

Many of the elements of importance in fuel reprocessing are found within the sixth and seventh periods of the periodic system. In the sixth period the rare-earth fission products, lanthanum to dysprosium, are difficult to separate from each other by chemical means. Their close similarity in chemical properties is explained on the basis of their electronic configurations [SI ] as shown in Table 9.1.

The chemical properties of elements are determined by the behavior of electrons in the outermost shells. The chemical properties of cesium, an alkali metal, and barium, an alkaline earth, differ appreciably because of the different number of electrons in the 6d sheU. The 6s shell is filled for barium, and for lanthanum the next electron is apparently added to the previously empty 5d shell. Additional electrons for most of the elements of higher atomic numbers, extending through lutetium, are added to the 4/ shell, which is so deep within the atoms as to have little influence on the chemical properties of these elements. The 4f shell is filled for lutetium, and succeeding elements, hafnium, tantalum, tungsten, etc., add electrons to the 5d shell. The 5d electrons are not strongly bound, and each member of the sixth-period transition series, hafnium to tungsten, shows chemical properties quite distinct from those of its neighbors.

The series of 15 elements, lanthanium to lutetium, is known as the lanthanide series. These elements all form trivalent ions in solution; quadrivalent oxidation states of cerium, praseodymium, and terbium, and bivalent states of samarium and europium are also obtained.

The seventh period of the periodic table is occupied by a similar series called the actinide series. Beginning with actinium the 5/electron shell is populated in a manner analogous to filling the 4f electron shell in the lanthanide series. A suggested electronic configuration [K2, M6], is shown in Table 9.2. After the alkaline earth radium, additional electrons are added to the 6d and 5/ shells, beginning the actinide series. At the beginning of the actinide series electrons are added

to the 6d shell rather than to the 5/, so actinium and thorium behave chemically as homologs of lanthanum and hafnium, respectively. For gaseous elements the 5/shell is preferred beginning with protactinium. However, for metallic crystals at room temperature the 6d shell is preferred as far as uranium and possibly neptunium, which is consistent with the initial contraction of the metallic radii with increasing atomic number as shown in Table 9.3 [Al].

The chemical properties of the actinides are much less similar to each other than those of the lanthanides, because the additional electrons added to the 5/ and 6d are bound less tightly than those of the 4/and 5d shells of the lanthanides. As shown in Table 9.4, the lanthanides in aqueous solutions exist principally in a single, trivalent oxidation state, whereas four or more oxidation states are observed in the aqueous chemistry of uranium, neptunium, and plutonium. The actinide ions normally formed in solution by the oxidation states II through VI are M2+, M3+, M4*, M02*, M022+, respectively.

The oxidation states of the lanthanide and actinide elements are summarized [K2] in Table

9.4. The most stable oxidation states in the actinide series are italicized. Numbers in parentheses indicate unstable or unusual states of oxidation. Actinides of the same oxidation state are similar in chemical properties, but different oxidation states show appreciably different chemical properties.

The stability of TBP and its selectivity for tetravalent and hexavalent metal nitrates led to its consideration and selection for processing irradiated thorium to separate 233 U and thorium and decontaminate them from fission products. The form of the Thorex process first developed used aluminum nitrate as salting agent to enhance the distribution coefficients of uranyl and thorium nitrates. It was used in the early 1960s by E. I. duPont de Nemours and Company [R5] to process thorium irradiated in the U. S. AEC’s Savannah River production reactors. Because the aluminum nitrate used as salting agent added undesirably to the nonvolatile material in the high-level wastes, it was replaced by nitric acid in the acid Thorex process developed by Oak Ridge National Laboratory in the early 1960s [R1 ]. More details of the history and applications of the acid Thorex process will be given in Sec. 5.

2.2 Other Aqueous Processes

Culler and Blanco [C18] have summarized other aqueous processes that have been studied for processing power reactor fuels not readily handled by the standard Purex or Thorex processes. Many of these require reagents other than nitric acid to dissolve either the cladding or the fuel, but finally use solvent extraction with TBP to separate and purify fissile materials. Details of these other processes are given in references cited by Culler and Blanco [Cl8].

Special problems in reprocessing LMFBR fuels compared with LWR fuels are as follows:

Removal of greater decay heat Deactivation of sodium

Voloxidation of mixed U02 — Pu02 fuel of high Pu02 content

More complete retention of iodine, if LMFBR fuel cooling time is reduced

More difficult dissolution

The higher plutonium concentration in the first Purex cycle

These problem areas are discussed in order in the following pages.

Another problem is the greater rate of solvent degradation caused by the higher specific activity of the fuel. This makes the use of short-contact-time contactors even more necessary than with LWR fuels. A final problem is control of criticality, which is vital throughout all fuel reprocessing and is discussed in general terms in Sec. 8.

In one respect, requirements for reprocessing fuel from an LMFBR is less demanding than from an LWR. It is not necessary to purify the uranium and plutonium products so completely. The uranium is not converted to UF6 and does not have to meet the strict purity requirements of feed for a uranium enrichment plant. When the plutonium is recycled to an LMFBR it is diluted with uranium, so complete separation from uranium is not necessary.

The final disposal technique depends on the type of waste. For the extremely long-lived HLW a concept has been accepted worldwide, namely, storage in a stable geologic formation deep underground, eventually nonretrievable. For non-high-level waste the storage philosophy is less uniform. Alpha waste is almost unanimously regarded a potential hazard similar to HLW and will probably be disposed of in a similar way. Other non-high-level waste will probably be handled differently. In less densely populated countries, including the United States, shallow burial is considered adequate for non-alpha, non-high-level waste. Heavily populated countries such as West Germany have rather decided to put any solidified radioactive waste eventually into a deep underground facility. However, as the design of a geologic waste repository will largely be letermined by the heat-generation rate of the waste, it will be simpler and cheaper to build a safe repository for non-high-level waste.

The United States, West Germany, and Canada have the most active programs in the field of geologic disposal. The U. S. program focuses on two pilot plants to be operational in 1986. The Canadian program has a similar time schedule. A German pilot plant in the abandoned salt mine Asse (Fig. 11.25) has been operated with LLW and MLW for about 10 years. Asse will also be used as an experimental facility for HLW. A preliminary site decision has already been made in West Germany for a full-scale repository. In other countries, e. g., the United Kingdom, France, Italy, the Netherlands, Belgium, and Spain, research programs on geologic disposal are also in progress.

As a means for intermediate storage of solidified HLW, engineered surface facilities are studied. These facilities shall be designed to contain and control liquid, solid, and gaseous waste resulting from normal and abnormal operations of the facility and from exposure to natural phenomena.

On a cascade operating with specified concentrations, xf in the raffinate and yf in the entering organic, and for a specified flow ratio (F + S)/E, an increase in the number of extracting stages results in an increase in the concentration xf/+1 of the solute in the aqueous stream entering the extracting section. For an infinite number of extracting stages, xf/+i occurs at the intersection of the extracting equilibrium and operating lines, as illustrated in Fig. 4.16a. Similarly, for an infinite number of scrubbing stages, xfj occurs at the intersection of the scrubbing equilibrium and operating lines.

If there are infinite numbers of both extracting and scrubbing stages, the equilibrium lines intersect the operating lines at the operating-line intersection xF, as illustrated in Fig. 4.16b; i. e.,

*m=xn+ i=xF (4.59)

where the asterisks denote the limiting conditions of Fig. 4.16b. The slopes of the operating lines in Fig. 4.16b are

Figure 4.16 Limiting flow ratios for extracting-scrubbing cascade, (a) Infinite number of stages in extracting section only; (b and c) infinite number of stages in extracting and scrubbing sections;

(c) double intersection in extracting section.

any one of the extractable components, they will be satisfied for all of the extractable components. For specified feed concentrations and for specified terminal concentrations or overall recoveries of two of the extractable components, these limiting conditions result in the minimum flow rates E*/F and S*/F of organic and scrub solution, relative to the feed rate, to achieve those recoveries.

For the separation of feed components A and B, and assuming zero concentration of these components in the entering organic (yf = 0), Eq. (4.60) written for component A is

(4.62)

By writing Eq. (4.62) for components A and В and eliminating the flow ratios:

^-4й-=д£-4й — (4.63)

XA XB

The performance of an extracting-scrubbing cascade may be defined in terms of the fractional recovery p of one of the desired components and the decontamination factor f of the organic product relative to the feed. These quantities are defined in terms of feed and product flow rates and composition as follows:

Equations (4.69) and (4.70) are the limiting flow ratios in the scrubbing and extracting sections, respectively. Equation (4.69) shows that, for the limiting conditions of an infinite number of extracting stages, no scrubbing is required, that is, 5* = 0, when

The limiting condition of infinite numbers of extracting and scrubbing stages can achieve complete recovery, that is, pA =1, provided that the extracting section flow ratio given in Eq. (4.70) is adjusted to

E* _ 1 F+S*

The above derivations are limited to the unique case of Fig. 4.16h. The solutions in terms of the limiting (asterisked) quantities are valid whenever the equilibrium curve intersects the operating lines only at x =xF. If the equilibrium line is sufficiently curved, as illustrated in Fig. 4.16c, one of the operating lines may first intersect its equilibrium line at some point other than xF, thereby making invalid the above limiting-condition equations for the triple-point intersection.

As a numerical example of these equations, consider an aqueous feed solution containing

to be extracted with 60 v/o TBP in kerosene, as considered in the simple extraction example of Sec. 6.2 (cf. Table 4.6).

From Table 4.4, the distribution coefficient for zirconium is

TfZr = 1-20

Нигё and Saint James [H4] have found that the zirconium-hafnium separation factor for this mixture is 10, so that

= 0.12

Suppose that we wish to recover 98 percent of the zirconium and to obtain a zirconium-hafnium decontamination factor of 200. The limiting ratio of scrub to solvent, from (4.69), is

S* _ (200) (0.12) — 1,20 =QU46

E* 199 ‘

The limiting ratio of solvent to scrub plus feed, from (4.70), is

E* = (0-98) (199) =

F + S* (1.2)(199) — (200)(1.2)(0.02)

1.15 Steps in Producing Refined Uranium Compounds

The steps in producing refined uranium compounds from crude uranium ores may be conveniently classified into concentration, purification, and conversion to the chemical form finally wanted. Concentration consists in separating uranium from most of the nonuraniferous
diluents that accompany uranium in nature. It increases the uranium oxide content from a few tenths of a percent in the ore to 85 to 95 percent in the concentrate, while rejecting most of the diluents as tailings. Concentration is usually carried out in uranium mills within a few miles of where the uranium is mined, to avoid the high cost of transporting the nonuraniferious bulk of the ore.

Purification consists in removing from the impure uranium the rest of the nonuraniferous contaminants and producing a pure uranium compound. In most uranium refineries purification is carried out before conversion of uranium to the chemical form finally wanted. This sequence of refining operations will be described in Secs. 9.2 through 9.6. However, in the process used by the Allied Chemical Company for producing UF6 from uranium concentrates, the sequence is reversed, with conversion to UF6 preceding purification of the impure UF6 by fractional distillation. This process will be described in Sec. 9.7.