Leach testing and its role in the waste acceptance process

The most important requirement for a waste form is its chemical durability, expressed as a dissolution rate. It should be noted that for some radionu­clides, solubility limits the dissolution rate while others are completely soluble, e. g. "Tc, 129I, or 135Cs. These soluble radionuclides are released at the maximum forward rate of dissolution. For the production of durable nuclear waste forms, it is desirable for the waste forms to be highly insoluble in the long term to minimize release to the environment, i. e. to have the slowest forward dissolution rate possible. Since no ‘durability test’ can be carried out on these geologic timescales, dual approaches are taken:

1. Durability test parameters such as surface area (SA), time (t), tempera­ture (T), or a combination such as (SA)^(t) are used to ‘accelerate’ dissolution as long as the acceleration parameter(s) used does not change the dissolution mechanism. To ensure that the mechanism is not ‘altered’ by the acceleration modes of the experiments, natural analogs are usually tested simultaneously.

2. Models are used to predict waste form dissolution from parameters that can be measured such as the activation energy of dissolution, forward rate of dissolution, and from an understanding of the dissolution mecha­nisms. Predictive and/or transport models for waste form performance on extended time scales (1,000-1,000,000 years) has led to various ther­modynamic and kinetic models (see [160], [187]).

Thus, there are no ‘waste form-specific’ durability tests, but a suite of tests that must be performed to understand the leaching mechanism(s) of a waste form and to derive the parameters necessary for the particular predictive or transport model(s) being applied.

In order to determine if a particular waste form is acceptable, it must be demonstrated that the waste form performance in the disposal system is adequate. Such evaluations in the US are known as total system perform­ance assessments (TSPA) for HLW and performance assessments (PA) for shallow land disposal of immobilized LAW known as ILAW in the inte­grated disposal facility (IDF). The TSPA or PA includes all of the testing and performance modeling that has been gathered on the waste form and the TSPA is intended to provide a technical basis that a waste form is acceptable for deep geological disposal.

For HLW in many countries the geological disposal sites have not been determined while wasteform producers have already made many canisters of vitrified waste (see Table 6.1). Due to the mismatch in timing between the need to stabilize HLW and when a geological repository will be chosen and ready to receive the wasteforms, the US devised a strategy to addresses vitrified waste acceptance based on production control. Production control is intended to determine how the production of a waste form material affects (or controls) its performance and identify the ranges for processing variables that result in an acceptable waste form. The primary role of most of the waste acceptance product specifications (WAPS) developed in the US for vitrified HLW waste forms verify that the properties of a specific waste form product are consistent with the existing regulations and thus will be acceptable for disposal, either by direct measurement or through process control.

Therefore, waste acceptance testing is, for the most part, focused on comparing a specific waste form product to the range of waste forms that are (1) considered to have acceptable performance based on performance modeling and (2) produced within the production control limits. What will be acceptable with respect to waste form performance and processability will depend on the disposal site and engineered system and cannot be com­pletely quantified at the time the waste form is made. The range of accept­able waste form compositions will depend on the required performance [167].

While the predicted long-term durability of a waste form is a necessity for its ‘qualification for shallow land burial’ or ‘deep geologic disposal’, there is also a need for short-term testing that can be related to acceptable performance by the following linking relationships [168]:

process control ^ composition control ^ dissolution rate control ^ performance control ^ acceptable performance.

This approach allows a waste form producer to ensure that the waste form that they are producing on a tonnage per year basis will be acceptable to long-term performance instead of having to test each and every canister or form produced. For HLW glass (alkali borosilicate glass) in the US, the manner in which this was done is given below in a brief stepwise fashion and explained in more detail in Refs [11, 169-173]:

1. Develop an acceptable waste form durability based on HLW perform­ance modeling (fractional dissolution rates between 10~4 to 10“6 parts per year (i. e., the glass waste form would take 10,000 to 1,000,000 years to totally dissolve [174]).

2. The middle of the range determined by HLW performance modeling was adopted as the waste form specification; if the long-term fractional dissolution rate of a wasteform was <10-5 parts per year for the most soluble and long-lived radionuclides, then borosilicate glass would provide acceptable performance for any repository site or concept.

3. Develop an understanding of the glass durability mechanisms from a combination of the test protocols (ASTM C1220 which was previously known as MCC-1, ASTM C1285 which is known as the Product Consist­ency Test (PCT) [175, 176] , ASTM C1662 which is the SPFT test, and ASTM C1663 which is the Vapor Hydration Test or VHT).

4. Develop a glass standard, the Environmental Assessment (EA) glass [177, 178] that bounded the upper release rate found to be acceptable from the HLW repository modeling from step 1 above.

5. Generate databases for modeling the maximum radioactive release rate(s) by relating the release of "Tc, 129I, and 135Cs to the release of non-radioactive species such as Na, Li, and B which leach at the same rate (congruently); this is part of the ASTM C1285 (PCT) test protocol.

6. Develop a short-term test and process control strategy for ensuring that every glass produced has a dissolution rate less than that of the EA glass at the l95% confidence level based on Na, Li, B which in turn ensures acceptable performance control.

7. Continue to qualify that the radionuclide response of production glasses verify that production glass radionuclide releases are consistent with the releases predicted by Na, Li, and B.

Therefore, a suite of the existing durability tests (those for affinity control, solubility control, and/or diffusion control) must be performed on a waste form to determine the mechanisms, and determine the parameters neces­sary for the mechanistic model(s) being developed, e. g. the transition state theory (TST) models used in the TSPA for HLW geological disposal or the PAs for shallow land burial. Different durability tests are used for a diffu­sion model, for example for cement. However, one cannot apply a glass standard that leaches by an affinity limited mechanism to cement that leaches by diffusion, nor can one apply a borosilicate glass standard to non — borosilicate-type glasses since it is not known whether the radionuclides in non-borosilicate glasses leach by the same degradation mechanism and whether the leaching of Na, Li, and B remain congruent with the leaching of the radionuclides. In these cases, new standards need to be developed and qualified and the leaching mechanisms understood.

For glasses, the advances in the measurement of medium range order (MRO) in glass waste forms has led to the understanding that the molecular structure and composition of a glass, like the molecular structure and com­position of minerals, controls the waste form durability by establishing the distribution of ion exchange sites, hydrolysis sites, and the access of water to those sites. During the early stages of glass dissolution, a ‘gel’ layer resembling a membrane forms through which ions exchange between the glass and the leachant (Fig. 6.7). The hydrated gel layer exhibits acid/base properties which are manifested as the pH dependence of the thickness and nature of the gel layer. Advances in the understanding of the dissolution mechanisms of borosilicate glasses proposed for nuclear waste solidification were extensively studied in the 1980s-1990s [22, 179-186] and such mecha­nisms are still being studied [160, 187-190]. At least four operative mecha­nisms have been shown to control the overall glass durability as shown in Fig. 6.7. These four mechanisms are ion exchange, matrix dissolution, accel­erated matrix dissolution, and surface layer formation (possibly of a protec­tive or passivating nature).

One can bound or model the shorter term durability of a glass using kinetic or thermodynamic models to describe the impacts of ion exchange and matrix dissolution or hydrolysis by examining either time-temperature data (Fig. 6.8) or release vs time or accelerated release, expressed as SA/VHime (Fig. 6.9), but these underlying mechanisms become modified if surface layers form and/or if, over very long periods of time, the gel layer ages in situ into clay or zeolite minerals or the leachate becomes satu­rated with respect to a clay or zeolite phase. If zeolite mineral assemblages (higher pH and Al3+ rich glasses) form, the dissolution rate increases (Fig. 6.9) which is undesirable for long-term performance of glass in the environment.

The current theories of glass dissolution [ 159] suggest that all glasses typically undergo an initial rapid rate of dissolution denoted as the ‘forward rate’ (Figs 6.8 and 6.9). However, as the contact time between the glass and the leachant lengthens, some glasses come to ‘steady state’ equilibrium and corrode at a ‘steady state’ rate, while other glasses undergo a disequilibrium reaction with the leachant solution that causes a sudden change in the solu­tion pH or the silica activity in solution [191] [ The ‘return to the forward rate’ (Fig. 6.9) after achieving ‘steady state’ dissolution is undesirable as it


6.7 ( a) Schematic diagram of glass dissolution mechanisms (ion exchange and matrix dissolution) in aqueous solution, coupled with both hydrated amorphous surface layer formation and crystallization/ precipitation from solution [179, 402]. (b) Schematic diagram of the glass dissolution mechanism known as ‘accelerated matrix dissolution.’ In this mechanism, the excess strong base in the leachate released by the ion exchange mechanisms attacks the glass surface layers, including the gel layer, and makes the glass appear to have little or no surface layer.

can cause a glass to return to the rapid dissolution characteristic of initial dissolution.

The initial rate is often referred to as Stage I dissolution in the US litera­ture, but it encompasses zones where multiple mechanisms are operative including regimes that are interdiffusion controlled, hydrolysis controlled, and a rate drop that is diffusion or affinity controlled [159] . The ‘steady


6.8 A temperature-time plot of the incongruent corrosion mechanisms exhibited by British Magnox waste glass in deionized water, showing that corrosion in deionized water at a constant temperature begins immediately with an instantaneous surface dissolution followed by a diffusion controlled ion exchange phase. As corrosion progresses, the impact of hydrolysis becomes significant with comparable contributions from both ion exchange and hydrolytic reactions.

Finally, glass corrosion in deionized water is fully controlled by hydrolysis [36].

state’ rate (also known as the residual or final rate) that signals the end of the alteration phase and/or a pseudo-equilibrium between the alteration and re-condensation reactions [159, 192] is known as Stage II dissolution, and the return to a forward rate (or resumption of alteration) is known as Stage III dissolution. Diffusion controlled dissolution of network modifiers and/or radionuclides during Stage I and Stage II normally follow a math­ematical function related to the square root of the test duration as observed in many burial studies [190], while other radionuclides are solubility limited, entrapped in the gel layer, or complexed in secondary alteration phases that form from the leachate solution.

A reaction zone is formed as the leached layer solution interface progresses into the glass (Fig. 6.7 a). The front of the reaction zone repre­sents the region where the glass surface sites interact with the ions in solu­tion [193]. The top of the gel reaction zone represents the leached layer-glass interface where a counter-ion exchange occurs [193] . The glass dissolution rate is modified by the formation of the hydrated amorphous gel layers and/ or secondary precipitates, e. g., metal hydroxo and/or metal silicate com­plexes that have reached saturation in the leachate and can precipitate on the surface of the gel layer [22, 179, 181, 182, 194, 195] . These ‘back reac­tions’ have been attributed to formation of silanol bonds as surface



6.9 Parabolic behavior of the diffusion profile of soluble species out of a waste glass through an increasingly thick surface layer [159]. Acceleration of glass durability tests using glass surface area (SA), leachant volume (V), and time. Acceleration appears to follow parabolic diffusion kinetics until SA/V is -20,000 m-1, when the glass dissolution mechanism appears to change reverting to a rate similar to the forward rate but likely controlled by precipitation of secondary phases.

adsorption sites which were modified by changes in solubility of the species in solution and surface (zeta potential) considerations [22, 196].

The gel layer may, under certain conditions, act as a selective membrane [194, 197] or as a protective/passivating layer [22, 159, 180-182, 184-186, 192, 198]. The slowing of glass dissolution to a steady state rate by solution saturation (affinity) of glass matrix elements or reaction through a surface layer has been referred to as Stage II dissolution including residual rate dissolution, steady state dissolution, or the final dissolution rate. Recent mechanistic modeling of glass durability, including the slowing of the dis­solution rate due to affinity and/or surface layer effects, was first modeled by Grambow and Muller [199] and is referred to as the GM2001 model. The GM2001 model combines the effect of glass hydration by water diffu­sion with ion exchange and affinity-controlled glass network corrosion (Figs 6.8 and 6.9). The slowing of dissolution due to the effect of a growing surface gel layer is represented by a mass transfer resistance for silica by this layer. At the interface between the glass and the gel layer, a different ‘gel layer’ is assumed to be hydrated glass that allows diffusion of H2O in and boron and alkali atoms out of the glass (similar to Fig. 6.7). A 2003 modification of the GM2001 model, known as the GM2003 model [159], treats silica dis­solution and silica diffusion through the gel separately from water diffusion, and boundary conditions are specified at the gel/diffusion layer and the gel/ solution interfaces. Recently, the GRAAL (glass reactivity with allowance for the alteration layer) model [187, 189] has been proposed, which is dependent on the composition and the passivating nature of the gel layer, called the passivating reactive interphase (PRI). The leached layer has been found experimentally to be zoned (5-7 zones) and the GRAAL model assigns various mechanisms to different zones within the PRI.

The resumption of alteration (Stage III) causes the long-term dissolution rate to reaccelerate to a rate that is similar to the initial forward dissolution rate for some glasses. This unexpected and poorly understood return to the forward dissolution rate has been shown to be related to the formation of the Al3+-rich zeolite, analcime, and/or other calcium silicate phases. Moreo­ver, the presence of Al3+ and Fe3+ in the HLW glass, in the leached layer, and in the leachant has been shown to influence whether a glass maintains Stage II dissolution or reverts to the forward rate of dissolution, e. g., Stage III dissolution. Van Iseghem and Grambow 3 191] demonstrated that an Al3+-rich zeolite (analcime) formed on certain glasses during dissolution but not on others. Van Iseghem and Grambow also demonstrated that a change in solution pH accompanied the return to the apparent forward rate when analcime formed. Likewise, Inagaki et al. [200] demonstrated that solution pH and solution concentrations of Na and K were also involved in the formation of undesirable analcime versus Na-bedellite (a smectite clay). Other zeolites and smectite clays that are rich in Fe3+ compared to Al3+ do not appear to accelerate glass corrosion [191, 201, 202].

Since many long-term durability models are still being refined and an international study group [203] is actively working on a refined understand­ing of the PRI, a variety of leaching tests are being used to facilitate an integrated understanding of these stages of durability.

Добавить комментарий

Ваш e-mail не будет опубликован. Обязательные поля помечены *