Waste Management 26 (2006) 689–698 www.elsevier.com/locate/wasman
Overview of waste stabilization with cement B. Batchelor
*
Environmental and Water Resource Engineering, Civil Engineering Department, Texas A&M University, 3136 TAMU, College Station, TX 77843, USA Accepted 31 January 2006 Available online 13 March 2006
Abstract Cement can treat a variety of wastes by improving physical characteristics (solidification) and reducing the toxicity and mobility of contaminants (stabilization). Potentially adverse waste-binder interactions are an important consideration because they can limit solidification. Stabilization occurs when a contaminant is converted from the dissolved (mobile) phase to a solid (immobile) phase by reactions, such as precipitation, sorption, or substitution. These reactions are often strongly affected by pH, so the presence of components of the waste that control pH are critical to stabilization reactions. Evaluating environmental impacts can be accomplished in a tiered strategy in which simplest approach would be to measure the maximum amount of contaminant that could be released. Alternatively, the sequence of release can be determined, either by microcosm tests that attempt to simulate conditions in the disposal zone or by mechanistic models that attempt to predict behavior using fundamental characteristics of the treated waste. Ó 2006 Elsevier Ltd. All rights reserved.
1. Introduction Solidification/stabilization (s/s) was used to treat nuclear wastes in the 1950s and then was widely applied to hazardous wastes in the early 1970s (Conner, 1990). It has been used both to treat wastes currently being produced, as well as soils and sediments contaminated by previous improper disposal. S/S has been identified by the US EPA as the Best Demonstrated Available Technology for 57 regulated hazardous wastes (Shi and Spence, 2004) and it is one of the most commonly applied technologies at Superfund sites in the US, being used at 24% of the sites between 1982 and 2002 (US EPA, 2004). Most applications of s/s are cement-based, in that they rely on Portland cement as the primary binder. However, it can be combined with lime as well as fly ash, blast furnace slag, and other similar materials. The interactions of these binders with waste components determine the extent of treatment. Treatment by s/s combines two interrelated processes that occur simultaneously to produce a material that will have reduced environmental impact when dis*
Tel.: +1 979 845 1304; fax: +1 979 862 1542. E-mail address:
[email protected].
0956-053X/$ - see front matter Ó 2006 Elsevier Ltd. All rights reserved. doi:10.1016/j.wasman.2006.01.020
posed or reused. One of these processes is called solidification and it is the process of producing a solid product with improved physical properties, primarily strength. The other process is stabilization and it is the process of converting the contaminant of concern to less mobile and less toxic forms. This paper will provide an overview of the processes of solidification/stabilization and the methods of characterizing the treated product in ways that facilitate determination of environmental impacts. 2. Solidification Solidification results in changes in primarily physical properties of the waste material so that a well solidified waste will no longer contain free liquids and will have improved strength. As such, it will be much more easily handled, particularly if the untreated waste was a liquid or sludge. Furthermore, a solidified waste will have less impact on the environment when disposed. It will not contain free liquids that can be more easily transported to contaminate the environment. It will usually be formed into solids forms of larger size than the untreated waste. This will result in smaller area/volume ratios that will result in lower rates of release of contaminants. The treated waste
690
B. Batchelor / Waste Management 26 (2006) 689–698
will typically have much reduced permeability, which reduces the advective flow through the waste. If the treated material has a substantially lower permeability than the material surrounding it in the disposal zone, fluids will flow around rather than through it, resulting in substantially reduced release of contaminants. A major factor in applying cement-based s/s to wastes is the extent to which components of the waste interfere with cement hydration reactions. Many of the compounds that are reported to accelerate or retard cement hydration (Taylor, 1990; Conner, 1990; Lea, 1971) are found to cause similar problems in wastes (Taylor and Fuessle, 1994; Trussell and Spence, 1994; Means et al., 1995; Hills and Pollard, 1997). These include organics, particularly polar organics, halides, metals and sulfate. Avoiding adverse waste-binder interactions is a major focus of developing s/s technologies for some wastes. Approaches to managing these problems include adding an accelerator to counteract a retarder and vice-versa. For example, soluble silicate has been used extensively as a s/s admixture to reduce effects of retarders. The primary method of characterizing the extent of solidification is to measure the unconfined compressive strength of the treated material. This parameter has been incorporated into some specifications for s/s treatment by regulatory bodies; however, the degree of strength required depends on the disposal scenario. Another important application for strength tests is to estimate the long-term stability of the waste form. A stronger waste form will tend to retain its integrity better than a weaker one, although future characteristics cannot be insured by measurement of current characteristics. Long-term integrity is important because a waste form that degrades to small sized particles will release contaminants more rapidly. Wet-dry and freeze-thaw durability tests have also been applied to waste forms to predict durability. Although they have a history of use to predict longer term performance of construction materials, their applicability to waste forms is in doubt because disposal conditions for many wastes are not similar to those simulated in these tests. Permeability to water is another important characterization parameter for wastes treated by s/s, because it predicts the extent to which water will tend to pass through the waste form in a disposal environment. Permeabilities below those associated with clay liners in possible in well-treated wastes (Stegemann et al., 1997; Conner, 1990). The pore structure in a waste treated by cement-based s/s is important in determining the extent of contaminant release, because contaminant transport within the waste form will be primarily by diffusion in the pores. A waste form with highly tortuous, unconnected pore structure will leach contaminants more slowly. Characterization of a waste treated by s/s is usually conducted shortly after treatment, but the parameters measured are often used to predict performance over long periods into the future. Therefore, the stability of these parameters is of concern. Hydrated cements and concretes change their physical properties over time and wastes treated by cement-based s/s should also be expected to change in sim-
ilar ways (Klich et al., 1999). Hydration reactions can continue over time and result in higher strength and lower porosity. A combination of Portland cement with pozzolanic materials usually results in slower hydration kinetics. Carbonation can also cause similar changes. Reaction with sulfate is known to degrade concrete and the same can be expected for treated wastes. The wide range of constituents in a waste would be expected to cause more long-term changes than are typically observed for construction materials, which are made with more clearly defined constituents. A waste form also interacts with its environment by leaching binder constituents and waste constituents. This can result in reduced strength and increased connected porosity. Contaminants will leach through the modified pore structure, so it should be characterized to predict release rates. 3. Stabilization Waste stabilization is the result of chemical changes in contaminants and their environment that cause the contaminants to be less mobile or less toxic. Changes in mobility are primarily due to a contaminant being converted from the dissolved phase to a solid phase. In the dissolved phase, a contaminant is free to diffuse down a pore to the external environment. In a solid phase, the contaminant is substantially mobile. Precipitation is a major reaction that results in immobilization during cement-based s/s. The high pH resulting from cement hydration results in many metal contaminants forming hydroxide or mixed hydroxide solids. Sulfide precipitates are also possible, particularly sulfides are added directly or indirectly through the use of blast furnace slag or similar binders that can produce sulfides (Taylor, 1990; Lea, 1971). Calcium and calcium hydroxide solid phases are possible for anionic contaminants. Although contaminant stabilization is typically considered to be by formation of pure solids, solid solutions can also be formed. The complex mixture of compounds present in wastes and binders may result in substantial formation of solid solutions. A chemical equilibrium model based on solid solutions of cement hydration products was found to do a good job of describing changes in aqueous phase concentrations (Abdel-Wahab and Batchelor, 2005). Formation of solid solutions can be considered a form of substitution. For example, immobilization of chromate has been reported as being the result of a chromate ion substituting for sulfate in ettringite (Ca6Al2(OH)12(SO4)3 Æ 26H2O) (Poellmann et al., 1993; Palmer, 2000; Zhang and Reardon, 2003). This could also be considered a solid solution of ettringite and the chromate analog of ettringite (Ca6Al2(OH)12(CrO4)3 Æ 26H2O) (Perkins and Palmer, 2000). Immobilization of boron, molybdenum and selenium has also been attributed to a similar mechanism with ettringite (AFt) and hydrocalumite (AFm) (Zhang and Reardon, 2003; Baur and Johnson, 2003a). The pore structure of wastes treated by cement-based s/s also provides substantial amounts of surface area to
B. Batchelor / Waste Management 26 (2006) 689–698
promote adsorption of contaminants. When high cement doses are used, specific surface areas in wastes treated by s/s can be expected to approach those for hydrated cements, i.e., 200–300 m2/g (Lea, 1971). Since adsorption of cationic metals on oxy-hydroxide surfaces is promoted at high pH, this mechanism can be important. Adsorption has been shown to be an effective mechanism for a variety of contaminants on different cement minerals (Baur and Johnson, 2003b; Ziegler et al., 2001; Schlegel et al., 2004). When adsorption results in the release of an equivalent concentration of co-ions, the process can be called ion exchange. Oxidation–reduction reactions can be important immobilization mechanisms for those contaminants that exist in multiple redox states and have substantially different chemical or toxicological behaviors in the different redox states. The classic example is chromium, which is much more toxic and more mobile in its oxidized state (CrO2 4 ) than in its reduced state (Cr3+). In contrast, arsenic is more toxic when it is in its more reduced state as arsenite than as arsenate. Redox reactions are also important for their effect on mobility. Trivalent chromium is more likely to be precipitated or sorbed than hexavalent chromium under most conditions. Technetium is another contaminant that can be immobilized by conversion to the reduced form (Spence et al., 1989). Cements generally provide a moderately oxidizing environment, but the addition of blast furnace slag can produce reducing conditions by releasing sulfide and other reduced sulfur compounds (Taylor, 1990; Lea, 1971). Furthermore, reductants, such as ferrous iron can be added to promote reduction of contaminants to promote immobilization. Complexation of metal contaminants with organic compounds can reverse the effects of stabilization reactions by increasing the total concentration of the metal in the mobile phase (Zomeren and Comans, 2004). Organic complexing agents that are present in the waste or that enter after disposal could result in mobilization of metals. Stabilization of organic contaminants can occur through a number of processes, including sorption. Although there are substantial amounts of surface area in wastes treated by cement-based s/s, the surfaces are polar and are not as suitable for removal of organics as solids with more non-polar surfaces. Activated carbon and a variety of other organic sorbents can be added to promote immobilization by sorption (Conner, 1990). The high pH environment can also promote degradation of some organic contaminants by base-catalyzed hydrolysis. This can be important for some chlorinated organics and pesticides (Jeffers et al., 1989; Mabey and Mill, 1978; Lyman et al., 1990; Schwarzenbach et al., 2003). This process may be overlooked, because contaminants that are not hydrolyzed effectively at environmental pH can be effectively degraded at the pH values found in pore waters of wastes treated by cement-based s/s. Rate constants for base catalyzed hydrolysis are proportional to the concentration of hydroxide ion, so their values in treated wastes can be many orders of magnitude
691
greater than those observed at environmental pH. Hydrodechlorination is a reaction by which a compound can be dechlorinated by removal of hydrogen and chlorine atoms with production of a double bond. For example, 1,1,2,2-tetrachloroethane can be hydrodechlorinated to produce trichloroethylene and the reaction is faster at high pH (Jeffers et al., 1989). Reducing environments in wastes treated by cement-based s/s can lead to abiotic reductive dechlorination of chlorinated organics. Reducing conditions in cement-based systems can be achieved by addition of ferrous iron and good dechlorination has been observed in these systems (Hwang and Batchelor, 2000b; Hwang and Batchelor, 2001; Hwang and Batchelor, 2002; Son, 2002; Hwang et al., 2005; Kang et al., 2003; Jung and Batchelor, 2004). Optimum values of pH for these reductive dechlorination reactions is often between pH 12 and 13, which makes them suitable for application to cement-based s/s. The combination of contaminant immobilization by conventional s/s and degradation has been called degradative solidification/stabilization. The most critical parameter in most environmental systems for determining the distribution of inorganic contaminants between dissolved and mobile phases is pH. The same is true in wastes treated by cement-based s/s, because the precipitation, adsorption and redox reactions that immobilize contaminants are all strongly influenced by pH. Stabilization reactions for organic contaminants are also strongly pH dependent. Therefore, the success of s/s is dependant on the interaction between binder reactions that determine pH in the treated waste and pH-dependant reactions that determine the extent of contaminant stabilization. The ability of cement to control pH can be expressed by its acid neutralizing capacity (ANC) and the dependence on the ANC pH. The acid neutralizing capacity is a measure of the amount of base present that can accept hydrogen ions from a strong acid. Since the extent of these acid base reactions is dependent on pH, the ANC is a function of pH. However, a single value of ANC is often reported by choosing a specific pH for determination. The ANC of a waste treated by cement-based s/s will depend on the acids and bases contained in the waste, as well as the ANC of the binders. Portland cement will tend to result in high ANC, while replacement of cement with pozzolanic binders will result in lower ANC. The ANC of the mixture of waste and binders can be estimated by assuming they are conservative quantities (Ramabhadran, 1996), although there is no fundamental basis for that assumption. The ANC of the treated waste depends on the ANC of the solid phase (SANC), as well as the ANC resulting from hydroxide ions in porewater. The SANC is due to solid phases, such as Portlandite, calcium-silicate hydrate and other hydration products that can release hydroxide. The ANC as a function of pH can be measured by contacting known amounts of the treated waste with different concentrations of strong acid and measuring pH after equilibrium is achieved. The SANC can be calculated as the difference of the total
692
B. Batchelor / Waste Management 26 (2006) 689–698
14 12 10
pH
8 6 4 2 0
0
2
4
6
8
10
Acid added (eq/kg) Fig. 1. Acid neutralizing capacity as a function of pH (Kim and Batchelor, 2001) for ash waste treated by cement-based s/s (Kosson et al., 1993).
100
y = 0.21x 3 - 4.19x 2 + 31.75x - 55.06 R2 = 0.94
SANCP (eq/l)
80
60
40
20 Experiment Simulation 0 2
4
6
pH
8
10
12
Fig. 2. Solid phase acid neutralizing capacity as a function of pH (Kim and Batchelor, 2001) for ash waste treated by cement-based s/s (Kosson et al., 1993).
ANC measured and the hydroxide ion in solution. Fig. 1 shows the ANC as a function of pH, and Fig. 2 shows SANC as a function of pH (Kim and Batchelor, 2001) using data for a waste ash treated with cement-based s/s (Kosson et al., 1993). 4. Evaluation Characterization of the effectiveness of treatment of wastes should be based on evaluation of their environmental impact after disposal or reuse, rather than meeting an arbitrary criterion. Such an evaluation would investigate the entire pathway of exposure including release of con-
taminants from the waste, transport through the environment and impacts on human health or environmental quality. The step of contaminant release is the one that is determined by the behavior of the waste form and it should be the focus of any procedure for evaluating effectiveness of treatment and suitability for disposal or reuse. The simplest approach to characterizing a treated waste for its potential to release contaminants to the environment is to measure the total amount of contaminant that can be expected to be released. The most conservative approach would be to measure the total amount of contaminant in the waste by aggressive digestion/extraction methods. Although conservative, this approach can greatly overestimate the amount of contaminant that would be released under environmental conditions and can overlook important differences in waste characteristics that actually influence release under environmental conditions. More realistic test procedures have been developed using environmental conditions (pH 4 and 8) to measure the amount of contaminant that is realistically available under environmental conditions (NNI, 1994). The availability test uses relatively high liquid/solid ratio, small particle sizes and sufficiently long leaching times to measure the total amount of contaminant that can be expected to be released and expresses it per mass of waste. Test procedures that attempt to measure release under more specific conditions have also been used. The toxicity characteristic leaching procedure (TCLP) is a widely used regulatory parameter in the US that was developed to simulate leaching conditions in a landfill that receives both hazardous and municipal solid wastes. The output of this test procedure is the concentration of contaminant in the leaching solution at the end of the test, not a total amount of contaminant per amount of waste. In all of these characterization tests, some assumptions must be made to determine the concentration of contaminant in environmental media that can be used to evaluate impacts. When the TCLP test is used to determine if a waste is to be regulated as a hazardous waste, the assumption is made that a dilution/attenuation factor of 100 can be applied to account for environmental factors that would reduce contaminant concentration. The maximum contaminant level (MCL) for drinking water is multiplied by 100 to set the regulatory level in the TCLP that determines if a material is considered to be hazardous. An alternative to use of simple single-point leaching tests is to attempt to describe the actual process of contaminant release. This process can be considered to be potentially limited by either the equilibrium behavior of the contaminant or the kinetic characteristics of its release. Which of these factors controls, depends on the specific characteristics of the release environment. In a laboratory setting where batch leaching tests are often conducted, the controlling factor is time. During early leaching times, the concentration of contaminant in the leaching fluid depends on the kinetics of release, i.e., the kinetics of mobilization and transport within the waste. If sufficient time is
B. Batchelor / Waste Management 26 (2006) 689–698
However, this requires special equipment and is not a typical procedure. For most inorganic contaminants, the effect of pH on the distribution of contaminant between mobile and immobile forms at equilibrium is critical. This behavior can be measured by repeating the equilibrium leaching tests at multiple pH values achieved by addition of strong acids/bases (Kosson et al., 2002; European Committee for Standardization, 2001). Such an equilibrium test can also be used to evaluate release independently of any mechanistic models (van der Sloot and Dijkstra, 2004; Kosson et al., 2002). The information on chemical equilibrium behavior of contaminants can be combined with that for physical transport in a reactive transport model to predict leaching. Data on chemical equilibrium in these models can be described by empirical relationships (Cote, 1986; Kim and Batchelor, 2001) or by geochemical equilibrium models (Sabharwal and Batchelor, 1993; Park and Batchelor, 2002; Dijkstra et al., 2004). In both cases, information must be obtained on behavior of contaminants and on behavior of components of the treated waste that are responsible for its acid neutralizing capacity. This approach has been reasonably successful in predicting leaching behavior as shown in Fig. 3 (Kim and Batchelor, 2001) and Fig. 4 (Park and Batchelor, 2002), but it has not been applied to a wide range of wastes. The descriptions of internal concentration profiles of these release models provide insight into the leaching process (Fig. 5). When contaminants are stabilized by pH-dependent reactions, their release is controlled by the leaching characteristics of the components of the waste that provide ANC. As these compounds leach, the pH in the pore water decreases and a pH profile is established from
Simulation Negative Logarithm
18
Observed Diffusivities (m2/sec)
provided, the contaminant concentration will approach an equilibrium concentration in a batch leaching test. For most contaminants, this equilibrium concentration will be strongly affected by the pH of the system. Release of contaminants in the environment often occurs under conditions that are different from batch laboratory tests, because the release is affected by flow. However, the impact of flow on release can be simulated in batch leaching tests by conducting them at a range of liquid/solid (L/S) ratios. The laboratory results for a specific L/S can be applied to a flow system by considering a time period over which the cumulative volume passing the waste gives the same L/S as the laboratory test. This provides an estimate of a time-averaged contaminant concentration leaving the disposal zone. A ‘‘microcosm’’ approach can be used to evaluate other effects on contaminant release by attempting to more closely duplicate conditions in the disposal zone. Column tests are common ‘‘microcosm’’ tests and they can be conducted with flow velocities and waste characteristics that are expected in be encountered after disposal or reuse. Concentrations of the effluent from the column can be used to directly predict concentrations in water after contact with the waste. This approach has the advantage of directly measuring the desired parameter and of simulating environmental conditions. However, it is limited by the ability to accurately predict those conditions and by the need to repeat the test to evaluate alternative conditions. An alternative approach to microcosm methods is to develop more mechanistic models of contaminant release that can predict release under a variety of conditions. These models require laboratory tests to determine the fundamental characteristics of the waste that are required by the model. Characteristics are needed to describe chemical transformation and physical transport of contaminants. In many situations, chemical transformations can be reasonably assumed to be at local equilibrium, so that chemical kinetic factors can be ignored (Batchelor, 1998). In these cases, the kinetics of release are determined solely by the kinetics of physical transport. An important advantage of this approach is that contaminant release can be evaluated in a number of different disposal scenarios without repeating characterization experiments. However, this approach does not directly measure the parameters of interest and is limited by the ability of the model to adequately describe the mechanisms of contaminant release. Equilibrium characteristics of contaminant behavior can be determined by well designed batch leaching tests. These tests should be conducted with small particle sizes and long leaching times to insure that equilibrium is achieved. They also should be conducted with low L/S to insure that the contaminant is not completely removed from the waste. An alternative to leaching tests is to extract the pore water from the treated waste and to characterize it directly (Barneyback and Diamond, 1981; Page and Vennesland, 1983; Glasser and Marr, 1984; Taffinder and Batchelor, 1993; Kyi and Batchelor, 1994; Trussell and Batchelor, 1996).
693
16
Cd 14 Cr Cu 12 Pb Zn 10 10
12 14 16 Experimental Negative Logarithm
18
Observed Diffusivities (m2/sec) Fig. 3. Model predicted and measured values of observed diffusivity (Kim and Batchelor, 2001) for ash waste treated by cement-based s/s (Kosson et al., 1993).
694
B. Batchelor / Waste Management 26 (2006) 689–698
18 Zn
Leachability Index by SBLEM
17
16 Pb 15
Cd
14
Si
13
Ca
12
Al
11 11
12
Cu
13
14
15
16
17
18
Leachability Index by Experiment Fig. 4. Model predicted and measured values of observed diffusivity (Park and Batchelor, 2002) for ash waste treated by cement-based s/s (Kosson et al., 1993).
14 13
Ks=10-3 pH
12 11 10
-5
Ks=10
9
1
pH
Dimensionless Total Concentration
2
8 -7
7
Ks=10
6
ANC
5 4
0
3 0
Dimensionless Distance
1
Fig. 5. Concentration profiles predicted by multicomponent leach model for contaminants stabilized by pH-dependent sorption reactions (Ks is a coefficient measuring the strength of adsorption).
the bath to the interior of the waste. As the ANC components continue to be released, this pH profile moves toward the interior of the solid. Every point in the solid experiences a porewater pH that decreases with time. Most inorganic contaminants are stabilized at high pH, so as pH drops they will be converted to dissolved compounds. As they are released, they will experience a concentration gradient from that point to the bath and will diffuse down that gradient. However, the concentration of many contaminants in the pore water will also decrease in the direction toward
the interior of the solid because the pH is higher in that direction. Therefore, a portion of the contaminant that is released will also diffuse toward the interior of the solid where it will be re-stabilized by reactions that convert it to a solid phase. This results in total contaminant (mobile and immobile) concentrations that are higher than the initial values. This contaminant peak gradually moves inward. Therefore, the amount of contaminant released is determined by the balance of contaminant that diffuses outward and the amount that diffuses inward. The amount of ANC component in the waste will affect how fast the pH profile moves through the waste and therefore, how fast the contaminants are released. It is important to recognize that the chemical and physical factors that have the greatest affect on contaminant release are those that characterize the leached zone. Characterization of the unleached material will be insufficient to adequately predict release. An alternative approach is to use a lumped-parameter model that is calibrated by directly measuring contaminant release in sequential batch leaching tests. These tests are conducted by replacing the leaching solution at different time intervals and the amount of contaminant released during the leaching interval is determined. The lumped-parameter in the model is the observed diffusivity. This is often identified as the ‘‘effective diffusivity’’, but this paper uses that term to describe the diffusivity that is reduced from the molecular diffusivity solely by physical effects of pore tortuosity. This paper will apply the term ‘‘observed diffusivity’’ to the parameter that is affected by both chemical reactions of the contaminant as well as its physical transport. A variety of simple contaminant release models can be developed to describe leaching with the following assumptions: (1) no reactions; (2) homogeneous distribution of contaminant; (3) infinite bath (contaminant concentration equals zero at bath-solid boundary); and (4) infinite solid (concentration in most interior portion of solid remains constant at initial condition). The model can be solved for the concentration distribution as a function of time, which can be integrated to calculate the fraction of contaminant leached at any time: 0:5 Mt 4De t ¼ ð1Þ M0 pL2 where Mt, mass of contaminant leached; M0, mass of contaminant initially in waste; De, effective diffusivity; and L, maximum distance from interior of waste to the leaching solution. This model can be used to describe leaching of contaminants from wastes forms with other geometries by generalizing the definition of L to be the ratio of the waste volume to the area of the waste-leachate interface (L = V/A). This is can only approximate leaching from non-rectangular waste forms, but it will not result in substantial error when applied to early leaching times in which a small fraction of contaminant has been released. When simple reactions are assumed to stabilize the contaminant, the same general release model can be used except that it is expressed in
B. Batchelor / Waste Management 26 (2006) 689–698
have also been observed to follow this model, so it can be assumed to be generally applicable to contaminant release that is controlled by internal diffusion. Design of sequential leach tests that can accurately measure observed diffusivities for many different contaminants in different wastes is challenging, but standard procedures have been developed (American Nuclear Society, 1986; NNI, 1995; Kosson et al., 2002). Two major considerations are the L/S ratio and the time interval between replenishment of the leaching solution. If the L/S is too high or the time interval too low, the concentration of strongly stabilized contaminants will be very low and difficult to measure. If the L/S is too low or the time interval too high, the concentration in the bath will rise to levels that reduce the leaching rate. In this condition, the leach test does not provide a value of observed diffusivity that can be generally used, because it is determined by the specific leaching conditions in the test. This problem has been demonstrated by measurements as shown in Fig. 7 (Schwantes and Batchelor, 2005) and by leach models as shown in Fig. 8 (Kim and Batchelor, 2001; Park and Batchelor, 2002). In both the mechanistic and direct measurement modeling methods, the total amount of contaminant that can leach must be determined. The availability test provides a good method for determining this value. Contaminant degradation can also occur in wastes after disposal or reuse. This can be promoted by addition of reactants, such as those that promote reductive dechlorination, or it can be the result of reactions that do not need additional reactants, such as base-catalyzed hydrolysis or radioactive decay. Contaminant release in such systems becomes a competition between diffusive release and contaminant degradation. A combined degradation/release model has been developed to describe this situation (Hwang and Batchelor, 2000a). It assumes that the contaminant is degraded by a first-order reaction and that it
terms of an observed diffusivity (Dobs), rather than the effective diffusivity. The definition of the observed diffusivity depends on the types of simple reactions that are assumed. Explicit relationships have been determined for following reactions that are assumed to be at local equilibrium: sorption/desorption, precipitation/dissolution and dissolution by reaction with a component of the leaching solution (Batchelor, 1990). Furthermore, this release model has been found to apply to contaminant release that is controlled by more complex multi-component interactions. Fig. 6 shows results of a multi-component leach model that fits the square-root release model well. Many leaching experiments
Fraction Leached for MK14
0.030
0.025
0.020
0.015
0.010
0.005
0.000 0
5
10
15
695
20
Square Root of Dimensionless Time Fig. 6. Fraction leached as function of square root of time as predicted by multicomponent leach model assuming immobilization of metal contaminant by formation of hydroxide precipitates.
Ratio of Leached to Available Cd
8.0E-02
SIDR Y = 7.44E-3X - 8.11E-3 2 R = 0.969
6.0E-02
4.0E-02
ANS 16.1 Y = 9.64E-4X + 2.67E-3 2 R = 0.994
2.0E-02
0.0E+00
0
1
2
3
4
5
6
7
8
9
Square Root of Time (day) Fig. 7. Fraction leached measured in conventional sequential batch method and in method in which ion exchange resins maintain bath concentration at negligible levels (Schwantes and Batchelor, 2005).
696
B. Batchelor / Waste Management 26 (2006) 689–698
4.0E-03 Finite-Zn
3.5E-03
Waste Forms
Cumulative Fraction Leached
Infinite-Zn
3.0E-03 2.5E-03 2.0E-03 1.5E-03 Ql C l
1.0E-03 5.0E-04 Q gw C gw,in
0.0E+00 0
0.5
1
1.5
2
2.5
Square Root of Time
Q gw + Q l C gw,out
Fig. 10. Schematic of disposal scenario where leachate contacts waste forms and mixes with groundwater flow (Batchelor, 1997).
Fig. 8. Leach model predictions for assumption of infinite bath and finite bath (Kim and Batchelor, 2001).
can sorb. The material balance equation can be solved by analogy to heat transfer (Carslaw and Yaeger, 1959) and the fraction leached can be expressed in terms of dimensionless variables. Fig. 9 demonstrates how the ultimate amount of contaminant that is released decreases as the second Damko¨hler number increases. The second Damko¨hler number represents the relative importance of characteristic diffusion time to characteristic reaction time. When it is large enough, the degradation rate is fast enough to destroy the contaminant before it is released. An example of the application of release models to evaluating environmental impact of waste disposal is shown schematically in Fig. 10 (Batchelor, 1998). The scenario
described by this figure is one in which a waste is disposed in the ground and contacts a leachate flow that could be the result of percolation. Water contacts the surfaces of the waste forms and contaminants are released to it. The concentration of contaminant in the leachate can be controlled by equilibrium factors or kinetic factors. If the flow of leachate is low, the concentration will approach that expected at equilibrium, which can be determined by an equilibrium leach test. If the flow of leachate is fast enough, the release will be controlled by the kinetics of release, which can be described by a release model. The leachate concentrations controlled by each process can be compared and the one that is lower would be the one expected to occur. 5. Summary
1.00
Fraction Leached (Mt/M0)
kt/R=0.1
0.80
D aII = 0.3
kt/R=1.0 kt/R=10
D aII = 3.0
0.60
0.40
D aII = 30
0.20
0.00 0
0.5 Square Root of Dimensionless Time
1
Fig. 9. Fraction leached as function of square root of time for contaminant degraded by first order rate at various values of second Damko¨hler number. Symbols represent points associated with different values of first Damko¨hler number (kt/R).
A wide range of modifications of cement-based s/s methods have been applied to treat waste materials. The relative importance of producing treated waste that is strong (solidification) and producing a treated waste that has contaminants with low mobility and toxicity (stabilization) depends on the particular waste and disposal/reuse scenario. Waste-binder interactions can be critical to determining the extent of solidification and adverse interactions can be mitigated with many of the admixtures studied systems with only Portland cement. Contaminants are stabilized by a wide variety of different types of reactions, but most of them are strongly affected by pH. Therefore, contaminant release, which begins by reversal of the stabilization reactions, is strongly affected by pH changes. These changes are controlled by release of products of binder reactions that provide acid neutralizing capacity to the waste, so contaminant release is often controlled by release of these acid neutralizing components. Characterization of the effectiveness of s/s treatment should be based on
B. Batchelor / Waste Management 26 (2006) 689–698
determining the environmental impact of the treated waste after it is disposed or reused. The simplest method to do this is to estimate the amount of contaminant that might be released to the environment, either through aggressive testing (e.g., total waste analysis), more environmentally based tests (e.g., availability), or tests using specific conditions (e.g., TCLP). More detailed characterization methods can be used that either attempt to simulate the environmental conditions in the test procedure (microcosm tests) or predict performance by models that are based on more fundamental characteristics of the waste material. Modeling approaches can be based on separately describing chemical conversion of contaminants from immobile to mobile forms as functions of pH and their physical transport from the solid, or by combining the physical and chemical processes into one lumped parameter (observed diffusivity). These models can be applied to a variety of disposal/reuse scenarios to predict environmental impact. References American Nuclear Society, 1986. Measurement of the Leachability of Solidified Low-level Radioactive Wastes by a Short-term Test Procedure, ANSI/ANS-16.1-1986, American Nuclear Society, LaGrange Park, IL, USA. Abdel-Wahab, A., Batchelor, B., 2005. An equilibrium model for chloride removal from recycled cooling water using ultra-high lime with aluminum process. Water Environment Research 77 (6), 3059–3065. Barneyback, R.S., Diamond, S., 1981. Expression of pore fluids from hardened cement paste and mortars. Cement and Concrete Research 11, 279–285. Batchelor, B., 1990. Leach models: theory and application. Journal of Hazardous Materials 24 (2+3), 255–266. Batchelor, B., 1997. A framework for risk assessment of disposal of wastes treated by solidification/stabilization. Environmental Engineering Science 14 (1), 3–13. Batchelor, B., 1998. Leach models for contaminants immobilized by phdependent mechanisms. Environmental Science and Technology 32 (11), 1721–1726. Baur, I., Johnson, C.A., 2003a. The solubility of selenate-AFt (3CaO Æ Al2O3 Æ 3CaSeO4 Æ 37.5H2O) and selenate-AFm (3CaO Æ Al2O3 Æ CaSeO4 Æ xH2O). Cement and Concrete Research 33 (11), 1741–1748. Baur, I., Johnson, C.A., 2003b. Sorption of selenite and selenate to cement minerals. Environmental Science and Technology 37 (15), 3442–3447. Carslaw, H.S., Yaeger, J.C., 1959. Conduction of Heat in Solids, second ed. Oxford University Press, London. Conner, J.R., 1990. Chemical Fixation and Solidification of Hazardous Wastes. Van Nostrand Reinhold, NY, USA. Cote, P. 1986. Contaminant leaching from cement-based waste forms under acidic conditions, Ph.D. Thesis, McMaster University, Hamilton, Ont., Canada. Dijkstra, J.J., Meeussen, J.C.L., Comans, R.N.J., 2004. Leaching of heavy metals from contaminated soils: an experimental and modeling study. Environmental Science and Technology 38 (16), 4390–4395. European Committee for Standardization 2001. prEN14429 Characterisation of waste: leaching behavior tests – pH dependence test with initial acid/base addition, Technical Committee CEN/TC 292 WG6. Glasser, F.P., Marr, J., 1984. The effect of mineral additives on the composition of cement pore fluids. British Ceramic Proceedings 35, 419–429. Hills, C.D., Pollard, S.J.T., 1997. The influence of interference effects on the mechanical, microstructural and fixation characteristics of cement-
697
solidified hazardous waste forms. Journal of Hazardous Materials 52 (1997), 171–191. Hwang, I., Batchelor, B., 2000a. Degradative solidification/stabilization technology for chlorinated hydrocarbons. In: Wickramanayake, G.B., Gavaskar, A.R. (Eds.), Physical and Thermal Technologies: Remediation of Chlorinated and Recalcitrant Compounds. Batelle Press, Columbus, OH, USA. Hwang, I., Batchelor, B., 2000b. Reductive dechlorination of tetrachloroethylene by Fe(II) in cement slurries. Environmental Science and Technology 34 (23), 5017–5022. Hwang, I., Batchelor, B., 2001. Reductive dechlorination of tetrachloroethylene in soils by Fe(II)-based degradative solidification/stabilization. Environmental Science and Technology 35, 3792–3797. Hwang, I., Batchelor, B., 2002. Reductive dechlorination of chlorinated methanes in cement slurries containing Fe(II). Chemosphere 48 (10), 1019–1027. Hwang, I., Park, H.-J., Kang, W.-H., Park, J.-Y., 2005. Reactivity of Fe(II)/cement systems in dechlorinating chlorinated ethylenes. Journal of Hazardous Materials 118 (1–3), 103–111. Jeffers, P.M., Ward, L.M., Woytowitch, L.M., Wolfe, L.N., 1989. Homogeneous hydrolysis rate constants for selected chlorinated methanes, ethanes, ethenes and propanes. Environmental Science and Technology 23 (8), 965–969. Jung, B.M., Batchelor, B., 2004. Transformation of 1,1,1-trichloroethane by Fe(II) in cement slurries. In: Proceedings of the Fourth International Conference on Remediation of Chlorinated and Recalcitrant Compounds. Batelle Press, Monterey, California. Kang, W.-H., Hwang, I., Park, J.-Y., 2003. Kinetics of trichloroethylene degradation by steel converter slag amended with Fe(II). Preprints of extended abstracts presented at the ACS National Meeting, American Chemical Society, Division of Environmental Chemistry 43(1), 117–122. Kim, I., Batchelor, B., 2001. Empirical partitioning leach model for solidified/stabilized wastes. Journal of Environmental Engineering 127 (3), 188–195. Klich, I., Batchelor, B., Wilding, L.P., Drees, L.R., 1999. Mineralogical alterations that affect the durability and metals containment of aged solidified and stabilized wastes. Cement and Concrete Research 29, 1433–1440. Kosson, D.S., van der Sloot, H.A., Sanchez, F., Garrabrants, A.C., 2002. An integrated framework for evaluating leaching in waste management and utilization of secondary materials. Environmental Engineering Science 19 (3), 159–204. Kosson, D.S., Kosson, T.T., van der Sloot, H.A., 1993. Evaluation of solidification/stabilization treatment processes for municipal waste combustion residues. US Environmental Protection Agency, EPA/600/ R-93/167,Cincinnati, OH, USA. Kyi, A.A., Batchelor, B., 1994. An electrical conductivity method for measuring the effects of additives on effective diffusivities in Portland cement pastes. Cement and Concrete Research 24 (4), 752–764. Lea, F.M., 1971. The Chemistry of Cement and Concrete, third ed. Chemical Publishing Co., NY, USA. Lyman, W.J., Reehl, W.F., Rosenblatt, D.H., 1990. Handbook of Chemical Property Estimation Methods: Environmental Behavior or Organic Compounds. American Chemical Society, Washington, DC, USA. Mabey, W., Mill, T., 1978. Journal of Physical Chemical Reference Data 7 (2), 383–415. Means, J.L., Smith, L.A., Nehring, K.W., Brauning, S.E., Gavaskar, A.R., Sass, B.M., Wiles, C.C., Mashni, C.I., 1995. The Application of Solidification/Stabilization to Waste Materials. Lewis Publishers, Boca Raton, FL, USA. NNI (Dutch Standardization Institute), 1994. NEN 7341 Leaching characteristics of soil, construction materials and wastes – leaching tests – determination of the availability of inorganic constituents for leaching from construction materials and waste materials. Delft, The Netherlands. NNI (Dutch Standardization Institute), 1995. NEN 7345 leaching tests – determination of the leaching behaviour of inorganic components from
698
B. Batchelor / Waste Management 26 (2006) 689–698
building monolithic waste materials with the diffusion test. Delft, The Netherlands. Page, C.L., Vennesland, O., 1983. Pore solution composition and chloride binding capacity of silica-fume cement pastes. Materiaux et Constructions 16 (91), 19–25. Palmer, C.D., 2000. Precipitates in a Cr (VI)-contaminated concrete. Environmental Science and Technology 34 (19), 4185–4192. Park, J.-Y., Batchelor, B., 2002. A multi-component numerical leach model coupled with a general chemical speciation code. Water Research 36 (1), 156–166. Perkins, R.B., Palmer, C.D., 2000. Solubility of Ca6[Al(OH)6]2(CrO4)3 Æ 26H2O, the chromate analog of ettringite; 5–75 °C. Applied Geochemistry 15 (8), 1203–1218. Poellmann, By H., Auer, St., Kuzel, H.J., Wenda, R.P., 1993. Solid solution of ettringites. Part II: incorporation of tetrahydroborate(1) and chromate ðCrO2 4 Þ in ettringite (3CaO Æ Al2O3 Æ 3CaSO4 Æ 32H2O). Cement and Concrete Research 23 (2), 422–430. Ramabhadran, S., 1996. Solidification/stabilization of simulated uranium and nickel contaminated sludges. M.S.Thesis, Texas A&M University, College Station, TX, USA. Sabharwal, K., Batchelor, B., 1993. A multi-component leaching model for solidified/stabilized wastes and soils. In: Presented at Third International Symposium on Stabilization/Solidification of Hazardous, Radioactive and Mixed Wastes, Williamsburg, Virginia. Schlegel, M.L., Pointeau, I., Coreau, N., Reiller, P., 2004. Mechanism of europium retention by calcium silicate hydrates: an EXAFS study. Environmental Science and Technology 38 (16), 4423–4431. Schwantes, J.M., Batchelor, B., 2005. Simulated infinite-dilution leach test. Environmental Engineering Science 23 (1), 4–13. Schwarzenbach, R.J., Gschewend, P.M., Imboden, D.M., 2003. Environmental Organic Chemistry, seocnd ed. John Wiley & Sons, Hoboken, NJ, USA. Shi, C., Spence, R., 2004. Designing of cement-based formula for solidification/stabilization of hazardous, radioactive and mixed wastes. Critical Reviews in Environmental Science and Technology 34, 391–417. Son, S., 2002. Reductive Dechlorination of chlorinated hydrocarbons using Fe(II) and modified green rusts in degradative solidification/ stabilization. Ph.D. Dissertation, Texas A&M University, College Station, TX, USA.
Spence, R.D., Bostick, W.D., McDaniel, E.W., Gilliam, T.M., Shoemaker, J.L., Tallent, O.K., Morgan, I.L., Evans-Brown, B.S., Dodson, K.E., 1989. Immobilization of technetium in blast furnace slag grouts. In: 3rd International Conference on the Use of Fly Ash, Silica Fume, Slag & Natural Pozzolans in Concrete, Trondheim, Norway. Stegemann, J.A., Caldwell, R.J., Shi, C., 1997. Variability of field solidified waste. Journal of Hazardous Materials 52, 335–348. Taffinder, G.G., Batchelor, B., 1993. Measurement of effective diffusivities in solidified wastes. Journal of Environmental Engineering 119 (1), 17– 33. Taylor, H.F.W., 1990. Cement Chemistry. Academic Press, London. Taylor, M.A., Fuessle, R.W., 1994. Effects of accelerators on stabilization of KO61 hazardous waste. Hazardous Waste and Hazardous Materials 11 (4), 529–539. Trussell, S., Spence, R.D., 1994. A review of solidification/stabilization interferences. Waste Management 14 (6), 507–519. Trussell, S.A., Batchelor, B., 1996. Chemical characterization of pore water of a solidified hazardous waste. In: Gilliam, T.M., Wiles, C.C. (Eds.), Stabilization/solidification of hazardous, radioactive and mixed wastes, 3rd vol., ASTM STP 1240. ASTM, West Conshohocken, PA, USA. US EPA, 2004. Treatment Technologies for Site Cleanup, eleventh edition. EPA-542-R-03-009. Office of Solid Waste and Emergency Response, Washington, DC. van der Sloot, H.A., Dijkstra, J.J., 2004. Development of Horizontally Standardized Leaching Tests for Construction Materials: A Material Based or Release Based Approach, Identical leaching mechanisms for different materials, Energy Research Centre of the Netherlands, ECNC–04-060. Zhang, M., Reardon, E.J., 2003. Removal of B, Cr, Mo, and Se from wastewater by incorporation into hydrocalumite and ettringite. Environmental Science and Technology 37 (13), 2947–2952. Ziegler, F., Giere, R., Johnson, C.A., 2001. Sorption mechanisms of zinc to calcium silicate hydrate: sorption and microscopic investigations. Environmental Science and Technology 35 (22), 4556–4561. Zomeren, A.van, Comans, R.N.J., 2004. Contribution of natural organic matter to copper leaching from municipal solid waste incinerator bottom ash. Environmental Science and Technology 38 (14), 3927– 3932.