| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
im
nekb
a Performance Assessments Unit, Belgian Nuclear Research Centre (SCK·CEN), Boeretang 200, B-2400 Mol, Belgium
b Dep. of Environmental Sciences, Univ. of California, Riverside, CA 92521
c USDA-ARS, U.S. Salinity Lab., 450 W. Big Spring Rd., Riverside, CA 92507
* Corresponding author (djacques{at}sckcen.be).
All rights reserved. No part of this periodical may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, recording, or any information storage and retrieval system, without permission in writing from the publisher.
Received 27 April 2007.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONCoupled Reactive Transport in...Case Study: Uranium Leaching...ConclusionsAppendixREFERENCES |
|---|
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONCoupled Reactive Transport in...Case Study: Uranium Leaching...ConclusionsAppendixREFERENCES |
|---|
An integrated approach is necessary for understanding and unraveling the complex interplay of different soil processes. Integration of all pertinent processes into a numerical transport modeling tool is indispensable for evaluating the various coupled effects involved (e.g., Mayer et al., 2002; Metz et al., 2002; Davis et al., 2004b; Lichtner et al., 2004; Appelo and Postma, 2005; Nützmann et al., 2005; Nuclear Regulatory Commission, 2006; Carrillo-González et al., 2006). Steefel et al. (2005) briefly discussed some historical steps in the development of such tools. Many reactive transport codes of different levels of complexity have been developed during the last few decades for saturated porous media (e.g., van der Lee and De Windt, 2001). More recently, similar codes started to consider also flow and transport in the unsaturated zone. Examples are 3DHYDROGEOCHEM (Yeh and Cheng, 1999), CORE2D (Samper et al., 2000), MIN3P (Mayer et al., 2002), Pollutrans (Kuechler and Noack, 2002), UNSATCHEM-2D (imnek and Suarez, 1994), RETRASO (Saaltink et al., 2004), HYDRUS-1D (imnek et al., 2005), and HP1 (Jacques et al., 2006b). Thus far, most applications of these multicomponent transport models to both saturated and unsaturated conditions have been limited to steady-state flow scenarios.
Transient flow in soils can substantially affect geochemical conditions for both equilibrium and kinetic reactions. Experimental evidence shows that alternations in infiltration (due to precipitation and irrigation) and evaporation periods, or changes in temperature, can strongly influence geochemical conditions near the soil surface. Such effects have been demonstrated during experimental cycles in laboratory setups (e.g., Öztürk and Özkan, 2004), seasonal variations in the field (e.g., Berner et al., 1998), and for artificial recharge of groundwater (Greskowiak et al., 2005). The interaction of discharging groundwater and shallow groundwater dynamics with temporal variations in atmospheric conditions also affects the soil water composition (e.g., de Mars et al., 1997; de Mars and Garritsen, 1997; Joris, 2005). Slattery and Ronnfeldt (1992) showed that seasonal variations in pH were partly related to ionic strength. Gonçalves et al. (2006) reported an experimental study to quantify salinization and alkalization risks in lysimeters exposed to natural atmospheric conditions (rainfall and evapotranspiration) and manually irrigated with waters of different quality. Overall salinity and Na concentrations changed significantly during the year. Sodium concentrations near the soil surface peaked during dry periods after irrigation events, but subsequently decreased again to initial values during rainfall periods. The experimental data were successfully simulated using HYDRUS-1D, which couples an ion chemistry module with a water flow and solute transport model.
A similar behavior for Na was found by Jacques et al. (2008) in a hypothetical simulation of the 30-yr migration of heavy metals under atmospheric conditions. During periods with a precipitation deficit (i.e., when potential evaporation is larger than precipitation), the pH near the soil decreased at the same time as the water content. The pH decrease was caused by increased ionic strength (i.e., higher concentrations) and the accumulation of anions such as Cl– near the soil surface due to upward flow. In addition, the total amount of Cd in soil water also increased during the summer due to (i) increased competition of monovalent cations for the exchange sites (especially with Na, which is a more mobile element than K and thus accumulates near the soil surface during upward flow), and (ii) increased aqueous complexation of Cd and Cl.
Water content variations also influence the rate of kinetic reactions due to their effects on the wetted reactive area. The reactive surface area is a critical parameter affecting the rate of mineral dissolution. It is not straightforward to link reactive surface area to experimentally (e.g., BET mineral surface areas) or geometrically determined values. Brantley (2003) reported several reasons for this discrepancy: (i) the difference in reactivity of different surface sites (e.g., edge vs. basal plane sites for phyllosilicates); (ii) aging effects; (iii) the occurrence of deep etch cavities inducing transport-limited microenvironments for dissolution; and (iv) coatings of natural mineral surfaces with organic matter, Fe, Al, and Si. The critical parameter during transient unsaturated flow is the specific solid–water interfacial area, which varies as a function of the water content. In a theoretical study, Saripalli et al. (2006) showed that the interfacial area changed only slightly near saturation since the larger pores entail only a small percentage of the interfacial area, whereas a much stronger decrease in interfacial area occurs under dry conditions. Simulations for steady-state flow conditions involving different degrees of saturation illustrated that changing interfacial areas can have an important effect on solute fluxes. This effect of the degree of saturation was recently illustrated by Kuechler and Noack (2007) for pyrite–calcite dissolution during unsaturated flow. Only 5% of the batch-measured reactive surface area for pyrite was required to describe unsaturated column experiments at a degree of water saturation of 11%.
These examples illustrate the important effects transient water flow can have on multicomponent solute transport. Our main objective is to further demonstrate these types of interactions between transient flow and biochemical processes in soils. We first discuss various feedbacks between transient flow and other processes, and then present a hypothetical example involving the migration of several chemical species during a 200-yr time period. The governing vadose zone flow and chemical equations of the coupled HP1 multicomponent reactive transport code (Jacques and imnek, 2005) are given next. Our focus is especially on nonlinear feedbacks between different processes relevant for soil systems. Then we show the important effects of the upper boundary condition (steady-state vs. transient flow) on the long-term migration of the main constituents (Ca and P) and a trace element (U) of an inorganic P fertilizer applied annually for 200 yr to a soil profile. The example demonstrates how transient water contents and fluxes affect soil pH, and hence the retention, bioavailability, and fluxes of various solute components.
|
Coupled Reactive Transport in the Vadose Zone |
|---|
|
TOPABSTRACTINTRODUCTIONCoupled Reactive Transport in...Case Study: Uranium Leaching...ConclusionsAppendixREFERENCES |
|---|
imnek et al., 2005) and the PHREEQC-2 biogeochemical model (Parkhurst and Appelo, 1999). More information about HP1, including the code itself, can be found at www.sckcen.be\hp1 (verified 18 Apr. 2008).
Processes and Reactions
Physical transport is described using the conservation equations for momentum, fluid mass, solute mass, and energy. Typically for soils and other porous materials, momentum conservation is described with the Darcy–Buckingham equation in combination with the conservation equation for water to yield the Richards equation as given by Eq. [1.1] in Table 1
(symbols are defined in the Appendix). We assume that the gas phase is continuous and at atmospheric pressure, which simplifies the description of fluid flow from a multiphase to a single-phase flow problem. To solve the Richards equation, the soil water content, 
[L3 L–3], as a function of the soil water pressure head, h [L], and the unsaturated hydraulic conductivity, K [L T–1], as a function of or h must be defined (e.g., Brooks and Corey, 1966; van Genuchten, 1980). The Richards equation describes uniform water flow in soils, but not any preferential or nonequilibrium flow. Alternative formulations for such conditions include various dual-porosity or dual-permeability models (see imnek et al., 2003; Köhne et al., unpublished data, 2008; imnek and van Genuchten, 2008).
|
imnek et al., 2005), including the soil water content. Heat and water are also simultaneously transported as water vapor, which provides strong coupling between soil water and heat dynamics (e.g., Saito et al., 2006). The latter process is not included in HP1.
The conservation of solute mass combined with Fick's law of diffusion results in the advection–dispersion equation as given by Eq. [1.3] in Table 1. This equation is not only applicable to dissolved chemical elements, but also to colloids (e.g., van Genuchten and imnek, 2004; DeNovio et al., 2004; imnek et al., 2006) and microorganisms (e.g., Taylor and Jaffé, 1990; Jin et al., 2002; Rockhold et al., 2004, 2005; Gargiulo et al., 2007). Note that Eq. [1.3] in Table 1 is expressed in terms of the total concentration C [mol L–1] of a given component j:
 | [1] |
ji is the stoichiometric coefficient of the jth component in the reaction equation for the ith species, and Nsa is the number of secondary species. Equation [1.3] is based on the assumption that diffusion coefficients of all species are equal (see also Lichtner, 1996; Mayer, 1999). The Ro,j term in Eq. [1.3] represents sources–sinks for components in the aqueous phase, accounting for various heterogeneous equilibrium and kinetic reactions and homogeneous kinetic reactions. Depending on the type of component, different reactions and processes will be involved.
Heterogeneous reactions involve chemical elements in other than the liquid phase and include adsorption reactions described by adsorption isotherms (linear, Freundlich, or Langmuir isotherms), cation exchange (e.g., Voegelin et al., 2001; Bruggenwert and Kamphorst, 1982), or surface complexation reactions (e.g., see Goldberg et al. [2007] for a review of soil systems), mineral dissolution–precipitation, and gas dissolution–exsolution. Except for adsorption reactions described by isotherms, the other three types of reactions, together with aqueous speciation, account for multicomponent interactions. Their equilibrium state is determined by mass-action laws such as given by Eq. [1.4–1.7]
in Table 1. See Morel and Hering (1993), Langmuir (1997), and Appelo and Postma (2005) for background on activities, exchange conventions, and surface complexation models. Heterogeneous equilibrium is obtained by inserting the number of moles of the species (derived from the mass-action equations) into the mole- and charge-balance equations. Heterogeneous reactions can also be treated kinetically (e.g., Dang et al., 1994; Tebes-Stevens et al., 1998).
Kinetic homogeneous reactions are reactions that occur solely in the liquid phase and include radioactive decay, biologically mediated degradation, and solute uptake by roots (Barber, 1995; Tinker and Nye, 2000; Hopmans and Bristow, 2002; Nowack et al., 2006). Examples of complex kinetic rate equations are mineral dissolution and precipitation rates based on transition-state theory (e.g., Lasaga, 1995; Mayer et al., 2002), Monod-type rate equations (e.g., Schäfer et al., 1998; Salvage and Yeh, 1998), and biofilm models (e.g., Cunningham and Mendoza-Sanchez, 2006) for biodegradation and microbial growth such as those given by Eq. [1.8–1.10] in Table 1.
Additional heterogeneous reactions may occur for colloids and microorganisms, such as attachment–detachment processes at the solid phase, straining, filtration, partitioning of colloids to the air–water interface, and size exclusion (e.g., Wan and Tokunaga, 1997, 2002; Bradford et al., 2003, 2006; imnek et al., 2006; Gargiulo et al., 2007). Biomass production and decay are other important processes that need to be taken into account when predicting the fate of microorganisms in soil systems.
Feedback
The heat transport equation (Eq. [1.2]) is coupled to the Richards equation (Eq. [1.1]) through the effects of the fluid flux on heat advection and the apparent thermal conductivity. Conversely, temperature affects surface tension, viscosity, and fluid density and, consequently, the soil hydraulic properties (e.g., Zhang et al., 2003).
Integrating the water flow and advection–dispersion equations into one numerical tool allows one to account for the effects of water flow on solute fate and transport. Solute advection and hydrodynamic dispersion are directly influenced by water fluxes, while many chemical and biological reactions depend strongly on water contents. Water contents also indirectly influence soil O2 and CO2 concentrations, which have a direct effect on pH and consequently on many chemical and biological reactions and processes. Although CO2 transport and production processes are considered in HYDRUS-1D, they have not yet been incorporated into HP1. Solute concentrations affect the density of the liquid phase and can also influence root water uptake due to salinity stress (van Genuchten, 1987; Hopmans and Bristow, 2002). Soil temperatures may be affected by heat produced or consumed during various geochemical and biological reactions. The reverse feedback (i.e., the effect of temperature on solute transport and reactions) is, however, much more important. Solute transport and reactions additionally are strongly affected by the temperature dependence of the diffusion coefficient and by thermodynamic constants and reaction rates.
How and what processes are incorporated into the reaction term may have a considerable effect on the predicted mobility of various components. Many unsaturated flow and transport models coupling water flow and solute and heat transport incorporate reaction terms for single components to account for such reactions as equilibrium or nonequilibrium sorption by using linear, Freundlich, or Langmuir isotherms and first- or zero-order kinetics. In these types of models, the concentration of one element has no effect on the fate of other elements except when considering a sequential decay reaction. By including interactions between different components, one can calculate aqueous speciation, competitive adsorption described by cation exchange or surface complexation reactions, or complex kinetic reactions (see Table 1). Aqueous speciation can enhance the apparent solubility, adsorption potential, and mobility of some of the elements due to cation–ligand complexation. Examples of such complexation reactions are U complexes (Waite et al., 1994; McKinley et al., 1995; Morrison et al., 1995; Lenhart and Honeyman, 1999; Lenhart et al., 2000; Barnett et al., 2002) or the formation of soluble Al– or Fe–organic matter complexes (Jansen et al., 2002; Carrillo-González et al., 2006). As discussed above, variably saturated flow affects many geochemical variables (e.g., pH, concentrations) that consequently influence geochemical equilibrium. A linear equilibrium single-component adsorption model in which the solute distribution between the liquid and solid phases is described using the distribution coefficient, Kd, leads to completely different concentration patterns than those predicted using a competitive cation exchange model (e.g., Gonçalves et al., 2006) or a surface-complexation model (Zhu, 2003). This is due to the inability of the Kd approach to account for spatial and temporal variations in the soil water chemistry and water contents. These variations lead to changes in the ratio between exchanger and fluid content and consequently affect competition between cations for the exchange complex (Steefel et al., 2005).
An important feedback of geochemical reactions on transport processes is the effect of mineral dissolution and precipitation on not only the overall porosity of the system, but also on the pore-size distribution. This feedback affects all transport-related soil properties. These properties or parameters can be updated in various ways. For example, diffusion coefficients can be evaluated using Archie's law by introducing a cementation factor (Soler and Lasaga, 1998; van der Lee et al., 2003), while saturated hydraulic conductivities can be modified using macroscopic or grain-scale models (e.g., Le Gallo et al., 1998), and the unsaturated hydraulic properties using various physical models (Freedman et al., 2004) or pedotransfer functions (Finke, 2006).
In addition to geochemical reactions, microbial colonization and microbially produced compounds can also significantly impact the hydraulic regime. Microbial processes, for example, may lower the capillary fringe or induce locally dry zones (Yarwood et al., 2006). Microbial processes can affect not only the physical properties of a porous medium by clogging pores, producing capsules, or generating gases, but also the air–liquid interfacial tension or the solid–liquid contact angle due to the production of surface-active compounds (Yarwood et al., 2006). Rockhold et al. (2002) modeled such changes by using pressure head scaling factors. Several of the processes and feedbacks discussed here (notably the effects of chemical precipitation–dissolution and microbial growth on porosity and pore size distribution and CO2 transport and production) are not yet included into the current version of the HP1 simulator (Jacques and imnek, 2005).
|
Case Study: Uranium Leaching from Soils Following Long-Term Mineral Phosphorus Fertilization |
|---|
|
TOPABSTRACTINTRODUCTIONCoupled Reactive Transport in...Case Study: Uranium Leaching...ConclusionsAppendixREFERENCES |
|---|
 | [2] |
 | [3] |
 | [4] |
i
et al., 1992). Uranium concentrations in P-bearing fertilizers have been reported to be in the range of 300 to 3000 Bq kg–1 of fertilizer (European Commission, 1999), or 1700 to 9200 Bq kg–1 of fertilizer for both 238U and 234U (Cogné, 1993).
The environmental and radiological impacts of the application of inorganic P fertilizers are, in general, relatively modest. Spalding and Sackett (1972) attributed the increase of U in North American rivers to applications of large amounts of P fertilizers to agricultural lands. Rothbaum et al. (1979) found that most U originating from applied superphosphate was retained within the plow layer of arable soils or in the organic surface layers under grassland. Barii et al. (1992) concluded that >20% of the annually deposited U by fertilizers is transported to drainage channels in the Kanovci area of the Republic of Croatia. Their data indicated that high U concentrations in surface waters and shallow groundwater were caused by phosphate fertilizer applications. On the other hand, Zielinski et al. (1997) found that leaching of U from inorganic P fertilizers contributed much less to high U concentrations in drainage water than natural or irrigation-enhanced leaching from soils. The radiological impact expressed as dose is also quite limited (Saueia and Mazzilli, 2006).
Zielinski et al. (1997) pointed out that the mobility and the nature of P and U fixation and precipitation depend on soil conditions, such as pH, moisture, mineralogy, and texture. Guzman et al. (2002) provided a number of processes controlling U mobility. The list included sorption on solid surfaces, formation of complexes with Ca, PO42–, and CO32–, and (co-)precipitation of U–P, Ca–U, Al–P, or Fe–P (acid soils) or Ca–P (alkaline soils) minerals (see also Zielinski et al., 1997; Brady, 1990).
To model the subsurface transport of U, either naturally occurring in soils or applied with mineral fertilizers, the interaction of U with other elements and the solid phase needs to be considered. The amount of U adsorption to the solid phase depends on the specific composition of both the solid phase (e.g., clay minerals, Fe and Al oxides, and organic matter) and the aqueous phase (e.g., pH, complexing agents such as Cl–, SO42–, PO42–, or the presence of competing cations for cation exchange or surface complexation). The amount of Fe oxides in the soil strongly affects the pH dependence of U sorption (Barnett et al., 2000). Changing chemical conditions significantly influences the mobility of U in soils and sediments (Kohler et al., 1996). Here we use HP1 to obtain insight into the complex system of interacting geochemical processes that govern Ca, P, and U mobility in the soil when an inorganic P fertilizer is applied annually. Our study is an update of preliminary calculations by Jacques et al. (2006a). We focus especially on the effects of the imposed upper boundary condition (constant vs. transient atmospheric boundary conditions) on the migration of U in an undisturbed acid soil profile. The analysis of migration of Ca, P, and U is performed in terms of temporal variations in the distribution coefficient Kd to illustrate the limitation of such an empirical approach, which was illustrated in previous studies by Bethke and Brady (2000), among others.
Problem Definition
Phosphorus fertilizers are often applied annually to agricultural fields, partly in inorganic form. In Flanders (northern Belgium), the amount of applied P that originates from mineral fertilizers decreased from 13.9 x 106 kg P in 1990 to 2.38 x 106 kg P in 2003 (MIRA Team, 2004). To simulate the long-term migration of P and U through the soil, the following assumptions were made: (i) an average P fertilization of 1 g P m–2 (or 6.35 x 106 kg P for all of Flanders) is applied each year on 1 May in the form of Ca(H2PO4)2; and (ii) 1 kg of Ca(H2PO4)2 contains 10–3 mol of U. Based on a specific activity 2.96 x 106 Bq mol–1 U, 2960 Bq of alpha activity exclusively due to 238U is present in 1 kg of Ca(H2PO4)2. Thus, a fertilization of 1 g P m–2 corresponds to 1.61 x 10–2 mol of Ca(H2PO4)2 containing 3.77 x 10–6 mol of U.
The soil profile studied here is a dry Spodosol located at the "Kattenbos" site (Lommel, Belgium) consisting of seven soil horizons in the top 1 m. The soil has a typical leached E horizon between 7 and 19 cm below the surface and enriched (in organic matter and Fe oxides) Bh horizons between 19 and 28 cm. We note that the top horizons are not typical for an agricultural soil, but rather reflect an undisturbed profile covered by natural vegetation (i.e., heather). The profile was selected for our modeling purposes because it was well characterized in terms of physical and geochemical properties. Table 2 gives the locations of the different soil horizons and their soil hydraulic parameters as determined on undisturbed soil samples of 100 cm3 (Seuntjens, 2000; Seuntjens et al., 2001a).
|
Three types of geochemical equilibrium reactions were taken into account: (i) aqueous speciation reactions; (ii) cation exchange reactions on organic matter; and (iii) surface complexation reactions on Fe oxides. The thermodynamic data for the aqueous species were taken from the WATEQ4F database (Ball and Nordstrom, 1991), complemented with data for U from Langmuir (1997).
The cation exchange complex was assumed to be associated solely with organic matter. Organic matter typically consists of an assemblage of different functional groups and types of adsorption sites. A multisite cation exchange complex allows simulations of the acid–base properties of the organic matter and the corresponding increase in the cation exchange capacity with increasing pH (e.g., Appelo et al., 1998). Appelo et al. (1998) included proton exchange on a multisite cation exchange complex to mimic the presence of different functional groups. Six cation exchangers were assumed to be present, each with a different Ke value for the half reaction for H+:
 | [5] |
|
FeOH as a neutral Fe oxide surface) as follows: FeOH2+, FeO–, FeH2PO4, FeHPO4–, FePO42–, FeSO4–, FeOHSO42–, FeF, FeHF–, FeOMg+, FeOCa+, and FeOUO2+. Equilibrium constants for weak sites were taken from Dzombak and Morel (1990) (see also Parkhurst and Appelo, 1999). The capacity of the surface was derived from the amount of Fe2O3 as measured in the soil profile assuming 0.875 reactive sites per mole of Fe (Waite et al., 1994). The cation exchange capacity and the reactive sites of each horizon are given in Table 4
.
|
imnek and Suarez, 1993; Suarez and imnek, 1993). Meteorological data from two weather stations (from Brogel for years 1969–1985 and from Mol/Geel for 1986–1998) were used to define precipitation, P [L T–1], and potential evaporation, Ep [L T–1], rates, the latter based on the Penmann equation (Penmann, 1948). Statistical information about the imposed climatological time series is given in Jacques et al. (2008). To obtain a 200-yr time series for modeling purposes, the 30-yr period was simply repeated several times. A free-drainage boundary condition was assumed at the bottom of the soil profile.
Element concentrations in the rainwater were obtained from Stolk (2001) for Station 231 located in Gilze-Rijen (the Netherlands), which is sufficiently close to the investigated site. The solution composition was based on the average of 13 measurements during 1999 (Table 5 ). The U concentration of the rainwater was set equal to an arbitrary low value of 10–24 mol L–1 to have the required non-zero initial concentration for PHREEQC simulations. The P fertilizer was assumed to be applied every year on 1 May in the form of Ca(H2PO4)2 in 1 cm of irrigation water with the same composition as rainwater. The fertilizer was assumed to be completely dissolved in the irrigation water, thus neglecting any possible kinetics in the dissolution process of the fertilizer grains. Before the 200-yr simulation, a 5-yr steady-state flow simulation was performed without fertilizer applications to establish chemical initial conditions in topsoil horizons in equilibrium with the chemistry of the inflowing rainwater.
|
Surface speciation was also calculated for the same pH range, the rainwater composition, the soil air partial pressures of O2 and CO2, and the total aqueous U concentration of 10–9 mol L–1. The Na concentration was adapted to obtain the required pH and charge balance. Figure 1a shows the most abundant surface species on the Fe oxides. The surface of the Fe oxides was mostly protonated. Below pH 5.5, FeH2PO4 was the second most abundant surface species. The amounts of FeOH, FeHPO4–, and FeOUO2+ increased substantially across the considered pH range. For example, FeOUO2+ increased up to seven orders of magnitude. The amount of adsorbed U consequently also increased drastically. Figure 1b shows the ratio of total adsorbed U to total U in the system together with the distribution of U between the two sorption processes involved (i.e., cation exchange on organic matter and surface complexation on Fe oxides). Almost all U was in solution at low pH since both the cation exchange complex and the surface sites were mostly protonated. While about 50% of the total U was adsorbed around pH 3, almost all U was adsorbed at pH 4. The relative contribution of U adsorbed by cation exchange compared with surface complexation decreased with increasing pH. This is consistent with the experimental findings of McKinley et al. (1995). We found that U adsorption on the cation exchange complex increased up to pH 4.7 (results not further shown). Since the Na concentration in solution increases with increasing pH owing to the Na addition for the purpose of obtaining charge balance, Na will increasingly replace the other cations on the exchange complex at higher pH. Note that above pH 6, U adsorption will decrease again when dissolved CO2 is present due to the formation of U(VI)–CO3 complexes (Davis et al., 2004a; Waite et al., 1994). Almost all P was adsorbed in the studied pH range, with 99.3 and 99.8% P being adsorbed at pH 2 and 6, respectively (results not shown).
|
|
Different solution compositions affect the relative mobility between Ca, P, and U. For rainwater without P fertilizer, U was less mobile than P at pH values above 4.3 to 4.5, depending on the soil horizon; however, U was always more mobile when the rainwater contained P fertilizer. The relative mobility of P and Ca did not depend only on the water composition, but also on the soil horizon. In the A horizon with a high exchange capacity and low surface complex capacity, Ca was less mobile than P at pH values >4.3 for rainwater without P fertilizer and 3.5 for rainwater with P fertilizer. In the Bh2 horizon with a relatively high surface complex capacity, the Kd of P was always larger that the Kd of Ca.
Calcium, Phosphorus, and Uranium Depth Profiles
Figure 3
shows distributions vs. depth of total Ca, P, and U per cubic meter of soil at selected times for both the steady-state and transient water flow simulations. Total concentrations were obtained by summing the amounts in the aqueous phase, on the cation exchange complex (for Ca and U), and on the surface complex (for Ca, U, and P). Note the different concentration scales for P in plots for the first 50 and the last 150 yr. Calcium accumulated mainly in the upper horizon during the first 50 yr, although more accumulated in the Bh1 horizon under transient-flow conditions. During the next 100 yr, the amount of Ca in the two Bh horizons gradually increased. This occurred somewhat faster for the transient than the steady-state simulation. After 200 yr, however, the amount of Ca in the top 50 cm was quite similar for the two cases.
|
Uranium showed the largest differences between the two simulations, especially during the first 50 yr. Uranium migrated relatively fast, similar to P, through the A and E horizons. As discussed with Fig. 2, U was more mobile than P when the pH was <4.3, especially at relatively high Ca and P concentrations. Much U accumulated in the Bh horizon during the last 150 yr. Uranium transport was again slightly faster for the transient simulation than during steady-state flow.
Temporal Variations in the Linear Distribution Coefficient
Figures 4
and 5
show time series of pH and the Kd values for Ca, P, and U at depths of 5 (A horizon) and 25 (Bh2 horizon) cm, respectively. A long-time trend in Kd was observed for both steady-state and transient flow simulations. This trend is related to long-time changes in the amount of Ca, P, and U in the soil solution due to application of the P fertilizer. Steady-state solute transport conditions are reached when Kd becomes constant for the steady-state flow simulation. This condition was reached faster for P and U (at about 40–50 yr) than for Ca (at about 100 yr) at 5-cm depth. This is similar to the conclusion drawn from Fig. 3. Since the total concentrations increased from the initial condition, the Kd at 5-cm depth decreased by almost one order of magnitude for Ca, by one-third for P, and slightly more than one order of magnitude for U. The increase in competition between elements for the surface sites had a larger effect on the long-time sorption potential than the long-time pH increase on the Kd values. A different picture emerged for the Bh2 horizon at a depth of 25 cm (Fig. 5). The increase in pH during the 200-yr simulation caused here an increase in adsorption potential for U (increasing Kd). The competition with P for the surface complex was limited since P migrated more slowly through the Bh horizons than U (see Fig. 3). After 200 yr, the peak in U occurred at a depth of about 25 cm, whereas the peak in P was at 20 cm. The Kd of P at the 25-cm depth did not show a pronounced long-time trend until 175 yr, which further reflects the limited transport of P to this depth.
|
|
The seasonal variations in pH and Kd are pronounced close to the soil surface and for early simulation times when P and U are not yet at their long-time steady state (at 5-cm depth and for the first 40 yr). The Kd values of Ca and U changed up to two orders of magnitude during the first 20 yr. Although the difference in total amounts of the three elements between the steady-state and transient flow simulations was relatively small (see Fig. 3), the aqueous concentrations differed dramatically. This may have a profound effect on processes that are dependent on aqueous concentrations, such as nutrient uptake by roots.
Fluxes in the Soil and at the Bottom of the Soil Profile
The solute mass flux density, Js [M L–2 T–1], is the sum of the mass flux due to advection of the dissolved solute and the mass flux due to hydrodynamic dispersion and molecular diffusion. In terms of the resident fluid concentration, Clr [M L–3], Js may be written as (Jury and Roth, 1990)
 | [6] |
|
|
Conclusions |
|---|
|
TOPABSTRACTINTRODUCTIONCoupled Reactive Transport in...Case Study: Uranium Leaching...ConclusionsAppendixREFERENCES |
|---|
We illustrated the feedbacks between the different processes, as well as the multiple interactions that may occur between the chemical species present, by using the HP1 reactive transport simulator to predict the migration of Ca, P, and U following inorganic P fertilization through a layered soil. The numerical simulations showed that the distribution coefficients for Ca, P, and U depend strongly on pH, the composition of the aqueous phase, the mineralogical properties of the soil horizons, and the amount and type of competing elements. Simulations also showed that P is a stronger competitive element for U than Ca. In horizons with large amounts of P (e.g., the A horizon), the sorption potential and Kd of U did not follow the long-term increase in pH due to the stronger counter effect of P sorption on the surface complex. On the other hand, when U and Ca reached a certain horizon (e.g., the Bh2 horizon in our example), while P did not, the Kd for U closely followed the increase in pH. Furthermore, yearly additions of Ca, P, and U in the form of P fertilizer resulted in a long-term decrease in Kd of the upper horizons. Short-term temporal variations in the water contents and fluxes, when transient flow was considered, caused significant pH variations, which produced very significant variations in the Kd values for Ca and U by up to two orders of magnitude. Leaching of Ca and U from the soil profile occurred faster in the transient flow simulation than in the steady-state simulation as a result of the interplay between changing hydrologic and geochemical soil conditions.
|
Appendix |
|---|
|
TOPABSTRACTINTRODUCTIONCoupled Reactive Transport in...Case Study: Uranium Leaching...ConclusionsAppendixREFERENCES |
|---|
|
|
ACKNOWLEDGMENTS |
|---|
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONCoupled Reactive Transport in...Case Study: Uranium Leaching...ConclusionsAppendixREFERENCES |
|---|
i, D., S. Luli, and P. Mileti. 1992. Radium and uranium in phosphate fertilizers and their impact on the radioactivity of waters. Water Res. 26:607–611.imnek, M. Bettahar, M.Th. van Genuchten, and S.R. Yates. 2003. Modeling colloid attachment, straining, and exclusion in saturated porous media. Environ. Sci. Technol. 37:2242–2250.[Medline]imnek, M. Bettahar, M.Th. van Genuchten, and S.R. Yates. 2006. Significance of straining in colloid deposition: Evidence and implications. Water Resour. Res. 42:W12S26, doi:10.1029/2005WR004791.imnek, S. Sauvé, and D. Adriano. 2006. Mechanisms and pathways of trace elements in soils. Adv. Agron. 91:111–178.imnek, P. Ustohal, H. Vereecken, and E. Klumpp. 2007. Bacterial transport and deposition under unsaturated conditions: The role of the matrix grain size and the bacteria surface protein. J. Contam. Hydrol. 92:255–273.[CrossRef][Web of Science][Medline]imnek, T.B. Ramos, J.C. Martins, M.J. Neves, and F.P. Pires. 2006. Multicomponent solute transport in soil lysimeters irrigated with waters of different quality. Water Resour. Res. 42:W08401, doi:10.1029/2005WR004802.imnek. 2005. User manual of the multicomponent variably saturated transport model HP1: Description, verification and examples. Version 1.0. BLG-998. SCK·CEN, Mol, Belgium.imnek, D. Mallants, and M.Th. van Genuchten. 2006a. Modelling uranium leaching from agricultural soils to groundwater as a criterion for comparison with complementary safety indications. Mater. Res. Soc. Symp. Proc. 932:1057–1064.imnek, D. Mallants, and M.Th. van Genuchten. 2006b. Operator-splitting errors in coupled reactive transport codes for transient variably saturated flow and contaminant transport in layered soil profiles. J. Contam. Hydrol. 88:197–218.[CrossRef][Web of Science][Medline]imnek, D. Mallants, and M.Th. van Genuchten. 2008. Modelling coupled water flow, solute transport and geochemical reactions affecting heavy metal migration in a podzol soil. Geoderma (in press), doi:10.1016/j.geoderma.2008.01.009.imnek, and B.-P. Mohanty. 2006. Numerical analysis of coupled water, vapor, and heat transport in the vadose zone. Vadose Zone J. 5:784–800.imnek, J., C. He, L. Pang, and S.A. Bradford. 2006. Colloid-facilitated solute transport in variably saturated porous media: Numerical model and experimental verification. Vadose Zone J. 5:1035–1047.imnek, J., N.J. Jarvis, M.Th. van Genuchten, and A. Gärdenäs. 2003. Review and comparison of models for describing non-equilibrium and preferential flow and transport in the vadose zone. J. Hydrol. 272:14–35.[CrossRef]imnek, J., and D.L. Suarez. 1993. Modeling of carbon dioxide transport and production in soil: 1. Model development. Water Resour. Res. 29:487–497.[CrossRef]imnek, J., and D.L. Suarez. 1994. Two-dimensional transport model for variably saturated porous media with major ion chemistry. Water Resour. Res. 30:1115–1133.[CrossRef]imnek, J., and M.Th. van Genuchten. 2008. Modeling nonequilibrium flow and transport processes using HYDRUS. Vadose Zone J. 7:782–797 (this issue).imnek, J., M.Th. van Genuchten, and M. ejna. 2005. The HYDRUS-1D software package for simulating the one-dimensional movement of water, heat, and multiple solutes in variably-saturated media. Version 3.0. HYDRUS Softw. Ser. 1. Dep. of Environmental Sciences, Univ. of California, Riverside.imnek. 1993. Modeling of carbon dioxide transport and production in soil: 2. Parameter selection, sensitivity analysis, and comparison of model predictions to field data. Water Resour. Res. 29:499–513.[CrossRef]imnek. 2004. Integrated modeling of vadose zone flow and transport processes. p. 37–69. In R.A. Feddes et al. (ed.) Unsaturated-zone modelling: Progress, challenges and applications. Frontis, Wageningen, the Netherlands.This article has been cited by other articles:
|
|
 |
|
  J. C. da Silva, E. do Amaral Vargas Jr., and O. Sracek Modeling Multiphase Reactive Transport in a Waste Rock Pile with Convective Oxygen Supply Vadose Zone J., November 17, 2009; 8(4): 1038 - 1050. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. Simunek and S. A. Bradford Vadose Zone Modeling: Introduction and Importance Vadose Zone J., May 27, 2008; 7(2): 581 - 586. [Full Text] [PDF]
|
|
|
|
|
|
J. Simunek, M. Th. van Genuchten, and M. Sejna Development and Applications of the HYDRUS and STANMOD Software Packages and Related Codes Vadose Zone J., May 27, 2008; 7(2): 587 - 600. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. Szegedi, D. Vetterlein, H. Nietfeld, R. Jahn, and H.-U. Neue New Tool RhizoMath for Modeling Coupled Transport and Speciation in the Rhizosphere Vadose Zone J., May 27, 2008; 7(2): 712 - 720. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| The SCI Journals | Agronomy Journal | Crop Science | |||
| Journal of Natural Resources and Life Sciences Education |
Soil Science Society of America Journal | ||||
| Journal of Plant Registrations | Journal of Environmental Quality |
The Plant Genome | |||
(
jm
js