HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Figures Only Full Text (PDF) Free Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in ISI Web of Science Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via ISI Web of Science (5) Citing Articles via Google Scholar Google Scholar Articles by Keijzer, Th. J. S. Articles by Loch, J. P. G. Search for Related Content PubMed Articles by Keijzer, Th. J. S. Articles by Loch, J. P. G. GeoRef GeoRef Citation Agricola Articles by Keijzer, Th. J. S. Articles by Loch, J. P. G. Related Collections Soil Chemistry Soil Physics

DIVISION S-1—SOIL PHYSICS

Chemical Osmosis in Compacted Dredging Sludge

Dep. of Geochemistry, Faculty of Earth Sciences, Utrecht Univ., P.O. Box 80021, NL–3508 TA Utrecht, the Netherlands

* Corresponding author (keijzer{at}geo.uu.nl)

	ABSTRACT

TOP ABSTRACT INTRODUCTION THEORY MATERIALS AND METHODS Experimental Design and... RESULTS AND DISCUSSION CONCLUSIONS REFERENCES

 
Sediments of high clay content are known to exhibit semipermeable properties when compaction is sufficiently high, resulting in osmotically induced water and solute transport. The semipermeability of these materials is quantified by the reflection coefficient, . In the design of large dredging sludge depots in the Netherlands, osmotic transport is rarely taken into account. In a flexible wall permeameter, a bentonite and two dredging sludge samples were subjected to a chemical potential gradient to monitor water transport and to obtain values for . Chemical osmosis was observed in the bentonite and one of the sludge samples (56% clay), but was absent in the sludge with a relatively low clay content (26%). The measured are low in comparison with values obtained with the Fritz-Marine Membrane Model (FMMM) and a model presented by Bolt. The discrepancy between the theoretical and experimental values is explained by the assumptions made in both models. Several of these assumptions do not hold for the samples. If the observed semipermeability of the sludge is applied to an existing depot, chemical osmosis induces a water flux close to the maximum allowed advective flux laid down in Dutch legislation.

Abbreviations: AWy, Ankerpoort Wyoming bentonite • BK, Beerkanaal • BMR, Beneden Merwede River • CEC, cation-exchange capacity • FMMM, Fritz–Marine membrane model • TMS, Teorell–Meyer–Siever model • , reflection coefficient

	INTRODUCTION

TOP ABSTRACT INTRODUCTION THEORY MATERIALS AND METHODS Experimental Design and... RESULTS AND DISCUSSION CONCLUSIONS REFERENCES

 
IT IS WELL ESTABLISHED that clays can act as semipermeable membranes and are therefore capable of inducing osmotic water transport (e.g., Kemper, 1961; Hanshaw, 1962; Milne et al., 1964; Young and Low, 1965; Kharaka and Berry, 1973; Kharaka and Smalley, 1976; Graf, 1983; Fritz, 1986; Demir, 1988; Keijzer et al., 1999). Semipermeability is defined as the ability of a material to prevent the passage of a solute without affecting the passage of the solvent (e.g., Mitchell, 1993). The ability of a clay to act as a semipermeable membrane arises from the presence of diffuse double layers on the particle surfaces. Their share in the total pore solution increases by compaction of the clay, or the clayey material, under an overburden load. At very high loads, they may make up the whole pore solution and overlap. Compaction results in a higher concentration of cations and a lower concentration of anions in the double layer with respect to the equilibrium solution. At sufficient compaction, the water film present in the narrow pore between the clay layers is dominated by the double layers and the electrical restrictions they impose. Anions attempting to migrate through the narrow aqueous film are repelled by the negative charge of the clay platelets. This effect is known as negative adsorption or Donnan exclusion (Mitchell, 1993; Horseman et al., 1996). To maintain electrical neutrality in the external solution, cations will tend to remain with their co-ions. Thus, their movement across the clay will also be restricted (Fritz, 1986). Water and noncharged solutes are freely admitted to the membrane.

The above mechanism holds for charged membranes such as clays. However, there are also uncharged semipermeable membranes. In these uncharged membranes, rejection of a solute is on the basis of size or mobility. The result is that smaller components are less restricted during their passage through the membrane than larger ones. Semipermeable properties of clays with essentially no charge, like kaolinite, arise predominantly from their size-exclusion properties (e.g., Olsen, 1969, 1972). On the basis of pore-size distribution, Whitworth (1993) concluded that chalks may prove to be relatively efficient geological membanes.

The osmotic pressure difference induced by a chemical potential gradient across a semipermeable membrane can be calculated using Eq. [1]. The ability of the membrane to restrict solute transport is expressed as . For ideal membranes which restrict the passage of all charged solutes, is 1; for porous media without membrane properties (e.g., sand), equals 0 (Katchalsky and Curran, 1965). The value of depends on the type of clay, its surface charge and exchangeable cations (Kharaka and Berry, 1973; Kharaka and Smalley, 1976; Bolt, 1982b), porosity (Kharaka and Berry, 1973; Fritz, 1986), mean pore water concentration (Berry, 1969; Kharaka and Berry, 1973), and temperature (Haydon and Graf, 1986). The thickness of the membrane has no effect on the value of (Benzel and Graf, 1984).

Laboratory experiments on compacted, monomineral clay membranes, mostly bentonite and kaolinite, have resulted in the description of water transport as the result of coupling of different driving forces, such as electro-osmosis (e.g., Mitchell, 1993; Grundl and Michalski, 1996), thermoosmosis (e.g., Dirksen, 1969), chemical osmosis (e.g., Kemper, 1961; Kemper and Rollins, 1966; Olsen, 1969; Fritz and Marine, 1983), and reverse osmosis (e.g., McKelvey and Milne, 1962; Kharaka and Berry, 1973; Kharaka and Smalley, 1976; Graf, 1983; Benzel and Graf, 1984). Experimental data on the semipermeable behavior of naturally occurring clayey materials are limited to shales and siltstones (Young and Low, 1965; Yearsley, 1989). Field evidence on these transport mechanisms is also limited. Marine and Fritz (1981) reported anomalously high pore water pressures as a result of different pore water compositions separated by a shale. Hanshaw and coworkers (Berry and Hanshaw, 1960; Hanshaw and Hill, 1969) also used chemical osmosis to explain abnormal pore water pressures in aquifers as a mechanism to induce faulting. Only recently, Neuzil (2000) reported on an in situ chemical osmotic experiment in the Pierre Shale of South Dakota (USA) presenting evidence of chemical osmosis on a field scale.

Based on the laboratory observations and the theoretical framework of the diffuse double layer, we postulated that in coastal regions, where large differences in salt concentrations may exist across clay layers or clay lenses, water and solute transport should be governed not only by hydraulic but also by osmotic pressure gradients (Loch and Keijzer, 1996). In the western parts of the Netherlands, aquifers containing either fresh or saline water are separated by layers of high clay content. The hydrology of this region may well be influenced by chemical osmosis (De Haven, 1982). In the predictions of water and contaminant fluxes from waste storage sites (e.g., dredging sludge depots in coastal regions as existing or planned in the Netherlands), these gradients are neglected. One of these depots, De Slufter (Fig. 1), is constructed in a salt water aquifer without a bottom liner. The clay rich dredging sludge itself is a low permeable layer restricting the advective transport of both inorganic and organic contaminants. However, the ability of the clay layer to act as a semipermeable membrane and its capability of inducing water transport and thus advective transport was not taken into account in the assessment of contaminant emissions.

View larger version (19K): [in this window] [in a new window]   Fig. 1. Locations of the dredging sludge depot De Slufter near the Port of Rotterdam, the Netherlands, and the locations of the sampling sites in the Beerkanaal (BK) and Beneden Merwede river (BMR).

	THEORY

TOP ABSTRACT INTRODUCTION THEORY MATERIALS AND METHODS Experimental Design and... RESULTS AND DISCUSSION CONCLUSIONS REFERENCES

 
Nonhydraulically driven fluid flow is generally denoted as osmosis. Chemical osmosis refers to the flow of water induced by a chemical potential difference across a semipermeable membrane (e.g., a compacted clay layer). Chemical osmosis will occur until a counteracting hydraulic pressure difference is established equal to the osmotic pressure difference between the two solutions. The osmotic pressure difference, ₀ in kPa, between two solutions seperated by a membrane can be calculated using:

(1)

in which R is the gas constant (8.31451 J mol^-1K^-1), T is the absolute temperature, _w is the partial molar volume of water (m³ mol^-1), and a_w is the activity of the water at the low (a^low_w) and high (a^high_w) concentration side of the membrane. Thus, the actual driving force is not the concentration difference of the salt across the membrane but the activity gradient of the water across the membrane. The activity of water at different concentrations below 1 mol L^-1 can be found using (Alexander, 1990):

(2)

here is an integer representing the number of ions into which the electrolyte dissociates (e.g., for NaCl, equals 2), and m is the molality of the solution in mol kg^-1.

As discussed above, natural semipermeable membranes are seldom ideal. Their ideality is expressed as , which is formally defined as the ratio between the osmotic pressure and the hydraulic pressure (P) after equilibrium (i.e., at zero water flux):

(3)

Its value determines the maximum expected osmotic water flux induced by a chemical potential gradient (Staverman, 1952).

Two models readily available in the earth and soil science literature were used to predict the theoretical of the studied materials: (i) the FMMM (Marine and Fritz, 1981; Fritz and Marine, 1983; Fritz, 1986) based on the Donnan concept of the ionic equilibrium distribution adjecent to a clay particle surface (Mitchell, 1993); and (ii) a model by Bolt based on the Gouy–Chapman concept of the diffuse double layer, modified for an immobile part in the double layer adjacent to the clay particle surface (Groenevelt and Bolt, 1969, 1972; Bolt, 1982b).

The FMMM is the most accessible model for the prediction of for natural materials exhibiting semipermeable properties. It is proven to be applicable to deep sedimentary basins where semipermeability of geological formations like shales is expected (Marine and Fritz, 1981; Alexander, 1990; Horseman et al., 1996; Osborne and Swarbrick, 1997). The FMMM is a refinement of the Teorell–Meyer–Siever model (TMS) for a diffusion potential across a charged membrane (Hanshaw, 1962; Kharaka and Berry, 1973). The FMMM is modified for the mineralogy and porosity found under geological conditions (Fritz, 1986). The expression derived for in this model is:

(4)

K_s is a measure for the salt exclusion of the semipermeable membrane given by the ratio of the anion concentration in the membrane pores (_a) to the average concentration of solute external to the membrane (C_s). _c is the cation concentration in the membrane pores given by (Hanshaw, 1962; Marine and Fritz, 1981; Fritz, 1986):

(5)

where _a is:

(6)

CEC is the cation-exchange capacity in mol_c g^-1, _p is the particle density of the membrane material in g cm^-3, and _w is the volumetric water content or the porosity of the membrane.

Equation [4] contains three friction ratio coefficients, R. Firstly, R_m is the ratio of the friction coefficient (f) of the cations (c) and anions (a) with the solid membrane structure (m); thus R_m equals f_cm/f_am. It relates the tendencies of the charged species to be retarded by frictional resistance by the membrane walls. A value of R_m between 1.2 and 1.8 for the NaCl + H₂O system can be used (Kharaka and Berry, 1973; Fritz, 1986; Marcus, 1997). Secondly, R_wm is the ratio of the friction coefficients of the anions with the matrix of the membrane and with the water (w) in the membrane, R_wm equals f_am/f_aw. In the FMMM, it is assumed that for a highly compact membrane, R_wm > 1, whereas R_wm is taken to be 0 when there is little frictional resistance as in loosely compacted membranes. For geological membranes with a porosity above 40%, it is assumed that the value for R_wm is equal to 0 (Fritz and Marine, 1983). Finally, R_w is the ratio of the friction coefficients for the ions with the water in the membrane, R_w equals f_cw/f_aw. If it is assumed that the frictional resistance of the cation and anion with the water in the membrane structure is equivalent to that in the free solution, the reciprocal of the transport number (t₀) of the cation equals R_w + 1 (Fritz and Marine, 1983). The transport number is a measure for the percentage of charge that is transported by the cation or anion when the mobilities of the ions in solution differ. For the NaCl + H₂O system, the value for t^Na₀ equals 0.38 throughout the concentration range of 0 to 3 mol L^-1, thus R_w is approximately 1.63.

The dependency of on the porosity for various clay minerals with different cation-exchange capacities is clearly illustrated by the FMMM (Fig. 2). The model also shows the change in ideality for a membrane with decreasing average solute concentration across the membrane as was experimentally found by several investigators (e.g., Berry, 1969; Kharaka and Berry, 1973; Kharaka and Smalley, 1976).

View larger version (15K): [in this window] [in a new window]   Fig. 2. The reflection coefficient, , according to the Fritz-Marine Membrane Model (FMMM); top as a function of the porosity at an average concentration of 0.1 mol L-1, and bottom as a function of the average concentration at a given porosity of 0.5. Curves were calculated using Eq. [4] for monomineral membranes made of kaolinite, illite or montmorillonite with values for the cation-exchange capacity (CEC) of respectively 15, 25, and 100 cmolc kg-1. For the frictional coefficient, Rwm, an initial value of 0.1 was taken which decreased linearly in the porosity range between 0 and 0.4. Above a porosity of 0.4 the value of Rwm was taken to be zero. Values of the other frictional coefficients are given in the text.

(7)

for which the analytical solution can be derived (Bolt, 1982b):

(8)

in which ₀ is the Debije-Hckel reciprocal thickness of a diffuse double layer in m^-1 for a given equilibrium solution and d_l is the thickness in m of the nontruncated double layer on a charged surface (Fig. 3). The value for ₀ is given by (ßI₀)^1/2 where ß is 10.80 x 10¹⁵ mmol^-1 at 25°C, and I₀ is the ionic strength of the equilibrium solution in mol m^-3 (Bolt, 1982b). Furthermore, t represents the mathematical expression (u^1/2 - 1)/(u^1/2 + 1), in which u is the Boltzmann accumulation factor for monovalent counterions in the diffuse double layer. At x equals x_s and x equals x_t; the values, t_s and t_d, are given by:

(9)

(10)

View larger version (23K): [in this window] [in a new window]   Fig. 3. Schematic representation of the distribution of the ions in a normal (dotted lines) and truncated (solid lines) double layer. The position of the immobile and mobile water adjacent to the clay platelet for the truncated double layer according to the model by Bolt is also indicated.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION THEORY MATERIALS AND METHODS Experimental Design and... RESULTS AND DISCUSSION CONCLUSIONS REFERENCES

 
Samples
A Wyoming bentonite, (AWy), or Na-montmorillonite commercially available as Ankerpoort Colclay A90 was used (AWy; Table 1). This bentonite is used in the Netherlands in both sand-bentonite and cement-bentonite liners as a low permeable barrier to prevent the transport of water and contaminants from landfill sites.

	Experimental Design and Procedures

TOP ABSTRACT INTRODUCTION THEORY MATERIALS AND METHODS Experimental Design and... RESULTS AND DISCUSSION CONCLUSIONS REFERENCES

 
An instrument was designed to study the semipermeable properties of compacted clay rich samples under near surface conditions by directly measuring the water transport induced by a chemical potential gradient. This resulted in an instrument based on a triaxial cell (Daniel et al., 1984) where two stainless steel solution reservoirs are connected to a flexible wall clay permeameter (Fig. 4). In the permeameter, a sample of 50-mm diameter and 100-mm maximum thickness can be fitted, and is placed in a cylindrical cell. The cell can be filled with deaired water and pressurized to ensure the contact between the flexible wall and sample. The bottom end of the sample is connected to a reservoir containing a low concentration solution, referred to as the fresh water reservoir. The top end of the sample is connected to a reservoir containing a saline solution, referred to as the salt water reservoir. The reservoirs each contain a magnetically coupled gear pump, a manifold which connects each reservoir to calibrated glass standpipes, and pressure transducers (Fig. 4). Each reservoir also contains an electrical conductivity cell. The whole assembly is placed in a temperature controlled bench top chamber at 25.0 ± 0.2°C. A more detailed instrumental description can be found in Keijzer (2000).

View larger version (47K): [in this window] [in a new window]   Fig. 4. Schematic drawing of the experimental instrument showing the cell in which the flexible wall permeameter is located and both water reservoirs. The close-up shows the location of the sample between the porous stones and the latex membrane within the cell.

The hydraulic conductivity of the sample was measured prior to the osmotic experiment using a falling head permeability test (Daniel et al., 1984; Benson and Daniel, 1990). Under identical conditions but without a hydraulic gradient present, the sample was subjected to a chemical gradient. The fresh water reservoir was filled with a 0.01 M NaCl solution and the salt water reservoir with a 0.10 M NaCl solution, corresponding to an osmotic pressure difference of 421 kPa or 43 m H₂O at 25°C (Eq. [1]). The solutions were deaired prior to the filling of the reservoirs. The flux of water was measured using the calibrated standpipes and the induced hydraulic head was monitored using the pressure transducers. Diffusion or convection of salt was monitored using the electrical conductivity in the fresh water reservoir.

	RESULTS AND DISCUSSION

TOP ABSTRACT INTRODUCTION THEORY MATERIALS AND METHODS Experimental Design and... RESULTS AND DISCUSSION CONCLUSIONS REFERENCES

 
Sample Preparation and Hydraulic Conductivity
The samples obtained after saturation had a porosity generally around 0.5 (Table 2). This porosity is representative for naturally occurring clays under near surface conditions. Porosities are not as low as can be expected under geological conditions where they are generally below 0.3.

Model Predictions
The two previously described models were used to predict of the bentonite and the dredging sludge in the experimental setting.

The as a function of porosity for the samples, is shown in Fig. 5. The is calculated over the whole porosity range using Eq. [4] for the sample properties (Table 1) and the experimental conditions during the chemical osmosis experiment (Table 2). The porosity of a sample during the experiment being known, can be easily derived from these curves (Table 3).

View larger version (13K): [in this window] [in a new window]   Fig. 5. Reflection coefficient, , as a function of the porosity for the three samples according to the FMMM. The curves were calculated using Eq. [4] with a value of s of 0.055 mol L-1. For the frictional coefficient, Rwm, an initial value of 0.1 was taken which decreased linearly in the porosity range between 0 and 0.4. Above a porosity of 0.4 the value of Rwm was taken to be zero.

View larger version (16K): [in this window] [in a new window]   Fig. 6. The reflection coefficient, , of AWy as a function of the normalized mobile water layer, 0dl, according to Bolt. The curves were calculated using Eq. [8]. In the inset the curves for all the samples are shown between 5 0dl 10. As the value of 0 for the samples are of the same order of magnitude—0.02 for AWy, 0.04 for BK and 0.06 for BMR—the curves plot close to each other.

View larger version (10K): [in this window] [in a new window]   Fig. 7. The development of a differential hydraulic pressure, P = Psalt – Pfresh, with time across the AWy bentonite as a result of chemical osmosis. The line is the best exponential fit from which the maximum value for P was obtained and used to calculate the experimental reflection coefficient, of the sample.

View larger version (14K): [in this window] [in a new window]   Fig. 8. The increase in conductivity during the first 72 h as the result of diffusion of salt across the samples.

View larger version (11K): [in this window] [in a new window]   Fig. 9. The development of a differential pressure, P = Psalt – Pfresh, with time across the BK dredging sludge as a result of chemical osmosis. The line is the best exponential fit from which the maximum value for P was obtained and used to calculate the experimental reflection coefficient, , of the sample.

View larger version (12K): [in this window] [in a new window]   Fig. 10. The transport of water during the first 12 h for the Beerkanaal (BK) dredging sludge showing the flow of water from the fresh to the salt water reservoir.

(11)

in which J_w is the measured water flux in m³ s^-1, K_h is the hydraulic conductivity in m s^-1, A is the area of the sample in m² and ₀ is the absolute osmotic pressure difference in m H₂O across the sample with thickness, x in m.

The experimental values for obtained by the two different methods yield values which are in good agreement (Table 4). They are, however, considerably lower than the theoretical values for predicted by the two models (Table 3). The only exception is calculated for the BK sludge by the Bolt model. For this sample, the measured value is of the same order of magnitude as predicted. The most striking contrast between theory and practice is observed for the BMR sludge which, in this experiment, did not show any semipermeable properties whereas high values of were predicted.

View this table: [in this window] [in a new window]   Table 4. Observed reflection coefficients () for AWy and the two dredging sludges.

The FMMM was supported by both laboratory experiments on bentonite (Fritz and Marine, 1983) using hyperfiltration or reverse osmosis and field observations in the Dunbarton Triassic Basin, USA (Marine and Fritz, 1981). The authors found a good correspondence with their experimental work even for porosities of the bentonite around 0.5. However, there is a fundamental difference between their experiments with reverse osmosis and the direct measurement of osmotic water transport as used in this study. In a typical reverse osmosis experiment, the salt solution is forced through the clay membrane against the osmotic gradient; and the osmotic pressure difference between the two solutions has to be overcome. Subsequently, the semipermeability of the sample can be derived. In these experiments, the sample is generally confined in a rigid wall permeameter and subjected to an axially applied overburden pressure to ensure a good wall–sample contact. The experiment is rapid which made it popular in research of osmotic phenomena (e.g., Kharaka and Berry, 1973; Kharaka and Smalley, 1976; Fritz and Marine, 1981; Benzel and Graf, 1984; Demir, 1988). But because reverse osmosis normally is applied in a short period of time it reduces the effects of salt diffusion into the sample. Therefore, the produced is a maximum value. As the clay is not an ideal membrane, salt will diffuse causing the overlap between adjacent double layers to reduce, and the clay to flocculate, making the membrane more permeable for the salt. As the experiment progresses in time, the diffusion front extends further into the clay, reducing the overall of the sample. Thus, the will decrease with time in the experiment (Fig. 11). In reverse osmosis experiments, this deterioration of the samples as a semipermeable membrane has been overlooked and not incorporated into the theoretical framework for the prediction of . The fact that is not an intrinsic value for a given sample or geological membrane was also observed by Whitworth and Fritz (1994).

View larger version (9K): [in this window] [in a new window]   Fig. 11. Deterioration of the reflectance coefficient, for the Beerkanaal (BK) dredging sludge during the experiment. The values for were calculated using the water flux from the fresh to the salt water reservoir.

Complete dispersion of the clay platelets also does not hold in our samples. The specific surface area (A_s) needed to calculate the should be the total area (Bolt, 1982b) as can be determined with the EGME–method (Churchman et al., 1991). When using the total A_s for the samples (Table 1), Bolt's model results in a for AWy of 0.95, for BK of 0.69 and for BMR of 0.92. These values correspond closely with those found with the FMMM listed in Table 3. However, the values calculated in Table 3 with the Bolt model were obtained using the external area determined with the N₂–BET method (Aylmore et al., 1970; Feller et al., 1992). This is the surface area of assemblies or aggregates of particles. The external area is a more realistic value to use as the samples. Especially the sludge samples are not completely dispersed, and most certainly contain assemblies of clay-sand minerals. For the AWy and the BK, the clay content was high enough to result in semipermeable properties but the clay content of BMR (approximately 26%) was too low for a working membrane. Additionally, the clay minerals present in the BMR sludge, mainly illite and kaolinite, are less favorable than those present in the BK sludge, which include smectite (Table 1). Clays of low CECs are inherently less ideal semipermeable membranes than those with of high CEC (Kharaka and Berry, 1973; Fritz, 1986). Furthermore, both models seem to overestimate the influence of the porosity on the value of as can be concluded from the discrepancy between the model values for the BMR sludge. It was, however, this sample which did not show any semipermeable properties.

Despite all assumptions that do not hold for the samples tested, the experimental values, especially for the BK sludge, correspond closely with predicted values found with the model by Bolt. This is the result of the possibilities it offers to adjust the input values, e.g., the A_s, needed to calculate a . The FMMM offers less possibilities to do so, and is too specifically derived for geological membranes. Finally, it should be noted that it can be shown that the friction coefficients in the FMMM vary only slightly for membranes at low and medium porosities and therefore have only a limited effect on the predicted value of the (Keijzer, 2000).

Implications for Dredging Sludge Depots
As dredging sludge is capable to act as a semipermeable membrane, it is possible to determine the possible osmotically induced water flux through a depot and compare these values with other driving forces.

As it is customary to reduce the hydraulic gradient across the sludge, the hydraulic induced water flux is generally in the order of 2 mm yr^-1. As a result, the maximum allowed advective water flux from a sludge in depot laid down in Dutch legislation should not exceed 2 mm yr^-1 (VROM, 1993).

Assuming that the hydraulic conductivity of the sludge in the Slufter depot is higher than under the experimental conditions as a result of the lower consolidation grade, a K_h-value of 10^-9 m s^-1 is taken. Using the found for the BK sludge (Table 4) and an approximated thickness of the sludge layer currently present in the depot of 20 m, a water flux of 1.5 mm yr^-1 is calculated. Thus, the osmotically induced water flux is of the same order of magnitude as the maximum allowed water flux from a sludge in depot.

Therefore, a minimization of the hydraulic gradient only will not automatically lead to a minimization of the water flux and of transport of contaminants. As a consequence, the design criteria laid down in Dutch legislation will be exceeded.

	CONCLUSIONS

TOP ABSTRACT INTRODUCTION THEORY MATERIALS AND METHODS Experimental Design and... RESULTS AND DISCUSSION CONCLUSIONS REFERENCES

 
To study osmotic transport in materials high in clay content, AWy and two dredging sludges from the Rotterdam harbor were subjected to a salt concentration gradient in the laboratory. Chemical osmosis was observed in the AWy and in one of the dredging sludge samples, quantified by of 0.015 to 0.030, and 0.019 to 0.022, respectively. No chemical osmosis was observed in a second dredging sludge sample of low clay content. The s were found to be significantly lower than predicted by the FMMM (Fritz, 1986), and by a model based on the diffuse double layer theory presented by Bolt (1982b). The discrepancy between the theoretical and experimental values can be explained by the assumptions made in both models, which do not apply to the samples used in this study. Neither model is suitable for the complex mineralogical and ionic composition of the studied materials. The two models also seem to overestimate the influence of the porosity on the value of . It is therefore essential that before applying one of the models for the prediction of the semipermeability of natural materials of high clay content, a thorough analysis is made of the applicability of their underlying assumptions.

The implications for dredging sludge as a semipermeable membrane in the transport of water and solutes are relevant when hydraulic gradients are eliminated. The semipermeability found for the BK sludge should be treated as a minimum value, but is still capable to induce a substantial water flux at salt concentrations present in coastal systems.

	ACKNOWLEDGMENTS

 
The authors thank Dr. G. Kamerling at the Directorate General for Public Works and Water Management (Rijkswaterstaat) for partly supporting this work. The experimental setup benefited from the comments and adjustments made by Pieter Kleingeld, Tony van de Gon Netscher and Hans Bliek. The manuscript was improved by the comments of Prof. Dr. C.H. van der Weijden. This is contribution 20010103 of the Netherlands Research School of Sedimentary Geology.

Received for publication March 20, 2000.

	REFERENCES

TOP ABSTRACT INTRODUCTION THEORY MATERIALS AND METHODS Experimental Design and... RESULTS AND DISCUSSION CONCLUSIONS REFERENCES

 

• Alexander, J. 1990. A review of osmotic processes in sedimentary basins. British Geological Survey, report no. WE/90/12, Nottingham, UK.
• Aylmore, L.A.G., I.D. Sills, and J.P. Quirk. 1970. Surface area of homoionic illite and montmorillonite clay minerals as measured by the sorption of nitrogen and carbon dioxide. Clays Clay Miner. 18:91–96.
• Barbour, S.L., and D.G. Fredlund. 1989. Mechanisms of osmotic flow and volume change in clay soils. Can. Geotech. J. 26:551–562.
• Benson, C.H., and D.E. Daniel. 1990. Influence of clods on hydraulic conductivity of compacted clays. J. Geotech. Eng. 116:1231–1248.
• Benson, G.H., and J.M. Trast. 1995. Hydraulic conductivity of thirteen compacted clays. Clays Clay Miner. 43:669–681.[Abstract]
• Benzel, W.M., and D.L. Graf. 1984. Studies of smectite membrane behavior: Importance of latter thickness and fabric at 20°C. Geochim. Cosmochim. Acta 48:1769–1778.
• Berry, F.A.F. 1969. Relative factors influencing membrane filtration effects in geological environments. Chem. Geol. 4:295–301.
• Berry, F.A.F., and B.B. Hanshaw. 1960. Geological field evidence suggesting membrane properties of shales. p. 209. In International Geological Congress, Norden. DET Berlingske Bogtrykkeri, Copenhagen, Denmark.
• Bolt, G.H. 1955. Analysis of the validity of the Gouy–Chapman theory of the electric double layer. J. Colloid Sci. 10:206–219.
• Bolt, G.H. 1982a. Surface interaction between the soil solid and the soil solution. p. 43–53. In G.H. Bolt (ed.) Soil chemistry: A. basic elements. Vol. 5a. Elsevier, Amsterdam, the Netherlands.
• Bolt, G.H. 1982b. Electrochemical phenomena in soil and clay systems. p. 387–432. In G.H. Bolt (ed.) Soil chemistry: B. Physico-chemical models. Vol. 5b. Elsevier, Amsterdam, the Netherlands.
• Churchman, G.J., C.M. Burke, and R.L. Parfitt. 1991. Comparison of various methods for the determination of specific surfaces of subsoils. J. Soil Sci. 42:449–461.
• Daniel, D.E., S.J. Trautwein, S.S. Boynton, and D.E. Foreman. 1984. Permeability testing with flexible-wall permeameters. Geotech. Test. J. 7:113–122.
• De Haven, J.C. 1982. Lowland underground water as a physicochemical system. p. 300–315. In Tweede Nota Waterhuishouding. SDU, The Hague, the Netherlands.
• Demir, I. 1988. Studies of smectite membrane behavior: Electrokinetic, osmotic, and isotopic fractionation. Geochim. Cosmochim. Acta 52:727–737.
• Dirksen, C. 1969. Thermo-osmosis through compacted saturated clay membranes. Soil Sci. Soc. Am. J. 33:821–826.
• Feller, C., E. Schouller, F. Thomas, J. Rouiller, and A.J. Herbillon. 1992. N2–BET specific surface areas of some low activity clay soils and their relationships with secondary constituents and organic matter contents. Soil Sci. 153:293–299.
• Fritz, S.J. 1986. Ideality of clay membranes in osmotic processes: A review. Clays Clay Miner. 34:214–223.[Abstract]
• Fritz, S.J., and I.W. Marine. 1983. Experimental support for a predictive osmotic model of clay membranes. Geochim. Cosmochim. Acta 47:1515–1522.
• Graf, D.L. 1983. Chemical osmosis, reverse chemical osmosis and the origin of subsurface brines. Geochim. Cosmochim. Acta 46:1431–1448.
• Groenevelt, P.H., and G.H. Bolt. 1972. Permiselective properties of porous as calculated from diffuse double layer theory. p. 241–255. In First Symposium of Fundamental Transport Phenomena, Haifa, Israel.
• Groenevelt, P.H., and G.H. Bolt. 1969. Non-equilibrium thermodynamics of the soil-water system. J. Hydrol. 7:358–388.
• Grundl, T., and P. Michalski. 1996. Electroosmotically driven water flow in sediments. Water Res. 30:811–818.
• Hanshaw, B.B. 1962. Membrane properties of compacted clays. Ph.D. thesis, Harvard University.
• Hanshaw, B.B., and G.A. Hill. 1969. Geochemistry and hydrodynamics of the Paradox Basin region Utah, Colorado and New Mexico. Chem. Geol. 4:263–294.
• Haydon, P.R., and D.L. Graf. 1986. Studies of smectite membrane behavior: Temperature dependence, 20–80°C. Geochim. Cosmochim. Acta 50:115–121.
• Horseman, S.T., J.J.W. Higgo, J. Alexander, and J.F. Harrington. 1996. Water, gas and solute movement through argillaceous media. Nucl. Energy Agency, Organization for Economic Co–operation and Development, report no. CC–96/1, Paris, France.
• Katchalsky, A., and P.F. Curran. 1965. Nonequilibrium thermodynamics in biophysics. Harvard University Press, Cambridge, MA.
• Keijzer, Th.J.S. 2000. Chemical osmosis in natural clayey materials. Geologica Ultraiectina 196. Ph.D. thesis, Utrecht University.
• Keijzer, Th.J.S., P.J. Kleingeld, and J.P.G. Loch. 1999. Chemical osmosis in compacted clayey material and the prediction of water transport. Eng. Geol. 53:151–159.
• Kemper, W.D. 1961. Movement of water as effected by free energy and pressure gradients: II. Experimental analysis of porous systems in which free energy and pressure gradients act in opposite directions. Soil Sci. Soc. Am. Proc. 25:260–265.
• Kemper, W.D., and J.B. Rollins. 1966. Osmotic efficiency coefficients across compacted clays. Soil Sci. Soc. Am. Proc. 30:529–534.
• Kharaka, Y.K., and F.A.F. Berry. 1973. Simultaneous flows of water and solutes through geological membranes: I. Experimental investigation. Geochim. Cosmochim. Acta 37:2577–2603.
• Kharaka, Y.K., and W.C. Smalley. 1976. Flow of water and solutes through compacted clays. AAPG Bull. 60:973–980.[Abstract]
• Loch, J.P.G., and Th.J.S. Keijzer. 1996. Chemical osmosis in contaminated clayey materials. p. 367. In 1996 V.M. Goldschmidt Conference, vol. 1. Heidelberg, Germany. Cambridge Publications, Cambridge, UK.
• Marcus, Y. 1997. Ion properties. Marcel Dekker, New York, NY.
• Marine, I.W., and S.J. Fritz. 1981. Osmotic model to explain anamalous hydraulic heads. Water Resour. Res. 17:73–82.
• McKelvey, J.G., and I.H. Milne. 1962. The flow of salt solution through compacted clay. Clays Clay Miner. 9:248–259.
• Milne, I.H., J.G. McKelvey, and R.P. Trump. 1964. Semi-permeability of benonite membranes to brines. Soil Sci. Soc. Am. Bull. 48:103–105.
• Mitchell, J.K. 1993. Fundamentals of soil behavior. John Wiley, New York, NY.
• Neuzil, C.E. 2000. Osmotic generation of ‘anomalous’ fluid pressures in geological environments. Nature 403:182–184.[Medline]
• Olsen, H.W. 1969. Simultaneous fluxes of liquid and charge in saturated kaolinite. Soil Sci. Soc. Am. Proc. 33:338–344.
• Olsen, H.W. 1972. Liquid movement through kaolinite under hydraulic, electric and osmotic gradients. Soil Sci. Soc. Am. Bull. 56:2022–2028.
• Osborne, M.J., and R.E. Swarbrick. 1997. Mechanisms for generating overpressure in sedimentary basins. Soil Sci. Soc. Am. Bull. 81:1023–1041.
• Staverman, A.J. 1952. Non-equilibrium thermodynamics of membrane processes. Trans. Faraday Soc. 48:176–185.
• Van Olphen, H. 1963. Clay colloid chemistry. Wiley, London, UK.
• VROM. 1993. Beleidsstandpunt verwijdering baggerspecie. Rep. No. vergaderjaar 1993–1994, 23450 nr. 1. Ministerie van VROM, Tweede Kamer der Staten Generaal, The Hague, the Netherlands.
• Whitworth, T. M. 1993. Hyperfiltration-induced isotopic fractionation: Mechanisms and in the subsurfac. Ph.D. thesis, Purdue University.
• Whitworth, T.M., and S.J. Fritz. 1994. Electrolyte induced solute permeability effects in compacted smectite membranes. Appl. Geochem. 9:533–546.
• Yearsley, E.N. 1989. Transport properties and coupled flow phenomena in Salton Sea Scientific Drilling Program Cores. M.S. thesis, Colorado School of Mines.
• Young, A., and P.F. Low. 1965. Osmosis in agrillaceous rocks. AAPG Bull. 49:1004–1008.

This article has been cited by other articles:

				C. D. Shackelford, C. D. Shackelford, and J.-M. Lee THE DESTRUCTIVE ROLE OF DIFFUSION ON CLAY MEMBRANE BEHAVIOR Clays and Clay Minerals, April 1, 2003; 51(2): 186 - 196. [Abstract] [Full Text] [PDF]

This Article Abstract Figures Only Full Text (PDF) Free Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in ISI Web of Science Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via ISI Web of Science (5) Citing Articles via Google Scholar Google Scholar Articles by Keijzer, Th. J. S. Articles by Loch, J. P. G. Search for Related Content