Tracers and Modelling in Hydrogeology (Proceedings of the TraM'2000 Conference held at Liège, Belgium, May 2000). IAHS Publ. no. 262, 2000. 303 Laboratory experiments for diffusion transport processes in fractured chalk K. WITTHÛSER, H. HÔTZL Applied Geology, University of Karlsruhe, Kaiserstrasse e-mail: [email protected] 12, D-76128 Karlsruhe, Germany B. REICHERT Geological Institute, University of Bonn, Nufiallee 8, D-53115 Bonn, Germany W. STICHLER GSF-Institute of Hydrogeology, Ingolstaedter Landstr. 1, D-85764 Neuherberg, Germany R. NATIV Hebrew University of Jerusalem, Seagram Center of Soil and Water, Faculty of Agriculture, PO Box 12, Rehovol 76100, Israel 2 1 Abstract Effective diffusivities of various artificial tracers of 0.59-2.78 E-6 cm s" for Maastrichtian chalk (Danish), and of 0.919-6.22 E-7 cm s" for Eocene (Israeli) chalk samples were obtained with the help of through-diffusion experiments. The variations can be explained by Archie's law with a distinct cementation factor for the chalk varieties tested. The Eocene chalk shows small relative diffusivity resulting in an extremely large m value of 3.3 due to its narrow and wide range of pore-throat sizes. The rock capacity factors obtained indicate retardation of pyranine in all chalk samples and batch experiments performed in parallel confirmed the sorptive behaviour of pyranine in chalk. 2 1 INTRODUCTION Diffusive mass transfer between mobile and immobile phases is an important transport mechanism and a central factor for sorptive uptake and release of solutes in fractured media. The importance of matrix diffusion as a retardation process for contaminant transport in fractured media is well known from risk analyses for nuclear and underground waste deposits. Most of the studies deal with diffusion of radioactive substances in the crystalline host rocks tested. Little data is available for other substances and/or other rock types. In the framework of an international joint project (EC: ENV4-CT97-0441) an interdisciplinary research group is investigating contaminant transport, monitoring techniques, and remediation strategies in fractured chalk across Europe. Our work encompasses the investigation of diffusion processes in different chalk varieties, compared to other fractured media, by means of laboratory experiments, as well as with analytical and numerical modelling. This paper deals mainly with throughdiffusion experiments with chalk samples from Denmark and Israel performed with various substances frequently used as artificial tracers in field investigations. The tracers being compared are lithium, bromide, uranine, pyranine, sodiumnapthionate, amidorhodamine G, sulphorhodamine B and deuterium. 304 K. Witthûser et al. EXPERIMENTAL SETUP Chalk samples collected from boreholes in Sigerslev, Denmark (Maastrichtian chalk) and in the Negev Desert, Israel (Eocene chalk), were saturated and then used for measuring the diffusion coefficients using diffusion cell devices (e.g. Feenstra et al., 1984). As far as possible the samples were taken perpendicular to fracture surfaces. We ensured a complete saturation by evacuating the samples in a desiccator before leaving them for at least two weeks in degassed water. BASIC EQUATIONS The one-dimensional diffusive flux J i of a solute in a porous media due to a gradient in concentration c can be described by Fick's first law: c dc S dc (1) dx ' x dx where D is the effective diffusion coefficient, D is the aquatic diffusion coefficient, 8 is the constrictivity and x is the tortuosity. Note that the definition of the effective diffusion coefficient includes the transport porosity s . The change in concentration at a point in the porous media due to diffusive transport of a sorbing solute is described by Fick's second law: -A aq 2 e aq t / ^ \dc 2 ^ dc (2) where e is the total porosity, p is the bulk density and Kd is the sorption coefficient for a linear isotherm. The term in parentheses can be summarized to the rock capacity factor a. An analytical solution of equation (2) for a porous slab (Carslaw & Jaeger, 1959) is used to describe the observed concentration change (Fig. 1) in the receiving cell: C(0 2 e n V 2 DE%t AdC a 6 n rr, s 2 a d (3) Co denotes the concentration of tracer reservoir cell, d the thickness (1 cm) and A the surface area (15.5 cm ) of the rock sample. The two fit parameters, the effective diffusivity D and the rock capacity factor a, were initialized with values determined from the slope of the linear part of the breakthrough curve (steady-state method, e.g. Bradbury & Green, 1985) and then estimated with a least-square routine (LevenbergMarquardt-algorithm). This procedure takes into account the transient phase (Fig. 1) of the breakthrough curve (i.e. the time to develop a linear concentration gradient in the rock sample) for the determination of the diffusion coefficient and the rock capacity factor. The resulting parameter and the total porosity (mercury porosimetry) of the samples are given in Tables 1 and 2. The determined effective diffusion coefficients were used to calculate the formation factor F (relative diffusivity D ) and the exponent m (cementation factor) of Archie's law (Archie, 1942; Klingenberg, 1951), both depending only on the 2 e Laboratoiy experiments for diffusion transport processes in fractured chalk 305 properties of the rock samples: 1 F = D = D ID e aq = s'" (4) The goal of the application of Archie's law is to establish a reasonable relationship for the investigated chalk varieties, which can be used for the prediction of the diffusion coefficient for any kind of tracer or substance of interest, thus avoiding timeconsuming diffusion experiments. T a b l e 1 Results of diffusion cell e x p e r i m e n t s for M a a s t r i c h i a n chalk s a m p l e s (Sigerslev, D e n m a r k ) . 2 1 D (cm s' ) cc(-) D„ (cm s~') c 2 2 1 D (cm s" ) s(-) m F{-) m (-) Uranine Lithium Pyranine Eosin Amidorhodamine G Sulphorhodamine 1.79 E-6 2.625 6.82 E-7 5.90 E-7 1.071 5.51 E-7 6.79 E-7 2.566 2.64 E-7 9.39 E-7 1.196 7.85 E-7 7.57 E-7 2.895 2.61 E-7 4.54 E-6 0.5143 0.395 1.4 6.3 E-6 0.5143 0.094 3.6 3.81 E-6 0.443 0.178 2.1 3.29 E-6 0.4992 0.286 1.8 3.7 E-6 0.5143 0.204 2.4 Bromide Deuterium B Sodiumnaphtionate 8.45 E-7 1.412 5.98 E-7 1.31 E-6 1.587 8.24 E-7 2.78 E-6 10.087 2.76 E-7 4.23 E-6 1.245 3.51 E-6 3.6 E-6 0.4992 0.234 2.1 5.69 E-6 0.443 0.230 1.8 1.19 E-5 0.4992 0.234 2.1 2.15 E-5 0.5143 0.196 2.5 T a b l e 2 Results of diffusion cell e x p e r i m e n t s for E o c e n e chalk s a m p l e s ( N e g e v desert, Israel). White chalk (24 m depth): Uranine Lithium A , ( c n r V ) <*(-) 2 1 D„ (cm s" ) s(-) F(-) m(-) 9.19 E-8 0.386 2.38 E-7 0.352 0.0202 3.73 6.22 E-7 0.432 1.44 E-6 0.389 0.0987 2.45 Pyranine Grey chalk (113 m depth): Uranine Pyranine 1.07 E-7 1.654 6.44 E-8 0.3084 0.0280 3.04 3.46 E-7 0.967 3.58 E-7 0.452 0.07622 3.24 2.78 E-7 2.40 1.16 E-7 0.438 0.0730 3.17 K. Witthûser 306 et al. RESULTS Although assumed as comparable non-reactive tracers, remarkable differences have been observed in the diffusion experiments. A range of effective diffusion coefficients for Maastrichtian (Danish) chalk samples of 0.59-2.78 E-6 cm s" (Table 1) was obtained, with the highest values for bromide and uranine and the lowest value for the lithium cation. The Eocene (Israeli) chalk samples show effective diffusion coefficients of between 0.919 and 6.22 E-7 cm s" (Table 2), and contrary to the Danish samples, uranine had the lowest diffusion coefficients and lithium the highest. In general our results for Danish and Israeli chalk samples are in accordance with Hill (1984) and Gooddy et al. (1996), reporting apparent diffusion coefficients in UK chalk samples, for several anions, in the range 0.28-3.23 E-6 cm s" , and 2.5-8.8 E-6 cm s" respectively. The observed variation of effective diffusion coefficients can be explained to a large extent by the different aquatic diffusion coefficients and porosity of the chalk varieties tested (Fig. 2). Figure 3 shows the relationship between the formation factor and porosity for the different chalk samples. Porosity variations and differences in the aquatic diffusion coefficients determine the variation of effective diffusion coefficients significantly. The effective diffusion coefficient can be estimated according to equation (4) with a cementation factor of m equal to 2.2 for the Maastrichtian chalk and 3.3 for the Eocene chalk. We used only the results of the anionic tracers to calculate Archie's law for the different chalk samples since we can assume different interactions with the matrix for cationic tracers. So far we have not found a general trend for differences of cationic vs anionic tracer diffusivities and more experiments related to this question are on the way. It is obvious from Fig. 3 that no unique relation between porosity and formation factor for the different chalk samples exists. The lower formation factors and respectively higher cementation factors of the Israeli chalk samples can be explained 2 2 1 1 2 2 1 1 Laboratory experiments for diffusion transport processes in fractured 307 chalk 0,40-4 W h i t e chalk, Israel m 0,35 4 0,30 4 • Grey chalk, Israel A Danish chalk, Sigerslev A Archie's law, m=2.2 Archie's law, m=3.3 A A A ..-••A" I 0,15-j 0,10-4 0,05 H 0,00 | 0,30 . | . i i i . i i i i i I 0,32 0,34 0,36 0,38 0,40 0,42 0,44 0,46 0,48 0,50 0,52 0,54 porosity e [-] F i g . 3 Plot of f o n n a t i o n factor against porosity for the chalk s a m p l e s . by the smaller and wider range of pore-throat sizes (Fig. 2) compared to the Maastrichtian chalk samples. Adler et al. (1992) found an m value of 1.64 for Fontainbleau sandstone, and Grathwohl (1998) m values of 1.8-2.4 for several sedimentary rock samples. Calculating the Archie's exponent for the diffusion coefficients in uncracked UK chalk samples, Hill (1984) reports values between 2.1 and 2.6. While the exponent of Archie's law for the Maastrichtian chalk is in the range of the reported values, the m value of 3.3 of the Israeli chalk tested exceeds values in the literature (e.g. Gooddy et al, 1996, Grathwohl, 1998, Hill, 1984). The Eocene chalk is generally characterized by a small relative diffusivity. Therefore it exhibits only limited retardation capacity with respect to diffusive mass transfer of tracers as well as of inorganic contaminants compared to the Danish and English chalk. However, diffusion is only one of the possible retardation factors. As a result of the relative high organic content of the Grey chalk (1.3%), strong sorption of organic contaminants can be expected and is currently under investigation. The rock capacity factor a, calculated from the chalk diffusion experiments indicates distinct retardation of bromide and pyranine. Both tracers, but in particular the anion bromide, has been proven to behave as a nearly ideal tracer with respect to its interaction with the solid matrix, in many tracer studies. Batch experiments performed in parallel confirmed the sorptive behaviour of pyranine (K<t = 30.09 ml g" ) but not of bromide (3.4 ml g"'). As usual the distribution coefficients which result are one magnitude higher due to the larger accessible surface area. More experiments are necessary to prove the behaviour of bromide in chalk, while pyranine can only be regarded as a reactive tracer in chalk. 1 Acknowledgements The work was supported by the EC under ENV4-CT97-0441. Dr J. Bloomfield, British Geological Survey, UK, performed the mercury injection porosity measurements. 308 K. Witthûser et al. REFERENCES Adler, P. M., Jaquin, C. G. & Thovert, J.-F. (1992) The formation factor of reconstructed porous media. Wat. Resour. Res. 28, 1571-1576. Archie, G. E. (1942) The electrical resistivity log as an aid in determining some reservoir characteristics. Trans. Am. Inst. Mech. Engrs 146, 54-61. Bradbury, M. H. & Green, A. (1985) Measurement of important parameters determining aqueous phase diffusion rates through crystalline rock matrices. J. Hydrol. 82, 39-55. Carslaw, H. S. & Jaeger, J. C. (1959) Conduction of Heat in Solids. Clarendon, Oxford, UK. Feenstra, S., Cherry, E. A., Sudicky, E. A. & Haq, Z. (1984) Matrix diffusion effects on contaminant migration from an injection well in fractured sandstone. Ground Water 22, 307-316. Gooddy, D. C , Kinniburgh, D. G. & Barker, J. A. (1996) Development of a rapid method for determining apparent diffusion coefficients for chloride in chalk. BGS Technical Report No. WD/95/66. Grathwohl, P. (1998) Diffusion in Natural Porous Media: Contaminant Transport, Sorption/Desorption and Dissolution Kinetics. Kluwer Academic Publishers, Dordrecht, The Netherlands. Hill, D. (19S4) Diffusion coefficients of nitrate, chloride, sulphate and water in cracked and uncracked Chalk. J. Soil Sci. 35, 27-33. Klingenberg, L. J. (1951) Analogy between diffusion and electrical conductivity in porous rocks. Bull. Geol. Soc, Am. 62 559-564.
© Copyright 2026 Paperzz