Comment on: “On the relationship between induced polarization and surface conductivity: Implications for petrophysical interpretation of electrical measurements” (A. Weller, L. Slater, and S. Nordsiek, G eophysics , 78, no. 5, D315–D325)
2014; Society of Exploration Geophysicists; Volume: 79; Issue: 2 Linguagem: Inglês
10.1190/geo2013-0300.1
ISSN1942-2156
AutoresA. Revil, Andreas Weller, Lee Slater, Sven Nordsiek,
Tópico(s)NMR spectroscopy and applications
ResumoPreviousNext No AccessGEOPHYSICSVolume 79, Issue 2Comment on: “On the relationship between induced polarization and surface conductivity: Implications for petrophysical interpretation of electrical measurements” (A. Weller, L. Slater, and S. Nordsiek, Geophysics, 78, no. 5, D315–D325)Authors: André RevilAndreas WellerLee SlaterSven NordsiekAndré RevilColorado School of Mines, Department of Geophysics, Golden, Colorado, USA; and Université de Savoie, LGIT, UMR C5559, 73376 Le Bourget-du-lac Cedex, France. E-mail: ., Andreas WellerTechnische Universität Clausthal, Institut für Geophysik, Clausthal-Zellerfeld, Germany., Lee SlaterRutgers-Newark, Department of Earth and Environmental Sciences, Newark, New Jersey, USA. E-mail: ., and Sven NordsiekTechnische Universität Braunschweig, Institut für Geophysik und extraterrestrische Physik, Braunschweig, Germany.https://doi.org/10.1190/geo2013-0300.1 SectionsAboutFull TextPDF/ePub ToolsAdd to favoritesDownload CitationsTrack CitationsPermissions ShareFacebookTwitterLinked InRedditEmail REFERENCESBinley, A., L. D. Slater, M. Fukes, and G. Cassiani, 2005, Relationship between spectral induced polarization and hydraulic properties of saturated and unsaturated sandstones: Water Resources Research, 41, W12417, doi: 10.1029/2005WR004202.WRERAQ0043-1397CrossrefWeb of ScienceGoogle ScholarBolève, A., A. Crespy, A. Revil, F. Janod, and J. L. Mattiuzzo, 2007, Streaming potentials of granular media: Influence of the Dukhin and Reynolds numbers: Journal of Geophysical Research, 112, B08204, doi: 10.1029/2006JB004673.JGREA20148-0227CrossrefWeb of ScienceGoogle ScholarBörner, F. D., 1992, Complex conductivity measurements of reservoir properties: Proceedings of the third European Core Analysis Symposium, 359–386.Google ScholarCarroll, S., R. S. Maxwell, W. Bourcer, S. Martin, and S. Hulsey, 2002, Evaluation of silica-water surface chemistry using NMR spectroscopy: Geochimica et Cosmochimica Acta, 66, no. 6, 913–926, doi: 10.1016/S0016-7037(01)00827-4.GCACAK0016-7037CrossrefWeb of ScienceGoogle ScholarComparon, L., 2005, Etude expérimentale des propriétés électriques et diélectriques des matérieux argileux consolidés: Ph.D. thesis, Institut de Physique du Globe de Paris, 400.Google ScholarDukhin, S. S., and V. N. Shilov, 2002, Nonequilibrium electric surface phenomena and extended electrokinetic characterization of particles: in Delgado, A.V., ed., Interfacial electrokinetics and electrophoresis, Surfactant Science Series, 106, 55–85.Google ScholarGimmi, T., and G. Kosakowski, 2011, How mobile are sorbed cations in clays and clay rocks?: Environmental Science and Technology, 45, 1443–1449, doi: 10.1021/es1027794.ESTHAG0013-936XCrossrefWeb of ScienceGoogle ScholarGonçalvès, J., P. Rousseau-Gueutin, and A. Revil, 2007, Introducing interacting diffuse layers in TLM calculations: A reappraisal of the influence of the pore size on the swelling pressure and the osmotic efficiency of compacted bentonites: Journal of Colloid and Interface Science, 316, 92–99, doi: 10.1016/j.jcis.2007.07.023.JCISA50021-9797CrossrefWeb of ScienceGoogle ScholarKoch, K., A. Kemna, J. Irving, and K. Holliger, 2011, Impact of changes in grain size and pore space on the hydraulic conductivity and spectral induced polarization response of sand: Hydrology and Earth System Sciences, 15, 1785–1794, doi: 10.5194/hess-15-1785-2011.HESSCF1027-5606CrossrefWeb of ScienceGoogle ScholarKoch, K., A. Revil, and K. Holliger, 2012, Relating the permeability of quartz sands to their grain size and spectral induced polarization characteristics: Geophysical Journal International, 190, 230–242, doi: 10.1111/j.1365-246X.2012.05510.x.CrossrefWeb of ScienceGoogle ScholarLeroy, P., and A. Revil, 2009, Spectral induced polarization of clays and clay-rocks: Journal of Geophysical Research, 114, B10202, doi: 10.1029/2008JB006114.JGREA20148-0227CrossrefWeb of ScienceGoogle ScholarLeroy, P., A. Revil, A. Kemna, P. Cosenza, and A. Ghorbani, 2008, Spectral induced polarization of water-saturated packs of glass beads: Journal of Colloid and Interface Science, 321, 103–117, doi: 10.1016/j.jcis.2007.12.031.JCISA50021-9797CrossrefWeb of ScienceGoogle ScholarLesmes, D. P., and K. M. Frye, 2001, Influence of pore fluid chemistry on the complex conductivity and induced polarization responses of Berea sandstone: Journal of Geophysical Research, 106, 4079–4090, doi: 10.1029/2000JB900392.JGREA20148-0227CrossrefWeb of ScienceGoogle ScholarLorne, B., F. Perrier, and J.-P. Avouac, 1999, Streaming potential measurements. 1. Properties of the electrical double layer from crushed rock samples: Journal of Geophysical Research, 104, 17857–17877, doi: 10.1029/1999JB900156.JGREA20148-0227CrossrefWeb of ScienceGoogle ScholarRevil, A., 2012, Spectral induced polarization of shaly sands: Influence of the electrical double layer: Water Resources Research, 48, W02517.WRERAQ0043-1397CrossrefWeb of ScienceGoogle ScholarRevil, A., 2013b, On charge accumulations in heterogeneous porous materials under the influence of an electrical field: Geophysics, 78, no. 4, D271–D291, doi: 10.1190/geo2012-0503.1.GPYSA70016-8033AbstractWeb of ScienceGoogle ScholarRevil, A., 2013a, Effective conductivity and permittivity of unsaturated porous materials in the frequency range 1 mHz–1 GHz: Water Resources Research, 49, 306–327.WRERAQ0043-1397CrossrefWeb of ScienceGoogle ScholarRevil, A., J. D. Eppehimer, M. Skold, M. Karaoulis, L. Godinez, and M. Prasad, 2013b, Low-frequency complex conductivity of sandy and clayey materials: Journal of Colloid and Interface Science, 398, 193–209, doi: 10.1016/j.jcis.2013.01.015.JCISA50021-9797CrossrefWeb of ScienceGoogle ScholarRevil, A., and N. Florsch, 2010, Determination of permeability from spectral induced polarization data in granular media: Geophysical Journal International, 181, 1480–1498.GJINEA0956-540XWeb of ScienceGoogle ScholarRevil, A., K. Koch, and K. Holliger, 2012, Is it the grain size or the characteristic pore size that controls the induced polarization relaxation time of clean sands and sandstones?: Water Resources Research, 48, W05602.WRERAQ0043-1397CrossrefWeb of ScienceGoogle ScholarRevil, A., and M. Skold, 2011, Salinity dependence of spectral induced polarization in sands and sandstones: Geophysical Journal International, 187, 813–824, doi: 10.1111/j.1365-246X.2011.05181.x.GJINEA0956-540XCrossrefWeb of ScienceGoogle ScholarRevil, A., M. Skold, S. S. Hubbard, Y. Wu, D. Watson, and M. Karaoulis, 2013a, Petrophysical properties of saprolites from the Oak Ridge Integrated Field Research Challenge site, Tennessee: Geophysics, 78, no. 1, D21–D40, doi: 10.1190/geo2012-0176.1.GPYSA70016-8033AbstractWeb of ScienceGoogle ScholarRevil, A., W. F. Woodruff, and N. Lu, 2011, Constitutive equations for coupled flows in clay materials: Water Resources Research, 47, W05548, doi: 10.1029/2010WR010002.WRERAQ0043-1397CrossrefWeb of ScienceGoogle ScholarSchmutz, M., A. Revil, P. Vaudelet, M. Batzle, P. Femenia Viñao, and D. D. Werkema, 2010, Influence of oil saturation upon spectral induced polarization of oil bearing sands: Geophysical Journal International, 183, 211–224.CrossrefWeb of ScienceGoogle ScholarScott, J., and R. Barker, 2003, Determining pore-throat size in Permo-Triassic sandstones from low-frequency electrical spectroscopy: Geophysical Research Letters, 30, 1450, doi: 10.1029/2003GL016951.GPRLAJ0094-8276CrossrefWeb of ScienceGoogle ScholarSlater, L., and D. Lesmes, 2002, Electrical-hydraulic relationships observed for unconsolidated sediments: Water Resources Research, 38, no. 10, 1213, doi: 10.1029/2001WR001075.CrossrefWeb of ScienceGoogle ScholarTong, M., L. Li, W. Wang, and Y. Jiang, 2006, Determining capillary-pressure curve, pore size distribution and permeability from induced polarization of shaley sand: Geophysics, 71, no. 3, N33–N40, doi: 10.1190/1.2195989.GPYSA70016-8033AbstractWeb of ScienceGoogle ScholarTournassat, C., Y. Chapron, P. Leroy, M. Bizi, and F. Boulahya, 2009, Comparison of molecular dynamics simulations with triple layer and modified Gouy-Chapman models in 0.1 M NaCl-montmorillonite system: Journal of Colloid and Interface Science, 339, 533–541, doi: 10.1016/j.jcis.2009.06.051.JCISA50021-9797CrossrefWeb of ScienceGoogle ScholarVan Olphen, H., and M. H. Waxman, 1958, Surface conductance of sodium bentonite in water: Proceedings of the Fifth National Conference, Clays and Clay Minerals, NAS-RRC Pub., 566, 61–80.CrossrefGoogle ScholarVaudelet, P., A. Revil, M. Schmutz, M. Franceschi, and P. Bégassat, 2011, Induced polarization signature of the presence of copper in saturated sands: Water Resources Research, 47, W02526, doi: 10.1029/2010WR009310.WRERAQ0043-1397CrossrefWeb of ScienceGoogle ScholarVinegar, H. J., and M. H. Waxman, 1984, Induced polarization of shaly sands: Geophysics, 49, 1267–1287, doi: 10.1190/1.1441755.GPYSA70016-8033AbstractWeb of ScienceGoogle ScholarWang, M., J. Liu, and S. Chen, 2007, Electrical potential distribution in nanoscale electroosmosis. From molecules to continuum: Molecular Simulation, 33, no. 15, 1273–1277.MOSIEA0892-7022Web of ScienceGoogle ScholarWeller, A., L. Slater, and S. Nordsiek, 2013, On the relationship between induced polarization and surface conductivity: Implications for petrophysical interpretation of electrical measurements: Geophysics, 78, no. 5, D315–D325, doi: 10.1190/geo2013-0076.1.GPYSA70016-8033AbstractWeb of ScienceGoogle ScholarWe thank André Revil for his positive assessment of the overall contribution of Weller et al. (2013). Given that he has made important contributions in developing mechanistic models for the interpretation of complex conductivity data, we are pleased that he recognizes the broad significance of this paper to hydrogeophysics and well logging. However, we feel that Revil has partly misunderstood the primary objective of Weller et al. (2013) and that his Discussion focuses on a relatively minor issue that does not detract from the main findings. Furthermore, some of the material presented by Revil in his Discussion paper, and in Revil (2013a, 2013b) that is extensively referenced in his Discussion paper, causes us to emphasize the need for empirical studies that are based on a careful and appropriate analysis of existing experimental data. This is the strength of Weller et al. (2013).We first respond to a general comment that appears in the opening sentence of this Discussion, where Revil states that Weller et al. (2013) attempted “to connect the surface and quadrature conductivities of porous media using the POLARIS model developed by Revil (2012).” We wish to stress that the POLARIS model is not the focus of our work. Weller et al. (2013) presented an exhaustive (at the time of submission) database conclusively defining the form of the relationship between the surface conductivity and quadrature conductivity over a wide range of sandstones and unconsolidated sediments. The major findings of Weller et al. (2013) are observational and independent of the POLARIS model, or any other model (e.g., the new model presented in Revil (2013a, 2013b) for the complex conductivity of porous media.We next address the specific criticisms of our work as raised by Revil. The first point raised in his Comment concerns our choice of the mobility ratio, i.e., the ratio of the mobility of the counterions in the Stern layer β(+)S to the mobility of the counterions in the diffuse layer (β(+)) that appears in the recently proposed POLARIS model (Revil, 2012). Revil suggests that we are unaware that this mobility ratio might be different for silicates versus aluminosilicates. Weller et al. (2013) demonstrate that the POLARIS model produces predictions that are consistent with the observations reported from their compiled database. In doing so, they find that the POLARIS model describes their extensive data set with a wide range of clay contents (including clean sands) when the mobility ratio between the counterions in the Stern layer and the diffuse layer is equal to 0.0029 as proposed by Revil (2012). Weller et al. (2013) contains the following restricting statement, “It should be noted that this ratio is only valid for sodium cations in clayey material (Revil, 2012).” There is no statement in Weller et al. (2013) that the value of the ratio of mobility of the counterions in the Stern layer and the mobility in the diffuse layer is the same for silicates and aluminosilicates. We note that Revil’s differentiation between silicates and aluminosilicates (see Discussion paper) is confusing because the aluminosilicates are phyllosilicates that also belong to the silicate group. Later in this reply, we will differentiate between “clean sand” and “clayey material.”The second point raised by Revil concerns the implications of the variations in this mobility ratio on the relationship between surface conductivity and quadrature conductivity. However, equations 6–8 in this Discussion are from a new (or modified) model presented in Revil (2013b) and are not the equations presented in the original POLARIS model the previous year (Revil, 2012). In fact, the updated models were not referred to as the POLARIS model in Revil (2013a, 2013b). Yet, this Discussion paper seems to extend such terminology to these updated models, which causes confusion. Our intent was to compare our results against the most recent mechanistic model for complex conductivity at the time of submission.However, if we do now consider the new model presented in Revil (2013b), the ratio between normalized chargeability (see equation 8 in the Discussion paper) and surface conductivity (equation 6 in the Discussion paper) results in the property R′ (as referred to in the Discussion paper), equivalent to lmn (as referred to in our paper), which can be expressed by R′=lmn=Mnσsurf≅β(+)Sβ(+)f.(10)Here, Mn is the normalized chargeability, σsurf is the surface conductivity, and f is the partitioning coefficient describing the fraction of counterions in the Stern layer. Equation 10 is obviously different from equation 3 in the Discussion paper. We cannot see how the equality R′=R (R=l=σ′′/σsurf) is justified for all material including clayey material as stated by Revil. Instead, according to the explanation presented in the Discussion paper, such equality should only be valid in the case of clean sands with β(+)S=β(+) (see equation 5 in the Discussion paper). The equality between R and R′ is not supported by our data. Equation 10 shows that the product of mobility ratio and partitioning coefficient has an effect on the ratio lmn. As discussed in our paper, a change in the mobility ratio can be compensated by a change in the partitioning coefficient to keep the ratios lmn=R′ or l=R more or less constant.We wish to emphasize here that a full discussion of the uncertainty in the parameters of the POLARIS model was beyond the scope of Weller et al. (2013). Instead, we showed that the POLARIS model suggested by Revil (2012) is consistent with the large set of experimental findings that we present. The POLARIS model is in no way central to the major findings or relevance of Weller et al. (2013). Consequently, it was never the objective of Weller et al. (2013) to offer a solution to the problem of assigning representative values to β(+)S/β(+) and f as criticized in the third point of the Discussion paper. Furthermore, the statement that our experimental findings are in agreement with the original version of the POLARIS model does not automatically implicate that the validity of this model is free of doubt.Revil implies that we suggest our empirical observations represent a universal trend. In fact, we did not use the term “universal” in our paper. Instead, we identified a more or less fixed ratio between quadrature conductivity and surface conductivity for our samples that span a wide range of mineral composition (including clayey, silty, and clean material). We did not determine any ratios of mobilities. Even if these mobility ratios would be distinctively different for clay and quartz, they clearly do not strongly affect the ratios between quadrature conductivity and surface conductivity. We cannot identify any splitting into two different trends based on mobility ratio differences in our graphs (see Figures 3 and 4 in Weller et al., 2013) as proposed by Revil in his Discussion. All our samples, consolidated and unconsolidated, incorporate mixtures of quartz grains and other silicate minerals, including a varying content of clay minerals. It is hard to imagine that these natural mixtures could be represented by a binary binning into clean sands and clayey material. Where would we define such a sharp boundary between the two groups? What concentration of clay minerals would be tolerated in a sandstone for it to be referred to as “clean sand”? In our opinion, the mixture should be characterized by an effective mobility ratio representing a weighting between the different minerals.These considerations have led us to doubt some of the experimental evidence of the graphs provided in the Discussion and related papers of Revil (2013a, 2013b). Figure 1b in the discussion paper, which is a repetition of Figure 14 in Revil (2013b), does not prove the statement of two distinct trends. The tendency of the clayey material is clearly related to the shaly sandstones of Vinegar and Waxman (1984). The other two references are not related to the data shown in the graph. Neither Lorne et al. (1999) nor Bolève et al. (2007) provide any data on quadrature conductivity and cation exchange capacity (CEC) for their investigated material (crushed Fontainebleau sandstone and glass beads) in their original papers. The sources of these experimental data should be clearly specified to understand the experimental and processing procedures.Furthermore, our extensive database includes a careful set of measurements conducted on a clean sandstone sample. The Fontainebleau sandstone was investigated by Börner (1992) (sample F3 in Weller et al., 2013). As shown in Table 1 of Weller et al. (2013), the ratio l=R=0.028 for this clean sandstone sample is clearly below the value of R=0.5 that would be predicted from the Revil (2013b) model for sands and close to the value R=0.026 predicted for clays in the same paper. Again, according to our data, a separation into clean sands and clayey material is not justified.Equation 7 in the Discussion paper provides a relation between imaginary part of conductivity and CEC σ′′≈(1Fϕ)ρSfβ(+)SCEC(11)with F being the formation factor, ϕ the porosity, and ρs the grain density. This equation provides the basis for Figure 1b of the Discussion paper. Unfortunately, we do not have CEC data for most of our samples as would be needed to test this equation. However, we do have reliable data of specific internal surface for our samples. Using the relation between the CEC and the specific surface per unit pore volume Spor (Weller et al., 2010; Revil, 2013a), CEC=Qsϕ1−ϕ1ρsSpor(12)with Qs being the mean charge density, equation 11 can be written as σ′′≈(1F(1-ϕ)Qsfβ(+)S)Spor.(13)Equation 13 corresponds to the relation that was carefully investigated in Weller et al. (2010). The comprehensive set of more than 100 samples used in Weller et al. (2010) contains some clean sandstones (including four samples of Fontainebleau sandstone). The study identified a robust, single linear trend between imaginary part of conductivity and Spor for the entire database. Consequently, the factor of proportionality, which corresponds to the term in brackets in equation 13, which is referred to as specific polarizability (Weller et al., 2011), should be nearly constant. Assuming the validity of the updated model (Revil, 2013a, 2013b), a considerable increase in the ion mobility in the Stern layer for clean sands (on the order of a factor 100 as indicated in Figure 1b in the Discussion paper) would result in a similar strong increase in imaginary part of conductivity that would be strikingly obvious in our data and in the graph. The data set presented in Weller et al. (2010) does not support any sharp differentiation into clean sands and clayey materials. On the contrary, the sand-clay mixtures (Slater et al., 2006) exhibit a slightly higher specific polarizability compared to the sandstones as can be seen in Figure 3 of Weller et al. (2010). The only materials that show a considerably higher specific polarizability than the silicate samples are the mixtures of sand and iron or magnetite (Slater et al., 2006). The specific polarizability of these mixtures is caused by the much stronger effect of electrode polarization that is widely applied in ore prospection. It is hard to imagine, and beyond our extensive experimental experience, that pure silica sands could cause an even stronger polarization effect than mixtures of sand and magnetite.In order to prepare our response to this Discussion paper, we spent considerable time carefully reviewing the data presented in this Discussion and also the data set presented in Revil (2013a, 2013b) that is heavily cited in this Discussion paper. In doing so, we became concerned about how some of the available data have been used by Revil to confirm his model. This reply is not the place to detail all these concerns, and we will instead focus these efforts on a future paper where we will test the Revil (2013a, 2013b) model and earlier derivatives of this model (e.g., the POLARIS model) using our own data sets that are all derived from the original sources. However, we feel it appropriate to highlight one example, as it is central to addressing the criticisms raised by Revil in this comment. This concerns data used in Revil (2013b) to argue the apparent importance of the differences in the mobility ratio between “clean sand” and “clayey materials” in controlling the relationship between imaginary (quadrature) conductivity and the specific surface area of soils and rocks (see Figure 11 of Revil, 2013b). We are perplexed as to why Revil (2013b) characterizes samples of the Sherwood Sandstone from the study of Binley et al. (2005) as clean sand. As described by Binley et al. (2005), these samples contained fining upward sequences with clay contents (by grain size and scanning electron microscopy mineralogy analysis) from 2%–4%. Furthermore, the reported quadrature conductivities for the Sherwood sandstone samples of Binley et al. (2005) reported in Figure 11 of Revil (2013b) are physically unrealistic, being about two orders of magnitude higher than what could be expected for this type of sandstone samples. This error is clearly propagated from the Binley et al. (2005) paper, where their Figure 10a (which Revil [2013b] used as the source) is obviously in error, as can be seen by checking the range of values of amplitude and phase of conductivity as presented in Figure 7 of Binley et al. (2005). Disturbingly, this erroneous figure is critical to defining the trend for clean sandstone reported in Figure 11 of Revil (2013b) and thus the central objection raised by Revil regarding Weller et al. (2013).Figure 5 shows the relation between mass normalized specific internal surface Sm and quadrature conductivity of sandstone and clean sand samples as derived from experiments performed at six different labs. The clean sands and sandstones are generally characterized by lower specific internal surface and lower quadrature conductivity, and there is no evidence for two separate trends as proposed by Revil. Using the correct imaginary part of conductivity of the Sherwood Sandstone samples (Binley et al., 2005), the data points are very well integrated into the single trend displayed for clean sands and clayey materials. The fitting of a power law function results in the following relation: σ′′=0.085(Sm)0.83(14)with σ′′ in mS/m and Sm in m2/g, with a coefficient of determination R2=0.864. Figure 6 uses the same samples to examine the relationship between the specific surface per unit pore volume Spor and the quadrature conductivity. The relation can be well described by a similar power law equation σ′′=0.0107(Spor)1.02(15)with σ′′ in mS/m and Spor in 1/μm. The exponent is close to one and the high coefficient of determination R2=0.940 supports the validity of models predicting a linear relation between Spor and quadrature conductivity as proposed by Börner et al. (1996) and confirmed by Weller et al. (2010). Our analysis reveals a better relation between σ′′ and Spor than between σ′′ and Sm, which is opposite to what is presented in Figure 7 of Revil (2012) based only on ten shaly sandstones and omitting the two Fontainebleau samples of the original sample set of Börner (1992). Given that our analysis is based on a wider range of material types (31 shaly sandstones, four clean sandstones, and seven samples of clean sand), this leads us to question the assumptions of the POLARIS model in terms of the stronger control of Sm on σ′′ (see Equation 45 of Revil, 2012) relative to Spor. We find that a single, approximately linear relation between σ′′ and Spor extends from the clean (clay free) Fontainebleau sandstones and clean sands to the clayey Sherwood sandstones. Furthermore, the clean sand samples are characterized by a slightly lower specific polarizability compared to the general trend, consistent with Weller et al. (2010). This strongly contradicts Figure 11 in Revil (2013b) that indicates an increased specific polarizability of the order of a factor 100 for the clean sand samples.Figure 5. Relation between imaginary part of conductivity (σ′′) measured at a frequency of about 1 Hz and mass normalized specific internal surface (Sm) for 42 samples originating from six laboratories. The data of the clean sand samples (F33, F34, F36, KH, and filter gravel) was kindly provided by Alexander Huisman, Forschungszentrum Jülich, Germany. The samples have been saturated with natural water (Binley et al., 2005; samples F33, filter gravel) or a sodium chloride solution (all other samples) with a fluid conductivity of about 100 mS/m.Figure 6. Relation between imaginary part of conductivity (σ′′) measured at a frequency of about 1 Hz and the specific surface per unit pore volume (Spor) for 42 samples originating from six laboratories. The data of the clean sand samples (F33, F34, F36, KH, and filter gravel) was kindly provided by Alexander Huisman, Forschungszentrum Jülich, Germany. The samples have been saturated with natural water (Binley et al., 2005; samples F33, filter gravel) or a sodium chloride solution (all other samples) with a fluid conductivity of about 100 mS/m.Our investigations conducted to prepare our reply to this discussion raised additional concerns regarding data analysis by Revil. The mass normalized specific surface area Sm of most samples included in Figure 11 of Revil (2013b) was determined by nitrogen adsorption method as can be concluded from the original papers (e.g., Börner and Schön, 1991; Lesmes and Frye, 2001; Binley et al., 2005; Weller et al., 2011). The specific surface areas Sm of the clean sand samples of Koch et al. (2011); which were actually calculated from the median grain diameter as described in the text of this paper, do not account for the small-scale grain surface roughness. Assuming that the specific internal surface is characterized by a fractal behavior (e.g., Pape et al., 1987, 1999), a deviation by more than a factor of 10 in the resulting calculated surface area can easily arise due to the different scale lengths, i.e., the resolution of a nitrogen molecule (∼0.4 nm) for the nitrogen adsorption method versus several micrometers for the grain-size-based estimate. The direct comparison of specific surface estimates using such different methods is therefore problematic. The use of grain size results in an unrealistically low Sm as the surface roughness is ignored.We further note that the set of Sherwood sandstone samples (Binley et al., 2005) is also used in Figure 3 of the Discussion paper to demonstrate the trend of clean sands in a diagram showing the relation between time constant and pore throat size. Although the Sherwood sandstone samples cannot be classified as clean sand, the samples follow the same trend as the clean sands of Koch et al. (2011). Furthermore, the shaly sandstone samples investigated by Tong et al. (2006) are also used to demonstrate the trend of clean sands. This unintended erroneous classification again weakens the argument that the diffusion coefficient is characterized by two distinct values, i.e., one for clean sand and another for clayey materials.Finally, the comparison of the diffusion coefficients derived from time constants based on different models is problematic. Several studies demonstrate that the time constants differ significantly between models for the same complex conductivity spectra (e.g., Nordsiek and Weller, 2008; Tarasov and Titov, 2013). Figures 3 and 4 in the Discussion paper repeat the content of Figures 15 and 16 in Revil (2013a). In both papers, it is not clear how the “main relaxation time,” “characteristic relaxation time” (Revil, 2013a), the “Cole-Cole relaxation time,” the “Cole-Cole time constant,” or the “time constant,” all labelled as τ0, have been determined. The required information needed to evaluate the significance of the presented data does not appear in the text, figure caption, legends, or the labelling of the axes. Such ambiguity limits the value of the presented data in efforts to test model performance.Our examination of the data analysis performed by Revil in this Discussion and the associated papers referenced above leads us to conclude that the mechanistic models that include a variety of parameters, which are poorly constrained, require further validation. We recognize that it is currently challenging to determine the mobility of the ions in the Stern layer. Consequently, the mobility ratio β(+)S/β(+), which is a key parameter in all the presented mechanistic models, is not as certain as presented by Revil. The data and graphs presented by Revil in his recent papers are partly inconsistent with our own extensive spectral induced polarization measurements performed over the last two decades. This causes us to question either the general validity of the mechanistic models and/or the proposed values of the ion mobility. Specifically, we are not convinced that the mobility of the ions in the Stern layer of clays is more than two orders of magnitude lower than in clean sands. Furthermore, we do not believe that such a high contrast in mobility is needed to explain the dependence of currently available complex conductivity data on specific internal surface.In summary, although we fully recognize the merit of mechanistic models to improve the understanding of the underlying physics controlling induced polarization, we feel that this Discussion paper has primarily highlighted the need for more studies that rely on interpretation of a broad range of data sets where (1) the original source data sets have been acquired and carefully analyzed for reliability, and (2) acquisition and processing procedures are comparable for all data used. We feel that this is the major contribution of Weller et al. (2013), which involved an exhaustive exercise to acquire original data sets for all samples included and adoption of uniform processing procedures. In taking this observational approach, we have provided evidence for a single empirical relationship between imaginary conductivity and surface conductivity that satisfies (to first order) a wide range of sample types, and can be used to significantly improve estimates of formation factor and water salinity in well logging and hydrogeophysical studies. Assuming that the existing mechanistic models continue to be updated at the same fast pace as in recent years, we would encourage further careful testing of such models using well constrained, original data sets subject to uniform data processing methods. REFERENCES Binley, A., L. D. Slater, M. Fukes, and G. Cassiani, 2005, Relationship between spectral induced polarization and hydraulic properties of saturated and unsaturated sandstone: Water Resources Research, 41, W12417, doi: 10.1029/2005WR004202.WRERAQ0043-1397 CrossrefWeb of ScienceGoogle Scholar Bolève, A., A. Crespy, A. Revil, F. Janod, and J. L. Mattiuzzo, 2007, Streaming potentials of granular media: Influence of the Dukhin and Reynolds numbers: Journal of Geophysical Research, 112, B08204, doi: 10.1029/2006JB004673.JGREA20148-0227 CrossrefWeb of ScienceGoogle Scholar Börner, F. D., 1992, Complex conductivity measurements of reservoir properties: Proceedings of the Third European Core Analysis Symposium, 359–386. Google Scholar Börner, F. D., and J. H. Schön, 1991, A relation between the quadrature component of electrical conductivity and the specific surface area of sedimentary rocks: The Log Analyst, 32, 612–613.LGALAS0024-581X Google Scholar Börner, F. D., J. R. Schopper, and A. Weller, 1996, Evaluation of transport and storage properties in the soil and groundwater zone from induced polarization measurements: Geophysical Prospecting, 44, 583–601, doi: 10.1111/j.1365-2478.1996.tb00167.x.GPPRAR0016-8025 CrossrefWeb of ScienceGoogle Scholar Koch, K., A. Kemna, J. Irving, and K. Holliger, 2011, Impact of changes in grain size and pore space on the hydraulic conductivity and spectral induced polarization response of sand: Hydrology and Earth System Sciences, 15, 1785–1794, doi: 10.5194/hess-15-1785-2011.HESSCF1027-5606 CrossrefWeb of ScienceGoogle Scholar Lesmes, D. P., and K. M. Frye, 2001, The influence of pore fluid chemistry on the complex conductivity and induced-polarization responses of Berea sandstone: Journal of Geophysical Research, 106, 4079–4090, doi: 10.1029/2000JB900392.JGREA20148-0227 CrossrefWeb of ScienceGoogle Scholar Lorne, B., F. Perrier, and J.-P. Avouac, 1999, Streaming potential measurements. 1. Properties of the electrical double layer from crushed rock samples: Journal of Geophysical Research, 104, 17 857–17 877, doi: 10.1029/1999JB900156.JGREA20148-0227 CrossrefWeb of ScienceGoogle Scholar Nordsiek, S., and A. Weller, 2008, A new approach to fitting induced-polarization spectra: Geophysics, 73, no. 6, F235–F245, doi: 10.1190/1.2987412.GPYSA70016-8033 AbstractWeb of ScienceGoogle Scholar Pape, H., L. Riepe, and J. R. Schopper, 1987, Theory of self-similar network structures in sedimentary and igneous rocks and their investigation with microscopical methods: Journal of Microscopy, 148, 121–147, doi: 10.1111/j.1365-2818.1987.tb02861.x.JMICAR0022-2720 CrossrefWeb of ScienceGoogle Scholar Pape, H., C. Clauser, and J. Iffland, 1999, Permeability prediction based on fractal pore-space geometry: Geophysics, 64, 1447–1460, doi: 10.1190/1.1444649.GPYSA70016-8033 AbstractWeb of ScienceGoogle Scholar Revil, A., 2012, Spectral induced polarization of shaly sands: Influence of the electrical double layer: Water Resources Research, 48, W02517.WRERAQ0043-1397 CrossrefWeb of ScienceGoogle Scholar Revil, A., 2013a, Effective conductivity and permittivity of unsaturated porous materials in the frequency range 1 mHz–1GHz: Water Resources Research, 49, 306–327.WRERAQ0043-1397 CrossrefWeb of ScienceGoogle Scholar Revil, A., 2013b, On charge accumulations in heterogeneous porous materials under the influence of an electrical field: Geophysics, 78, no. 4, D271–D291, doi: 10.1190/geo2012-0503.1.GPYSA70016-8033 AbstractWeb of ScienceGoogle Scholar Slater, L. D., D. Ntarlagiannis, and D. Wishart, 2006, On the relationship between induced polarization and surface area in metal-sand and clay-sand mixtures: Geophysics, 71, no. 2, A1–A5, doi: 10.1190/1.2187707.GPYSA70016-8033 AbstractWeb of ScienceGoogle Scholar Tarasov, A., and K. Titov, 2013, On the use of the Cole-Cole equations in spectral induced polarization: Geophysical Journal International, 195, 352–356, doi: 10.1093/gji/ggt251.GJINEA0956-540X CrossrefWeb of ScienceGoogle Scholar Tong, M., L. Li, W. Wang, and Y. Jiang, 2006, Determining capillary-pressure curve, pore size distribution and permeability from induced polarization of shaley sand: Geophysics, 71, no. 3, N33–N40, doi: 10.1190/1.2195989.GPYSA70016-8033 AbstractWeb of ScienceGoogle Scholar Weller, A., L. Slater, S. Nordsiek, and D. Ntarlagiannis, 2010, On the estimation of specific surface per unit pore volume from induced polarization: A robust empirical relation fits multiple data sets: Geophysics, 75, no. 4, WA105–WA112, doi: 10.1190/1.3471577.GPYSA70016-8033 AbstractWeb of ScienceGoogle Scholar Weller, A., K. Breede, L. Slater, and S. Nordsiek, 2011, Effect of changing water salinity on complex conductivity spectra of sandstones: Geophysics, 76, no. 5, F315–F327, doi: 10.1190/geo2011-0072.1.GPYSA70016-8033 AbstractWeb of ScienceGoogle Scholar Weller, A., L. Slater, and S. Nordsiek, 2013, On the relationship between induced polarization and surface conductivity: Implications for petrophysical interpretation of electrical measurements: Geophysics, 78, no. 5, D315–D325, doi: 10.1190/geo2013-0076.1.GPYSA70016-8033 AbstractWeb of ScienceGoogle ScholarFiguresReferencesRelatedDetailsCited byExperimental study on hydrate saturation evaluation based on complex electrical conductivity of porous mediaJournal of Petroleum Science and Engineering, Vol. 208Field-scale estimation of soil properties from spectral induced polarization tomographyGeoderma, Vol. 403Low-Frequency Electrical Conductivity of Aqueous Kaolinite Suspensions II: Counterion Effects and Estimating Stern Layer Mobilities of Counterions12 February 2018 | Clays and Clay Minerals, Vol. 66, No. 1Ultra-broad-band electrical spectroscopy of soils and sediments—a combined permittivity and conductivity model3 June 2017 | Geophysical Journal International, Vol. 210, No. 3Low-frequency electrical conductivity of aqueous kaolinite suspensions: surface conductance, electrokinetic potentials and counterion mobility2 January 2018 | Clay Minerals, Vol. 52, No. 3Textural control on the quadrature conductivity of porous mediaQifei Niu, Manika Prasad, André Revil, and Milad Saidian18 July 2016 | GEOPHYSICS, Vol. 81, No. 5Induced polarization and pore radius — A discussionAndreas Weller, Zeyu Zhang, Lee Slater, Sabine Kruschwitz, and Matthias Halisch28 July 2016 | GEOPHYSICS, Vol. 81, No. 5Evaluation of low frequency polarization models using well characterized sintered porous glass samplesJournal of Applied Geophysics, Vol. 124Induced polarization dependence on pore space geometry: Empirical observations and mechanistic predictionsJournal of Applied Geophysics, Vol. 123Predicting permeability from the characteristic relaxation time and intrinsic formation factor of complex conductivity spectra27 August 2015 | Water Resources Research, Vol. 51, No. 8On the specific polarizability of sands and sand-clay mixturesAndreas Weller, Lee Slater, Johan Alexander Huisman, Odilia Esser, and Franz-Hubert Haegel17 March 2015 | GEOPHYSICS, Vol. 80, No. 3Microseisms and its impact on the marine-controlled source electromagnetic signal23 December 2014 | Journal of Geophysical Research: Solid Earth, Vol. 119, No. 12 Volume 79Issue 2Mar 2014Pages: 1MA-Z52ISSN (print):0016-8033 ISSN (online):1942-2156 publication data© 2014 Society of Exploration GeophysicistsPublisher:Society of Exploration Geophysicists HistoryReceived: 07 Aug 2013Accepted: 11 Nov 2013Published Online: 17 Feb 2014Published in print: 01 Mar 2014 CITATION INFORMATION André Revil, Andreas Weller, Lee Slater, and Sven Nordsiek, (2014), "Comment on: “On the relationship between induced polarization and surface conductivity: Implications for petrophysical interpretation of electrical measurements” (A. Weller, L. Slater, and S. Nordsiek, Geophysics, 78, no. 5, D315–D325)," GEOPHYSICS 79: X1-X10. https://doi.org/10.1190/geo2013-0300.1 Plain-Language Summary Keywordsinduced polarization (IP)porositypermeabilityrock physicsPDF Download Metrics Loading ...
Referência(s)