Previous   |    Next

doi:10.2204/iodp.proc.304305.204.2009

Background and methods

Electrical properties of porous media are sensitive to fluid content and to alteration and can thus be used to detect conductive features in a resistive matrix (e.g., Walsh and Brace, 1984; Pezard and Luthi, 1988). Modes of electrical conduction are identified in the laboratory from resistivity measurements at various saturating fluid salinities and provide information on the porosity structure and degree of alteration. The method was initially proposed for sandstones and clays (Waxman and Smits, 1968; Revil and Glover, 1998) but has been successfully applied to various low-porosity igneous rocks such as basalts (e.g., Pezard, 1990; Einaudi et al., 2000), gabbros (Pezard et al., 1991; Ildefonse and Pezard, 2001), peridotites (Ildefonse et al., 1999), and granites (Pape et al., 1985; Pezard et al., 1999; Belghoul, 2007).

In porous media composed of a matrix considered as infinitely resistive and a connected pore space saturated with a conductive electrolyte, two main mechanisms are responsible for electrical conduction: (1) an electrolytic conduction mechanism in pore volumes and (2) a surface conduction mechanism at the interface between the electrolyte and minerals. Electrolytic conduction is related to the nature and salinity of the saturating fluid; surface conduction is related to the presence of charges along pore surfaces, which is related at the first order to alteration (e.g., Revil and Glover, 1998). The measured electrical conductivity can be written as follows (Waxman and Smits, 1968):

C₀ = (C_W/F) + C_S,

where

• C₀ = total conductivity of the pore space,
• C_W = conductivity of water (saturating fluid),
• F = electrical formation factor, and
• C_S = surface conductivity.

C_S is related to claylike silicates, hence to alteration through the cation exchange capacity (Waxman and Smits, 1968). This model works well at high salinity; however, it tends to overestimate C_S at low salinity. Alternatively, we can use the nonempirical statistical approach proposed by Revil and Glover (1998) based on the microgeometry of the porous space. The conductivity of the sample is then given by a complex model that has two simpler forms at high and low salinities (see equations in Revil and Glover, 1998; Ildefonse and Pezard, 2001).

To characterize their electrolytic and surface conduction components, each sample (minicore) was analyzed by measuring electrical resistivity at 1 kHz and variable saturating fluid salinity (six series of measurements, from 0.06 to 60 g/L) (Fig. F1; Tables T1, T2). Measurements were taken with two electrodes; the polarizing effect at the electrode/sample interface was reduced by using paper filter to separate the electrodes from the sample (see Pezard, 1990, for a detailed description of the experimental protocol).

The contribution of surface conduction to the total measured electrical conductivity can be estimated by using β (Ildefonse and Pezard, 2001):

β = (F × C_S)/[(F × C_S) + C_W],

where C_W is 5000 mS⁄m (for seawater at 24°C).

The intrinsic F and the surface conductivity are extracted from the high- and low-salinity parts of the curves, respectively, using both the Waxman and Smits (1968) and Revil and Glover (1998) models.

F depends only on the microstructural characteristics of the rock and is classically considered to characterize the three-dimensional topology of the pore space (e.g., Guéguen and Palciauskas, 1992). An empirical relationship between F and porosity (ϕ) was proposed by Archie (1942) with:

F = ϕ^–m.

In the oil industry the exponent m is called the “cementation index.” The exponent m typically varies from ~1.5 to 2.5 in crystalline rocks (e.g., Guéguen and Palciauskas, 1992).

The relation between F and ϕ can also be expressed in terms of degree of connectivity of the inner pore space, characterized by electrical the tortuosity (τ) (Walsh and Brace, 1984; Pezard, 1990; Pezard et al., 1991; Guéguen and Palciauskas, 1992):

F = τ/ϕ.

Whereas m describes the nonuniformity of the section of the conductive channels, τ relates to the complexity of the path followed by the electrical current (e.g., Guéguen and Palciauskas, 1992) or, in a more general sense, the efficiency of electrical flow processes (Clennell, 1997). In igneous low-porosity rocks, the average electrical tortuosity is generally on the order of 10 (e.g., Pezard et al., 1991; Ildefonse and Pezard, 2001).

Density and porosity were measured using the classic triple weighing method, with an OHAUS precision scale (10^–4 g accuracy). Samples were first weighed after being dried in an oven at ~50°C and resaturated for “wet” weight measurements in air and immersed.

Bulk density (ρ_b) and grain density (ρ_g) are given by:

ρ_b = [M_sat/(M_sat – M_imm)] × ρ_water

and

ρ_g = [M_dry/(M_dry – M_imm)] × ρ_water,

where

• M_dry = dry sample mass,
• M_sat = saturated sample mass,
• M_imm = the mass of the immersed saturated sample, and
• ρ_water = 1.02 g/cm³ for a 30 g/L salinity at 0.1 MPa and 20°–25°C.

The closed, unconnected porosity is supposed to be negligible; the connected porosity (in percent) is:

ϕ = 100 × [(M_sat – M_dry)/(M_sat – M_imm).

Top of page   |    Previous   |    Next