Influence of Wettability on Two- and Four-Electrode Resistivity Measurements on Berea Sandstone Plugs

Abstract

Summary Two- and four-electrode resistivity measurements were conducted on water- and oil-wet Berea sandstone plugs for different saturation directions with a steady-state flooding procedure. The data from nine plugs indicated that reliable resistance values can be obtained on water-wet cores with both two- and four-electrode methods. On oil-wet cores, however, two-electrode measurements yielded higher resistance values than those from the four-electrode method because of end effects and contact resistance, especially for low water saturations. In addition, the saturation history substantially influenced the resistivity measurements of oil-wet cores. The four-electrode technique is better for oil-wet core resistivity measurements because it ensures that correct Archie saturation exponents are derived. Introduction The quality of the determination of the Archie saturation exponent depends on the accuracy of resistance measured on samples fully or partially saturated with brine. Ideally, resistance should be measured on a sample that has a uniform saturation profile throughout the whole core.1 In practice, it is impossible to obtain such an excellent saturation uniformity in a sample plug of limited size. Such currently used desaturation processing as flooding displacement, centrifuging, and the porous plate technique will inevitably lead to a heterogeneous fluid distribution because of viscous fingering at the inlet and boundary effects at the outlet, i.e., end effects.2-4 Although several measures have been recommended to avoid perturbation,5 this phenomenon cannot be fully eliminated. Sometimes, the saturation accuracy of the core is sacrified to obytain improvement. Both two and four electrodes can give satisfactory results when properly used.4,6 Note, however, that this conclusion was based on water-wet cores in which the saturation exponent was ˜2. The applicability of this conclusion on oil-wet cores should be checked because of the significance of the wettability property on core electrical behavior.7 That the saturation history crucially influences the fluid distribution in porous media saturated with two or three unmixed phases is well-known. As a nonwetting phase in oil-wet cores, brine distribution is dependent on the saturation history, and therefore, the electrical conductivity behavior is closely related to the saturation processing. The purpose of our work has been to investigate the saturation-history effect on the water-saturation/resistivity-index relationship and to examine the potential influence of end effects and the contact resistance on resistivity measurements of water- and oil-wet cores with a steady-state flooding technique. The results of the saturation-history effect are described elsewhere.8 This paper reports the potential influence of end effects and the contact resistance. Experimental Setup Material. Core plugs with a 3.7-cm diameter were drilled out from two blocks of standard Berea sandstone. After cutting rough ends with a diamond saw, the final core length was about 12 cm. The plugs were cleaned with Soxhlet extraction by acetone, methanol, toluene, and methanol, sequentially. Then, the cores were dried in an oven at 100°C for 24 hours. Routine analyses of the two sandstone blocks yielded porosities of 24.3% and 19% and permeabilities of 780 and 150 md. Some of the dried cores were ready to be used in the experiment as water-wet samples. The remaining dried cores were rendered oil-wet after being treated with a solution of 2 vol% Quilon-S™ in distilled water.9 The wettability of the cores was tested with The Amott wettability method.10Table 1 shows the results. Rechecking the cores with the same method after 1 year demonstrated that the wettability property was stable. The fluids used in the experiments were kerosene, filtered through a column of silica gel, and a 36 g/L NaCl solution, with a resistivity of 0.198 O·m at 21°C. Oil and brine densities were 0.775 and 1.018 g/cm3 at 21°C, respectively. Apparatus. Fig. 1 shows the apparatus used in the experiment. The essential advantage of the apparatus was the ability to use a digital balance to monitor saturation changes in the core continuously during the resistivity measurements. Careful adjustment of teflon tubes and wires ensured that the balance sensitivity was not disturbed. Normally, the readings could be taken within an absolute accuracy of ±0.01 g. In this way, the error that would arise from disrupting the tests, thus violating the continuous state, was avoided. The brine-saturation error resulting from the dead volume of the core holder was <3%. Fig. 2 shows the electrode pattern. Two end electrodes constructed with plastic plugs were made from stainless steel. Two inner electrodes wound around the core were tin-coated copper wires. One piece of stainless-steel screen and one piece of filter paper were put between the plug and the core end to iimprove the electrolyte contact. Three kinds of resistance could be measured with this arrangement: R4, the resistivity measured with all four electrodes (the standard four-electrode method); R2, the resistivity measured with the two end electrodes (the standard two-electrode method); and Rin and Rout, which were measured between one end electrode and the nearest inner electrode as a two-electrode method. The measurement instrument was a four-channel frequency-response analyzer, Solartron™ 1254. An AC power source of 500 Hz and 200 mV was imposed on the system, including the reference resistor of 1004 O and the measured core in series. The maximum current density in the core was 0.14 A/m2. The voltage drop ratio of the reference resistor to any portion of the core was read directly from the instrument screen. Material. Core plugs with a 3.7-cm diameter were drilled out from two blocks of standard Berea sandstone. After cutting rough ends with a diamond saw, the final core length was about 12 cm. The plugs were cleaned with Soxhlet extraction by acetone, methanol, toluene, and methanol, sequentially. Then, the cores were dried in an oven at 100°C for 24 hours. Routine analyses of the two sandstone blocks yielded porosities of 24.3% and 19% and permeabilities of 780 and 150 md. Some of the dried cores were ready to be used in the experiment as water-wet samples. The remaining dried cores were rendered oil-wet after being treated with a solution of 2 vol% Quilon-S™ in distilled water.9 The wettability of the cores was tested with The Amott wettability method.10Table 1 shows the results. Rechecking the cores with the same method after 1 year demonstrated that the wettability property was stable. The fluids used in the experiments were kerosene, filtered through a column of silica gel, and a 36 g/L NaCl solution, with a resistivity of 0.198 O·m at 21°C. Oil and brine densities were 0.775 and 1.018 g/cm3 at 21°C, respectively. Apparatus. Fig. 1 shows the apparatus used in the experiment. The essential advantage of the apparatus was the ability to use a digital balance to monitor saturation changes in the core continuously during the resistivity measurements. Careful adjustment of teflon tubes and wires ensured that the balance sensitivity was not disturbed. Normally, the readings could be taken within an absolute accuracy of ±0.01 g. In this way, the error that would arise from disrupting the tests, thus violating the continuous state, was avoided. The brine-saturation error resulting from the dead volume of the core holder was <3%. Fig. 2 shows the electrode pattern. Two end electrodes constructed with plastic plugs were made from stainless steel. Two inner electrodes wound around the core were tin-coated copper wires. One piece of stainless-steel screen and one piece of filter paper were put between the plug and the core end to iimprove the electrolyte contact. Three kinds of resistance could be measured with this arrangement: R4, the resistivity measured with all four electrodes (the standard four-electrode method); R2, the resistivity measured with the two end electrodes (the standard two-electrode method); and Rin and Rout, which were measured between one end electrode and the nearest inner electrode as a two-electrode method. The measurement instrument was a four-channel frequency-response analyzer, Solartron™ 1254. An AC power source of 500 Hz and 200 mV was imposed on the system, including the reference resistor of 1004 O and the measured core in series. The maximum current density in the core was 0.14 A/m2. The voltage drop ratio of the reference resistor to any portion of the core was read directly from the instrument screen.