Low-Resistivity Hydrocarbon-Bearing Sand Reservoirs

Abstract

Summary It has been found that significant amounts of water-free hydrocarbons can be produced from sandstone reservoirs that have water saturations greater than 50%. From comprehensive core and log analyses, the sands were found to have high surface areas that immobilize large amounts of water. The actual water saturations can be high, but water-free hydrocarbons will be produced. Introduction The movement of oil exploration into new areas normally is accompanied by new problems. Well log interpretation is not excluded. Exploration for oil and gas in the Pleistocene sediments of the Gulf of Mexico is one such case in point. By careful monitoring of the mud log and by obtaining sidewall cores in the interval of the slightest hydrocarbon anomaly, production from low-resistivity pay sands was discovered. In particular, in one well where the problem was investigated with a concentrated research effort, sands with resistivities on the order of 0.5 to 0.6 omegam and lower were found to produce water-free hydrocarbons. Log-calculated water saturations of these sands ranged from 50 to 80%. Fig. 1 is a suite of logs from one such well-Well No. 1. Water saturation calculated from resistivity exponents measured on cores is generally shown to be greater than 50%. The two resistivity anomalies in the example do not result from hydrocarbons but are the result of two "hard" zones. The less than 50% water saturation calculated over the interval 2427 to 2429 m [7,964 to 7,969 ft] is probably a result of the difference in thin-bed resolution of the density log compared to that of the induction log. In reality, the water saturation probably is greater than 50% over the entire interval of interest. Cumulative water production for an initial 26-month period was less than 0.02%, while hydrocarbon production was in excess of 3.14 × 10(6) m3 [1/2 million bbl]. Such results are typical for a number of Pleistocene sands in the Gulf of Mexico. Several explanations for this behavior were suggested. One view was that the logging tools were not reading correctly because of the presence of unknown minerals. Another possibility was that the logs were correct but that new interpretation techniques were needed to reduce calculated water saturations to the more reasonable levels usually associated with water-free production (i.e., 20 to 30% or at least less than 50 %). Both views reflect the years of oil exploration and producing experience that have taught that production of water-free hydrocarbons is not expected from reservoirs with water saturation greater than 50%. This paper is a summary of research that shows that the log readings and analyses are correct. We found that these reservoirs have extremely high internal surface areas (74.3 to 18.5 m/g [800 to 200 ft/g]) because of the presence of relatively small amounts of mixed-layer expandable clays (1.9 to 11.5 wt%). The large surface areas immobilize great amounts of water. The resistivity log responds to the total water present (bound and free). The calculated and actual water saturations, therefore, can be high, but water-free hydrocarbons will be produced. A number of papers have been published on the low-resistivity problem. Some address it as the "shaly-sand" problem, which basically is the correction of the resistivity log for the presence of conductive minerals. Without this correction, the calculated water saturation is higher than the true water saturation. This is one class of low-resistivity reservoir. The correction to the resistivity log is most effective where water salinities are less than 60,000 ppm NaCl. A second class, and the one dealt with here, consists of reservoirs where the actual water saturation is greater than 50 % but dry oil is produced. The mechanism responsible for the high water saturation usually is described as being caused by microporosity. Others have not dealt with the mechanism, but instead have depended on core data to make predictions about production. In this paper, we are dealing with salinities on the order of 150,000 to 180,000 ppm NaCl. Water saturations from log data are calculated from the simple Archie equation. From apparent water resistivity calculations in a water zone, water resistivity, Rw, was found to be 0.028 omegam. With water at this high salinity and with the relatively small amounts of clay minerals, the various shaly-sand equations are ineffective. Core Analyses To explain this apparent paradox of production from sands with water saturation greater than 50%, two 6-m [20-ft] rubber-sleeved cores were obtained in an offset well, Well No. 2. Fig. 2 shows the cored interval. Although not as extreme an example as Fig. 1, in that water saturation is calculated to be as low as 28 %, intervals with water saturations greater than 50% were tested, and no water was produced. A project was initiated to correlate the results of core studies, well logs, and well tests to verify whether log data (induction resistivity and density porosity) were reliable; i.e., to find out whether normal values of the cementation and saturation exponents could be used to calculate water saturations. If the log data were correct, the problem would be to explain how water-free hydrocarbons could be produced from formations with water saturations greater than 50 %. Types of Analyses. Twelve samples were selected, by careful examination of the core material, as representative of the observed log variations. The measurements were made for the reservoir properties of porosity, permeability, capillary-pressure curves (centrifuged and nonrestored state), saturations, and bulk and grain densities. Measurements also were made for resistivity and saturation exponents, sieve analyses, scanning electron microscope (SEM) photomicrographs, nuclear magnetic resonance (NMR) surface areas, and X-ray diffraction (XRD). These measurements were done in a sequence such that the same material was used for all of them. Results of Analyses. On the basis of macroscopic examinations of the cores and binocular microscope studies of selected samples, the zone consists of a regressive-transgressive couplet whose origin is attributable to shallow water; i.e., deltaic sedimentation. The measured resistivity exponent was m = 1. 87, and the measured values for the saturation exponent, n, ranged from 1.72 to 1.92 with a mean of n = 1.84 0.08. These differ from the value of 2 that one would usually use, but still must be considered normal. Furthermore, no unusual minerals that could cause problems for logging tools were detected by XRD. We conclude, therefore, that the logs are reading correctly. The resistivity indices are normal; therefore, the water saturations are indeed high. The core analyses show that the sands are not "clean"; i.e., that clay/silt-size material is present. In particular, the results show thatirreducible water saturations are higher than normal from capillary-pressure curves;SEM photomicrographs indicate "dirtiness";surface areas are large as measured by NMR; andthe samples consist of 14 to 34 wt% less than 30- mu m material. Two core samples-one dirty (Sample 5) and one clean (Sample 9) illustrate the manifestation of the first three points. Capillary-pressure curves for these two samples are shown in Fig. 3. The irreducible water saturations are 75 and 23% (at 345-kPa [50-psig] equivalent air/brine) for the dirty and clean samples, respectively. Fig. 4 shows SEM photomicrographs of the same two samples. It is evident from these photomicrographs that even for the clean sample (a), material of some sort is coating the individual grains. SPEFE P. 515^