Water Saturation Measurements Research Articles

Characterization and saturation determination of reservoir metagraywacke from The Geysers corehole SB-15-D (USA), using Nuclear Magnetic Resonance Spectrometry and X-ray Computed Tomography

Novel pressure-coring technology, along with nuclear magnetic resonance (NMR) spectrometry and X-ray computed tomography (CT) of cores retrieved from scientific drillhole SB-15-D in the Sulphur Bank area of The Geysers steam field, have enabled the first measurements of indigenous water saturation in a high-temperature, vapor-dominated geothermal system. Two cores for the analyses were obtained using a wireline-retrievable core barrel sealed and pressurized at simulated reservoir pressure during ascent to the surface, and there immediately frozen in dry ice and secured in insulated receptacles to immobilize contained fluids. Immediately upon being thawed at the laboratory, the cores were measured for total water volume by NMR. These values, combined with rock porosities later determined by Boyles Law gas expansion, yielded total water saturations. For the deeper of the two cores (485.2–488.3 m), water saturation was determined by these methods to be 3–13%, but these values probably include some drilling-fluid filtrate. For the shallower core (432.9–435.8 m), filtrate incursion was measurable by CT, since a tritium tracer had been added to the drilling fluid during coring operations. Filtrate accounted for a fifth to three-fifths of the total water in this core, and corrected water saturations ranged from 8 to 25%. Results of subsequent CT-based experiments on the cores suggest that invasion probably took place as a result of drilling-fluid vaporization and convective migration of the vaporized filtrate. This finding (1) implies that much of the hot (about 235°C) cored rock sequence is relatively dry; and (2) independently supports the low water saturation measurements obtained by NMR. The dryness and low water saturations of the SB-15-D cores doubtless reflect the 40-year production history of the Sulphur Bank area, and cannot be considered representative of The Geysers as a whole. As indigenous water saturation will remain a critical variable in the longevity of this resource, it is recommended that our new pressure-coring and saturation-measurement techniques be judiciously applied in other (particularly deeper) portions of the reservoir.

EXPERIENCE WITH DETERMINATION OF WATER SATURATION FROM CORE IN THE COSSACK FIELD

Water saturation measurements were made on core in two Cossack field appraisal wells to investigate the discrepancy between the water saturation calculated from logs and that observed from the capillary pressure data in the Cossack-1 discovery well. The core measurements resulted in a more accurate water saturation parameter which was then used to estimate the volume of hydrocarbons in place. The core measurements also provided valuable information about the wettability of the reservoir rocks. The results of this exercise also highlighted the uncertainty attached to the water saturation determination from logs in what was an apparently simple case of a thick, clean sandstone reservoir.

Wettability and Its Effect on Oil Recovery

wide acceptance of the conclusion that most reservoirs are at wettability conditions other than VSWW. This conclusion has led to a resurgence of interest in satisfactory proce­ dures for measuring reservoir wettability and determining its effect on oil recovery, especially with respect to waterflooding. De­ termination of reservoir wettability and its effect on oil recovery by methods that in­ volve core samples will be referred to as ad­ vanced core analysis for wettability (ACAW). Reservoir wettability is not a simply de­ fined property. Classification of reservoirs as water-wet or oil-wet is a gross oversim­ plification. Various procedures for meas­ uring wettability have been proposed. Two methods of quantifying wettability based on rocklbrine/oil displacement behavior, the modified Amott test 2 and the USBM test, 3 are in common use. Each method depends on water saturation measurements and re­ lated capillary pressures or flow conditions to define a wettability scale. The tests show that reservoir wettability can cover a broad spectrum of wetting conditions that range from VSWW to very strongly oil-wet. With­ in this range, complex mixed-wettability conditions given by combinations of preferentially water-wet and oil-wet surfaces have been identified. In this paper, the 'adopted scales of reservoir wettability and their relationships to interface boundary con­ ditions are considered together with the dra­ matic effects that wettability can have on oil recovery.

Pressure Core Measurements in Tight Sandstone Lenses During the Multiwell Experiment

SummaryA pressure-coring operation was performed in two tight sandstone lenses of the Mesaverde group during the drilling of one well of DOE's Multiwell Experiment (MWX). The goals were (1) to obtain accurate water saturation data; (2) to measure the amount and composition of gas on a foot-by-foot basis; and (3) to recover fluids for water-chemistry analysis. This application of pressure coring had not been tried before; it was felt to be a novel approach for accurate water-saturation measurements in tight sandstone core.Several changes to usual pressure-coring procedures were adopted, including the development of an organic, noninvasive coring fluid compatible with the oil-based drilling fluid used to help preserve the original water saturation. The pressure-core assembly was modified so that gas samples could be taken directly from the inner barrel and would allow the capture of any gas not trapped during the process; however, none was found. Three 15-ft [4.6-m] -long pressure-core barrels were taken from two sandstone lenses 20 and 30 ft [6.1 and 9.2 m] thick.Water-saturation data were obtained on the pressure core by both Dean-Stark distillation and vacuum-oven drying. The analyses of all recovered core suggest that water-extraction procedures should be carried on for a long time and that the Dean-Stark procedure may not be the water-extraction procedure of choice for tight sandstone core. The failure of one core barrel to hold pressure turned out to provide a bonus, because identical analyses were made on pressurized and unpressurized core from the same sandstone. These analyses showed no appreciable water-saturation difference on rock with the same porosities. No fluid was recovered during thawing alone; the rock was too tight. Gas was obtained in the pressure-coring process and measured on a foot-by-foot basis.

Comparison of Conventional Core Data, Well Logging Analyses, and Sidewall Samples

Conventional core data and analyses of sidewall samples have been compared with data derived from well logs in reservoir rocks of varying lithologies. Correlation of porosity, hydrocarbon and water saturation, and permeability variations makes it possible to use one data set-taking special precautions when only sidewall samples are available-in the absence of other measurements. Introduction The comparison of core data from both conventional and sidewall samples with log-derived data makes possible a more realistic and reliable evaluation of possible a more realistic and reliable evaluation of the reservoir. Unfortunately, published comparisons of rock properties obtained from core and sidewall samples are scarce. Included in the many generalized conclusions in those papers are thatpercussion sample porosities in softer, looser sands are only slightly higher than those of conventional cores;sidewall sample permeabilities are decreased in higher permeability formations; andwater saturations of the sidewall cores are higher and oil saturations slightly lower than those of conventional cores. Although similar conclusions can be drawn from this study, our investigation uncovered additional complications. Comparison of Conventional and Sidewall Samples An example from Offshore Louisiana shows the comparison between conventional cores using a rubber-sleeved core barrel and sidewall samples (Figs. 1 and 2) in a Pliocene sand. Handling of this Pliocene reservoir rock was complicated by its semiconsolidated nature and by the presence of hydratable clays, which gave up water even at room temperatures if exposed for long periods of time. Since the core-gamma ray log showed multiple sand and shale partings corresponding to lithology variations over short distances, it was difficult in many cases to find comparable porosities within a 6-in. interval of the original core sample. Windows were cut in the rubber sleeve, test plugs were removed and the rubber sleeve was plugs were removed and the rubber sleeve was rescaled to preserve the remaining core. As the study progressed it became apparent that the Dean-Stark extraction method for determining saturations was releasing clay-bound water along with pore water, resulting in too high a measurement of water saturation; using the summation-of-fluids method with varying water plateaus gave more reasonable results. Finally, the plateaus gave more reasonable results. Finally, the remaining rubber sleeve was slit lengthwise and a continuous colored photographic log was taken of the core. Such photographs show clearly the varying degrees of both lenticularity and consolidation. The effect of these two core analyses on the porosity and permeability data in this shaly pay sand as reported permeability data in this shaly pay sand as reported by a commercial core laboratory are shown in Tables 1 and 2. We do not mean to imply by the tables that fluid saturations by "summation of fluid" (SOF) is generally or usually better than the Dean-Stark method. In fact, we prefer the latter method over the SOF method almost exclusively. However, in this particular case, the SOF method under closely controlled particular case, the SOF method under closely controlled conditions (not routine procedures) resulted in data that were reasonable and acceptable. A Miocene Gulf Coast formation was also cored using rubber-sleeve coring equipment. JPT P. 1409

Coring for Reservoir Connate Water Saturations

Coring with oil-base muds is generally accepted as the most reliable method for determining connate water saturations above the transition zone in an oil reservoir. The basis for using less expensive, indirect methods, such as capillary pressure tests rests largely on empirical agreement with cores taken using oil-base mud. Hence, testing the validity of this coring procedure is vital. procedure is vital. In coring for connate water with oil-base mud, the mechanism most likely to produce inaccurate determinations is flushing of water from the rock matrix by high pressure gradients near the bit. We know that similar flushing can occur when coring for residual oil saturations with water-base muds. However, wetting and nonwetting residual phase saturations are held by the rock in distinctly different ways. In preferentially water wet rocks, connate water is held as preferentially water wet rocks, connate water is held as hydraulically isolated volumes retained in finer pores by capillarity. Residual oil, however, is trapped in the larger pores as discontinuous globules which are not strongly attached to rock surfaces. Thus, we might expect that, during coring, flushing of connate water would be considerably less severe than flushing of residual oil. To test this hypothesis, we conducted two flooding experiments. In the first, a Berea sandstone core containing normal waterflood residual oil was further waterflooded at differential pressures sufficiently high that significant amounts of additional oil were flushed from the core. Results of this experiment are presented as the oil curve in Fig. 1. In the second experiment, a separate Berea core containing normal oilflood residual water was further oil flooded over the same range of pressure gradients employed in the first experiment. The results the brine curve in Fig. 1showed that comparatively little additional water was flushed from the core. To further test the hypothesis, we conducted two microbit coring experiments to compare degrees of residual phase flushing when coring for residual oil and for residual water. (A comprehensive study of coring for residual oil saturation with water-base muds has been reported in Ref. 1, which contains experimental details.) We prepared one rock section for coring by flooding with brine at the relatively low pressure gradient of 100 psi/ft until no more oil was pressure gradient of 100 psi/ft until no more oil was produced. We then cored the rock section, with a produced. We then cored the rock section, with a 200-psi borehole pressure differential, using brine as the coring fluid. Similarly, we flooded a second rock section with oil, at 100 psi/ft, to residual brine saturation, then cored, with oil as the coring fluid, with a 200-psi borehole differential pressure. During coring of the rock section containing residual water, pore pressure measurements along the rock section indicated that the entire rock was exposed to pressure gradients of about 1,500 psi/ft. Although pressure gradients of about 1,500 psi/ft. Although this gradient represents an extremely severe coring condition, water saturation in the core was reduced only 17 percent. In the comparable coring test using the rock section at residual oil saturation, the rock was exposed to pressure gradients of 900 psi/ft and residual oil saturation was reduced by 40 percent. Results of these laboratory flooding and coring tests show that, under similar conditions, coring for immobile water saturation is more reliable than coring for residual oil saturation. We previously demonstrated the feasibility of accurately coring for residual oil saturation by controlling the borehole to formation overbalance pressure. We conclude that controlled coring with oil-base mud should produce a satisfactory measurement of water saturation above the transition zone. P. 932

Saturation From Logs Laboratory Measurements of Logging Parameters

Abstract While logs provide the only in situ measurement of water saturation, the measurement is not direct. Calculated saturation depends upon the value assumed for the saturation exponent n. The literature furnishes evidence that n varies and that it depends upon the type of rock and fluids, wettability, confining pressure, etc. Hence, a laboratory technique was developed for measuring m and n under a confining pressure and at formation temperature. The technique involves mounting a sample in a Hassler-type sleeve, using an oil-brine flow system and making two- and four-electrode measurements. Results of measurements on Frio, Miocene and Berea sandstones and Cretaceous limestones are presented with applications to log analysis. Introduction Decisions of whether or not to complete and how to complete, as well as reservoir engineering calculations, all require accurate water saturation data. Narrowing profit margins and development of marginal fields place strong demands on accuracy of data. Logs provide the only in situ technique for determining water saturation. However, the measurement is not direct since the logs only measure formation resistivities. While equations relating resistivities to saturations were developed more than 20 years ago, these equations are too general for accurate application to all fields or formations. Also, the equations contain constants which are known to vary with rock type, stress, water resistivity, wettability, etc. Until more is known about these basic relationships, the recent advances in log quality and application of computers to log analysis cannot be fully utilized. The basic equations for resistivity log analysis are (1) where phi = fractional porosityF = formation factor (the ratio of the resistivity ofa brine-filled core, Ro, to the resistivity ofthe brine, Rw)m = m factor or cementation factora = an empirical constant (2) where Ro = resistivity of brine-filled coreRt = resistivity of core at saturation, Swn = saturation exponent. Constants a, m and n ideally should be determined for every reservoir. Currently, if no laboratory-measured values are available, the values a = 1.0 and m 2 are commonly used for limestone reservoirs, and a 0.62 and m = 2.15 are used for sandstone reservoirs. A value of 2.0 is assumed for n. Values for n ranging from 1.0 to 4.3 for natural core, and up to 54 for synthetic cores, are reported in the literature (Table 1). Also, several references show that the measuring technique is very important and that many of the problems associated with relative permeability measurements might be expected. Even allowing for measurement problems, there is ample evidence to show that n does vary. JPT P. 251ˆ

The Effect of Gypsum on Core Analysis Results

Published in Petroleum Transactions, AIME, Volume 216, 1959, pages 221–224. Abstract The presence of gypsum in samples subjected to standard core analysis introduces serious errors in the measurement of water saturation and porosity. The magnitude of these errors, depending upon the type of analysis procedure used, has been evaluated experimentally and theoretically. Water-saturated pore volumes determined by vacuum distillation or retort procedures may be too high by as much as 48 per cent of the volume of gypsum present. The maximum error in this value determined by Dean-Stark extraction with toluene is 36 per cent of the gypsum volume. If pore space is measured by a Boyle's law type porosimeter the maximum error due to gypsum is 38 per cent of the original gypsum volume present. The conscientious core analyst may eliminate the errors caused by gypsum by either of two methods. He may apply suitable corrections for the water of crystallization removed from gypsum, or take special precautions to prevent dehydration of gypsum during core cleaning and analysis procedures. These alternatives are discussed and methods for obtaining reliable core analysis data are proposed. Introduction Gypsum (CaSO42H2O) occurs as a common mineral in many of the world's sedimentary basins. In addition to its presence as primary bedded evaporitic deposits, it is often found as secondary replacement zones and pore-fillings in host rocks. That it may cause false indications in porosity, both on neutron logs and in standard core analysis, has been generally recognized. False porosity interpretations from neutron logs result because the chemically bound water in gypsum attenuates neutrons as effectively as free water. Errors in core analysis result when gypsum is dehydrated prior to or during the analysis. Vacuum retort procedures used by many laboratories to determine fluid saturation and porosity effect a complete removal of water of crystallization from gypsum. The crystal water, distilled and collected with free water from the sample, is counted as water-saturated pore space.