Enhanced recovery response prediction

Abstract

Methods of combining mineral composition and laboratory test results for reservoir rock samples to predict future responses to secondary and tertiary oil recovery treatments are disclosed. Particular, SEM and EDS will be combined to produce a mineral map, including mineral distribution around the rock's pore space, for comparison with laboratory data to predict and/or interpret how certain mineral distributions will respond to various fluid-rock interactions.

Claims

1. A method for improving enhanced oil recovery (EOR) from a hydrocarbon reservoir, said method comprising: a) obtaining one or more reservoir rock sample(s) from a reservoir of interest; b) obtaining scanning electron microscopy (SEM) image data in back-scattered electron (BSE) mode of said reservoir rock sample(s); c) obtaining energy dispersive spectral (EDS) data of said reservoir rock sample(s); d) generating a mineral map by combining said SEM image data and said EDS data on an individual pixel basis; e) assigning a chemical mineralogy of each pixel in said mineral map; f) determining a spatial profile of chemical mineralogy of one or more pore walls in said reservoir rock sample(s); g) evaluating said reservoir rock sample(s) in one or more laboratory EOR test(s); h) identifying a composition of one or more critical region(s) of pore walls by comparing said spatial profile with results from said evaluating step g; i) forecasting oil recovery for a plurality of EOR methods using a reservoir model for one or more regions in said reservoir having a same critical region of pore walls composition as said reservoir rock sample(s) from said identifying step h; j) selecting an optimal EOR method based on a best forecasted oil recovery from step i; and k) using said optimal EOR method from selecting step j to produce hydrocarbon from said reservoir.

2. The method of claim 1, wherein said one or more laboratory EOR test(s) are secondary recovery test(s).

3. The method of claim 1, wherein said one or more laboratory EOR test(s) are tertiary recovery test(s).

4. The method of claim 1, wherein said spatial profile includes chemical mineralogy of an interface between solid mineral grains and a void space that defines a pore in said pore walls.

5. The method of claim 1, wherein said one or more laboratory EOR test(s) are selected from wettability distribution, formation damage, electrical conductivity, permeability, and single or multiple fluid phase transport properties.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

(1) FIG. 1 depicts an exemplary SEM backscattered electron image of a rock where the variations in gray scale correspond with atomic number or grain density with the empty pores being black and mineral grains of increasing density defined by progressively lighter shades of gray.

(2) FIG. 2 depicts the process by which individual spectra of characteristic X-Rays captured at each grid point on the sample surface is interpreted as a discrete mineral phase by comparison to a set of standard spectra.

(3) FIG. 3 depicts a mineral map where individual mineral phases identified from the characteristic X-Ray spectra are gray-scale-coded, where black represents the pore space, and various shades of gray represent different mineral phases.

(4) FIG. 4 depicts a correlation between the total clay mineral abundance in a set of sandstone samples as determined by traditional bulk mineralogy methods, and the improvement in oil recovery in a low-salinity waterflood as noted by changes in water saturation (Sw). The absence of a strong correlation indicates that a measure of total clay mineral content is not a good predictor of enhanced oil recovery by low-salinity waterflood methods.

(5) FIG. 5 depicts a correlation between the amount of clay minerals in the same set of sandstones from FIG. 4 that is adjacent to the pore space as defined by the mineral map and the amount of improved oil recovery in a low-salinity waterflood as noted by changes in water saturation (Sw). The improved correlation indicates that the addition of spatial information on the mineral composition of the pore walls yields an improved prediction tool for enhanced oil recovery by low-salinity waterflood methods.

DESCRIPTION OF EMBODIMENTS OF THE DISCLOSURE

(6) The disclosure provides novel methods of predicting potential responses of a rock sample or reservoir to various enhanced oil recovery processes. Greater insights in the mechanisms involved in the fluid/rock interactions are also expected.

(7) Methods of using scanning electron microscopy coupled with an energy dispersive spectrometer to determine the qualitative mineral content of a sample are well known. While modal mineral abundances are estimated by these observational methods, a quantitative measurement of the relative proportion of each mineral phase with respect to each other and the pores is missing from these approaches. Further, no method has combined the mineral content information with actual laboratory data to quantitatively interpret fluid-rock interactions for critical regions in a rock sample.

(8) The presently disclosed method combines the spatial information of a mineral distribution of a sample, obtained from SEM and EDS analytical techniques, with laboratory or field-scale testing results to determine to determine how a specific mineralogy reacts. This can then be used to identify compositions of the critical regions of a sample's pore wall. This information can be applied to other samples having similar critical regions or mineral distribution to predict how the sample will respond to future test and to predict the effectiveness of a secondary or tertiary oil recovery treatment.

(9) The present methods includes any of the following embodiments in any combination(s) of one or more thereof:

(10) In one embodiment, this disclosure describes a method of predicting fluid-rock interactions for reservoir rock, comprising: a) obtaining scanning electron microscopy (SEM) image data in back-scattered electron (BSE) mode and energy dispersive spectral (EDS) data of a reservoir rock sample; b) assigning chemical mineralogy of each pixel using the EDS spectra compared to mineral standards; c) generating a mineral map by combining said SEM image data and said EDS-based mineral assignments on an individual pixel basis; d) determining a spatial profile of chemical analysis-based mineralogy of one or more pore walls in said reservoir rock sample; e) evaluating said reservoir rock sample's response to one or more enhanced oil recovery (EOR) tests, wherein said one or more EOR tests focus on interactions between a fluid and said reservoir rock sample; f) comparing said spatial profile with results from said evaluating step e) to identify the composition of one or more critical region of pore walls; and g) predicting fluid-rock interactions for reservoir rocks having the same critical region of pore walls composition as said reservoir rock sample.

(11) In another embodiment, this disclosure describes a method of correlating mineral composition of a reservoir rock at the pore-wall interface with fluid interactions, comprising: a) obtaining SEM image data in a BSE mode and EDS spectral data of a reservoir rock sample; b) assigning chemical mineralogy of each pixel using the EDS spectra compared to mineral standards; c) generating a mineral map by combining said SEM image data and said EDS-based mineral assignments on an individual pixel basis; d) determining a spatial profile of chemical-analysis based mineralogy of one or more pore walls in said reservoir rock sample; e) evaluating said reservoir rock sample's response to an EOR test, wherein said EOR test focuses on interactions between a fluid and said reservoir rock sample; f) comparing said spatial profile with results from said evaluating step e) to identify the composition of one or more critical region of pore walls; and g) correlating said pore mineral composition to said critical region.

(12) In another embodiment, this disclosure describes a method for forecasting an enhanced oil recovery process for a reservoir comprising: a) acquiring a reservoir rock sample from a reservoir of interest; b) obtaining SEM image data in BSE mode and EDS spectral data of said reservoir rock sample; c) generating a mineral map by combining said SEM image data and said EDS spectral data on an individual pixel basis; d) assigning a chemical mineralogy of each pixel in said mineral map; e) determining a spatial profile of chemical mineralogy of one or more pore walls in said reservoir rock sample; f) evaluating said reservoir rock sample's response to one or more EOR laboratory test on said reservoir rock sample; g) comparing said spatial profile with results from said evaluating step e) to identify the composition of one or more critical region of pore walls; and h) forecasting enhanced oil recovery for one or more regions in said reservoir of interest having the same critical region of pore walls composition as said reservoir rock sample.

(13) In one embodiment, the EOR test can be a secondary recovery test or a tertiary recovery test.

(14) In one embodiment, the spatial profile includes chemical mineralogy of the interface between solid mineral grains and void space that defines the pore in the pore walls.

(15) In one embodiment, the response can be wettability distribution, formation damage, electrical conductivity, permeability, and other single or multiple fluid phase transport properties.

(16) The inventive methods are described in additional detail next:

(17) The objective of coring and core analysis is to reduce uncertainty in reservoir evaluation by providing data representative of the reservoir at in situ conditions. Core derived data have been integrated with other field data to minimize reservoir uncertainties that cannot be addressed by other data sources such as well logging, well testing or seismic. The quality and reliability of core data have become more important with the ever-increasing pressure to optimize field development.

(18) A drawback of core analysis is the time needed to properly evaluate the reservoir. Attempts have been made to limit core analysis by building prediction models incorporating mineral composition data for the core. However, correlation of mineral composition to fluid-rock interactions have been limited to qualitative descriptions of pore wall mineral compositions to models predicting formation damage, wettability alteration, and geochemical reactivity during a chemical stimulation EOR process.

(19) Thus, there is still a need for methods that reduce the core analysis needed yet provide better interpretation of reservoir responses and better predictive power.

(20) The present methods address one or more of these needs by combining mineral composition information with real core data to identify critical regions that influence the rock's response to recovery methods and assign compositional information to these regions. This information can then be applied to subsequent samples to predict how the rock will respond to given stimuli.

(21) In more detail, reservoir rock samples, preferably core samples, can be obtained through normal means. The cores will be cleaned and dried according to standard industry practice and routine properties such as porosity and permeability will be measured. Ubani et al. discuss many of these methods in “Advances in Coring and Core Analysis for Reservoir Formation Evaluation”.

(22) A small petrographic thin section of the sample core or rock will be removed for analysis using SEM and EDS. Typically, approximately 30 μm thick sample is needed. However, thicker samples may be necessary if destructive analytical techniques will be performed after the SEM/EDS analysis. The remaining core will then be prepared for laboratory testing.

(23) SEM/EDS spectral data will be collected for the sample reservoir rock. The SEM data will be images, an example of which is depicted in FIG. 1, of the rock sample, including pore geometry and distributions of clay and/or other authigenic material associated with the pore system of the rock. The gray-scale on a backscattered electron image provides information on the relative density or composition of each grain or particle in the image wherein empty pores being black and mineral grains of increasing density defined by progressively lighter shades of gray.

(24) In the case of similar compositions between two mineral phases it is difficult to separate the two based solely on gray-scale differences. For each pixel in the image, an EDS chemical spectrum will be available to provide elemental composition at that pixel. This will provide elemental composition of each individual particle or grain in the SEM image at the greatest resolution of the image as defined by the dimensions of a pixel in each image.

(25) High resolution identification of mineral phases allows for the assignment of small amounts of minerals found adjacent to the pores, often referred to as the composition of the pore wall. The characteristic X-ray spectrum at each pixel is then assigned a mineral phase identification based on the “goodness-of-fit” between the spectrum and the spectra of a set of mineral standards, as shown in FIG. 2. The mineralogical analysis can be color coded, and overlaid over the SEM image, as shown in FIG. 3.

(26) The characteristic x-rays in the EDS analysis can be used to generate a mineral map of the distribution of specific elements whose EDS spectral is known, as shown in FIG. 2. This information will be used to identify and assign the mineralogy for each pixel and to determine distinct features such as grains and pores and the like. This is in contrast to the methods of overlaying a series of element maps over the original image and assigning a mineral phase based on a simple Boolean operation of combining minerals. This latter approach works with simple ore deposits (gold vs. silver vs. platinum), but not for minerals that have similar composition. A spatial profile of chemical mineralogy, particularly on the pore walls, detailing these features will be prepared.

(27) Once a mineral map such as that in FIG. 3 is prepared, the distribution of mineral phases, more specifically the relative amounts of the various minerals that are adjacent to the pore space, can be compared with the responses obtained from the laboratory tests. Any correlation amongst the type of mineral phase observed on the pore wall, or the relative abundance of any particular mineral on the pore wall, or the size of the pore, and any measure of the change in fluid-rock behavior of the laboratory core test, including but not limited to relative permeability of one fluid relative to another, or changes in hydrocarbon production effectiveness as determined by changes in amount of production or rate of production, can be used to develop predictive models. In some embodiments, a larger, field size test is also performed. Results for either the lab-scale or field-scale procedures can be used with the analytical techniques data.

(28) The techniques for performing the laboratory or field-scaled test will depend on the properties being measured and the type of oil recovery technique being studied. Below is a generalized description of steps taken to perform a core analysis and is not intended to limit the presently disclosed method to any one technique for obtaining a reservoir rock's response to various enhanced oil recovery methods.

(29) For the e.g. laboratory test, dead crude oil (crude oil without dissolved gas) is flowed into the reservoir rock core, and the core is aged at reservoir temperature for at least six weeks to restore wettability toward reservoir conditions. The dead crude oil flowed through the core will serve to measure oil permeability at initial water saturation. The same test can be used with live crude oil (crude oil with dissolved gas) measured at temperatures and pressures approaching reservoir conditions.

(30) Once the core is prepared, it will be flooded with a “formation brine”. The formation brine core flooding will set the baseline oil recovery compared to modified brines that are used in later laboratory experiments. The composition of the formation brine is not limited and can contain other additives that are needed to establish a baseline recovery. The relative differences between flood compositions (i.e. formation brine and test solution) is expected to be representative of the order of magnitude of the impact the varying compositions would have when implemented in actual hydrocarbon recovery operations. Subsequently solutions containing chemicals and additives used in various enhance recovery methods will be flowed through the core and the core's response will be monitored for comparison with SEM/EDS mineral maps.

(31) Responses regarding the oil recovery from the brine, formation damage, wettability alteration, and/or geochemical reactivity will be collected for use with the SEM/EDS spectral data.

(32) The spatial profile of the rock sample, particularly around the pore, will be compared with the results from one or more laboratory tests. Correlations between the rock's responses in the tests and the spatial profile can be determined. From there, critical regions of the pore wall that influence the fluid-rock interactions can be identified and the mineral composition can be determined. These correlations can then be applied to additional reservoir samples to negate the need for separate laboratory or field scale testing. In other applications, the correlations can be used to predict how the reservoir rock sample may response to testing with similar flood compositions.

(33) Information on the spatial distribution of minerals in a reservoir rock, especially those found adjacent to the pores can be used to predict a variety of responses associated with mineral-rock interactions during EOR processes, including but not restricted to estimating the amounts of surfactant that adsorbs to the pore wall surface, alterations in wettability due to specific adsorption of certain hydrocarbon molecules to specific mineral surfaces (for example certain asphaltenes and resins are known to adsorb preferentially to clay mineral and/or carbonate surfaces).

(34) Correlations between mineral distribution information and the results of an enhanced oil recovery test are developed from the summary statistics of each measurement. Enhanced oil recovery tests often produce estimates of incremental oil recovery following a specific treatment, or the final saturation after said treatment, or the amounts of residual fluids after said treatment, or other estimates of the efficiency of said test. The enhanced oil recovery results for a given sample or set of samples from the same reservoir are compared to the mineral distribution results, in particular the relative percentage of each mineral of interest that is adjacent or near the pores observed in the SEM images of the sample. The correlation can be improved by using samples for the SEM-based mineral analysis that are selected directly from the core plugs used in the enhanced oil recovery or other fluid-flow SCAL tests. Often a small piece taken from one end of the core plug is used for the SEM-based mineralogy analysis.

(35) Once completely analyzed, the correlations and results from the one or more rock samples can then be applied to a reservoir model, and reservoir performance predicted. The best EOR method can then be applied to the reservoir itself.

(36) The present method is exemplified with respect to the enhanced oil recovery test for a waterflood described below. However, this is exemplary only, and the method can be broadly applied to any core analysis, both small and large scale. The following descriptions are intended to be illustrative only, and not unduly limit the scope of the appended claims.

Test 1

(37) The present method was applied to suites of rock samples obtained from a sandstone reservoir under consideration for an enhanced oil recovery project. The enhance recovery method being considered was a low-salinity waterflood. This test shows how a set of sandstones from a reservoir being considered for an enhanced oil recovery project were evaluated with the spatial mineralogy process in order to gain a better understanding of the role in clay minerals in affecting ultimate recovery from that reservoir.

(38) A set of core plugs saturated with dead oil and at initial water saturation conditions was flooded with high-salinity water under standard laboratory core flooding procedures. After the production of oil leveled off and an end point or residual oil saturation was determined, a low-salinity brine was then introduced to the core flood experiment. The amount of incremental oil that was produced in addition to the original production was measured by the change in water saturation. The incremental oil increase from the low-salinity processed ranged from 0.01 to 0.07 fractional saturation units.

(39) Many models attempt to predict the improvement in oil recovery from low-salinity waterflood by estimating the total amount of clay minerals in the reservoir rock sample, but the results to date are equivocal. As shown in FIG. 4, the improvement in oil recovery in a low-salinity waterflood, as measured by the change in water saturation (Sw), does not correlate well with the total amount of clay minerals as measured by conventional methods.

(40) An SEM/EDS analysis was performed and Table 1 displays the mineral concentration adjacent to a large number of pores that can be further specified by selecting a range of pore diameters that are considered as critical.

(41) TABLE-US-00002 TABLE 1 Relative Mineral composition adjacent to a number of pores observed in a sandstone Mineral % Adjacent to Pore Quartz 30.8 Glauconite 20.0 Illite 10.0 Siderite 33.0

(42) The addition of spatial information on the relationship of these same clay minerals to the surface of the pore walls, from Table 1, improves the correlation that can in turn be used to predict enhanced oil recovery from a low-salinity waterflood process.

(43) The correlation observed in FIG. 5 relative to the absence of any definable trend in FIG. 4 illustrates how the spatial mineralogy information can be used to develop models that can predict fluid-rock interactions, including in this instance an estimate of enhanced oil recovery.

(44) The correlation in FIG. 5 can be used to develop a single-variable linear model between the relative proportion of the pore walls associated with clay minerals and the improvement in oil recovery. These predictive models are not limited to single variables and linear trends, a multivariate, non-linear model can be constructed to predict a range of fluid-rock properties based on any number of experimental or theoretical parameters, including information on the spatial distribution of mineral phases relative to the pores.

(45) Such models can be applied to determine an optimal recovery technique and then that best EOR actually applied to the reservoir.

(46) The following references are incorporated by reference in their entirety. US20140117231 Ubani, et. al., “Advances in Coring and Core Analysis for Reservoir Formation Evaluation”, Petroleum & Coal, 54(1) 42-51, 2012. Available at: http://www.vurup.sk/sites/default/files/downloads/pc_1_2012_ubani_149_0.pdf SPE-168789 (2013) Lemmens H. et al., From SEM Maps and EDS Maps to Numbers in Unconventional Reservoirs. Howard, et al., “Mineral Distribution in Reservoir Oil Rocks and its Impact on Enhanced Oil Recovery”, Presentation at the International Symposium of the Society of Core Analysts held in Austin, Tex., USA, Sep. 18-21, 2011.