The probabilistic neural network (PNN) is functional in recognizing complex patterns without doing any pretraining of source data. However, for some data clusters, independence and colinearity characteristics of the variables in learning samples can seriously distort the window lengths of their probability density distributions, then leading to the incorrect or totally wrong calculated probability values and final recognition results. In view of such drawbacks, an improved PNN that incorporates two techniques of mean impact value (MIV) and correlation analysis is proposed in order to perfect the original PNN’s calculation mechanism by removing those interference and colinear variables from the source data. The data used to validate the method are from two wells in the Iara oilfield. Recognition accuracies of the improved network in four experiments are, 74.05%, 71.7%, 83.02% and 88.24%, respectively, each of which is the highest accuracy. The validation results demonstrate that the new network has the capability of recognizing complex carbonate lithofacies and the results are reliable enough to serve as the reference data for other geological efforts, such as analyzing sedimentary process and building a sequence framework.
In geological resarch, lithofacies identification is generally viewed as an important basic step because the results can provide remarkable revelations to other geological areas of study, such as analyzing sedimentary cycles, establishing sequence frameworks and constructing sedimentation models. For a particular well, in order to continuously obtain its lithofacies information, collecting rich source data in terms of lithology, electrical and petrophysical properties at each depth is essential, thus almost all the relevant identification methods are realized by processing logs (Baldwin et al., 1990; Carrasquilla et al., 2008). Crossplots are a classic tool used to predict lithofacies. The crossplot axes represent two well-log types that have significance in lithofacies classification, such as natural gamma ray (GR) and acoustic log (AC) logs, in the case of distinguishing sand and shale. With the advantage of crossplots, the lithofacies of noncored intervals can be predicted in accordance with the identification principles discovered from the data analysis of cored intervals (Busch et al., 1987; Dubois et al., 2007; Gifford and Agah, 2010; Grana et al., 2012). Nonetheless, crossplots methods require that each lithofacies has distinct characteristics on all the analyzed logs, which could not be recognized when dealing with complex lithofacies identification. As such, two other methods, statistics and neural networks, are rapidly being developed in place of the crossplot method (Tang and White, 2008; Insua et al., 2015).
Jácomo, Marta H. (Universidade de São Paulo) | Trindade, Ricardo I. F. (Universidade de São Paulo) | de Oliveira, Everton L. (Universidade de São Paulo) | Leite, Carlson de M. M. (Petrobras) | Montrazi, Elton Tadeu (Universidade de São Paulo) | Andreeta, Mariane (Universidade de São Paulo) | Bonagamba, Tito J. (Universidade de São Paulo)
Continuous chlorite coats are known to preserve the porosity in deeply buried sandstones by forming physical barriers to quartz early overgrowth. Sandstones from Água Grande Formation, Recôncavo Basin, Brazil, present anomalous porosity due to development of chlorite coating and is therefore suited for studying the nuclear magnetic response to this effect. Samples from this unit were classified into three groups according to their texture, composition and abundance of chlorite coatings: Group 1 with low amounts of coating, Group 2 with high amounts of coating and a nonreservoir sample (Group 3). Group 1 samples show wide NMR T2 distribution, while Group 2 samples present a bimodal T2 distribution. Nonreservoir Group 3 samples showed only a T2 peak in much shorter T2 times. To interpret the NMR results, transmittedlight optics, scanning electron microscopy, porosity (φ) and permeability (k) measurements, micro-CT, X-ray diffraction, magnetic susceptibility and hysteresis were used. We conclude that the longest T2 (> 0.1 s) peak of reservoir samples (Groups 1 and 2) is due to intergranular macropores, the intermediate peak is due to feldspar or clay intraclasts dissolution and the shortest peak (~0.01 s) is due to the chlorite coating itself, with minor contribution from secondary microporosity. The microporosity is predominant in the nonreservoir sample and relates to the clay-bound water. The shift to shorter times of longer T2 peaks in samples with higher contents of chlorite-bearing sandstones is likely related to diffusive coupling.
Wang, Haijing (Chevron U.S.A. Inc.) | Wang, Hanming (Chevron U.S.A. Inc.) | Toumelin, Emmanuel (Chevron U.S.A. Inc.) | Brown, Ronald L. (Chevron U.S.A. Inc.) | Crousse, Luisa (Chevron Latin America Business Unit)
Dielectric logging has evolved from a single-frequency mandrel tool in the 1970s to a multifrequency, fully articulated pad tool in the 2000s. Dielectric dispersion, the frequency-dependent dielectric property of sedimentary rocks, provides an additional dimension to petrophysical evaluation over broad frequency up to about 1 GHz. However, the interpretation of dielectric dispersion can be particularly difficult in organic-shale reservoirs, often due to a variety of polarization mechanisms and considerable uncertainties caused by complex mineralogy and organic matter.
In this paper, we present an integrated workflow including dielectric core analysis, processing of dielectric-dispersion logs, and petrophysical interpretation through core-log integration. We emphasize the need for accurate matrix-permittivity determination for all current interpretation methods, and explore the possibility to determine matrix permittivity directly from dielectric well logs. Dielectric core analysis is used to validate the interpretation model and calibrate dielectric well logs. For instance, matrix permittivity can be calibrated in the laboratory by optimizing the dielectric constant of each mineral and kerogen. This ensures that kerogen is lump-summed with the matrix for more accurate estimation of hydrocarbon volume. Multifrequency dielectric well-log data are then fitted with an appropriate mixing law or dispersion model to obtain petrophysical parameters, such as water-filled porosity, salinity, textural information, and flushed-zone resistivity. Inspired by the Pickett plot as a visual representation of the Archie equation, we propose a new graphical method that we call Complex-Domain Analysis (CDA) to solve dielectric-mixing-law equations without having to know matrix permittivity. This new method provides a simple way to determine a uniform matrix permittivity or matrix-permittivity endpoints, directly from dielectric log without a need for calculating it from mineralogy, thus very useful for quality control and dielectric interpretation immediately after logging. The integrated dielectric interpretation workflow and CDA method are demonstrated in two case studies in organic-shale reservoirs.
The Bakken Petroleum System (BPS) is composed of both conventional and unconventional units exhibiting significant variations in lithology, rock texture, clay content, total organic carbon (TOC), accompanied by high connate-water salinity, presence of disseminated pyrite grains, and low porosity. These petrophysical attributes lead to inconsistency in water-saturation estimates obtained from various subsurface logs, such as NMR log, resistivity log, dielectric dispersion log, and induction log.
An inversion-based interpretation method is applied to process dispersive electrical conductivity and dielectric-permittivity logs acquired at four dielectric-dispersion log-acquisition frequencies in the range of 10 MHz to 1 GHz. The Lichtenecker–Rother (LR), Stroud-Milton-De (SMD) models are coupled with the PS model, a mechanistic pyrite-clay dispersion model, to jointly process the dielectric-dispersion logs for simultaneous estimation of water saturation, water salinity, cementation index, and homogeneity index.
Using the proposed interpretation method, water-saturation estimate for a specific depth is obtained as a range of possible values within a desired accuracy. These estimates were compared against those obtained from the induction resistivity log, NMR log, Quanti-ELAN solver, the service company’s dielectric inversion, and Dean-Stark core measurements. Our estimates of water saturation and those obtained using the service company’s dielectric inversion exhibit a best match with Dean-Stark core water saturation in the Middle Bakken and Three Forks formations. The water-salinity estimates range from 150,000 to 350,000 ppm. The homogeneity index obtained using our method indicates the presence of layering and heterogeneity in the Lower Three Forks and Middle Bakken formations. The cementation index indicates high tortuosity and cementation in Upper and Lower Bakken formations.
Chiang, Wei-Shan (Aramco Research Center, National Institute of Standards and Technology, and University of Delaware) | LaManna, Jacob M. (National Institute of Standards and Technology) | Hussey, Daniel S. (National Institute of Standards and Technology) | Jacobson, David L. (National Institute of Standards and Technology) | Liu, Yun (National Institute of Standards and Technology and University of Delaware) | Zhang, Jilin (Aramco Research Center) | Georgi, Daniel T. (Aramco Research Center) | Kone, Jordan R. (Aramco Research Center) | Chen, Jin-Hong (Aramco Research Center)
Hydrocarbon production from shales using horizontal drilling and hydraulic fracturing has been the key development in the US energy industry in the past decade and has now become more important globally. Nevertheless, many fundamental problems related to the storage and flow of light hydrocarbons in shales are still unknown. It has been reported that the hydrocarbons in shale rocks are predominantly stored within the kerogen pores with characteristic length scale between 1 to 100 nm. In addition, it is possible that the 3D connectivity of these kerogen pores with fractures from the micrometer to centimeter scale, form the flow path for light hydrocarbons. Therefore, to better model the gas-in-place and permeability in shales, it is necessary to quantify the structural distribution of organic and inorganic components and fractures over a large breadth of length scales.
Simultaneous neutron and X-ray tomography offers a core-scale nondestructive method that can distinguish the organic matter, inorganic minerals, and open and healed fractures in 2.5 cm diameter shales with a resolution of about 30 µm and a field of view of about 3 cm. In the reconstructed neutron volume, the hydrogen-rich areas, i.e., organic matter, are brighter because hydrogen has a larger attenuation coefficient and attenuates neutron intensity more significantly. For the X-ray volume, the attenuation coefficient of an element is related to its atomic number Z and the brighter areas indicate the region containing more high-Z elements, such as some heavy minerals. Open fractures do not attenuate either neutrons or X-rays and therefore look dark in both reconstructed neutron and X-ray volumes.
In this study, two shale samples from different locations were investigated using simultaneous neutron and X-ray tomography for the first time. We were able to construct 3D images of the shales and isolate 3D maps of organic matter and minerals including high-Z elements. The distribution of kerogen and fractures can be used in the modeling of hydrocarbon flow in core scale, a 109 upscaling from current methods that model the flow based on SEM images.
Zhu, Linqi (Yangtze University) | Zhang, Chong (Yangtze University) | Guo, Cong (Yangtze University) | Jiao, Yifeng (Yangtze University) | Chen, Lie (Yangtze University) | Zhou, Xueqing (Yangtze University) | Zhang, Chaomo (Yangtze University) | Zhang, Zhansong (Yangtze University)
Shale reservoir exploration technology has attracted increasing attention, and total porosity is a parameter that characterizes the shale storage. Due to the complexity of mineral components and the large variety of pore types, the evaluation accuracy of total porosity of shale reservoirs is not satisfactory, at present. To address this problem, this paper proposes an evaluation method for shale reservoir total porosity based on a shale petrophysical model. We first established the petrophysical model for the calculation of total porosity and then eliminated the effect of gas saturation in the petrophysical model by combining density and neutron-porosity logging. After that, evaluations of matrix density, matrix neutron porosity, and organic matter were conducted using a combined method of elemental logging and conventional logging. Finally, the total porosity of the shale reservoir was calculated. The calculation results showed that by using the elemental logging method and based on actual conditions in the research area, the shale mineral composition could be obtained, and an accurate evaluation of matrix neutron porosity and matrix density could be realized. The total organic carbon (TOC) and organic matter (OM) in the shale reservoir can be accurately calculated according to conventional logging data. The evaluation accuracy of total porosity by this method was high, wherein the predicted relative error was only 0.4. Moreover, based on theoretical deduction, it is known that the proposed method has high applicability for shale reservoirs. If the inversion effect of matrix minerals can be guaranteed, an accurate calculation of shale total porosity can be obtained. In summary, the proposed method can accurately calculate the total porosity of shale reservoirs, which provides a reference for the exploration and exploitation of shale reservoirs.
Deep directional electromagnetic (EM) logging-while-drilling technology is the principal enabler of proactive well placement. The reservoir-scale measurements are used to map boundaries and fluid contacts more than 30 m away from the wellbore and to optimize well placement in increasingly more-complex scenarios. As the drilling progresses, deep directional resistivity (DDR) data are continuously inverted to estimate a 1D formation resistivity profile locally, and the inversion results are stacked to create a 2D reservoir map. This approach is adequate if the formation is layered and slowly varying laterally. However, in complex reservoir scenarios with locally 2D or 3D structures or where the formation changes abruptly, the 1D approximation used in real-time inversion may not be the most accurate solution.
We introduce a new, minimally biased pixel-based inversion to accurately map 2D complex reservoir structures. The inversion uses a 2.5D EM simulator, makes no assumption about the reservoir model, and is able to image non-1D geological structures, such as faults, sand injectites, shale lenses, and other complex geometries with arbitrary anisotropic resistivity distributions. An adaptive regularization ensures that the most plausible resistivity distribution with the least resistivity variation, consistent with the data, is found. The algorithm is parallelized to run on a cluster and is feasible for real-time application, provided relatively moderate computational resources are available.
The new reservoir maps derived from the 2D inversion enable more quantitative and informed well-placement decisions, deliver detailed insight about the reservoir structure, and allow a precise refinement of existing reservoir models.
Clastic laminated reservoirs have historically posed difficulties in formation evaluation. Difficulties are largely due to convoluted log responses, which preclude accurate assessment of key petrophysical properties, such as thin sand bed porosity and water saturation. In Southeast Asian (SEA) basins the abundance of silt in reservoir and nonreservoir rocks adds another layer of complexity and directly affects the design of appropriate data acquisition programs. This paper describes the silty thin-bed problem by assessing the efficacy and uncertainties of various log measurements to arrive at the correct petrophysical solution.
A review of rock physics literature is presented to highlight the underlying reasons for log behavior in silty facies. Generally, laminated rocks are evaluated from two different approaches: (1) high-resolution, or (2) bulk rock volume. High-resolution approaches include borehole image logs, deconvolution, and digital core imaging analysis. Bulk rock (or volumetric) approaches generally use Thomas-Stieber, multicomponent resistivity, and nuclear magnetic resonance (NMR) techniques. Three wells drilled in different sedimentary basins in SEA were selected to demonstrate the theory, challenges, and pitfalls of the most common approaches and techniques.
The Thomas-Stieber approach is often regarded as the most suitable for a binary sand-shale system and if conditional assumptions are met, results in a linear trend from which net-to-gross can be calculated. Adding a third component, such as silt, violates the assumptions and distorts this trend into a nonlinear “boomerang” shape.
Resistivity anisotropy, i.e., the vertical to horizontal resistivity ratio (Rv/Rh), provides further necessary input for accurate formation evaluation in laminated sand-silt-clay systems. Vertical resistivity is a key measurement as it is very sensitive to hydrocarbons in thinly laminated sands. Additional information, like borehole image and NMR data, contribute to reducing net-to-gross uncertainty and understanding the reservoir geometry. Where available, the saturation-height function results are compared to multicomponent resistivity results. One very silty to fine-grained sand reservoir in Vietnam displays anisotropy due to grain-size variation on a very fine level. In this example, the relevance of shale laminar volume is questioned and can only be justified with detailed core studies. It is, however, argued that reliable identification of hydrocarbon-bearing silt-rich sequences is only possible with multicomponent resistivity data. In addition, quantification of sand-lamina resistivity, Rsand, is possible in silty sands with variable amounts of irreducible water.
Although many papers discuss the thin-bed formation evaluation problem, very few publications address issues related to laminated sand-silt-clay reservoirs. This paper partly addresses this literature scarcity.
Hadibeik, Hamid (Halliburton) | Azari, Mehdi (Halliburton) | Kalawina, Mahmoud (Halliburton) | Ramakrishna, Sandeep (Halliburton) | Eyuboglu, Sami (Halliburton) | Khan, Waqar (Halliburton) | Al-Rushaid, Mona (Kuwait Oil Company) | Al-Rashidi, Hamad (Kuwait Oil Company) | Ahmad, Munir (Kuwait Oil Company)
Reservoir relative permeability as a function of saturation is critical for assessing reservoir hydrocarbon recovery, selecting the well-completion method, and determining the production strategy. It is a key input to reservoir simulation for predicting lifetime production of a well. Estimation of relative permeability curves at reservoir conditions is also a crucial task for successful reservoir modeling and history matching of production data. The relative permeability data estimated from core analysis may cause concern regarding representativeness, and adjustments are typically necessary for successful production forecasting. This paper proposes a new method to obtain relative permeability curves with downhole pressure-transient analysis of mini-drillstem tests (mini-DSTs) and well-log-derived saturations.
The new approach was based on performing mini-DSTs in the free water, oil, and oil-water transition zones. Analyses of the mini-DST buildup tests provided absolute formation permeability, endpoints of relative permeability to both oil and water, and curvature of the relative permeability data. Additionally, porosity and resistivity logs were used to determine irreducible water, residual oil, and transition zone saturations. Combining all of these downhole measurements provided the relative permeability curves.
When multiphase fluids flow in a reservoir, the flow rate of each phase depends on the effective permeability of that phase (Alkafeef et al., 2016). Effective permeability is obtained from absolute permeability of a reservoir multiplied by the relative permeability. Although absolute permeability is a function of reservoir pore geometry and does not change with fluid type, relative permeability is a fluid-dependent parameter and mainly depends on fluid saturation, pore geometry, viscosity, and surface tension (Goda and Behrenbruch, 2004).
Editor’s comment: This article is the third in a series of short “tutorial-like” notes styled to mentor users of digital well logs in becoming confident practitioners of petrophysics.When we left Shaly Sands Tutorial No. 1, Easy Money was riding high basking in the glow of having mastered the interpretation of low-contrast/low-resistivity shaly sandstone formations. Or so he thought. Only too soon, Easy Money crashed and burned by missing a low-resistivity pay zone in what appeared to be a clean sandstone. Since it looked to be shale-free, straight Archie was used resulting in too-high water saturation to be considered as pay in a clean sandstone. How, you ask, was this zone recognized as a productive hydrocarbon-bearing zone? It is obvious from the scenario explained above that a logs-only interpretation was a bust. However, two very important data streams had yet to be reconciled: First, the mud log showed a substantial gas show, coupled with a drilling break and faster drilling through the entire zone. Second, Archie’s directive to all his petrophysicists was “Don’t forget to look at the rocks!” So naturally a fast-drilling zone with gas shows had to be Fig. 8—The chemical structures of the “Big Four” clay mineral species using Grim’s (1968) notation. Left to right: kaolinite, chlorite, smectite, and illite. The legend shows the correspondence between color and atomic species.] investigated with sidewall samples. Solvent cuts at the wellsite indicated oil, and overnight we had cookouts that also indicated oil. Sample examination determined that the sandstone was moderately shaly! How could all of our usual logging methods missed that this rock was shaly? Well, Easy Money had more work to do to explain what happened and convince his team members that he would never miss this sort of zone again. I urge the reader to continue on with the remaining Shaly Sand Trilogy Tutorials and learn how Easy Money made sense out of these contradictory indicators by studying the chemical structure and physical chemical properties of clay minerals.