Wireline Log Data Research Articles

Enhancing Paleoreef Reservoir Characterization through Machine Learning and Multi-Attribute Seismic Analysis: Silurian Reef Examples from the Michigan Basin

Historically, Silurian pinnacle reef complexes in the Michigan Basin have been largely identified using 2D seismic with very little research on the reservoir characterization of these reefs using 3D seismic data. By incorporating a high-resolution 3D dataset constrained by a well-studied and data-rich paleoreef reservoir, the Puttygut reef, seismic attributes were correlated to petrophysical properties through machine learning and self-organizing maps (SOMs). A suite of structural and frequency-based attributes was calculated from pre-stack time migrated (PSTM) seismic data, with only a subset of them selected as SOM inputs. Structural attributes enhanced details in the reef but frequency attributes were overall more useful for correlating with reservoir quality. A strong relationship between certain combination percentages of attributes and certain sections of the reef with porosity and permeability was found after the SOM results were compared to wireline log and core analysis data. Areas with high permeability and porosity correlated with the average frequency and spectral decomposition at 29 and 81 Hz. Areas with high porosity and varying permeability correlated with the average frequency and spectral decomposition at 29, 57, and 81 Hz. Areas with intermediate porosity correlated with the average frequency and spectral decomposition at 29 and 57 Hz. The efficacy of the procedure was then demonstrated on two nearby reefs with very similar results.

Comparison of logging-while-drilling and wireline logging data from gas hydrate-bearing deep-sea sediments, the Ulleung Basin, East Sea

Gas hydrate exploration was conducted using logging-while-drilling (LWD) and wireline logging (WLL) at two sites (UBGH2-6 and UBGH2-10) in the Ulleung Basin, East Sea (Sea of Japan). Coring for gas hydrate sampling was also conducted. Seismic profiles revealed that the upper layers at the logged sites (∼180 m at UBGH2-6 and ∼150 m at UBGH2-10) are dominated by turbidite/hemipelagic sediments, whereas the lower-lying sediments are characterised by thick mass transport deposits. Except in the gas hydrate-bearing interval (140–160 metres below the seafloor) of UBGH2-6, the WLL velocity and resistivity data are a little higher (∼100 m/s and ∼0.4 ohm-m, respectively) than those indicated by LWD. Conversely, the density and porosity do not show a consistent pattern for the two methods. Discrepancies between LWD and WLL can be caused by the methodology, borehole condition, and gas hydrate occurrence type. The high resistivity values (∼100 ohm-m) in the gas hydrate-bearing zone of UBGH2-6 are typical of inclined fracture-filling gas hydrate. In this case, the LWD resistivity may be overestimated as a result of the data acquisition principle. The relationships between the physical properties are significantly controlled by the presence of gas hydrate. Therefore, in fracture-filling gas hydrate-bearing sediments, the WLL resistivity data are more reasonable than the LWD resistivity data. Both methods have advantages and disadvantages in terms of data quality, efficiency, and resolution. Therefore, it is necessary to compare and interpret data measured using both methods to identify the physical properties of gas hydrate-bearing sediments.

Identification of Porosity and Permeability While Drilling Based on Machine Learning

The predictions of porosity and permeability from well logging data are important in oil and gas field development. Currently, many scholars use machine learning algorithms to predict reservoir properties. However, few scholars have researched the prediction of reservoir porosity and permeability while drilling. This approach requires not only a high prediction accuracy but also short model processing and calculation times as new logging data are incorporated while drilling. In this paper, four machine learning algorithms were evaluated: the one-versus-rest support vector machine (OVR SVM), one-versus-one support vector machine (OVO SVM), random forest (RF) and gradient boosting decision tree (GBDT) algorithms. First, samples of wireline logging data from the Yan969 wellblock of the Yan’an gas field were chosen for model training. To improve the accuracy and reduce the input parameter dimensions and model training time as much as possible, data correlation analysis was performed. Second, we used the grid search method to approximate ranges of reasonable parameter values and then used k-fold cross-validation to optimize the final parameters and avoid overfitting. Third, we used the four classification models to predict porosity and permeability while drilling with data from logging while drilling (LWD) logs. Finally, we indicate the best porosity and permeability prediction models to use while drilling. To ensure that the prediction accuracy is as high as possible and that the model training time is as short as possible, the OVO SVM algorithm was suggested for porosity and permeability prediction. Therefore, appropriate machine learning algorithms can be used to predict porosity and permeability while drilling.

Predicting Reservoir Fluid Properties from Advanced Mud Gas Data

SummaryIn a recent paper, we published a machine learning method to quantitatively predict reservoir fluid gas/oil ratio (GOR) from advanced mud gas (AMG) data. The significant increase of the model accuracy compared to traditional modeling approaches makes it possible to estimate reservoir fluid GOR based on AMG data while drilling, before the wireline operation. This approach has clear advantages because of early access, low cost, and a continuous reservoir fluid GOR for all reservoir zones. This paper releases further study results to predict other reservoir fluid properties in addition to GOR, which is essential for geo-operations, field development plans, and production optimization.Two approaches were selected to predict other reservoir fluid properties. As illustrated by the reservoir fluid density example, we developed machine learning models for individual reservoir fluid properties for the first approach, similar to the GOR prediction approach in the previous paper. As for the second approach, instead of developing many machine learning models for individual reservoir fluid property, we investigated the essential properties for equation of state (EOS) fluid characterization: C6 and C7+ composition and the molecular weight and density of the C7+ fraction. Once these properties are in place, the entire spectrum of reservoir fluid properties can be calculated with the EOS model.The results of reservoir fluid property prediction are satisfactory with both approaches. The reservoir oil density prediction has a mean average error (MAE) of 0.039 g/cm3. The accuracy is similar to the typical density derived from the pressure gradient from wireline logging data. For the essential fluid properties required for EOS model prediction, the overall accuracy is less than the laboratory measurements but acceptable as the early phase estimations. The reservoir fluid properties predicted from the EOS model are similar to the predictions from individual machine learning models. We applied the field measured AMG data into the reservoir fluid property models and achieved good results, as illustrated by the reservoir fluid density example.The previous paper completed the methodology to predict all reservoir fluid properties based on AMG data. This work paves the way to generate a complete reservoir fluid log for all relevant reservoir fluid properties while drilling. The method has a significant business impact, providing full coverage of reservoir fluid properties along the well path in the early drilling phase. The advantage of providing reservoir fluid properties in all reservoir zones while drilling far outweighs the limitation of somewhat reduced reservoir fluid property accuracy.

Correlation-based data analytics of wireline logs for decoding and modeling shale gas resources

Correlation is widely used and its importance has become even more pronounced in big data. Generally, scientists and engineers look for highly correlated variables and ignore weakly correlated variables. We use a more holistic view of correlations in evaluating unconventional resources by integrating correlation with causal inference. In formation evalutions, many variables are involved and they generally have different effects on hydrocarbon accumulations, and they are inter-correlated. These phenomena often lead to weak correlations among causally related variables. We show that using weak or moderate correlations can help identify gas-prone lithofacies. This methodology is related to discerning a statistical phenomenon, termed Simpson's paradox. Although this paradox has been interpreted as a bias in the literature, we show that it can genuinely represent a weak correlation as an effect of multiple causes. More importantly, we will show that both a weak and strong correlation of wireline-log data can be used to decode gas-prone organic mudstones in evaluating and developing shale gas resources when the causal inference based on physical laws is invoked.

Quantify Coal Macrolithotypes of a Whole Coal Seam: A Method Combing Multiple Geophysical Logging and Principal Component Analysis

Coal macrolithotypes control the reservoir heterogeneity, which plays a significant role in the exploration and development of coalbed methane. Traditional methods for coal macrolithotype evaluation often rely on core observation, but these techniques are non-economical and insufficient. The geophysical logging data are easily available for coalbed methane exploration; thus, it is necessary to find a relationship between core observation results and wireline logging data, and then to provide a new method to quantify coal macrolithotypes of a whole coal seam. In this study, we propose a L-Index model by combing the multiple geophysical logging data with principal component analysis, and we use the L-Index model to quantitatively evaluate the vertical and regional distributions of the macrolithotypes of No. 3 coal seam in Zhengzhuang field, southern Qinshui basin. Moreover, we also proposed a S-Index model to quantitatively evaluate the general brightness of a whole coal seam: the increase of the S-Index from 1 to 3.7, indicates decreasing brightness, i.e., from bright coal to dull coal. Finally, we discussed the relationship between S-Index and the hydro-fracturing effect. It was found that the coal seam with low S-Index values can easily form long extending fractures during hydraulic fracturing. Therefore, the lower S-Index values indicate much more favorable gas production potential in the Zhengzhuang field. This study provides a new methodology to evaluate coal macrolithotypes by using geophysical logging data.

Resolving lithostratigraphic complexities in the Crayfish Group, Otway Basin using chemostratigraphy

The Otway Basin comprises a significant part of the eastern Australian Southern Rift System, a divergent passive continental margin formed during the Cretaceous separation of the Australian and Antarctic continents. Early rifting activity resulted in the development of many half grabens within the Otway Basin, which are largely infilled by sediments of the Casterton Formation and Crayfish Group. Despite over 20 years of exploration and hydrocarbon production from these units however, their lithostratigraphic characterisation and nomenclature remain ambiguous, with structural complexity and prevalent lateral facies changes leading to confusion in their basin-wide correlation. Deposited in a largely non-marine, fluvial/lacustrine environment, repeating cycles of sandstones and shales of the Crayfish Group can be difficult to resolve using petrology, palynology and wireline log data. The use of chemostratigraphy is favoured as an investigative tool in this situation since changes in provenance, lithic composition, facies, weathering and diagenesis are reflected in the mineralogy of the sediments, resulting in variations in their inorganic geochemistry. Uniform sedimentary successions can thus potentially be differentiated into unique sequences and packages based on their characteristic geochemistry, aiding in the resolution of complex structural relationships and facies changes. In this study, we present new inorganic geochemistry data for four key wells in the South Australian (SA) Penola Trough and interpret the geochemistry data consistent with, and building on, the chemostratigraphic schema of Forbes et al. to demonstrate its utility and robustness. We then undertake inter-well wireline log correlations across the SA Penola Trough using the wells with chemostratigraphic data as controls.

Hybrid wavelet transform with artificial neural network for forecasting of shear wave velocity from wireline log data: a case study

Shear wave velocity (Vs) is an important variable for performing geomechanical and geophysical modeling and reservoir studies. Field tests to measure this variable directly are high costs and time consuming. Due to the operational difficulties mentioned above, it is more convenient estimating Vs without direct measurements from conventional log data. In this research, the hybrid of wavelet transform with artificial neural network is utilized to estimate the Vs. To input variables (log gamma, log compressional wave velocity, and log bulk density), preprocessing is done using wavelet transform and then variables are entered to artificial neural network model. The estimation abilities of the hybrid artificial neural network with wavelet transform were substantiated using field data achieved from Marun reservoir, Iran. The results obtained in this study show a positive effect of input parameters’ preprocessing using wavelet transform in the estimation of Vs, and it has led to noticeable increase in the accuracy of model calculations.

Lithological attributes of hd-001 well, shallow offshore, Niger delta basin, Nigeria

Lithological analysis was carried out on 96 ditch cuttings samples from HD-001 well located within the shallow offshore Niger delta basin, Nigeria. Three lithofacies sequences were delineated by the integration of wireline logs textural/lithologic attritudes and the distribution of index accessory minerals. They are transitional paralic, paralic and marine paralic sequences. The lithologic, textural and wireline log data indicate that the entire interval studied in the HD-001 well belongs to the Agbada Formation. The Formation is made up of alternating sand and shale units which suggests rapid shoreline progradation. The grain size increases from essentially fine to medium-grained at the basal part of the well to dominantly coarser grain at the upper part. The index accessories recognize shallow marine to coastal deltaic settings environment of deposition. Sand bodies which represent sub-environments within those settings are deposited in sequences. Each sequence begins with a transgressive phase followed by significant regressions.  Â

The sedimentary architecture of hyperpycnites produced by transient turbulent flows in a shallow lacustrine environment

Hyperpycnal flows are river-derived extrabasinal turbidity currents transporting both sand and clay to lacustrine, coastal, shelf and deepwater sedimentary environments. Experimental research in the past twenty years has shown that the presence of clay in sediment-laden flows promotes a transitional behavior between fully turbulent flow and a quasi-laminar plug flow regime. However, to date, most work concerning the fluid dynamic interpretation of lacustrine hyperpycnal flows has been founded on concepts based on fully turbulent flows, and relatively little is known about the influence of clay on the rheological and turbulence characteristics of these flows. With the help of 3D seismic volume and wire-line logging data, the present study undertakes a description and interpretation of subsurface mixed sandstone–mudstone bedforms observed in cores from the Upper Cretaceous Heidimiao Sandstone, Nenjiang Formation, Songliao Basin, NE China. Six lithofacies and five lithofacies associations are recognized from the cores and interpreted as formed under both turbulent and transitional flows. Typical bedforms of transitional flows include large current ripples and low-amplitude bedwaves. The delta-fed shallow lacustrine hyperpycnal flow deposits are characterized by proximal sinuous channels (extending for c. 20 km) and lobe deposits (extending for >20 km), showing a fan-like geometry associated with distributary channel extension in the down-dip direction. Basal erosion in the proximal channels is produced by the initial turbulent and turbulence-enhanced transitional flows, with channel infill dominated by upper transitional plug flows or quasi-laminar plug flows. The lobe deposits possess a fan-like geometry with large aspect ratio, and were formed from turbulence-enhanced and lower transitional plug flows. The downdip transformation of sediment-laden flows in the lacustrine basin varies from an initial turbulent flow, via quasi laminar plug flows to upper and lower transitional plug flows, to turbulence enhanced transitional flows. These transitional flows are interpreted to have experienced gradual dilution and deceleration and eventually transformed to turbulent flows in their distal regions. A new model for delta-fed shallow lacustrine hyperpycnal flow deposits is presented incorporating decelerated transitional flows and flow transformations. This new model can aid understanding of the depositional processes of lacustrine transitional flows and the facies distribution of hyperpynal flows in both modern and ancient sediments.

Azimuthal P-P prestack amplitudes in the presence of oil-filled aligned porosity (fracture porosity)

A 3D P-P high-fold full-azimuth full-offset reflection survey was acquired and processed to characterize a naturally fractured carbonate reservoir. The reservoir is a thick carbonate, which will flow commercial oil with a sufficient fracture network. Extensive calibration data include (1) a horizontal borehole’s resistivity image log, (2) the first 24 months cumulative oil produced, by stage, as known from chemical frac tracer data, (3) pre- and postfrac job instantaneous shut-in pressures, (4) microseismic, and (5) wireline log data. We used the cumulative oil production to document the spatially varying amount of aligned vertical porosity (aligned compliance or fracture porosity) connected to the stage borehole location. The stages of high oil production exhibited, for the fracture-perpendicular azimuth, the more positive amplitude variation with angle (AVA) gradients, and dimmer near-angle (6°–15° angles of incidence) amplitudes, compared to the fracture-parallel azimuth. The azimuthal variation of the AVA gradient fit the cos 2θ curve well, indicating the presence of one set of vertical aligned fractures dominating the azimuthal amplitude signature. In a similar fashion, the azimuthal variation of the mathematical intercept, physically the near-angle amplitudes, also fit the cos 2θ curve well. We have constructed crossplots of the azimuthal near-angle amplitude versus the AVA gradient on a bin-by-bin basis: we observed a straight line at bins with elevated oil production (elevated fracture density). A straight line crossplot of the (AVA gradient, mathematical intercept) is the signature of change of the (sensed) porosity, as long as the lithology and pore fluid are held constant. In accord with industry knowledge, we found that porosity affects the P impedance and thus the near-angle amplitudes: the aligned porosity yields azimuthal P impedance (measured at the 6°–15° angles of incidence). Legacy high-fold 3D P-P surveys rich in the 6°–20° angles of incidence should be considered for reprocessing and reinterpretation using these techniques.

Application of machine learning tool to predict the porosity of clastic depositional system, Indus Basin, Pakistan

Porosity is one of the key factors of a reservoir system that is typically distributed in a spatially non-uniform and non-linear manner. Nevertheless, spatial distribution of porosity in stratigraphic traps (channel sand bodies) is challenging due to the variations in depositional environments, shale intercalation, and repetition of textural changes in the pore sizes. This paper presents a strategy of joint inversion that combines support vector machine (SVM) and particle swarm optimization (PSO) algorithms to predict the spatial distribution of porosity using well logs and seismic data. In addition, Gaussian simulation algorithms (sequential Gaussian and Gaussian indicator simulations) were used to predict lithology and porosity distribution using wireline logs and core data. Rock physics modeling was performed to bridge the gap between the elastic and petrophysical properties for seismic-inversion based reservoir characterization. The results show that the joint inversion strategy leads to the most stable prediction of AI and porosity distribution in the lower Goru heterogeneous reservoir of Sawan Gas Field, Pakistan. The calibration of the individual spatial distribution of lithology and porosity from wireline logs data using Gaussian simulation algorithms and post-stack seismic inversion using SVM and PSO; revealed favorable matching. In fact, the spatial pattern of low AI corresponds with the high porosity and sandstone lithofacies. The study results provide valuable insights for identifying the sweet spots in the heterogeneous lower Goru clastic packages of Sawan Gas Field through quantitative analysis of rock physics, petrophysics, seismic inversion, and fuzzy optimization. The proposed joint inversion strategy may be considered useful for the prediction of porosity in similar context of extremely heterogeneous stratigraphic traps in other basins of the world.

Machine Learning-Based Probabilistic Lithofacies Prediction from Conventional Well Logs: A Case from the Umiat Oil Field of Alaska

A good understanding of different rock types and their distribution is critical to locate oil and gas accumulations in the subsurface. Traditionally, rock core samples are used to directly determine the exact rock facies and what geological environments might be present. Core samples are often expensive to recover and, therefore, not always available for each well. Wireline logs provide a cheaper alternative to core samples, but they do not distinguish between various rock facies alone. This problem can be overcome by integrating limited core data with largely available wireline log data with machine learning. Here, we presented an application of machine learning in rock facies predictions based on limited core data from the Umiat Oil Field of Alaska. First, we identified five sandstone reservoir facies within the Lower Grandstand Member using core samples and mineralogical data available for the Umiat 18 well. Next, we applied machine learning algorithms (ascendant hierarchical clustering, self-organizing maps, artificial neural network, and multi-resolution graph-based clustering) to available wireline log data to build our models trained with core-driven information. We found that self-organizing maps provided the best result among other techniques for facies predictions. We used the best self-organizing maps scheme for predicting similar reservoir facies in nearby uncored wells—Umiat 23H and SeaBee-1. We validated our facies prediction results for these wells with observed seismic data.

Biostratigraphy of the Ulleung Basin, East Sea (Japan Sea)

An interpretation of both micropaleontological and palynological analyses was carried out for eleven wells from the Ulleung Basin, East Sea (Japan Sea). Based on bioevents of microfossil groups (calcareous nannofossils, foraminifera, diatoms, dinocysts, spores and pollen) twenty-three datums are established. The biostratigraphic and paleoenvironmental data indicate significant tectonic events such as subsidence (17 Ma), closing of the Korea Strait (15 Ma), and uplift in Dolgorae deformed region (12.5–10.2 Ma) of the basin. The high resolution sequence biostratigraphy, correlated with wireline log and seismic data, provided to be effective for constraining the interpretation of depositional systems tracts of the basin. The Paleogene reworked taxa in Neogene section suggest that the provenance of sediments of the basin was southern offshore basins of Korea, and/or southwestern Japan (Kyushu area) and the Fukue Basin.

Organic-rich shale caprock properties of potential CO2 storage sites in the northern North Sea, offshore Norway

Assessment of the geomechanical properties of organic-rich shale caprocks is critical for a successful CO2 storage into a saline aquifer. In this study, we investigated the geochemical properties of the organic-rich shale caprocks of the Draupne and Heather formations, overlying the potential sandstone reservoirs of Sognefjord, Fensfjord, and Krossfjord formations in the northern North Sea, offshore Norway. The caprock's depositional variations within the sub-basins are established by analyzing the gamma-ray shape and stacking patterns. The effect due to differences in depositional environments, on the caprock compaction behavior is investigated by integrating petrographical analysis of core and cutting samples from 3 wells and by rock physical analysis of wireline log data from 27 exploration wells. Three rock physics templates are used where the wireline log data are interpreted using the published background trends. The effect of kerogen type, maturation level, and deposition environment on caprock properties within the study area are also evaluated. Moreover, the caprock property, such as brittleness, is estimated by using four mineralogy and elastic property-based, empirical relations, which is a quantitative measure of caprock property with respect to changes in stress-state. Finally, the seismic inversion method is assessed for the possibility of extracting caprock properties from surface seismic data. Regardless of compaction processes, the results indicate that the Heather Formation is mechanically stronger than the Draupne Formation. However, both formations appear to be ductile in nature. The depositional environments control the mineralogical composition and fabric of the Draupne and Heather formations, which influence the caprock properties significantly. Results also show that the effect of TOC on caprock properties is insignificant in the study area. The brittleness of the organic-rich shale caprocks in the study area follows a different trend compared to the published trends. We also observed an excellent correlation between the log-derived elastic properties and geomechanical parameters. Still, it is difficult to assess the caprock elastic properties from seismic due to the overlap of data clusters. The evaluation of caprock geomechanical behaviors is challenging as these properties are site-specific and also influenced by other factors such as exhumation, in-situ stress conditions, the existence of natural fractures, and their orientations.

Lower Cretaceous Rodby and Palaeocene Lista Shales: Characterisation and Comparison of Top-Seal Mudstones at Two Planned CCS Sites, Offshore UK

Petroleum-rich basins at a mature stage of exploration and production offer many opportunities for large-scale Carbon Capture and Storage (CCS) since oil and gas were demonstrably contained by low-permeability top-sealing rocks, such as shales. For CCS to work, there must be effectively no leakage from the injection site, so the nature of the top-seal is an important aspect for consideration when appraising prospective CCS opportunities. The Lower Cretaceous Rodby Shale and the Palaeocene Lista Shale have acted as seals to oil and gas accumulations (e.g., the Atlantic and Balmoral fields) and may now play a critical role in sealing the Acorn and East Mey subsurface carbon storage sites. The characteristics of these important shales have been little addressed in the hydrocarbon extraction phase, with an understandable focus on reservoir properties and their influence on resource recovery rates. Here, we assess the characteristics of the Rodby and Lista Shales using wireline logs, geomechanical tests, special core analysis (mercury intrusion) and mineralogical and petrographic techniques, with the aim of highlighting key properties that identify them as suitable top-seals. The two shales, defined using the relative gamma log values (or Vshale), have similar mean pore throat radius (approximately 18 nm), splitting tensile strength (approximately 2.5 MPa) and anisotropic values of splitting tensile strength, but they display significant differences in terms of wireline log character, porosity and mineralogy. The Lower Cretaceous Rodby Shale has a mean porosity of approximately 14 %, a mean permeability of 263 nD (2.58 × 10−19 m2), and is calcite rich and has clay minerals that are relatively rich in non-radioactive phases such as kaolinite. The Palaeocene Lista Shale has a mean porosity of approximately 16% a mean permeability of 225 nD (2.21 × 10−19 m2), and is calcite free, but contains abundant quartz silt and is dominated by smectite. The 2% difference in porosity does not seem to equate to a significant difference in permeability. Elastic properties derived from wireline log data show that Young’s modulus, material stiffness, is very low (5 GPa) for the most shale (clay mineral)-rich Rodby intervals, with Young’s modulus increasing as shale content decreases and as cementation (e.g., calcite) increases. Our work has shown that Young’s modulus, which can be used to inform the likeliness of tensile failure, may be predictable based on routine gamma, density and compressive sonic logs in the majority of wells where the less common shear logs were not collected. The predictability of Young’s modulus from routine well log data could form a valuable element of CCS-site top-seal appraisals. This study has shown that the Rodby and Lista Shales represent good top-seals to the Acorn and East Mey CCS sites and they can hold CO2 column heights of approximately 380 m. The calcite-rich Rodby Shale may be susceptible to localised carbonate dissolution and increasing porosity and permeability but decreasing tendency to develop fracture permeability in the presence of injected CO2, as brittle calcite dissolves. In contrast, the calcite-free, locally quartz-rich, Lista Shale will be geochemically inert to injected CO2 but retain its innate tendency to develop fracture permeability (where quartz rich) in the presence of injected CO2.

Self-similar segmentation and multifractality of post-stack seismic data

Layering detection is an important step in petroleum engineering. Time series of post-stack seismic data and wire-line log data belong to subsurface layering. They exhibit multifractal properties with complex patterns because of the heterogeneity and different genetic properties in the earth layers. In a multifractal configuration, any piece of a series has a distinct Hurst exponent that reflects its nature and can be used for zone detection. Time series are post-stack seismic traces and wire-line log data near the well-bores. Self-similar Autoregressive Exogenous (SAE) model is a modified method which can place self-similar post-stack seismic and wire-line log segments across layers with the same lithology. The results satisfy the capability of layering identification from seismic data by SAE model.

A data-driven approach to predict compressional and shear wave velocities in reservoir rocks

Compressional and shear wave velocities (Vp and Vs respectively) are essential reservoir parameters that can be used to delineate lithology, calculate porosity, identify reservoir fluids, evaluate fracture and calculate mechanical properties of rocks. In this study, the potential application of intelligent systems in predicting Vp and Vs of reservoir rocks is presented. To date, considerable efforts are being carried out to obtain the best set of parameters capable of predicting Vp and Vs with a high degree of accuracy. Three intelligent models namely artificial neural network (ANN), adaptive neuro-fuzzy inference system (ANFIS) and least square support vector machine (LSSVM) were used in this study. The different models were based on the available information sourced from wireline log data. Parametric studies showed that measured depth, neutron porosity, gamma-ray, and density log data are vital in predicting both Vp and Vs. In developing the models, a comprehensive dataset available from one of the oil fields in the Norwegian North Basin was used. In evaluating the different models, two different statistical parameters namely Pearson’s correlation coefficient (R2) and root mean square error (RMSE) were considered. It was found that the LSSVM model is the most accurate technique for predicting both Vp and Vs. LSSVM model predicted the Vp with R2 and RSME of 0.9706 and 0.0893 respectively. In addition, the model showed an excellent accuracy level in the prediction of Vs with R2 and RMSE of 0.9991 and 0.0457 respectively. The proposed approach, if implemented, is crucial for geoscientists, reservoir and drilling engineers working on reservoir characterization and drilling operations.

Machine Learning Approach to Model Rock Strength: Prediction and Variable Selection with Aid of Log Data

Comprehensive knowledge and analysis of in situ rock strength and geo-mechanical characteristics of rocks are crucial in hydrocarbon and mineral exploration stage to maximize wellbore performance, maintain wellbore stability, and optimize hydraulic fracturing process. Due to the high cost of laboratory-based measurements of rock mechanics properties, the log-based prediction is a viable option. Nowadays, the machine learning tools are being used for estimation of the in situ rock properties using wireline log data. This paper proposes a machine learning approach for rock strength (uniaxial compressive strength) prediction. The main objectives are to investigate the performance of data-driven predictive model in determining this vital parameter and to select features of predictor log variables in the model. The backpropagation multilayer perception (MLP) artificial neural network (ANN) with Levenberg–Marquardt training algorithm as well as the least squares support vector machine (LS-SVM) with coupled simulated annealing (CSA) optimization technique is employed to develop the dynamic data-driven models. Capturing nonlinear, high dimensional, and complex nature of real field log data, the rock strength models’ performances are evaluated using statistical criteria to ensure concerning the model reliability and accuracy. The model predictions are compared and validated against the measured values as well as the results obtained from existing log-based correlations. Both the MLP-ANN and the CSA-based LS-SVM connectionist strategies are able to predict the rock strength so that there is a very good match between the model results and corresponding measured values. The input log parameters are ranked based on their contributions in prediction performance. The acoustic travel time and gamma ray are found to have the highest relative significance in estimating rock strength. New correlations are also developed to obtain the in situ rock strength of the siliciclastic sedimentary rocks using the most important log parameters such as dynamic sonic slowness, formation electron density, and shalyness effect. The developed correlations can be used to obtain quick estimation of dynamic uniaxial compressive strength profile using wireline logging data, instead of static data from the surface measurements or laboratory data. It is expected that the proposed models and tools will enable oil and gas engineers to better predict rock strength and thus enhance wellbore performance.

Sequence stratigraphic investigation of the Early Miocene Formations in the Zagros Folded belt using wireline well logs analysis

Based on the available core sample description and wireline log analysis, sedimentary facies, microfacies and lithofacies interpretation are evaluated to determine the large scale architecture in the Early Miocene formations, from close to the contemporaneous Lower Miocene depositional margin of the Zagros basin, as represented in outcrop sections, to productive oil field within the Kirkuk embayment equivalents ca. 100 km to the south-west and west. Sequence stratigraphic methods of sedimentary log interpretation and log correlation are used to construct a model of the overall regressive and transgressive or combined transgressive to regressive character of each of the formations, thus providing an indication of the gross reservoir stacking patterns over this distance. Based on outcrop descriptions and wireline log data the Dhiban Formation is interpreted as a lowstand system tract (LST) of a third order depositional sequence, lying between the Euphrates Formation regressive parasequence set that could represent a highstand systems tract (HST) of a third order sequence and the transgressive parasequence set (third order transgressive systems tract, TST) of the lower part of the Jeribe Formation. Several higher order transgressive and regressive cycles are evident within the Dhiban Formation in the subsurface sections. The transgressive cycles within the Dhiban formation are associated with the carbonate layers. Yellowish grey, friable anhydrite is recognized in the Dhiban Formation at outcrop, and thick anhydrite of low gamma ray and high density readings was identified in the subsurface sections. This lowstand evaporite system filling the basin centre can be confirmed as the relevant model for the mixed carbonate-evaporite basin at the scale of a third order cycle.