Lithological Interpretation Research Articles

Lithology Prediction Using K-Nearest Neighbors (KNN) Algorithm Study Case In Upper Cibulakan Formation

Abstract Knowledge in the characterization and evaluation of hydrocarbon formations by identifying the type of lithology is an important step in oil and gas exploration. However, this method is not efficient enough because it requires a lot of time and effort. This study focuses on lithology predictions using log data input by applying one of the machine learning algorithms, namely K-Nearest Neighbors (KNN). This KNN algorithm can classify data based on K values to predict lithology. The Hyperparameter K value is determined based on the number of nearest neighbors calculated using the Euclidean distance principle. The K value in KNN needs to be optimized in order to find out the best K value to get high accuracy. The accuracy value of the model is obtained by comparing the results of the actual qualitative interpretation of lithology with the prediction results of lithology with KNN resulting precision, recall and F1 as representative of model accuracy. In the data train wells KP-24, KP-54 and KP-57 as well as test data, namely the KP-34 well, the value of k = 5 has the highest accuracy value of 89.9%. With the dominance of lithology in the predicted results, namely limestone, according to the information on the well data area, namely the Cibulakan Formation above and into the well, it reaches the Baturaja Formation. These results indicate that the lithology predicted by the KNN algorithm from the train data is already representative of the actual lithology data. Comparison of the results of qualitative interpretation with the results of machine learning predictions on the KP-34 well can predict thin or small layers and data labels in a relatively short time.

Real-Time Lithology Prediction at the Bit Using Machine Learning

Real-time drilling analysis requires knowledge of lithology at the drill bit. However, logging-while-drilling (LWD) sensors in the bottom hole assembly (BHA) are usually positioned 2–50 m (7–164 ft) above the bit (called the sensor offset), leading to a delay in real-time drilling analysis. The current industry solution to overcome this delay involves stopping drilling to perform a bottoms-up circulation for cuttings evaluation—a process that is both time-consuming and costly. To address this issue, our study evaluates three methodologies for real-time lithology prediction at the bit using drilling and petrophysical parameters. The first method employs a petrophysical approach, which involves using bulk density and neutron porosity predicted at the bit. The second method combines unsupervised and supervised machine learning (ML) for prediction. The third method employs classification algorithms on manually labeled lithology data from mud log reports, a novel approach used in this work. Our results show varying degrees of success: the bulk density versus neutron porosity cross-plot method achieved an accuracy of 58% with blind-well test data; the ML approach improved accuracy to 66%; and the Random Forest (RF) classification with manual labeling significantly increased accuracy to 86%. This comparative analysis of three different methodologies for lithology prediction has not been previously explored in the literature. While clustering and classification methods have been regarded as the most effective, our study demonstrates that they do not always yield the best result. These findings demonstrate that ML models, particularly the manual labeling approach, substantially outperform the petrophysical method. This new algorithm, designed for real-time applications, uses selected input parameters to effectively minimize problems associated with the sensor offset of LWD tools. It rapidly adapts to changes, offering a quicker and more cost-effective interpretation of lithology. This eliminates the need for time-consuming bottoms-up circulation to evaluate cuttings. Ultimately, this approach enhances drilling efficiency and significantly improves the accuracy of lithology prediction, notably in identifying interbedded geological layers.

Improved elastic full‐waveform inversion of ocean bottom node data

AbstractElastic full‐waveform inversion enables the quantitative inversion of multiple subsurface parameters, significantly enhancing the interpretation of subsurface lithology. Simultaneously, with the ongoing advancements in ocean bottom node technology, the application of elastic full‐waveform inversion to marine ocean bottom node data is receiving increasing attention. This is attributed to the capability of ocean bottom node to acquire high‐quality four‐component data. However, elastic full‐waveform inversion of ocean bottom node data typically encounters two challenges: First, the presence of low S‐wave velocity layers in the seabed leads to weak energy of converted S‐waves, resulting in significantly poorer inversion results for S‐wave velocity compared to those for P‐wave velocity; second, the cross‐talk effect of multiple parameters further exacerbates the difficulty in inverting S‐wave velocity. To effectively recover the S‐wave velocity using ocean bottom node data, we modify the S‐wave velocity gradient in conventional elastic full‐waveform inversion to alleviate the impact of cross‐talk from multiple parameters on the inversion of S‐wave velocity. Furthermore, to invert for density parameters, we adopt a two‐stage inversion strategy. In the first stage, P‐wave and S‐wave velocities are updated simultaneously with a single‐step length. Because the initial density model is far from the true one, density is updated using an empirical relationship derived from well‐log data. In the second stage, velocities and density are updated simultaneously with multi‐step length to further refine the models obtained in the first stage. The high effectiveness of the improved elastic full‐waveform inversion is validated by numerical examples.

Geological Mapping of Kaso, Tambaksari, Ciamis, West Java Province Using Lidar Images Interpretation

Abstract Remote sensing nowadays is used to quickly and effectively obtain information about an object, area or phenomenon and can be applied to large areas that are sometimes difficult to reach. Satellites with multispectral imagery are commonly used in geological mapping, but have drawbacks when used in tropical areas such as Indonesia, with dense vegetation, fast sedimentation rates and highly weathered that will cover fresh rocks making it difficult to interpret geological condition through aerial photo. Light Detection And Ranging (LiDAR) images using laser beams to produce elevation data (x,y,z) with high spatial resolution and can penetrate the surface of vegetation are expected to be a solution in areas with dense vegetation that are sometimes difficult to reach. The study was conducted by interpreting the morphometric aspects using Digital Terrain Model (DTM) data processed from LiDAR imagery with a spatial resolution of 1 m/pixel to display the topographical surface in detail. The morphometric aspects used include slope, surface roughness, aspect, hillshade, and stream pattern. Field data validation through surface mapping is carried out to correct the lithology interpretation by comparing the results of LiDAR image interpretation. Field data validation results show that interpretation using LiDAR imagery has good results (79%) through overall accuracy percentage, with the lithology units being sandstones, siltstone, breccias, and andesitic igneous rocks. The lineament interpretation of the DTM data indicates the existence of a geological structure in the form of a normal and strike-slip fault in the study area.

Identification of Geothermal Prospect Zones in Bittuang Area Based on Remote Sensing and Geochemical Data

The research location is in the Bittuang area located in Tana Toraja Regency, South Sulawesi, Indonesia. The high level of risk of geothermal drilling at the exploration stage requires integrated geological and geochemical analysis. The integrated analysis is poured into the geothermal system conceptual hypothesis model used to identify geothermal prospect zones. This study focuses on Remote Sensing and Geochemistry methods. To identify the existence of geological structures and lithological interpretation as well as surface manifestations, Remote Sensing is used. Based on the results of Remote Sensing interpretation, the research area consists of 13 rock units of Tertiary to Quaternary age with geological structures in the form of normal faults and horizontal faults that are relatively north-northwest-southeast east and southwest-northeast. The manifestations present on the surface consist of traces of fumaroles and hot springs. The hot springs consist of the chloride type in the centre of the site and the bicarbonate type in the south of the study area. The results of geothermometer calculations of geochemical data obtained a subsurface temperature of 176 -198 °C, which is categorized as a moderate to high temperature geothermal system. The geothermal conceptual model built based on the integration of geological and geochemical data shows that the geothermal system is influenced by volcanic activity and controlled by the structure of the research area. The area of the prospect is about 4 km2 with the up-flow area in the middle of the area while the outflow is in the South.

Strata-constrained Gaussian window weighted-constrained long short-term memory network for logging lithology prediction

Precisely identifying rock lithology from logging curves is critically important for reservoir characterization and exploration risk assessment. Although traditional knowledge-based lithology interpretation by a well-logging interpreter has achieved success, the interpreter-dominated lithology prediction process, in turn, could lead to a biased prediction or erroneous decision making. Deep neural network indicates the most advanced performance in various domains such as medical science, computer vision, or even geosciences. Therefore, a potential strata-constrained long short-term memory (LSTM) strategy is developed. By combining Gaussian windows to characterize the weighted stratigraphic sequence information on the target formation, rock lithology can be intelligently identified from the input logging curves. This weighted stratigraphic sequence constrain-based LSTM workflow can predict the rock lithology precisely, even for the thinner layers. Here, the F1 score and confusion matrix demonstrate that considering the rock strata sequence features, the predicted lithology by the Gaussian window weighted-constrained LSTM model and rectangular window weighted-constrained LSTM model have superior performance than those of the conventional LSTM model.

Wasserstein distance-based full waveform inversion method for density reconstruction

Reliable density information is an essential factor in lithologic interpretation and reservoir evaluation. However, density reconstruction is hampered by a lack of long-wavelength information and the crosstalk between different parameters. The powerful nonlinearity in density inversion constrains the multi-parameter full waveform inversion application. This paper introduces the Wasserstein distance into the full waveform inversion for velocity and density based on the optimal transport theory. We construct an objective function with better convexity to avoid the cycle-skipping caused by the poor initial model. Furthermore, due to the sensitivity of the quadratic Wasserstein distance to low-frequency, we add the low-wavenumber component absent in the conventional density inversion. We acquire a long-wavelength density background using optimal transport inversion and then apply it as an initial model for least-squares inversion. Numerical examples show that our optimal transport inversion effectively builds a reasonable density background model for FWI. Density inversion can be improved by incorporating optimal transport and least-squares theory, leading to a reliable quantitative description for petroleum exploration and development.

A Pliocene lacustrine system in the Nellis basin, southern Nevada, USA: implications for the Colorado River drainage system

ABSTRACT Lacustrine basins and the sediments in them provide a critical component of regional tectonic development and climate history. Each sub-basin in the landscape records the interplay between active tectonism and shifting climate. Many of the basins in southern Nevada that contain deposits of the Muddy Creek Formation were closed fluvio-lacustrine systems during the latest Miocene and into the Pliocene. They were subsequently integrated into the regional Colorado River system. Lacustrine deposits of the Muddy Creek Formation are exposed along the trace of the Las Vegas Valley shear zone in the Nellis Dunes Recreation Area (NDRA), northeast of Las Vegas, Nevada. The Muddy Creek Formation in NDRA consists of approximately 130 m of mixed carbonates and clastics, and new tephrochronology suggests an early to late Pliocene age (∼ 4.7 to ∼ 2.6 Ma) for the lowermost lacustrine deposits and an unknown, but younger, age for the uppermost spring-fed lacustrine deposit. These sediments were deposited in an arid to semiarid lake (Lake Nellis) and alluvial floodplain system based on the interpretation of lithologies and facies associations. The stratigraphic succession coarsens upward and reflects establishment of a lacustrine carbonate system overlain by a clastic succession of peripheral lake sediments of alluvial and floodplain origin. The peripheral lake deposits consist of spring-fed and fluvial wetland mudflats of brown claystone and siltstone. This brown claystone is capped by a succession of yellow and red sandstones deposited by fluvial and minor eolian processes. The second, and youngest, freshwater limestone, likely disconformably, overlies the yellow and red sandstones, and is thought to be temporally and depositionally distinct from the underlying Muddy Creek deposits. The transition from the lower lacustrine carbonates to clastic fluvial channel and flood-plain deposits indicates desiccation of Lake Nellis and possibly occurs soon after ∼ 2.7 to ∼ 2.6 Ma based on the ages of tuffs occurring in the uppermost limestone and marl beds. This is consistent with an interpreted increase in regional aridity after ∼ 2.8 Ma (Smith et al. 1993). The second limestone deposit at the top of the section represents an even younger spring-fed lacustrine deposit of unknown age. Throughout the lower carbonate section, three tuffs were identified, and geochemically correlated, using a discriminant-function analysis, to the tuff of Napa (≤ 4.70 ± 0.03 Ma), the Putah Tuff (∼ 3.3 Ma), and the lower tuffs of the Badlands (∼ 2.7 to ∼ 2.6 Ma). The timing of deposition of Lake Nellis sediments postdates the integration of the three major lake systems in this region (Lake Grand Wash, Lake Hualapai, and Lake Las Vegas) into the Colorado River drainage (5.6–4.9 Ma), and also postdates the full integration of the river to sea level (4.8–4.63 Ma) (Howard et al. 2015; Crow et al. 2021). Lake Nellis represents an isolated lake basin that was one, and maybe the last, lacustrine system to be fully integrated into the Colorado River drainage in this region.

DDViT: Advancing lithology identification on FMI image logs through a dual modal transformer model with less information drop

Lithology is an essential topic in oil and gas reservoir studies. Lithological observation lays the foundation for assessing oil and gas prospects and guides future exploration and development. Currently, the prevailing approach in lithological observation heavily depends on the manual analysis of core samples. However, such an approach is highly subjective and time-consuming. With the development of deep learning, some automated deep learning-based methods have been proposed for lithology interpretation from logging curves. However, the Fullbore Formation MicroImager (FMI) image logging, while widely used in the oil field exploration and development process, is occasionally deployed and utilized for lithology identification. In this work, we proposed a Dual-modal Drop-less-information Vision Transformer (DDViT), an FMI image lithology identification model based on transformer architecture. DDViT uses a dual-modal architecture to identify lithology using two different image modalities, namely dynamic FMI images (FMI_DYN) and static FMI images (FMI_STAT). These modalities reflect the local and overall characteristics, respectively. Furthermore, DDViT uses a less-information dropping module to drop the blank band information inherent in the FMI images to make our model more rational and stable. DDViT achieved a 90.81% lithology identification accuracy on Fengxi Well A of the western Qaidam Basin, providing a new approach to lithology identification and demonstrating the great potential of deep learning in geological images.

Geological Background Prototype Learning-Enhanced Network for Remote-Sensing-Based Engineering Geological Lithology Interpretation in Highly Vegetated Areas

Geological Background Prototype Learning-Enhanced Network for Remote-Sensing-Based Engineering Geological Lithology Interpretation in Highly Vegetated Areas

Dual-Model Collaboration Consistency Semi-Supervised Learning for Few-Shot Lithology Interpretation

Dual-Model Collaboration Consistency Semi-Supervised Learning for Few-Shot Lithology Interpretation

Seismic facies characterization using Rock Physics Templates and brittleness indices: Stybarrow field

This paper proposes a method for identifying the lithological properties of the medium based on the joint analysis of the Lamé parameters, Young's Modulus, and Poisson's Ratio. Additionally, an analysis of brittleness was proposed to identify brittle/ductile intervals and areas of potential reservoirs. The petroelastic properties were analyzed at well and seismic scales using ternary rock physics templates. The ternary templates were built from a self-consistent micromechanical model. In addition, the analysis allows for preserving the conditions of the environment subsurface in the seismic and log information. A workflow for petroelastic lithology interpretation was coupled with a workflow of brittleness modeling. The results correlate well with conventional qualitative methodologies applied in previous studies. A brittleness analysis methodology was developed and tested to identify reservoirs associated with the Lower Cretaceous in the Stybarrow field in Australia; the results highlight the high brittleness zones attenuated by hydrocarbons ( ). The proposed seismic-based methodology is an improvement to conventional analysis trends for identifying lithologies and prospective hydrocarbon zones.

Hyperspectral Facies Analysis as a Lithological Interpretation Tool for Carbonate Rocks

In a pilot study, hyperspectral image analysis was applied to four boreholes from the North Midlands Block in Ireland. The selected holes are extremely well characterised lithologically and have a detailed micropalaeontology accurately constraining stratigraphic positions. Hyperspectral facies were defined using features extracted from the hyperspectral data and compared with existing litho- and biostratigraphic logs and samples. These were able to distinguish changes in the lithologies of the cores and were useful for defining unbiased lithological contacts and for regional correlations.

Optimizing Surface Lithology Interpretation from Global Gravity Model and Landsat 8 Satellite Imagery in Semeru Mountain, Indonesia

Conventional geological mapping has high data accuracy but has ineffectiveness such as incomplete sample sweep coverage due to extreme topography, densely populated environmental conditions, and weather constraints in the field. Remote sensing is an effective alternative geological mapping solution through a combination of satellite gravity methods and satellite imagery interpretation. This study aims to apply remote sensing to obtain a more detailed distribution of Mount Semeru lithological types from regional geological maps. Integration of rock density results from satellite gravity and lithology distribution from satellite imagery results in a more detailed lithological types interpretation with more specific physical conditions of rock density. Relatively low rock density values (1.9 - 2.3 gr/cm3) distribution is at the top and foot of the mountain in the outermost areas of the study area indicating relatively thick pyroclastic material with relatively deep bedrock depth, and unconsolidated material. On the side of the body of the mountain slope, there is a relatively high rock density value (2.9 - 3.5 gr/cm3) with an indication of the formation of pyroclastic material which is relatively thinner with the presence of shallower bedrock and has consolidated material. Geological mapping optimization forms the basis for further research related to the planning and development of natural resources, as well as a more effective and efficient advanced geological mapping automation process.

Application of High-Resolution Aeromagnetic and Gamma-ray Spectrometry Surveys for Litho-Structural Mapping in Southwest China

Airborne geophysical surveys serve as an effective tool for litho-structural mapping, providing extensive and high-resolution underground information. They offer vital data support for the interpretation and analysis of lithologies and structures, complementing surface geological mapping. In the study area of the Nanpanjiang-Youjiang metallogenic belt in southeast China, we obtained high-resolution aeromagnetic and gamma-ray spectrometry data covering an area of approximately 27,000 km2, which facilitated the conduct of litho-structural mapping. The total magnetic intensity, reduction to the pole, and directional derivative maps generated from the aeromagnetic data, efficiently identified concealed rocks and faults. Additionally, the total count, potassium, thorium to potassium ratio, and ternary maps generated from the airborne gamma-ray spectrometry data demonstrated advantages in distinguishing carbonate rocks from clastic rocks. They also provided more comprehensive geological information, refining the structural strike and location interpreted by the aeromagnetic data. The litho-structural map produced in this study significantly contributes to our understanding of the structures in the Nanpanjiang area and offers valuable guidance for successful mineral exploration endeavors.

Petrophysical Properties and Identification of Electrofacies from Well Log Data of Nahr Umr Formation in Subba Oilfield, Southern Iraq

The main purpose of this study is to identify electrofacies and evaluate the petrophysical properties of the Nahr Umr Formation in four wells: A, B, C, and D in the Subba oilfield, southern Iraq. The petrophysical properties, such as shale volume, porosity, water saturation, hydrocarbon saturation, and bulk water volume, were computed and interpreted using Techlog software. According to the interpretation of density, neutron, and gamma ray sonic logs and deep resistivity logs using the IPSOM technique, four electrofacies were identified as sandstone, shaly sand, sandy shale, and shale electrofacies. The formation was divided into three parts based on well log interpretation and petrophysical analysis: the upper part, middle part, and lower part Interpretation of formation lithology and petrophysical parameters shows that the lower part consists mainly sandy shale with layers of sandstone and shale high porosity and high oil saturation whereas middle part consists mainly of sandstone, alternating with thin bedded of sandy shale and shale with high porosity and medium oil saturation. The upper part consists mainly of a sandy shale alternating with shale and shaly sand with high porosity, hydrocarbon saturation, and lower water saturation.

Penentuan Litologi Batuan di Daerah Samboja Berdasarkan Analisis Pemodelan 3D Data Cutting dan Data Logging Geofisika

The interpretation of rock lithology is inaccurate if it only uses cutting data, so it must be assisted with other data, including logging data. The purpose of this study was to determine the lithological arrangement of rocks based on cutting data and logging data and to compare the cross section results of cutting data and logging data. This study uses cutting data and logging data (gamma ray log and density log). Cutting data and logging data will be interpreted and processed into 2D and 3D sections which are corrected with topographical data including coordinates, elevation values, strike dip and total depth. The results of processing from both methods, obtained lithology composition in the form of soil, silt, sand, coal, clay, carbon and carbont clay.

The use of a multi-sensor core scanner workflow as the backbone of a digital core repository

Abstract A core repository is a physical record of a country's or commercial organization's subsurface wealth. Some of the largest core repositories hold thousands of kilometres of core material and it is a challenge to turn this physical archive into an accessible digital resource for all. Non-destructive multi-sensor core logger, hyperspectral and X-ray imaging techniques offer a unique chance to rescue valuable data trapped within core samples, improving the way that a core repository delivers data to academic or industrial end users. Here we present a case study of an archived petroleum core acquired in 1985 at the Osprey Field, UK Continental Shelf. Data from the UK National Data Repository are augmented by a multi-sensor core logger, hyperspectral and X-ray dataset that is uploaded into a cloud-based digital repository. The data were analysed using a multi-variant analysis to reclassify the original lithological interpretations, uncovering a greater proportion of clay and cemented horizons than was previously interpreted. A workflow is established to optimize the use of legacy cores and exploit the abundance of data trapped within the core repository using continuous multi-sensor core scanning and imaging data, which are stored within the virtual environment for visualization and access to all.

High-resolution seismic imaging of shallow structure at proposed IODP drilling sites, Kane oceanic core complex, Mid-Atlantic Ridge

Two 500-m-deep holes are proposed to be drilled at the Kane oceanic core complex (OCC), one aimed at a serpentinized peridotite massif and the other at a gabbroic body. Ocean drilling is vital to validate lithological interpretation derived from seismic structures. Utilizing a long marine streamer, seismic imaging can effectively delineate the dominant lithologies within the OCCs. This study applied a suite of techniques, including downward-continued multi-channel seismic data, full waveform inversion, and reverse time migration, to obtain detailed fine-scale shallow structures beneath the Kane OCC. Through the downward continuation method, refracted seismic data at near offsets were utilized, thus doubling the resolution at shallow depths. Compared to previous findings, our results greatly enhance the understanding of shallow subseafloor structures, revealing a more precise morphology of the gabbroic bodies and well-imaged very shallow low velocities caused by seawater percolation in the shallow fissure due to footwall rotation extension. We present reference drilling hole velocities and link them to the likely composition, considering potential alteration processes such as serpentinization of mantle peridotite. Our investigation suggests that the lower crustal melt fluxes of the Kane OCC represent a transitional phase from low to medium accretion, providing valuable insights for future scientific ocean drilling projects in this region.

Chapter 12. Well sequence stratigraphy interpretation methodology

Abstract This chapter describes the methodology employed to recognize depositional sequences in well data, involving the integration of wireline logs, biostratigraphy, lithostratigraphy (and lithological interpretations), seismic and sedimentology (particularly facies analysis), with an emphasis on a pragmatic approach to sequence recognition in North Sea Jurassic–lowermost Cretaceous well successions. Wireline log profiles characterizing individual depositional sequences or parts of sequences are illustrated by reference to key North Sea well sections.