Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “LOGS”

Search indexed NASA NTRS and DOE OSTI research on propulsion, heat transfer, battery materials and energy systems. Follow report and document links to the original sources.

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 37 records · Page 2
All sourcesNASA NTRSDOE OSTI
DOE OSTI2016

Utah FORGE: Milford Digitized Geophysical Logs from Acord 1

This submission includes digitalized versions of the following: McCulloch Geothermal Corp Acord 1-26 Cover Letter McCulloch Geothermal Corp Acord 1-26 Drilling Plan McCulloch Geothermal Corp Acord 1-26 Bond Documents Division of Water Rights Permission to Drill Drillers Log Geothermal Data (Mud) Log Compensated Densilog - Neutron Log Dual Induction Focused Log BHC Acoustilog Differential Temperature Log Dual Induction Focused Log Gamma Ray Neutron Log Temperature Log Caliper Temperature Log (Run 3) Densilog Gamma Ray Neutron Log Temperature Log (Run 4) Compensated Densilog Sample Log (Page 1 of 2) Report of Well Driller Stratigraphic Report (J.E. Welsh) Photographs and Negatives of Acord 1-26 Well Site (7) Petrography Report (M.J. Sweeney) Cuttings Samples (21 Boxes at Utah Core Research Center)

15 GEOTHERMAL ENERGY↗
DOE OSTI2019

Generation of synthetic dielectric dispersion logs in organic-rich shale formations using neural-network models

Dielectric dispersion (DD) logs acquired in subsurface geologic formations generally are composed of conductivity ([Formula: see text]) and relative permittivity ([Formula: see text]) measurements at four discrete frequencies in the range of 10 MHz to 1 GHz. Acquisition of DD logs in subsurface formations is operationally challenging, and it requires a hard-to-deploy infrastructure. We developed three supervised neural-network-based predictive methods to process conventional, easy-to-acquire subsurface logs for generating the eight DD logs acquired at four frequencies. These predictive methods will improve reservoir characterization in the absence of a DD logging tool. The predictive methods are tested in three wells intersecting organic-rich shale formations of the Permian Basin and the Bakken Shale. The first method predicts the eight dispersion logs simultaneously using a single artificial neural network (ANN) model, whereas the second method simultaneously predicts the four conductivity dispersion logs using one ANN model, followed by simultaneous prediction of four permittivity dispersion logs using a second ANN model. The third method sequentially predicts the eight dispersion logs, one at a time using eight sequential ANN models, based on a predetermined ranking of the prediction accuracy for each of the eight DD logs when simultaneously generated. Considering that the conventional and DD logs are recorded more than 10,000 ft deep in the subsurface using logging tools that are run at different times in rugose boreholes for sensing the near-wellbore geologic formation, the data used in this predictive work is prone to noise and biases that tend to adversely affect the prediction performances of the proposed methods. In terms of normalized root-mean square error (Nrms error), the prediction performances of the second predictive method are 8.5% worse and 6.2% better for the conductivity and permittivity dispersion logs, respectively, as compared with those of the first predictive method. The third method has best prediction performance for permittivity dispersion logs, which is 0.089 in terms of the Nrms error.

Geochemistry & Geophysics↗
DOE OSTI2023

Improving multiwell petrophysical interpretation from well logs via machine learning and statistical models

Well-log interpretation estimates in situ rock properties along well trajectory, such as porosity, water saturation, and permeability, to support reserve-volume estimation, production forecasts, and decision making in reservoir development. However, due to measurement errors, variability of well logs caused by multiple measurement vendors, different borehole tools, and nonuniform drilling/borehole conditions, estimations of rock properties with original well logs without proper preprocessing may not be accurate, especially in the context of multiwell estimation. Well-log normalization techniques such as two-point scaling and mean-variance normalization are commonly used to improve the robustness of multiwell rock-property estimation. However, these techniques do not consider the correlation between well logs and require subjective knowledge for their effective implementation. To reduce uncertainties and processing time associated with multiwell rock-property estimation from well logs, we develop discriminative adversarial (DA) and linear constraint models for well-log normalization and rock-property estimation. The DA neural network model developed for well-log normalization and interpretation can perform linear and nonlinear well-log normalization while considering the joint distribution of each well log and rock properties. However, the linear constraint model uses an ensemble of predictions from linear models to constrain well-log normalization and rock-property estimation. We also develop a divergence-based type well identification method to select type (training) wells for a test well based on the statistical similarity of associated well-log distributions instead of the interwell distance. We apply the DA model to perform well-log normalization and prediction of permeability for the Seminole San Andres Unit carbonate reservoir. Compared with the permeability predicted with the classical machine learning model without well-log normalization and models with two-point scaling normalization, the DA model yields the most accurate permeability prediction by decreasing the mean-squared error of permeability prediction by 20%–50%.

Geochemistry & Geophysics↗
DOE OSTI2020

Assessing impacts of selective logging on water, energy, and carbon budgets and ecosystem dynamics in Amazon forests using the Functionally Assembled Terrestrial Ecosystem Simulator

Abstract. Tropical forest degradation from logging, fire, and fragmentation not only alters carbon stocks and carbon fluxes, but also impacts physical land surface properties such as albedo and roughness length. Such impacts are poorly quantified to date due to difficulties in accessing and maintaining observational infrastructures, as well as the lack of proper modeling tools for capturing the interactions among biophysical properties, ecosystem demography, canopy structure, and biogeochemical cycling in tropical forests. As a first step to address these limitations, we implemented a selective logging module into the Functionally Assembled Terrestrial Ecosystem Simulator (FATES) by mimicking the ecological, biophysical, and biogeochemical processes following a logging event. The model can specify the timing and aerial extent of logging events, splitting the logged forest patch into disturbed and intact patches; determine the survivorship of cohorts in the disturbed patch; and modifying the biomass and necromass (total mass of coarse woody debris and litter) pools following logging. We parameterized the logging module to reproduce a selective logging experiment at the Tapajós National Forest in Brazil and benchmarked model outputs against available field measurements. Our results suggest that the model permits the coexistence of early and late successional functional types and realistically characterizes the seasonality of water and carbon fluxes and stocks, the forest structure and composition, and the ecosystem succession following disturbance. However, the current version of FATES overestimates water stress in the dry season and therefore fails to capture seasonal variation in latent and sensible heat fluxes. Moreover, we observed a bias towards low stem density and leaf area when compared to observations, suggesting that improvements are needed in both carbon allocation and establishment of trees. The effects of logging were assessed by different logging scenarios to represent reduced impact and conventional logging practices, both with high and low logging intensities. The model simulations suggest that in comparison to old-growth forests the logged forests rapidly recover water and energy fluxes in 1 to 3 years. In contrast, the recovery times for carbon stocks, forest structure, and composition are more than 30 years depending on logging practices and intensity. This study lays the foundation to simulate land use change and forest degradation in FATES, which will be an effective tool to directly represent forest management practices and regeneration in the context of Earth system models.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2018

Defining a Metric Space of Host Logs and Operational Use Cases

Host logs, in particular, Windows Event Logs, are a valuable source of information often collected by security operation centers (SOCs). The semi-structured nature of host logs inhibits automated analytics, and while manual analysis is common, the sheer volume makes manual inspection of all logs impossible. Although many powerful algorithms for analyzing time-series and sequential data exist, utilization of such algorithms for most cyber security applications is either infeasible or requires tailored, research-intensive preparations. In particular, basic mathematic and algorithmic developments for providing a generalized, meaningful similarity metric on system logs is needed to bridge the gap between many existing sequential data mining methods and this currently available but under-utilized data source. In this paper, we provide a rigorous definition of a metric product space on Windows Event Logs, providing an embedding that allows for the application of established machine learning and time-series analysis methods. We then demonstrate the utility and flexibility of this embedding with multiple use-cases on real data: (1) comparing known infected to new host log streams for attack detection and forensics, (2) collapsing similar streams of logs into semantically-meaningful groups (by user, by role), thereby reducing the quantity of data but not the content, (3) clustering logs as well as short sequences of logs to identify and visualize user behaviors and background processes over time. Overall, we provide a metric space framework for general host logs and log sequences that respects semantic similarity and facilitates a wide variety of data science analytics to these logs without data-specific preparations for each.

Verma, Miki↗
DOE OSTI2016

TH-AB-201-12: Using Machine Log-Files for Treatment Planning and Delivery QA

Purpose: To determine the segment reduction and dose resolution necessary for machine log-files to effectively replace current phantom-based patient-specific quality assurance, while minimizing computational cost. Methods: Elekta’s Log File Convertor R3.2 records linac delivery parameters (dose rate, gantry angle, leaf position) every 40ms. Five VMAT plans [4 H&N, 1 Pulsed Brain] comprised of 2 arcs each were delivered on the ArcCHECK phantom. Log-files were reconstructed in Pinnacle on the phantom geometry using 1/2/3/4° control point spacing and 2/3/4mm dose grid resolution. Reconstruction effectiveness was quantified by comparing 2%/2mm gamma passing rates of the original and log-file plans. Modulation complexity scores (MCS) were calculated for each beam to correlate reconstruction accuracy and beam modulation. Percent error in absolute dose for each plan-pair combination (log-file vs. ArcCHECK, original vs. ArcCHECK, log-file vs. original) was calculated for each arc and every diode greater than 10% of the maximum measured dose (per beam). Comparing standard deviations of the three plan-pair distributions, relative noise of the ArcCHECK and log-file systems was elucidated. Results: The original plans exhibit a mean passing rate of 95.1±1.3%. The eight more modulated H&N arcs [MCS=0.088±0.014] and two less modulated brain arcs [MCS=0.291±0.004] yielded log-file pass rates most similar to the original plan when using 1°/2mm [0.05%±1.3% lower] and 2°/3mm [0.35±0.64% higher] log-file reconstructions respectively. Log-file and original plans displayed percent diode dose errors 4.29±6.27% and 3.61±6.57% higher than measurement. Excluding the phantom eliminates diode miscalibration and setup errors; log-file dose errors were 0.72±3.06% higher than the original plans – significantly less noisy. Conclusion: For log-file reconstructed VMAT arcs, 1° control point spacing and 2mm dose resolution is recommended, however, less modulated arcs may allow less stringent reconstructions. Following the aforementioned reconstruction recommendations, the log-file technique is capable of detecting delivery errors with equivalent accuracy and less noise than ArcCHECK QA. I am funded by an Elekta Research Grant.

60 APPLIED LIFE SCIENCES↗
DOE OSTI2016

Utah FORGE: Well Acord 1-26 Logs and Data: Roosevelt Hot Spring Area

This is a compilation of logs and data from Well Acord 1-26 in the Roosevelt Hot Springs area in Utah. This well is also in the Utah FORGE study area. Logs include: mud log (45'-12645'), compensated densilog (1102'-7923', 7900'-12644'), neutron log (1102'-7923'), dual induction focused logs (1100'-7923', 7904'-11447'), BHC acoustilog (7800'-11439'), differential temperature log (380'-11448'), gamma ray neutron logs (7900'-12148', 12000'-12647'), temperature logs (7900'-12144', 7900'-12145', 7800'-12655', 7900'-12655'), and caliper log (7800'-12655'), densilog (7900'-12655'). The file is in a compressed .zip format and there is a data inventory table (Excel spreadsheet) in the root folder that is a guide to the data that is accessible in subfolders.

15 GEOTHERMAL ENERGY↗
DOE OSTI2022

Feasibility of source-free DAS logging for next-generation borehole imaging

Characterizing and monitoring geologic formations around a borehole are crucial for energy and environmental applications. However, conventional wireline sonic logging usually cannot be used in high-temperature environments nor is the tool feasible for long-term monitoring. We introduce and evaluate the feasibility of a source-free distributed-acoustic-sensing (DAS) logging method based on borehole DAS ambient noise. Our new logging method provides a next-generation borehole imaging tool. The tool is source free because it uses ever-present ambient noises as sources and does not need a borehole sonic source that cannot be easily re-inserted into a borehole after well completion for time-lapse monitoring. The receivers of our source-free DAS logging tool are fiber optic cables cemented behind casing, enabling logging in harsh, high-temperature environments, and eliminating the receiver repeatability issue of conventional wireline sonic logging for time-lapse monitoring. We analyze a borehole DAS ambient noise dataset to obtain root-mean-squares (RMS) amplitudes and use these amplitudes to infer subsurface elastic properties. We find that the ambient noise RMS amplitudes correlate well with anomalies in conventional logging data. The source-free DAS logging tool can advance our ability to characterize and monitor subsurface geologic formations in an efficient and cost-effective manner, particularly in high-temperature environments such as geothermal reservoirs. Further validation of the source-free DAS logging method using other borehole DAS ambient noise data would enable the new logging tool for wider applications.

58 GEOSCIENCES↗
DOE OSTI2015

Species characterization and responses of subcortical insects to trap-logs and ethanol in a hardwood biomass plantation: Subcortical insects in hardwood plantations

1. We characterized subcortical insect assemblages in economically important eastern cottonwood (Populus deltoides Bartr.), sycamore (Platanus occidentalis L.) and sweetgum (Liquidambar styraciflua L.) plantations in the southeastern U.S.A. Furthermore, we compared insect responses between freshly-cut plant material by placing traps directly over cut hardwood logs (trap-logs), traps baited with ethanol lures and unbaited (control) traps. 2. We captured a total of 15 506 insects representing 127 species in four families in 2011 and 2013. Approximately 9% and 62% of total species and individuals, respectively, and 23% and 79% of total Scolytinae species and individuals, respectively, were non-native to North America. 3. We captured more Scolytinae using cottonwood trap-logs compared with control traps in both years, although this was the case with sycamore and sweetgum only in 2013. More woodborers were captured using cottonwood and sweetgum trap-logs compared with control traps in both years, although only with sycamore in 2013. 4. Ethanol was an effective lure for capturing non-native Scolytinae; however, not all non-native species were captured using ethanol lures. Ambrosiophilus atratus (Eichhoff) and Hypothenemus crudiae (Panzer) were captured with both trap-logs and control traps, whereas Coccotrypes distinctus (Motschulsky) and Xyleborus glabratus Eichhoff were only captured on trap-logs. 5. Indicator species analysis revealed that certain scolytines [e.g. Cnestus mutilates (Blandford) and Xylosandrus crassiusculus (Motschulsky)] showed significant associations with trap-logs or ethanol baits in poplar or sweetgum trap-logs. In general, the species composition of subcortical insects, especially woodboring insects, was distinct among the three tree species and between those associated with trap-logs and control traps.

60 APPLIED LIFE SCIENCES↗
DOE OSTI2017

Big Data Meets HPC Log Analytics: Scalable Approach to Understanding Systems at Extreme Scale

Today's high-performance computing (HPC) systems are heavily instrumented, generating logs containing information about abnormal events, such as critical conditions, faults, errors and failures, system resource utilization, and about the resource usage of user applications. These logs, once fully analyzed and correlated, can produce detailed information about the system health, root causes of failures, and analyze an application's interactions with the system, providing valuable insights to domain scientists and system administrators. However, processing HPC logs requires a deep understanding of hardware and software components at multiple layers of the system stack. Moreover, most log data is unstructured and voluminous, making it more difficult for system users and administrators to manually inspect the data. With rapid increases in the scale and complexity of HPC systems, log data processing is becoming a big data challenge. This paper introduces a HPC log data analytics framework that is based on a distributed NoSQL database technology, which provides scalability and high availability, and the Apache Spark framework for rapid in-memory processing of the log data. The analytics framework enables the extraction of a range of information about the system so that system administrators and end users alike can obtain necessary insights for their specific needs. We describe our experience with using this framework to glean insights from the log data about system behavior from the Titan supercomputer at the Oak Ridge National Laboratory.

Park, Byung↗
DOE OSTI2015

SU-E-T-392: Evaluation of Ion Chamber/film and Log File Based QA to Detect Delivery Errors

Purpose: Ion chamber and film (ICAF) is a method used to verify patient dose prior to treatment. More recently, log file based QA has been shown as an alternative for measurement based QA. In this study, we delivered VMAT plans with and without errors to determine if ICAF and/or log file based QA was able to detect the errors. Methods: Using two VMAT patients, the original treatment plan plus 7 additional plans with delivery errors introduced were generated and delivered. The erroneous plans had gantry, collimator, MLC, gantry and collimator, collimator and MLC, MLC and gantry, and gantry, collimator, and MLC errors. The gantry and collimator errors were off by 4{sup 0} for one of the two arcs. The MLC error introduced was one in which the opening aperture didn’t move throughout the delivery of the field. For each delivery, an ICAF measurement was made as well as a dose comparison based upon log files. Passing criteria to evaluate the plans were ion chamber less and 5% and film 90% of pixels pass the 3mm/3% gamma analysis(GA). For log file analysis 90% of voxels pass the 3mm/3% 3D GA and beam parameters match what was in the plan. Results: Two original plans were delivered and passed both ICAF and log file base QA. Both ICAF and log file QA met the dosimetry criteria on 4 of the 12 erroneous cases analyzed (2 cases were not analyzed). For the log file analysis, all 12 erroneous plans alerted a mismatch in delivery versus what was planned. The 8 plans that didn’t meet criteria all had MLC errors. Conclusion: Our study demonstrates that log file based pre-treatment QA was able to detect small errors that may not be detected using an ICAF and both methods of were able to detect larger delivery errors.

60 APPLIED LIFE SCIENCES↗
DOE OSTI2016

SU-F-T-469: A Clinically Observed Discrepancy Between Image-Based and Log- Based MLC Position

Purpose: To present a clinical case which challenges the base assumption of log-file based QA, by showing that the actual position of a MLC leaf can suddenly deviate from its programmed and logged position by >1 mm as observed with real-time imaging. Methods: An EPID-based exit-fluence dosimetry system designed to prevent gross delivery errors was used in cine mode to capture portal images during treatment. Visual monitoring identified an anomalous MLC leaf pair gap not otherwise detected by the automatic position verification. The position of the erred leaf was measured on EPID images and log files were analyzed for the treatment in question, the prior day’s treatment, and for daily MLC test patterns acquired on those treatment days. Additional standard test patterns were used to quantify the leaf position. Results: Whereas the log file reported no difference between planned and recorded positions, image-based measurements showed the leaf to be 1.3±0.1 mm medial from the planned position. This offset was confirmed with the test pattern irradiations. Conclusion: It has been clinically observed that log-file derived leaf positions can differ from their actual positions by >1 mm, and therefore cannot be considered to be the actual leaf positions. This cautions the use of log-based methods for MLC or patient quality assurance without independent confirmation of log integrity. Frequent verification of MLC positions through independent means is a necessary precondition to trusting log file records. Intra-treatment EPID imaging provides a method to capture departures from MLC planned positions. Work was supported in part by Varian Medical Systems.

60 APPLIED LIFE SCIENCES↗
DOE OSTI2014

Quantification of Wellbore Leakage Risk Using Non-destructive Borehole Logging Techniques

Well integrity is important at all potential CCS locations and may play a crucial role establishing leakage risk in areas where there is a high density of existing wells that could be impacted by the storage operations including depleted petroleum fields where EOR or CCS will occur. To address a need for risk quantification methods that can be directly applied to individual wells using borehole logging tools a study was conducted using data from five wells in Wyoming. The objectives of the study were: Objective 1: Develop methods to establish the baseline flow parameters (porosity and permeability or mobility) from individual measurements of the material properties and defects in a well. Objective 2: Develop a correlation between field flow-property data and cement logs that can be used to establish the flow-properties of well materials and well features using cement mapping tools. Objective 3: Establish a method that uses the flow-property model (Objective 2) to analyze the statistical uncertainties associated with individual well leakage that can provide basis for uncertainty in risk calculations. The project objectives were met through the logging of five wells in Carbon and Natrona County Wyoming to collect data that was used to estimate individual and average well flow properties and model the results using ultrasonic data collected during the logging. Three of the five wells provided data on point and average flow properties for well annuli. Data from the other two wells were used to create models of cement permeability and test whether information collected in one well could be used to characterize another well. The results of the in-situ point measurements were confirmed by the lab measurements sidewall cores collected near the same depths Objective 1 was met using the data collected through logging, testing, and sampling. The methods were developed that can establish baseline flow parameters of wells by both point and average test methods. The methods to estimate the flow properties modeling of point pressure tests, modeling of vertical interference tests, and laboratory measurement of cased-hole sidewall cores The wells were in sufficiently good shape to allow the development of the characterization methods while still having enough defects to study differences in results as they relate to well integrity. Samples and tests analyzed from three of five wells studied in showed the cements were largely intact and had not degraded from exposure native brines. Log results taken in conjunction with the core measurements indicate that interfaces and/or problems with cement placement due to eccentering provide preferential flow paths for fluids, which can increase the effective permeability of the barrier several orders of magnitude above the permeability of intact cement. The results of the maps created using logging tools indicating that the cement condition and bond are generally good identify a need for more research to understand how logs can be used to predicteffective well permeabilities such as those measured by the VITs in this study.

42 ENGINEERING↗
DOE OSTI2023

Estimating Compressional Velocity and Bulk Density Logs in Marine Gas Hydrates Using Machine Learning

Compressional velocity (Vp) and bulk density (ρb) logs are essential for characterizing gas hydrates and near-seafloor sediments; however, it is sometimes difficult to acquire these logs due to poor borehole conditions, safety concerns, or cost-related issues. We present a machine learning approach to predict either compressional Vp or ρb logs with high accuracy and low error in near-seafloor sediments within water-saturated intervals, in intervals where hydrate fills fractures, and intervals where hydrate occupies the primary pore space. We use scientific-quality logging-while-drilling well logs, gamma ray, ρb, Vp, and resistivity to train the machine learning model to predict Vp or ρb logs. Of the six machine learning algorithms tested (multilinear regression, polynomial regression, polynomial regression with ridge regularization, K nearest neighbors, random forest, and multilayer perceptron), we find that the random forest and K nearest neighbors algorithms are best suited to predicting Vp and ρb logs based on coefficients of determination (R2) greater than 70% and mean absolute percentage errors less than 4%. Given the high accuracy and low error results for Vp and ρb prediction in both hydrate and water-saturated sediments, we argue that our model can be applied in most LWD wells to predict Vp or ρb logs in near-seafloor siliciclastic sediments on continental slopes irrespective of the presence or absence of gas hydrate.

Naim, Fawz↗
DOE OSTI2011

Real-Time System Log Monitoring/Analytics Framework

Analyzing system logs provides useful insights for identifying system/application anomalies and helps in better usage of system resources. Nevertheless, it is simply not practical to scan through the raw log messages on a regular basis for large-scale systems. First, the sheer volume of unstructured log messages affects the readability, and secondly correlating the log messages to system events is a daunting task. These factors limit large-scale system logs primarily for generating alerts on known system events, and post-mortem diagnosis for identifying previously unknown system events that impacted the systems performance. In this paper, we describe a log monitoring framework that enables prompt analysis of system events in real-time. Our web-based framework provides a summarized view of console, netwatch, consumer, and apsched logs in real- time. The logs are parsed and processed to generate views of applications, message types, individual/group of compute nodes, and in sections of the compute platform. Also from past application runs we build a statistical profile of user/application characteristics with respect to known system events, recoverable/non-recoverable error messages and resources utilized. The web-based tool is being developed for Jaguar XT5 at the Oak Ridge Leadership Computing facility.

Oral, H Sarp↗
DOE OSTI2018

Fallon FORGE: Lithology Logs and Well 21-31 Drilling Data

This submission includes lithology logs for all Fallon FORGE area wells; determined from core, cuttings, and thin section. Wells included are 84-31, 21-31, 82-36, FOH-3D, 62-36, 18-5, 88-24, 86-25, FOH-2, 14-36, 17-16, 34-33, 35A-11, 51A-20, 62-15, 72-7, 86-15, Carson_Strat_1_36-32, and several others. Lithology logs last updated 3/13/2018 with confirmation well 21-31 data, and revisited existing wells. Also included is well logging data for Fallon FORGE 21-31. Well logging data includes daily reports, well logs (drill rate, lithology, fractures, mud losses, minerals, temperature, gases, and descriptions), mud reports, drilling parameter plots, daily mud loss summaries, survey reports, progress reports, plan view maps (easting, northing), and wireline logs (caliper [with GR], triple combo [GR, caliper, SP, resistivity, array induction, density, photoelectric factor, and neutron porosity], array induction with linear correlation [GR, SP, Array Induction, caliper, conductivity], and monopole compression dipole shear [GR, SP, Caliper, sonic porosity, delta-T compressional, and delta-T shear])

15 GEOTHERMAL ENERGY↗
DOE OSTI2021

Utah FORGE: Well 16A(78)-32 Logs

This dataset contains all well logs from Utah FORGE well 16A(78)-32. This includes the mud log, Sanvean Technologies logs, and Schlumberger logs. Please see the file descriptions below for information about each log.

15 GEOTHERMAL ENERGY↗
DOE OSTI2020

Prediction of gas hydrate saturation using machine learning and optimal set of well-logs

We report resistivity and acoustic logs are widely used to estimate gas hydrate saturation in various sedimentary systems using one of the two popular methods ((1) acoustic velocity and (2) electrical resistivity), but the limitations of these two methods are often overlooked, which include (i) well-specific calibration of empirical exponents in the electrical resistivity method, (ii) assumption of known pore morphology for gas hydrates in the acoustic velocity method, and (iii) presence of unknown mineralogy and bulk modulus terms in the acoustic velocity method. NMR-density porosity-derived gas hydrate saturation based on the analysis of the transverse magnetization relaxation time (T2) is considered the most precise method, but acquisition of NMR-based logs is limited at relatively recent drilled sites; additionally, its use in conventional oil and gas reservoirs is not that common due to higher cost and operational deployment limitations associated with acquiring NMR well-logs. This study proposes a new method that predicts gas hydrate saturation (S h ) for any well using porosity, bulk density, and compressional wave (P wave) velocity well-logs with neural network (or stochastic gradient descent regression) without any well-specific calibration and/or other aforementioned shortcomings of the existing methods. The method is developed by examining the underlying dependency between S h and different combinations of well-logs, chosen from 6 routine logs, with 12 different machine learning (ML) algorithms. The accuracy of the proposed method in predicting S h is ~ 84%, which is better than the accuracy of seismic and electrical resistivity methods (≤ 75%) per the results reported by three different studies. The robustness of the method in the specific case of permafrost-associated gas hydrates is demonstrated with well-log data from two wells drilled on the Alaska North Slope.

58 GEOSCIENCES↗