scholarly journals Synthesizing Nuclear Magnetic Resonance (NMR) Outputs for Clastic Rocks Using Machine Learning Methods, Examples from North West Shelf and Perth Basin, Western Australia

Energies ◽  
2022 ◽  
Vol 15 (2) ◽  
pp. 518
Author(s):  
Reza Rezaee

A nuclear magnetic resonance (NMR) logging tool can provide important rock and fluid properties that are necessary for a reliable reservoir evaluation. Pore size distribution based on T2 relaxation time and resulting permeability are among those parameters that cannot be provided by conventional logging tools. For wells drilled before the 1990s and for many recent wells there is no NMR data available due to the tool availability and the logging cost, respectively. This study used a large database of combinable magnetic resonance (CMR) to assess the performance of several well-known machine learning (ML) methods to generate some of the NMR tool’s outputs for clastic rocks using typical well-logs as inputs. NMR tool’s outputs, such as clay bound water (CBW), irreducible pore fluid (known as bulk volume irreducible, BVI), producible fluid (known as the free fluid index, FFI), logarithmic mean of T2 relaxation time (T2LM), irreducible water saturation (Swirr), and permeability from Coates and SDR models were generated in this study. The well logs were collected from 14 wells of Western Australia (WA) within 3 offshore basins. About 80% of the data points were used for training and validation purposes and 20% of the whole data was kept as a blind set with no involvement in the training process to check the validity of the ML methods. The comparison of results shows that the Adaptive Boosting, known as AdaBoost model, has given the most impressive performance to predict CBW, FFI, permeability, T2LM, and SWirr for the blind set with R2 more than 0.9. The accuracy of the ML model for the blind dataset suggests that the approach can be used to generate NMR tool outputs with high accuracy.

2015 ◽  
Vol 3 (1) ◽  
pp. SA77-SA89 ◽  
Author(s):  
John Doveton ◽  
Lynn Watney

The T2 relaxation times recorded by nuclear magnetic resonance (NMR) logging are measures of the ratio of the internal surface area to volume of the formation pore system. Although standard porosity logs are restricted to estimating the volume, the NMR log partitions the pore space as a spectrum of pore sizes. These logs have great potential to elucidate carbonate sequences, which can have single, double, or triple porosity systems and whose pores have a wide variety of sizes and shapes. Continuous coring and NMR logging was made of the Cambro-Ordovician Arbuckle saline aquifer in a proposed CO2 injection well in southern Kansas. The large data set gave a rare opportunity to compare the core textural descriptions to NMR T2 relaxation time signatures over an extensive interval. Geochemical logs provided useful elemental information to assess the potential role of paramagnetic components that affect surface relaxivity. Principal component analysis of the T2 relaxation time subdivided the spectrum into five distinctive pore-size classes. When the T2 distribution was allocated between grainstones, packstones, and mudstones, the interparticle porosity component of the spectrum takes a bimodal form that marks a distinction between grain-supported and mud-supported texture. This discrimination was also reflected by the computed gamma-ray log, which recorded contributions from potassium and thorium and therefore assessed clay content reflected by fast relaxation times. A megaporosity class was equated with T2 relaxation times summed from 1024 to 2048 ms bins, and the volumetric curve compared favorably with variation over a range of vug sizes observed in the core. The complementary link between grain textures and pore textures was fruitful in the development of geomodels that integrates geologic core observations with petrophysical log measurements.


2007 ◽  
Vol 50 (1) ◽  
pp. 25-36
Author(s):  
G. Holló ◽  
G. Nuernberg ◽  
P. Bogner ◽  
G. Kotek ◽  
K. Nuernberg ◽  
...  

Abstract. Title of the paper: Effect of feeding on the fatty acid composition of different fatty tissues of Hungarian Grey and Holstein Friesian bulls. II. 1H-Nuclear Magnetic Resonance (NMR) investigations In this attempt the relaxation times using 1H-NMR spectroscopy from three different (subcutaneous, perinephric and internal fat) fat depots of Hungarian Grey and Holstein Friesian extensive or intensive fattened young bulls were measured. The relaxation properties were compared with the analysis of fatty acid compostion. The different diets and the sample location have a higher influence on the relaxation times than the breed. In fat samples from extensive groups the T1-relaxation time was longer, while the T2-relaxation time was significantly shorter in intensive fed groups. The T2-relaxation time, as well as the relaxation time of T21- und T22-components were the shortest in extensive fed animals, while the proportion of T21-component was the highest in kidney fat, furthermore the difference was statistics proved. The T2-relaxation time showed a close negative relationship with the ratio of saturated fatty acids (SFA). The ratio of v21 and v22 depends on chemical composition of fat samples. In fat tissues with a high SFA percentage caused a higher proportion of v21. It is suggested that differences in fatty acid compositon of fat samples caused also alteration in the relaxation time.


Geophysics ◽  
2019 ◽  
Vol 84 (2) ◽  
pp. MR73-MR84 ◽  
Author(s):  
Fatemeh Razavirad ◽  
Myriam Schmutz ◽  
Andrew Binley

We have evaluated several published models using induced polarization (IP) and nuclear magnetic resonance (NMR) measurements for the estimation of permeability of hydrocarbon reservoir samples. IP and NMR measurements were made on 30 samples (clean sands and sandstones) from a Persian Gulf hydrocarbon reservoir. We assessed the applicability of a mechanistic IP-permeability model and an empirical IP-permeability model recently proposed. The mechanistic model results in a broader range of permeability estimates than those measured for sand samples, whereas the empirical model tends to overestimate the permeability of the samples that we tested. We also evaluated an NMR permeability prediction model that is based on porosity [Formula: see text] and the mean of the log transverse relaxation time ([Formula: see text]). This model provides reasonable permeability estimations for the clean sandstones that we tested but relies on calibrated parameters. We also examined an IP-NMR permeability model, which is based on the peak of the transverse relaxation time distribution, [Formula: see text] and the formation factor. This model consistently underestimates the permeability of the samples tested. We also evaluated a new model. This model estimates the permeability using the arithmetic mean of log transverse NMR relaxation time ([Formula: see text]) and diffusion coefficient of the pore fluid. Using this model, we improved estimates of permeability for sandstones and sand samples. This permeability model may offer a practical solution for geophysically derived estimates of permeability in the field, although testing on a larger database of clean granular materials is needed.


Sign in / Sign up

Export Citation Format

Share Document