Browse Prior Art Database

Applications of PCA for Characterization of Rock-Quality and Viscosity from NMR Data

IP.com Disclosure Number: IPCOM000099240D
Publication Date: 2005-Mar-14
Document File: 1 page(s) / 29K

Publishing Venue

The IP.com Prior Art Database

Related People

Pedro Romero: INVENTOR

Abstract

Abstract of the Invention: Note: The procedures explained in the following texts are basically similar: Application of PCA (Principal Components Analysis) applied first to rock-quality and second to viscosity characterization from NMR data LAPEC 2005 (not accepted) Rock quality classification based on Principal Components Analysis of T2-distribution curves from core plug samples. Pedro Romero, Baker Atlas GeoScience Each of the rock quality classes as Megaporosity, Macroporosity, Mesoporosity, Microporosity and Nanoporosity obtained by means of the Winland-Pittman equations and capillary pressure data have been shown to have corresponding T2-distribution curves, with maxima varying from higher to lower T2 values respectively. However, in order to identify and compare each rock quality class by its T2-distribution curve, it is necessary to define a characteristic set of variables, containing representative information of the whole curve; this can be done with the principal component analysis (PCA). Analyzing published T2-distribution data of water saturated core samples from El Furrial Field in Eastern-Venezuela using PCA was possible to obtain a well defined representation of the T2-distributions. The corresponding PCA-scree-plot yields that very few principal components account for the most of the variation in the T2-distribution, so that the others components can be thought as noise or their contribution as negligible. As a matter of fact, the first two or three principal components are representative enough for the classification of the data set. The Cartesian plot of the first two principal components PCA-1 vs. PCA-2 for the whole data set shows a characteristic U-shape curve, where the rock quality classes are varying continuously from one end to the other of the U-curve, following the sequence from Nanoporosity (less rock quality) to Megaporosity (best rock quality). This characteristic behavior shows that the PCA analysis can be a very useful tool in order to classify T2-distributions of core plug samples in terms of rock quality. SCA2005 (submitted) Application of Principal Component Analysis for Characterization of NMR T2 Distribution Curves from Heavy Oil Samples. Pedro Romero, Baker Atlas GeoScience The fluid typing has become a very important topic in the NMR applications. In order to classify the different T2 distributions of bulk crude oil and of oil-water saturated core plugs samples, we have applied the Principal Components and Non-Metric Multidimensional scaling analysis techniques to published data from the Melones heavy oil reservoir in East Venezuela. The analysis of the scatter diagrams and scree plots of T2-distributions of bulk oil and oil-water saturated samples show a good differentiation of the samples in terms of their Principal Components and Non-Metric MDS, which can be correlated to the viscosities of the samples at different temperatures. This methodology can be use to easily parameterize the T2 distributions by reducing the variables need to described the T2 curves to no more than three principal components.

This text was extracted from a Microsoft Word document.
At least one non-text object (such as an image or picture) has been suppressed.
This is the abbreviated version, containing approximately 53% of the total text.

Ref.:  584-40737-US

Title:  Applications for PCA for Characterization of Rock-Quality and Viscosity from NMR Data

Abstract of the Invention:

Note: The procedures explained in the following texts are basically similar: Application of PCA (Principal Components Analysis) applied first to rock-quality and second to viscosity characterization from NMR data LAPEC 2005 (not accepted) Rock quality classification based on Principal Components Analysis of T2-distribution curves from core plug samples. Pedro Romero, Baker Atlas GeoScience Each of the rock quality classes as Megaporosity, Macroporosity, Mesoporosity, Microporosity and Nanoporosity obtained by means of the Winland-Pittman equations and capillary pressure data have been shown to have corresponding T2-distribution curves, with maxima varying from higher to lower T2 values respectively. However, in order to identify and compare each rock quality class by its T2-distribution curve, it is necessary to define a characteristic set of variables, containing representative information of the whole curve; this can be done with the principal component analysis (PCA). Analyzing published T2-distribution data of water saturated core samples from El Furrial Field in Eastern-Venezuela using PCA was possible to obtain a well defined representation of the T2-distributions. The corresponding PCA-scree-plot yields that very few principal components account for the most of the variation in the T2-distribution, so that the others components can be thought as noise or their contribution as negligible. As a matter of fact, the first two or three principal components are representative enough for the classification of the data set. The Cartesian plot of the first two principal components PCA-1 vs. PCA-2 for the whole data set shows a characteristic U-shape curve, where the rock quality classes are varying continuously from one end to the other of the U-curve, following the sequence from Nanoporosity (less rock quality) to Megaporosity (best rock quality). This characteristic behavior shows that the PCA analysis can be a very useful tool in order to classify T2-distributions of core plug samples in terms of rock quality. SCA2005 (submitted) Application of Principal Component Analysis for Characterization of NMR T2 Distribution Curves from Heavy Oil Samples. Pedro Romero, Baker Atlas GeoScience The fluid typing has become a very important topic in the NMR applications. In order to classify the different T2 distributions of bulk crude oil and of oil-water saturated core plugs samples, we have applied the Principal Components and Non-Metric Multidimensional scaling analysis techniques to published data from the Melones heavy oil reservoir in

East Venezuela

. The analysis of the scatter diagrams and scree...