Abstract:
A method and apparatus for subsurface data processing includes determining a set of clusters based at least in part on measurement vectors associated with different depths or times in subsurface data, defining clusters in a subsurface data by classes associated with a state mode, reducing a quantity of the subsurface data based at least in part on the classes, and storing the reduced quantity of the subsurface data and classes with the state model in a training database for a machine learning process.
Abstract:
A liquid saturation may be identified from nuclear magnetic resonance (NMR) data having overlapping peaks indicative of two liquids by, generally, identifying a first endpoint based at least in part on the T2 NMR data for the first liquid, and identifying a second endpoint based at least in part on the T2 NMR data for the second liquid. Then, the liquid saturation is identified by relating a composition of the first liquid for an overlapping distribution region based at least in part on the first endpoint and the second endpoint. In some embodiments, the liquid saturation is identified based on an interpolation between the first endpoint and the second endpoint.
Abstract:
A method for generating a model of a formation property includes acquiring a formation property measurement. A petrophysical quantity is inverted from the formation property measurement. A model is generated based on the inverted petrophysical quantity.
Abstract:
A well-logging method for a geological formation having a borehole therein may include collecting a plurality of nuclear magnetic resonance (NMR) snapshots from the borehole indicative of changes in the geological formation and defining NMR data. The method may further include identifying a plurality of fluids within the geological formation based upon the NMR data, determining respective NMR signatures for the identified fluids based upon the NMR data, determining apparent volumes for the identified fluids based upon the NMR signatures, and determining adjusted volumes for the identified fluids based upon the apparent volumes.
Abstract:
Methods and systems for characterizing a subterranean formation using nuclear magnetic resonance (NMR) measurements are described herein. One method includes locating a downhole logging tool in a wellbore that traverses the subterranean formation, and performing NMR measurements to obtain NMR data for a region of the subterranean formation. The NMR data is processed by employing sparse Bayesian learning (SBL) to determine a multi-dimensional property distribution of the NMR data (e.g., T1-T2, D-T2, and D-T1-T2 distributions). The sparse Bayesian learning can utilize Bayesian inference that involves a prior over a vector of basis coefficients governed by a set of hyperparameters, one associated with each basis coefficient, whose most probable values are iteratively estimated from the NMR data. The sparse Bayesian learning can achieve sparsity because posterior distributions of many of such basis coefficients are sharply peaked around zero.
Abstract:
A method for generating a model of a formation property includes acquiring a formation property measurement. A petrophysical quantity is inverted from the formation property measurement. A model is generated based on the inverted petrophysical quantity.
Abstract:
A method for analyzing at least one characteristic of a geological formation may include obtaining measured data for the geological formation based upon a logging tool. Measured data may come from multiple passes or multiple depths of investigation. The method may further include generating a kernel describing a known linear mapping between the measured data and unknown data points representing at least one characteristic of the geological formation, and a redundant dictionary including a plurality of different basis functions expected to span the solution space of the unknown data points. The unknown data points representing the at least one characteristic of the geological formation may be determined from the measured data, the kernel and the redundant dictionary based upon an L1 minimization.
Abstract:
A method for determining volumetric data for fluid within a geological formation is provided. The method includes collecting first and second dataset snapshots of the geological formation based upon measurements from the borehole at respective different first and second times and generating a differential dataset based upon the first and second dataset snapshots. Multiple points are determined within the differential dataset, including a first point representing a first displaced fluid, a second point representing a second displaced fluid, and an injected fluid point that corresponds to properties of the injected fluid. A further third point is determined based on at least one other property of the displaced fluid, and a volumetric composition of the displaced fluids is determined based upon the differential dataset, the first point, and second point, and third point.
Abstract:
A method and apparatus for subsurface data processing includes determining a set of clusters based at least in part on measurement vectors associated with different depths or times in subsurface data, defining clusters in a subsurface data by classes associated with a state mode, reducing a quantity of the subsurface data based at least in part on the classes, and storing the reduced quantity of the subsurface data and classes with the state model in a training database for a machine learning process.
Abstract:
A method of interpreting petrophysical measurement data include arranging measurements of at least one physical property of formations into a matrix representing the measurements and selecting a range of number of unobserved factors or latent variables for factor analysis. Factor analysis is performed on the measurement matrix and comprises performing factorization of measurements matrix into a number of factorsand performing rotation of the factorization results. Whether the factor loadings for each factor have achieved a “simple structure” is determined and either each of the selected number of factors is associated with a physical parameter of the formations, or one is added to the number of factors and factor analysis and rotation are repeated until factor loadings of all factors have achieved “simple structure” such that the each of the number of factors is associated with a physical property of the formations.