Abstract: A neural network system includes a first neural network configured to predict a mean value output and epistemic uncertainty of the output given input data, and a second neural network configured to predict total uncertainty of the output of the first neural network. The second neural network is trained to predict total uncertainty of the output of the first neural network given the input data through a training process involving minimizing a cost function that involves differences between a predicted mean value of a geophysical property of a geological formation from the first neural network and a ground-truth value of the geophysical property of the geological formation. The neural network system further includes one or more processors configured to run a software module that determines aleatoric uncertainty of the output of the first neural network based on the epistemic uncertainty of the output and the total uncertainty of the output.

Abstract: Methods are provided for determining values for a set of parameters for multiple locations in a formation by inversion of formation data obtained from a plurality of different logging tools. The inversion of the formation data is constrained by certain formation data that characterizes each particular location in the formation as obtained from at least one of the plurality of different logging tools. In one embodiment, the set of parameters for each particular location in the formation includes an apparent cementation factor mn and a formation water saturation Sw, which can be derived by inverting dielectric data that characterizes the particular location in the formation as obtained from a dielectric logging tool. The methods can also be adapted to characterize a porous medium such as reservoir rock, particular with regard to laboratory analysis of porous media samples.

Abstract: A method and a non-transitory computer readable medium for performing a calculation in a neural network comprise: accepting a data set into the neural network; performing a calculations with the neural network using the data set, wherein the calculations use a loss function and provide an aleatoric and epistemic uncertainty that is correlated to a value; and displaying results of the calculations performed.

Abstract: Processes and systems for determining asphaltene equilibrium between two or more downhole geographic locations are provided. In some embodiments, the process can include measuring one or more fluid properties of a plurality of fluid samples at varying downhole depths to generate a one or more downhole fluid analysis measurement data points; selecting an asphaltene diameter distribution based on prior knowledge; utilizing the asphaltene diameter distribution to fit a first set of one or more equation of state curves to the one or more downhole fluid analysis measurement data points to define a first model of fitted equation of state curves and to determine one or more posterior distributions of asphaltene diameters; and determining if the varying downhole depths are in an asphaltene equilibrium by determining whether the one or more posterior distributions of asphaltene diameters is consistent with that of asphaltenes in equilibrium.

Abstract: Methods and systems are provided for characterizing fluids in a subterranean formation traversed by a borehole, where NMR data is measured by a downhole NMR tool that is conveyed in the borehole that traverses the formation. The NMR data is processed to derive T1-T2 maps at different depths of the formation. The T1-T2 maps at different depths of the formation can be processed to generate a cluster map of different fluids that are present in the formation, wherein the cluster map is a two-dimensional array of grid points in the T1-T2 domain with each grid point being assigned or classified to a specific fluid. The cluster map of different fluids that are present in the formation can be used to characterize properties of the formation, such as fluid volumes for the different fluids at one or a number of depths.

Abstract: A method, computer program product, and computing system for receiving downhole logging data for a porous media. A pore size distribution index may be estimated based upon, at least in part, nuclear magnetic resonance data (NMR) from the downhole logging data of the porous media. A relative permeability and capillary pressure curve may be generated with a feasible region of solutions based upon, at least in part, the pore size distribution index.

Abstract: Embodiments herein include a system and method for modeling and interpreting an evolution of fluids in an oilfield using artificial intelligence. Embodiments may include identifying, using at least one processor, one or more reservoir fluid dynamics processes or properties and generating a model for the one or more reservoir fluid dynamics processes or properties. Embodiments may include receiving, at the model, one or more parameter values corresponding to the one or more reservoir fluid dynamics processes or properties and displaying, at a graphical user interface, one or more results, based upon, at least in part, the model and the one or more parameter values.

Abstract: Methods and systems are provided to learn and apply a mapping function from data representing concentrations of atomic elements in a geological formation (or other data corresponding thereto) to mineral component concentrations in the geological formation (and/or from mineral component concentrations to reconstructed elemental concentrations in the geological formation). The mapping function can be derived from a trained neural network (such as an autoencoder). The output of the mapping function can be used to determine estimates of one or more formation properties, such as formation matrix density, formation porosity, matrix Sigma, formation saturation, other formation property, or combinations thereof.

Abstract: Embodiments herein include a method for characterizing a rock formation sample. The method for characterizing a rock formation sample includes obtaining a plurality of data sets characterizing the rock formation sample. The method further includes training a neural network to generate a computational model. Moreover, the method additionally includes using the plurality of data sets as input to the computational model, wherein the computational model may be implemented by a processor that derives an estimate of permeability of the rock formation sample.

Abstract: Methods and systems are provided for conducting formation analysis. Data from borehole logging tools is used to conduct a petrophysical analysis of the formation in order to determine (quantify) total porosity and formation matrix permittivity for an interval of the formation. Noninvaded zone water saturation and flushed zone water saturation for the interval of the formation is determined using a saturation model of the interval. The noninvaded zone water saturation is compared to the flushed zone water saturation, and the results of the comparison are used to determine that the interval of the formation contains movable hydrocarbon, immobile hydrocarbon or movable formation water.

Abstract: Methods and systems are provided characterizing a formation traversed by a wellbore, wherein the formation includes at least a flushed zone and an uninvaded zone, which involve obtaining well log data based on plurality of different well log measurements of the formation at multiple depths in the wellbore. The well log data is used to a computational model that solves for a set of petrophysical parameters that characterize a portion of the formation corresponding to the multiple depths in the wellbore, wherein the set of petrophysical parameters include a cementation exponent, a saturation exponent, and a flushed zone water resistivity. The solved-for set of petrophysical parameters can be used to determine a value of water saturation of the uninvaded zone for the portion of the formation corresponding to the multiple depths in the wellbore.

Abstract: A method, computer program product, and computing system for receiving downhole logging data for a porous media. A pore size distribution index may be estimated based upon, at least in part, nuclear magnetic resonance data (NMR) from the downhole logging data of the porous media. A relative permeability and capillary pressure curve may be generated with a feasible region of solutions based upon, at least in part, the pore size distribution index.

Abstract: Methods and systems for determining epistemic and aleatoric uncertainties using a dual neural network architecture are described.

Abstract: A methods are provided for investigating a sample containing hydrocarbons by subjecting the sample to a nuclear magnetic resonance (NMR) sequence using NMR equipment, using the NMR equipment to detect signals from the sample in response to the NMR sequence, analyzing the signals to extract a distribution of relaxation times (or diffusions), and computing a value for a parameter of the sample as a function of at least one of the relaxation times (or diffusions), wherein the computing utilizes a correction factor that modifies the value for the parameter as a function of relaxation time for at least short relaxation times (or as a function of diffusion for at least large diffusion coefficients).

Abstract: Methods and systems are provided for characterizing fluids in a subterranean formation traversed by a borehole, where NMR data is measured by a downhole NMR tool that is conveyed in the borehole that traverses the formation. The NMR data is processed to derive T1-T2 maps at different depths of the formation. The T1-T2 maps at different depths of the formation can be processed to generate a cluster map of different fluids that are present in the formation, wherein the cluster map is a two-dimensional array of grid points in the T1-T2 domain with each grid point being assigned or classified to a specific fluid. The cluster map of different fluids that are present in the formation can be used to characterize properties of the formation, such as fluid volumes for the different fluids at one or a number of depths.

Abstract: Apparatus and method of characterizing a subterranean formation including observing a formation using nuclear magnetic resonance measurements, calculating an answer product by computing an integral transform on the indications in measurement-domain, and using answer products to estimate a property of the formation. Apparatus and a method for characterizing a subterranean formation including collecting NMR data of a formation, calculating an answer product comprising the data, wherein the calculating comprises a formula K ? ( x ) ? ? 0 ? ? k ? ( t ) ? e - t / x ? dt . and estimating a property of the formation using the answer product.

Abstract: Methods are provided for determining values for a set of parameters for multiple locations in a formation by inversion of formation data obtained from a plurality of different logging tools. The inversion of the formation data is constrained by certain formation data that characterizes each particular location in the formation as obtained from at least one of the plurality of different logging tools. In one embodiment, the set of parameters for each particular location in the formation includes an apparent cementation factor mn and a formation water saturation Sw, which can be derived by inverting dielectric data that characterizes the particular location in the formation as obtained from a dielectric logging tool. The methods can also be adapted to characterize a porous medium such as reservoir rock, particular with regard to laboratory analysis of porous media samples.

Abstract: Methods and related systems are described for estimating fluid or rock properties from NMR measurements. A modified pulse sequence is provided that can directly provide moments of relaxation-time or diffusion distributions. This pulse sequence can be adapted to the desired moment of relaxation-time or diffusion coefficient. The data from this pulse sequence provides direct estimates of fluid properties such as average chain length and viscosity of a hydrocarbon. In comparison to the uniformly-spaced pulse sequence, these pulse sequences are faster and have a lower error bar in computing the fluid properties.

Abstract: Systems and methods are described herein for investigating a downhole formation using a nuclear magnetic resonance (NMR) tool. The tool includes multiple detectors spaced along an axis parallel to a length of the wellbore. While the tool is moving through the wellbore, NMR pulse sequences are applied to the formation and resulting NMR signals are detected by the detectors. The data obtained over a period of time by the multiple detectors are inverted together to obtain an indication of a parameter of the formation at multiple locations along the length of the wellbore.

Abstract: Methods are provided for determining values of a pore cementation exponent m and/or a saturation exponent n for locations in a formation having similar petrophysical properties. Formation porosity, formation water saturation, and an apparent cementation exponent mn are obtained for the locations and their values are utilized to find the exponents. In one embodiment, the apparent cementation exponent and the formation water saturation are obtained from a dielectric logging tool.