... relationship of pixel intensity values in images to quantify rock fabric. We compute GLCMs in rectangular sliding windows having the same horizontal dimension (image width) and varying vertical dimensions to extract rock textural features at multiple scales. Additionally, we calculate rock compositional...

... of the approaching heterogeneities by quantifying their dimensions and orientation with respect to the wellbore. The workflow has also been applied to multiple ultradeep directional resistivity simulated data sets, demonstrating the ability to map arbitrary 3D heterogeneities with high resolution and precision...

... requirements of the BiLSTM model and the two-dimensional CNN model, the size and dimension of the data need to be adjusted when entering the model. The Bi-LSTM model uses two-dimensional data, the first dimension is depth, the To generate virtual reservoir parameters, it is necessary to randomly generate...

... logging symposium resonance frequency spwla-2023-0062 upstream oil & gas cement and bond evaluation log analysis frequency apparatus eccentricity spwla 64 application excitation frequency algorithm frf lwc displacement dimension through-tubing lightweight cement SPWLA 64th Annual...

.... Artificial Intelligence Reservoir Characterization free water level Symposium resistivity log reservoir model Upstream Oil & Gas porosity hydrocarbon capillary pressure water saturation FWL core data reservoir electrical log machine learning dimension fractal function coastline spwla...

... of the resulting flood in terms of the dimensions and the variation of fluid saturations away from the injection point. In this paper, we describe initials trials of the technique to evaluate the performance of an EOR agent. The technique is executed under in-situ conditions downhole in a very short time...

... the smoothness of the solution where W is the second derivative in the form of 2 2 1 L i i= for L-dimensional inversion and reduces to 2 2 2 2 2T D for 2D inversion. This form of regularization has two drawbacks: (1) It does not distinguish the smoothness on individual dimensions, such as T2 and D in 2D...

... information from multi-shell NMR data while optimizing precision. This is achieved by performing a constrained simultaneous four dimensional (4D) inversion of all data acquired at each DOI. The first three dimensions are molecular diffusion rate (D), longitudinal relaxation time (T1) and transverse relaxation...

... for the NMR relaxation response in terms of average relaxation times. In addition to standard assumptions of pore isolation, constant surface relaxivity, and fast diffusion, a permeability averaging rule is implicitly assumed at the finest scale. New developments in MRI adding a Laplace dimension allow...

... venkataramanan diffusion Artificial Intelligence dimension experiment log analysis resolution projection Hürlimann spwla 47 dg 2 0 relaxation time kernel internal gradient University well logging attenuation inverse Laplace transform Symposium gradient permeability SPWLA 47th Annual...

... relaxation time distribution in the third dimension. These 2D NMR displays can be easily extended to multi-dimensional representations, where many correlating parameters are displayed. In this paper, we report our work on T1 relaxation measurements under magic angle spinning two- dimensional nuclear magnetic...

... been restricted. Available logs contained limited radial information; in addition shallow investigation laterolog devices and deep investigation induction tools respond differently to resistivity contrast in the radial and depth dimensions. These complications are resolved by a multiarray induction...

... the resoluticmof the seismic tool, particularlgat depth. If the exploration petrophysicist can determine that there is a difference in salinitgin the two wells from the we]llogs then it impossible to showat least one dimension of a trbp or that a trap does exist. The situation may occur (Figure 4) where...

... measurements in the VHF (30-300 MHz) and the UHF (300-1000 MHz) ranges (roughly covering the frequencies of interest at present in dielectric well logging: 20-1100 MHz) require solid samples to be precision-machined to fit sample accommodators with precision dimensions. However, the core samples usually...

... ABSTRACT Physical modeling has been used to study the behavior of induction tools in a number of bed geometries which will ultimately include dipping beds. Scaling has been used to allow reasonably-sized formation structures to be housed in a laboratory. Linear dimensions have been scaled down...

... dimension log analysis porous medium well logging sonic travel-time data Symposium time-average equation statistical method derive equation Spwla Fifteenth logarithmic time-average equation rock property probability density function bulk property equation Reservoir Characterization Upstream...