A platform for research: civil engineering, architecture and urbanism
A platform for research: civil engineering, architecture and urbanism
Carbonate reservoir characterization with lithofacies clustering and porosity prediction
One of the objectives in reservoir characterization is to quantitatively or semi-quantitatively map the spatial distribution of its heterogeneity and related properties. With the availability of 3D seismic data, artificial neural networks are capable of discovering the nonlinear relationship between seismic attributes and reservoir parameters. For a target carbonate reservoir, we adopt a two-stage approach to conduct characterization. First, we use an unsupervised neural network, the self-organizing map method, to classify the reservoir lithofacies. Then we apply a supervised neural network, the back-propagation algorithm, to quantitatively predict the porosity of the carbonate reservoir. Based on porosity maps at different time levels, we interpret the target reservoir vertically related to three depositional phases corresponding to, respectively, a lowstand system tract before sea water immersion, a highstand system tract when water covers organic deposits and a transition zone for the sea level falling. The highstand system is the most prospective zone, given the organic content deposited during this stage. The transition zone is also another prospective feature in the carbonate depositional system due to local build-ups.
Carbonate reservoir characterization with lithofacies clustering and porosity prediction
One of the objectives in reservoir characterization is to quantitatively or semi-quantitatively map the spatial distribution of its heterogeneity and related properties. With the availability of 3D seismic data, artificial neural networks are capable of discovering the nonlinear relationship between seismic attributes and reservoir parameters. For a target carbonate reservoir, we adopt a two-stage approach to conduct characterization. First, we use an unsupervised neural network, the self-organizing map method, to classify the reservoir lithofacies. Then we apply a supervised neural network, the back-propagation algorithm, to quantitatively predict the porosity of the carbonate reservoir. Based on porosity maps at different time levels, we interpret the target reservoir vertically related to three depositional phases corresponding to, respectively, a lowstand system tract before sea water immersion, a highstand system tract when water covers organic deposits and a transition zone for the sea level falling. The highstand system is the most prospective zone, given the organic content deposited during this stage. The transition zone is also another prospective feature in the carbonate depositional system due to local build-ups.
Carbonate reservoir characterization with lithofacies clustering and porosity prediction
Carbonate reservoir characterization with lithofacies clustering and porosity prediction
Abdulrahman Al Moqbel (author) / Yanghua Wang (author)
Journal of Geophysics and Engineering ; 8 ; 592-598
2011-12-01
7 pages
Article (Journal)
Electronic Resource
English
Spectral mixture modelling and spectral stratigraphy in carbonate lithofacies mapping
| Online Contents | 1996
Reservoir lithofacies analysis using 3D seismic data in dissimilarity space
| IOP Institute of Physics | 2013
Reservoir lithofacies analysis using 3D seismic data in dissimilarity space
| Online Contents | 2013
Permeability Prediction Using Different Methods in Carbonate Reservoir
| Springer Verlag | 2024