Artificial Neural Network as a Tool for Reservoir Characterization and its Application in the Petroleum Engineering
- Arvind Kumar (Indian School Of Mines)
- Document ID
- Offshore Technology Conference
- Offshore Technology Conference, 30 April-3 May, Houston, Texas, USA
- Publication Date
- Document Type
- Conference Paper
- 2012. Offshore Technology Conference
- 6.1.5 Human Resources, Competence and Training, 5.5.2 Core Analysis, 5.6.1 Open hole/cased hole log analysis, 5.1 Reservoir Characterisation, 2.1.3 Sand/Solids Control, 5.1.5 Geologic Modeling, 5.8.7 Carbonate Reservoir, 1.6.10 Coring, Fishing, 7.6.6 Artificial Intelligence, 1.10 Drilling Equipment
- 2 in the last 30 days
- 717 since 2007
- Show more detail
- View rights & permissions
Due to the increasing number of complicated problems and time consuminganalysis, the applications of advanced information technologies like fuzzyLogic, pattern recognition, intelligent networks and artificial neural networkhave gained momentum. Among all of them, Artificial Neural Network (ANN) provesto be having an edge on other computing applications for all types of datainterpretations and analysis work related to petroleum exploration as well asexploitation. Nowadays, ANN has been widely accepted as the most powerful andefficient tool especially for reservoir characterization. Reservoircharacterization mainly includes prediction of porosity, permeability,lithology, sand thickness, and well log data. This paper focuses on theapplication of ANN in the prediction of permeability and porosity of areservoir for a given well log data and seismic data. This paper discusses manyexamples which highlight the efficiency of ANN in obtaining nonlinear systemsand models for reservoir characterization problems. Well log data and seismicdata are the parameters which have been used in the prediction of porosity andpermeability using ANN in a carbonate reservoir.
Key Words: Reservoir Characterization, Artificial Intelligence, ArtificialNeural Network, Porosity, Permeability, Fluid Saturation, NeuronArchitecture.
We can understand neural network in the form of a massive parallel-distributionof processors which are made up of simple processing units called neurons. Now,each neuron has its own natural tendency and capacity for congregatingexperiential information and making it available for us whenever needed. Thisproperty of neurons has resulted in the tremendous propagation of their uses inthe simulation and other soft computing applications. A neuron can be comparedto a human brain as of due to following characteristics:
(i) A neuron acquires information from its environment through a definedlearning process.
(ii) The acquired information or knowledge is being stored by the interneuronconnection strengths, called as synaptic weights.
Nowadays neural networks are being used in all the major data interpretationindustries including aerospace, automotive, banking, defence, electronics,financial, insurance, manufacturing, medical, robotics, oil & gasexploration and exploitation, telecommunications etc. Today, petroleumengineers have been using the advanced technologies from different disciplineslike CT-scan, MRI, Microwave, simulation software, geological modeling devices,intelligent systems, and even expert systems for getting their odd, complex andtime consuming estimation or predictive problems. Also, artificial intelligenceseems to be no longer an exception in their uses as general or in the form ofneural networks (ANN). The main motto behind using artificial neural networksin the oil and gas exploration and exploitation industry, or in any otherdiscipline, is to have a clear observation, recognition and explanation ofproblem in the way understood by the neural networks defined by us. ANN seemsto be acting like a panacea for the oil industry, and is very advantageous insolving some of the problems which has not been solved yet using theconventional computing processes used earlier.
|File Size||979 KB||Number of Pages||12|