Carbonate Reservoir Characterization Using Sequential Hybrid Seismic Rock
Physics
and Artificial Neural-Network: A Case Study of Tiaka Field
Abstract
Tiaka Field has main production in carbonate reservoir. the characterization of carbonate reservoir is usually difficult due to the complexity of matrix, pore system and even chemical reactions of fluid to the pore's matrix and also caused by the complexity of seismic wave propagation in carbonate rock. In this paper, we present a latest technology for characterizing complex reservoir of carbonate using a hybrid of neural-network and statistical analysis, this methodology can integrate huge data ranging from core information, well-log, multi atribute of seismic both post stack and pre-stack, and seismic rock physics
. All of input data are trained in sequential way by combining statistics, neural-network and seismic rock
physics
, and then followed by prediction of some reservoir parameters, i.e: facies, porosity, fracture and fluid content. Seismic rock
physics
has ability to relate seismic waves and reservoir parameters. After collecting data of seismic rock
physics
, the direct prediction of reservoir parameter can be done and combine with seismic rock
physics
, statistical analysis (PCA and Bayesian) and neural network. We present also how neural network and statistics can tie various information from ‘core or thin slice size’ in centimeter scale, the well log information in feet scale and seismic wave in meter scale. In this paper, we show the application of hybrid seismic rock
physics
, statistics and neural network in predicting various reservoir parameter, i.e: facies prediction, fracture prediction, various porosity parameter and fluid content prediction. All of reservoir parameter prediction uses sequential algorithm based on lithofacies prediction. Therefore, this method can provide more accurate in reservoir parameter prediction because it consider facies information. Accuracy testing shows more than 90% match with reservoir parameter in available wells.
AAPG Datapages/Search and Discovery Article #90217 © 2015 International Conference & Exhibition, Melbourne, Australia, September 13-16, 2015