Vug carbonate reservoir modeling
|School||Southwest Petroleum Institute|
|Course||Mineral prospecting and exploration|
|Keywords||Carbonate Fracture Cave Reservoir Geostatistical Stochastic simulation Reservoir modeling Geological model Artificial neural networks|
The reservoir modeling is a new technique developed in recent years. Quantitative reservoir characterization can be realized, and heterogeneities in different scales can be characterized by reservoir modeling. Based on the case of the fracture-cave dominanted carbonate reservoir in Lunnan oilfield in Tarim basin, the establishment of single-well、 2D、 3D geological model of fracture-cave dominanted carbonate reservoir has been researched in detail by many knowledge’s combination.First of all, this paper analyses the theory foundation of reservoir modeling, including basic geostatistical content such as variable space analysis, Kriging evaluation theory and stochastic simulation theory. It not only defines theory of different modeling, but also analyses their strongpoint and shortcoming to form solid theory foundation for reservoir modeling research. Secondly, this article analyses Ordovician reservoir geological characteristics of Lunnan oil field in Tarim Basin. It takes sedimentology, exploitation geology, logging geology and stochastic theory as guidance, and comprehensively utilizes information in geology, logging, well core, seism and field outcrop. Through multi-subject comprehensive modeling and appropriate reservoir geological modeling method, this article combines deterministic modeling methods with stochastic modeling methods following the step of single well model→two-dimensional model → three-dimensional model. Finally, three-dimensional geological model of fracture-cave carbonate reservoir of this area is established.Through research above, we come to the following understanding and conclusion: (1) Geological modeling of fracture-cave carbonate reservoir is a rather comprehensive work, which request information of various categories and should follow the way of integrative modeling method; (2) Combining deterministic modeling with stochastic simulation, we can greatly reduce uncertainty and increase reservoir modeling precision; (3) An improved neural network is purposed in this paper to set up porosity and permeability model of pore and cave. The application result is rather favorable. It uses imaging data and core materials to scale and validate fracture parameter interpreted by routine logging curve, and improves interpretation precision of reservoir fracture parameter. This method provides an efficientand feasible approach for logging evaluation of carbonate reservoir. (4) Based on relative theory of reservoir engineering, this article presents a way to estimate reservoir parameter of big cave, which can effectively estimate and measure reservoir parameter of cave section without logging data; (5) By utilizing "Karst facies-controlling modeling principle" advanced in this article to mark stratum gridding, gridding division of fracture-cave carbonate reservoir modeling can be solved; (6) Reservoir geological modeling establishment is subject to person’s understanding about geological conditions of reservoir, and to person’s familiarity degree of modeling theory and software, among which geological knowledge is the most important and basic factor.