| Author's Name: LÜ Zhou, DU Xiao, WANG Youjing, ZHANG Jie, LI Nan, WANG Nai, WANG Jun, HONG Liang, HAO Jinjin |
| Porous bioclastic limestone is one of the main reservoir types in the Middle East. It usually shows the
characteristics of strong reservoir heterogeneity, large permeability difference, significant pore type difference and complex microscopic pore structure. Diagenesis is an important reason for the differences in the reservoir properties of
porous bioclastic limestone. In particular, the differential distribution of dissolution and cementation causes changes in
pore types and microscopic pore structure, which in turn affects reservoir permeability, resulting in obvious differences in permeability under the same porosity. The study of diagenetic facies is an important method to comprehensively analyze the influence of agenesis on the reservoir. In order to quantitatively characterize the distribution characteristics of diagenetic facies and establish the corresponding diagenetic facies geological model, this paper take the bioclastic limestone of the ramp facies of the Upper retaceous Hartha Formation in Iraq H Oilfield as the research object and carry out the characterization of core diagenesis, diagenetic facies division and analysis of pore throat characteristics, diagenetic facies logging response and neural network learning, diagenetic facies space constraint analysis and geological modeling
research work. The research results show that the reservoir of Hartha Formation in the study area is developed in the top of HA, with seawater cementation, burial cementation, quasi-syngenetic dissolution, burial dissolution, compaction and dolomitization. Based on the difference in diagenesis and the difference in pore types caused by diagenesis, it can be divided into five types of diagenetic facies, namely, primary pore diagenetic facies, dissolved pore diagenetic facies, moldbody cavity diagenetic facies, ntercrystalline pore diagenetic facies, and microporous diagenetic facies. After the five diagenesis types are summarized and merged into three distinct logging diagenetic facies types, neural network learning is
carried out with conventional logging curves based on lithology calibration, which can effectively identify the diagenetic facies distribution of non-coring wells. Based on the controlling effect of the three attributes of uranium / thoriumpotassium ratio, sequence interface distance, and longitudinal wave time difference on diagenetic facies and taking them as spatial constraints, a three-dimensional constrained probability field of diagenetic facies is established, single-well identification data is integrated,and a three-dimensional diagenetic facies model is established. Uncertainty analysis improves the prediction accuracy of geological models. The geological modeling of diagenetic facies in this study provides a geological basis for the classification of reservoir rock types, the characterization of permeability and the distribution prediction of dominant reservoirs. The research process is based on core-conventional logging diagenetic facies identification and neural network learning. And the estimation of the spatial distribution probability volume based on the influencing factors of diagenesis provides a method and modeling example for the geological modeling of the diagenetic facies of the porous bioclastic limestone. |
