The KT-II layer in the Zananor Oilfield,Caspian Basin,Kazakhstan,contains carbonate reservoirs of various types.The complex pore structure of the reservoirs have made it difficult to identify watered-out zones with tr...The KT-II layer in the Zananor Oilfield,Caspian Basin,Kazakhstan,contains carbonate reservoirs of various types.The complex pore structure of the reservoirs have made it difficult to identify watered-out zones with traditional logging interpretation methods.This study classifies the reservoirs on the basis of core analysis and establishes an identification model for watered-out layers in the field to effectively improve the interpretation accuracy.Thin section analysis shows that there are three types of pores in the reservoirs,i.e.,the matrix pore,fracture and dissolution vug.A triple porosity model is used to calculate the porosities of the reservoirs and the results are combined with core analysis to classify the reservoirs into the fractured,matrix pore,fracture-pore as well as composite types.A classification standard is also proposed.There are differences in resistivity logging responses from the reservoirs of different types before and after watering-out.The preewatering-out resistivities are reconstructed using generalized neural network for different types of reservoirs.The watered-out layers can be effectively identified according to the difference in resistivity curves before and after watering-out.The results show that the watered-out layers identified with the method are consistent with measured data,thus serving as a reference for the evaluation of watered-out layers in the study area.展开更多
文摘The KT-II layer in the Zananor Oilfield,Caspian Basin,Kazakhstan,contains carbonate reservoirs of various types.The complex pore structure of the reservoirs have made it difficult to identify watered-out zones with traditional logging interpretation methods.This study classifies the reservoirs on the basis of core analysis and establishes an identification model for watered-out layers in the field to effectively improve the interpretation accuracy.Thin section analysis shows that there are three types of pores in the reservoirs,i.e.,the matrix pore,fracture and dissolution vug.A triple porosity model is used to calculate the porosities of the reservoirs and the results are combined with core analysis to classify the reservoirs into the fractured,matrix pore,fracture-pore as well as composite types.A classification standard is also proposed.There are differences in resistivity logging responses from the reservoirs of different types before and after watering-out.The preewatering-out resistivities are reconstructed using generalized neural network for different types of reservoirs.The watered-out layers can be effectively identified according to the difference in resistivity curves before and after watering-out.The results show that the watered-out layers identified with the method are consistent with measured data,thus serving as a reference for the evaluation of watered-out layers in the study area.
