Study On Logging Evaluation Method Of Water Flooded Reservoir In Changqing Oilfield

Water injection development is an important means of oil and gas development today.In the process of water flooding,various reservoir characteristics such as petrophysical properties,rock pore structure,formation water properties and remaining oil distribution will change to varying degrees,resulting in complex changes of logging response characteristics,which causes great difficulties to reservoir development and evaluation.The water-flooded position of Chang 6 tight sandstone reservoir in Bai209 block of Changqing Oilfield is extremely uneven,and the information of the waterflooded zone reflected by the logging curve is not obvious,so it is extremely difficult to identify and interpret the water-flooded zone by logging.Therefore,it is urgent to conduct a comprehensive and systematic study on the evaluation technology of waterflooded zone logging,effectively improve the coincidence rate of water-flooded formation logging interpretation,and correctly guide and adjust reservoir development plans and measures.First of all,in this article,the geological situation and development status of the Bai209 block in Changqing Oilfield were fully investigated,a large number of experimental data were collected and the changes of various parameters after water flooding were carefully analyzed.Then,according to the water-flooding experiment combined with variable multiple material balance method,the numerical simulation is carried out to comprehensively analyzed the characteristics of reservoir changes in the study target area and the logging response characteristics of different water-flooded levels,which lays the foundation for the following research.Based on the research of the geophysical characteristics of the water-flooded layer,the adjacent well identification chart method and the inflection point saturation method are proposed to identify the water-flooded layer,and the probabilistic neural network method is applied to the recognition of the water-flooded layer with the idea of man-machine judgment,and a good effect is achieved.Finally,after correcting the actual logging data,the core calibration logging method is used to establish a geological model suitable for this block.Taking the theory of variable multiple material balance method as the core and combining with the dynamic and static logging interpretation model of the reservoir,a set of processing program is compiled to calculate the parameters of water-flooded layer,which is successfully linked with the logging software platform to form the water flooded layer data processing software.The research results show that after the reservoir is flooded,the formation resistivity will appear in the form of "L","U" or "S" with the increase of water saturation according to the difference of salinity of injected water.The variable multiple material balance method can simulate the complex process of rock electricity more accurately,and the difference between the simulation results and the core experimental data is small;the adjacent well identification chart method can be used to identify different types of reservoir flooding and different types of injection water conditions;PNN probabilistic neural network method has a better discrimination effect when the number of samples is large;the inflection point saturation method requires high data accuracy to identify watered-flooded layer,and it can better identify oil layer and waterflooded layer under the condition of accurate calculation of various data;the calculation of the key parameters of the water-flooded zone by the variable multiple material balance method combined with the logging interpretation model is basically consistent with the sealing coring experiment data and the oil test production data.The log interpretation coincidence rate of the water-flooded zone reaches 91.4 %,which provides a scientific basis for the oilfield development plan and has a good reference value and guidance function.