This paper describes the identification of waterflooded zones and the impact of waterflooding on reservoir properties of sandstones of the Funing Formation at the Gao 6 Fault-block of the Gaoji Oilfield, in the Subei Basin, east China. This work presents a new approach based on a back-propagation neural network using well log data to train the network, and then generating a cross-plot plate to identify waterflooded zones. A neural network was designed and trained, and the results show that the new method is better than traditional methods. For a comparative study, two representative wells at the Gao 6 Fault-block were chosen for analysis: one from a waterflooded zone, and the other from a zone without waterflooding. Results from this analysis were used to develop a better understanding of the impact of waterflooding on reservoir properties. A range of changes are shown to have taken place in the waterflooded zone, including changes in microscopic pore structure, fluids, and minerals.
28 February 2020 in Advances in Geo-Energy Research
The steady development of the oil field is facing severe challenges due to the problems of small-layer division, unclear genesis period and unclear river channel distribution in the 4-6 sand formation in the second district of Shengtuo Oilfield. Based on the processing and optimization of logging data, this paper firstly divided the isochronous strata and established the high-resolution isochronous stratigraphic framework. Using the geo-statistics method in the stratigraphic framework, the sand bodies in each small layer were divided according to the principle of equal time of fluvial facies. On this basis, the distribution pattern of the sand bodies in each stage was simulated by the magnetic random walk model. The magnetic random walk model has obtained robust simulation results, which is consistent with the anatomy of reservoir architectures by experienced geologists. The results also show that the number of channels in each small-layer is different, while the overall distribution of NE direction is reflected. At present, the model can well simulate the position of the main channel line, but it cannot reflect the variation of the river width. The method of quantitative fine description based on logging data has great potential application in fluvial reservoir, especially the magnetic random walk model that can reveal the distribution of sand body in every stage. At the same time, the model can also reflect certain randomness and facilitate the uncertainty analysis of geological factors.