Shale gas sweet spot prediction method based on BP neural network

Abstract

The invention provides a shale gas sweet spot prediction method based on a BP neural network, and belongs to the field of geological exploration. The method comprises: acquiring reservoir desserts, and storing various geological parameters in the desserts and pressure coefficient desserts, evaluating geological characteristic parameters related to the gas content of the shale gas, establishing a mapping relation corresponding to the gas content of the shale gas for the geological characteristic parameters through a BP neural network model, obtaining the optimal weight of each geological characteristic parameter, and obtaining a dessert distribution prediction model of the shale reservoir by quantitatively superposing each parameter planar graph. According to the prediction method, the influence of various geological parameters in reservoir desserts, storage desserts and pressure coefficient desserts on the shale gas content is systematically analyzed, the BP neural network is combined,the weights of the geological parameters of the desserts in different structural regions are defined, and the distribution of the shale gas dessert regions of the shale reservoir is quantitatively predicted.

Description

technical field [0001] The invention relates to the field of shale gas, in particular to a method for predicting sweet spots of shale gas based on BP neural network. Background technique [0002] Rich shale gas resources are expected to alleviate the energy crisis we are facing. However, due to the complexity of shale formations, its exploration and development are difficult, and the cost of shale gas drilling is much higher than that of conventional oil drilling. This requires accurate prediction and identification of "sweet spot" potential areas for future exploration when developing shale gas reservoirs. [0003] According to Jarvie et al. (2007), the reference indicators for North American shale sweet spots are mainly static parameters, including three types, shale distribution (burial depth, shale thickness and lateral distribution, fracture distribution); shale composition (organic matter type and abundance, clay minerals, brittle minerals); shale properties (thermal...

AI Technical Summary

Problems solved by technology

It can be seen that the existing North American shale evaluation methods are difficult to meet the prediction of marine shale sweet spots.

[0004] In addition, the currently existing methods for evaluation of sweet spot geological parameters and optimization of sweet spot areas all have problems such as qualitative description, simple subjective assignment and scoring, multi-parameter comprehensive evaluation only through simple superposition, and single preservation condition index evaluation, which is difficult to quantify.

This makes it difficult to quantify the geological characteristic parameters during the prediction process, and the prediction results of the superposition method are not comparable; at the same time, the weight of the impact of each geological characteristic parameter on the gas content is not considered, and the prediction plan obtained by the superposition method cannot judge the study area. A change in sweet spot area or trend within

Images