Deep Learning-based Electrical Imaging Monitoring Of Shale Gas Fracturing Using Steel Well Casings As Long Electrodes

Hydraulic fracturing requires a large volume of water to be injected into the earth.Monitoring the injection,flowback,retention,and absorption of the fracturing fluid over time is of great significance for the safe and sustainable large-scale production and development of shale gas.The fracturing fluid usually has an electrical conductivity higher than the surrounding formation,which allows the utilization of geophysical electrical and electromagnetic methods.However,it is difficult for surface electrical and electromagnetic methods conventionally carried out above the anomalous objects to detect signals of small-scale fracturing fluid in the deep shale layer.In addition,the downhole methods are expensive and may interrupt the normal operation of the wells.Thus,new surface electrical field monitoring methods for deep reservoirs are urgently needed.In recent years,many geophysical surveys have been using steel-cased wells as long electrodes to enhance deep signals.The international exploration geophysics community has also conducted many in-depth studies on the forward modeling of steel well casing’s effect in the oil fields.The long electrode electromagnetic method has gradually become a powerful tool for reservoir development monitoring.Compared with the conventional electromagnetic methods,the presence of steel casings as man-made infrastructure in extremely high conductivity mounts a challenge to the forward numerical simulation at the scale of geological stratum.And the requirement of fast fracturing imaging monitoring further increases the difficulty of inversion and imaging.At present,the research on long electrode electrical and electromagnetic methods with high spatial and temporal resolution is mostly focused on the forward modeling,but only very few have studied the imaging and inverse problem.Within the context of the hydraulic fracturing of unconventional oil & gas reservoirs scenario,this paper introduces the long electrode electrical method and deep learning technique into(1)high-resolution imaging monitoring of directional fracturing fluid flow and(2)well casing’s conductivity inversion.Firstly,a typical geoelectric model considering shale layer and a survey configuration using top-casing source and receivers around the wellhead are established in this paper.The 3D equivalent resistance network method(RESnet)is used to study the detectability of time-lapse differential surface electric field data for the fracturing fluid flow in the reservoir,as well as for the anomalous conductivity distribution reflecting the casing integrity.From the detectability studies,some qualita-tive knowledge and the condition of target anomaly detection are obtained.This paper then introduces the deep learning technique in the form of a fully connected network to the long electrode electric imaging for fracturing distribution and casing’s conductivity distribution with a specifically designed deep neural network algorithm as the inversion method.Compared with the conventional forward modeling methods,RESnet avoids mesh refinement for small-scale targets without loss of necessary accuracy.This feature is useful in the fast data generation for deep learning.The results of deep learning-based inversion of fracturing fluid conductivity distribution show that the deep neural network algorithm designed in this paper can effectively decode the information reflecting the fracturing fluid distribution at the horizontal well from surface data and obtain reasonable imaging results.The robustness experiments also demonstrate the stability of this algorithm against uncertain disturbances such as noise.Once trained,this deep learning-based inversion algorithm can meet the requirement of real-time imaging of fracturing fluid during hydraulic fracturing.In solving the inversion of casing’s conductivity distribution,tests on different geoelectrical models about casing integrity have shown that the trained network can recover the anomalous conductivity distribution caused by damage or corrosion on the casing from surface data.This paper is the first attempt to apply deep learning techniques to fracturing fluid imaging monitoring and casing integrity monitoring.The results show that the method of using massive numerical simulation for offline training provides a new technical route for fracturing fluid distribution and casing conductivity inversion.