Electrical imaging of hydraulic fracturing fluid using steel-cased wells and a deep-learning method

Li, Yinchu; Yang, Dikun

doi:10.1190/geo2020-0178.1

Cited by 17 publications

(10 citation statements)

References 59 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…First, the 3D DC problem on a 3D rectilinear mesh is solved using the equivalent resistor network (RESnet) method. [10] to obtain the steady-state current intensity values on all the edges in the mesh. Then, exploiting the orthogonality between the x-, y-and z-oriented current segments on mesh edges, we derive the simplified formula of Biot-Savart law to quickly calculate one certain component of the vector magnetic field at an observation point determined by adding up the contributions solely from the orthogonal current segments…”

Section: Modeling Methodsmentioning

confidence: 99%

3D Simulation of Semi-airborne Magnetometric Resistivity Data for the Monitoring of Water-rich Fracture Zones

Chen,

Yang

2023

J. Phys.: Conf. Ser.

Self Cite

View full text Add to dashboard Cite

Fracture zones, which are commonly saturated with water, pose significant risks during underground construction and the ongoing maintenance of infrastructure projects. Detecting such objects and monitoring their stability usually require surface geophysical methods over rough terrains that are slow and expensive, sometimes even impractical if the fracture zone is underwater. In this work, we propose using the semi-airborne magnetometric resistivity (SA-MMR) method to monitor the water content in the fracture zone rapidly. SA-MMR combines the advantages of efficiency from drone-borne geophysics and high data quality because of the measurements of a steady-state magnetic field by high-precision total-field magnetometers. We design our model based on an actual water gushing accident during the tunnel construction. The accident happened when the water from a reservoir above the route gushed into the tunnel through a weathered fault zone. In our synthetic survey, we utilize a set of electrodes that act as current sources, strategically placed along the strike of the fault; the total-field magnetometer, which is suspended 5 meters above the terrain using a drone that covers the entire survey area; two identical surveys are carried out before and after the fault’s water content increases. Our 3D forward modeling indicates that despite the low-resistivity water layer above the fracture zone, the SA-MMR data anomaly can be as large as 16 nT for a 10 A source current. The magnitude of the detected anomaly in question surpasses the noise level of contemporary airborne total-field magnetometers by multiple orders of magnitude, and our study has shown that SA-MMR has great potentials in quickly and economically mapping the variation of water content in water-rich fracture zones.

show abstract

Section: Modeling Methodsmentioning

confidence: 99%

3D Simulation of Semi-airborne Magnetometric Resistivity Data for the Monitoring of Water-rich Fracture Zones

Chen,

Yang

2023

J. Phys.: Conf. Ser.

Self Cite

View full text Add to dashboard Cite

show abstract

“…Marinenko et al (2019) proposed numerical experiments to evaluate the potential impact of metal pipes located above and below the surface on ERT's ability to identify low-resistivity permafrost thawing zones around gas production wells. Li and Yang (2021) developed a concept of 'edge conductivity' to include the steel well casings in the geo-electrical model for the application of hydraulic fracturing monitoring; the steel casings are no longer considered interferences; instead, they have been shown to be able to amplify weak anomalous signals due to injected fluid.…”

Section: Introductionmentioning

confidence: 99%

Electrical resistivity tomography through reinforced concrete floor

Yang,

Yuan

2023

Near Surface Geophysics

Self Cite

View full text Add to dashboard Cite

The electrical resistivity tomography (ERT) method is often challenged by the presence of reinforced concrete (RC) in urban and industrial environments, because the embedded metallic wire mesh can severely distort the distribution of subsurface currents. We investigate one typical scenario in real applications, in which an RC floor overlays the natural topsoil or rock. Our synthetic forward simulations show that the embedded wire mesh behaves like a local good conductor in data of small source‐receiver separations and acts like an equal‐potential object that keeps the potential from decaying at large source‐receiver separations. Routine ERT inversions that ignore the RC cannot work properly because the thin and highly conductive wire mesh may be manifested as large uninterpretable low‐resistivity anomalies in the imaging results. Two remedies are adopted to improve the ERT resolution in such cases. First, we find a top layer with high conductivity in our model to adequately represent the wire mesh; then, we initiate the inversion with the top‐layer model as the starting and reference model. This warm‐start approach overcomes the difficulty of recovering the large conductivity contrast between metallic objects and regular earth materials. Second, underground electrodes are added to the survey array, so more information from depth can be obtained to fight against the dominance of current channelling in the wire mesh. Finally, our strategies are used to invert a real ERT dataset from an indoor manufacturing plant, where RC covers the entire floor of the building and electrodes are in contact with the soil through open holes in the floor. Our simulation and field data inversion verify our findings and demonstrate the effectiveness of our solutions in improving the resolution of ERT when the survey is carried out over RC floor in urban and industrial environments.

show abstract

“…In recent years, deep‐learning (DL) inversion has drawn attention in seismic inversion as it overcomes some of the drawbacks of the traditional FWI mentioned above (Araya‐Polo et al ., 2018; Wu & Lin, 2019; Yang &, Ma, 2019; Zhang & Alkhalifah, 2019; Colombo et al ., 2020; Oh et al ., 2020; Mosser et al ., 2020; Zhang & Lin, 2020; Kazei et al ., 2021; Li & Yang, 2021; Kaur et al ., 2022). DL inversion requires training a neural network using realistic subsurface models and data.…”

Section: Introductionmentioning

confidence: 99%

Real‐time deep‐learning inversion of seismic full waveform data for CO₂ saturation and uncertainty in geological carbon storage monitoring

Alumbaugh

Lin

et al. 2022

Geophysical Prospecting

View full text Add to dashboard Cite

Deep learning inversion has recently drawn attention in geological carbon storage research due to its potential of imaging and monitoring carbon storage in real time, significantly improving efficiency and safety of carbon storage operations. We present a deep-learning full waveform inversion method that after the neural network has been trained can image CO2 saturation and its uncertainty in real time. Our deep learning inversion method is based on the U-Net architecture with the neural network trained on pairs of synthetic seismic data and CO2 saturation models. Accordingly, our training establishes a mapping relationship between seismic data and CO2 saturation models and once fully trained directly estimates CO2 saturation as a function of subsurface location. We further quantify uncertainties of CO2 saturation estimates using the Monte Carlo dropout method and a bootstrap aggregating method. For this proof-ofconcept study, the CO2 training models and data are derived from the Kimberlina 1.2 model, a hypothetical 3D geological carbon storage model that is constructed based on various geological and hydrological data from the Southern San Joaquin Basin, California. We perform deep-learning inversion experiments using noise-free and noisy training and test data sets and compare the results. Our modeling experiments show that 1) the deep-learning inversion can estimate 2D distributions of CO2 fairly well even in the presence of Gaussian random noise and 2) both CO2 saturation imaging and uncertainty quantification can be done in real time. Our results suggest that the deep-learning inversion method can serve as a robust real-time monitoring tool for geological carbon storage and/or other time varying reservoir/aquifer properties that result from injection, extraction, and/or other subsurface transport phenomena.

show abstract

Electrical imaging of hydraulic fracturing fluid using steel-cased wells and a deep-learning method

Cited by 17 publications

References 59 publications

3D Simulation of Semi-airborne Magnetometric Resistivity Data for the Monitoring of Water-rich Fracture Zones

3D Simulation of Semi-airborne Magnetometric Resistivity Data for the Monitoring of Water-rich Fracture Zones

Electrical resistivity tomography through reinforced concrete floor

Real‐time deep‐learning inversion of seismic full waveform data for CO₂ saturation and uncertainty in geological carbon storage monitoring

Contact Info

Product

Resources

About

Electrical imaging of hydraulic fracturing fluid using steel-cased wells and a deep-learning method

Cited by 17 publications

References 59 publications

3D Simulation of Semi-airborne Magnetometric Resistivity Data for the Monitoring of Water-rich Fracture Zones

3D Simulation of Semi-airborne Magnetometric Resistivity Data for the Monitoring of Water-rich Fracture Zones

Electrical resistivity tomography through reinforced concrete floor

Real‐time deep‐learning inversion of seismic full waveform data for CO2 saturation and uncertainty in geological carbon storage monitoring

Contact Info

Product

Resources

About

Real‐time deep‐learning inversion of seismic full waveform data for CO₂ saturation and uncertainty in geological carbon storage monitoring