Geophysical Prospecting - 9 Advanced Techniques, Methods and Applications for an Integrated Approach to the Geophysical Prospecting, 2023

Controlled source audio‐frequency magnetotelluric is frequently used in association with other geophysical methods, especially in complex geological areas to highlight geological structures such as water reservoir rock. Although it gives satisfactory results, combining several methods requires time and expense. In addition, despite this combination, several drilling locations proposed after geophysical investigations were inaccurate, resulting in many unsuccessful drillings. The latter occurs due to the difficulty to emphasize the fracture zones properly. To work around this problem, we proposed a novel approach called pseudostratigraphic to reduce the repercussion of unsuccessful drilled boreholes and to demarcate the water reservoir rock. The technique consists to discretize the resistivity of the inverted OCCAM2D model based on the true layer resistivities collected from the borehole log data (observed layers). The discrete resistivity model is known as the new resistivity model. It is used to generate the pseudostratigraphic log at each station by pseudo‐demarcating the thicknesses of the observed layers with a low margin of error. Moreover, the combination of multiple new resistivity models from different survey lines creates a three‐dimensional pseudostratigraphic map useful to emphasize the water reservoir rock. The pseudostratigraphic implementation is carried out in the Xingning area as a real‐world case study. The results show that the intersection of the main fault (F1) and the conductive zones (≤100 Ω m) indicates the potential water reservoir rock. Its thickness is estimated around 150–600 with an error equal to ± 7°m. Based on the three‐dimensional pseudostratigraphic map, the water found in the fracture zone located under the reservoir rock is considered much hotter due to the intense geothermal activity along F1, thereby making it a better place for hot water exploitation. Finally, the pseudostratigraphic technique could be an innovative and cheap strategy to find groundwater reservoirs in complex geological areas.

We present an automated approach for inferring surface geological structures from geophysical survey data. Our method employs machine learning, using mapped geological structures as labels and filtered geophysical surveys as reference maps. We compared the performance of the eight main machine learning algorithms and their 32 branches. Applied to the Geological Survey of Victoria's database for the Bendigo Zone, following an appropriate choice of geological features, the 3‐class classification model using subspace K‐nearest neighbour methods achieves a stable and validated 92% accuracy in around 1 min. The fault‐only classification model achieves a stable and validated 97% accuracy in around 6 min. This shows that geological structural features on the surface may be inferred from between one and three of the following geophysical data types: gravity, airborne total magnetic intensity and first vertical derivative of total magnetic intensity. It shows the prospect of machine learning in geological research and suggests that geophysical data combined with machine learning may be useful and efficient in determining the existence of geological structural features.

The mine roadway built in the same thick rock stratum will not change the lithologic medium in front of the roadway under the influence of the small structure of the mine, which makes the mine radar face many difficulties in structural interpretation. In the coal measure strata in southwest China, the coal seam roof is the weak aquifer of the Kayitou (T1k) group. The mine structure is the excellent drainage boundary of an aquifer, resulting in the water content of the structural belt being different from that of the surrounding rock. This paper proposes a method to identify hidden structures in mine roadways using the autoregressive and moving average power spectrum energy enveloping medium water content inversion technique. Based on the actual water content of strata in the study area, 48 roadway models are constructed. The finite‐difference time‐domain algorithm was used to obtain the forward data and calculate the volumetric water content of the model. The structure of the roadway model is accurately identified by the water content difference. In the numerical simulation study, the segmentation frequency of high‐frequency and low‐frequency envelope of the autoregressive and moving average spectrum of 100‐MHz radar antenna is determined to be 155 MHz, and the value range of water content calculation regulation parameter k is determined. The high‐ and low‐frequency envelope inversion characteristics of the autoregressive and moving average spectrum of different tunnel media models are summarized and analysed. To solve the calculation problem of energy envelope selection of autoregressive and moving average power spectrum of engineering data, the k‐nearest neighbour machine learning algorithm was used to learn and classify the power spectrum features of different tunnel models. The detection application is analysed through the whole‐rock medium and coal and rock medium roadway in the study area. The results show that the method can effectively identify the location and strike characteristics of the hidden structures in front of the roadway.

This paper deals with ground‐penetrating radar prospecting and exploits a semi‐heuristic strategy to account for inhomogeneous background media, empty cavities or topography of the surface. We assume here that no more than a commercial processing software is available. Customarily, commercial codes assume a homogeneous soil and a flat interface in order to achieve the focusing of the data. Therefore, this is also the model exploited here, whereas the data are referred to an inhomogeneous soil or to a non‐flat interface. The proposed strategy exploits the principle that ‘the data speak’, even if with some ambiguity and some reticence, and they can reveal or at least suggest important features of the underground scenario. On this basis, heuristically, we exploit features derived from the data themselves as a posteriori information, improving the available focusing and time–depth conversion even having at disposal a basic model of the background scenario.

We carried out an integrated interpretation of available geophysical data (gravity and electrical data) in Sidi Bou Aouane–Khadhkhadha Pb–Zn province, located in a Neogene basin south of the Alpine thrust‐belt front in northern Tunisia. The interpretive approach of gravity data was done based on the Fourier transformation and spectral analysis to extract the residual component. Furthermore, edge enhancement techniques (tilt angle, total horizontal derivative) were applied to image the underlying lineaments in the study area. The computed gravity maps reveal multiple NE–SW hidden faults considered potential mineralization targets, referring to geological information. Detrital formations, known as containers of disseminated Pb–Zn mineralization, are expressed by low gravity responses. Resistivity and chargeability 3D inversion was conducted concurrently through on an iterative approach based on the conventional least‐squares algorithm and “incomplete Gauss‐Newton” as an optimization method. The initial model was defined based on drill‐hole data and geological knowledge. Inverted resistivity confirms the basin architecture expressed by gravity data. The combined interpretation of inverted resistivity and chargeability correlated to borehole and outcrops allows the definition of different electrical ranges associated to different lithologies and mineralization type and contents. A new potential target, expressing the same electrical signature of mineralization, is evidenced northward Khadhkhadha old mine. In addition, a potential copper concentration area was proven based on its electrical responses south Sidi Bou Aouane mine. 2.5D gravity modelling supports the evidenced interpreted target. The revealed results provided evidence for future interpretations of further geological structures, along with evaluations of mineral resources.

Pore aspect ratio, together with porosity, is a structural parameter that represents the geometric property of rock reservoirs. We have adopted the theory of ellipsoid modelling in material mechanics to derive the dry rock modulus. Based on this derivation, the Gassmann equation, which is a constitutional equation for a fully saturated rock model, can be linearized in terms of the pore structure parameters and the elastic parameters. We have established a relationship between the seismic reflection coefficient and the pore structure parameters, where the equivalent pore aspect ratio and porosity are two key parameters. Based on this relationship, the equivalent pore aspect ratio can be inverted directly from seismic reflection data instead of being converted from the intermediate parameters of conventional seismic inversion. Therefore, seismic inversion is a simultaneous inversion in which seven parameters are inverted: the elastic moduli (matrix bulk modulus and shear modulus, fluid bulk modulus), the densities (matrix and fluid densities) and the pore structure parameters (the equivalent pore aspect ratio and porosity). We applied this direct inversion scheme to a carbonate reservoir to predict the fracture development zone with low porosity and low aspect ratio.

Although lithofacies routinely is featured by distinct logging responses from each other, many types of lithofacies in practical cases show similar measuring characteristics on logs, and then to achieve a desirable solution from logging‐based lithofacies prediction actually is challengeable. Since the mathematical essence of lithofacies prediction can be explained as an issue of pattern recognition, a light gradient boosting machine, a state‐of‐the‐art ensemble learning, specifically developed to address supervised classification, could be a potential solver. Nonetheless, due to an incompatibility of inherent exclusive feature bundling algorithm for logs and usage of a great deal of hyper‐parameters, a raw light gradient boosting machine might not be suitable or functional to predict lithofacies. Thus, continuous restricted Boltzmann machine and Bayesian optimization, respectively, employed to process original logging data and optimize hyper‐parameters, are adopted as technical assistants for the light gradient boosting machine, and accordingly a new ensemble learning‐based predictor called continuous restricted Boltzmann machine – Bayesian optimization – the light gradient boosting machine, is proposed for lithofacies. To validate classifying capability and robust nature of new predictor, three experiments are designed purposefully based on the application of a dataset collected from the wells located within pre‐salt lacustrine carbonate reservoirs of the Santos Basin. Simultaneously, to highlight validating effect, other three sophisticated classifiers are introduced as competitors in all experiments, including supper vector machine, random forest and extreme gradient boosting. According to the analysis and comparison of all experimental results, overall four points are verified: (1) the integration of continuous restricted Boltzmann machine and Bayesian optimization indeed is beneficial for the light gradient boosting machine in data preprocess and modelling stage; (2) the light gradient boosting machine cored predictor shows more capability of producing reliable predicted lithofacies information compared to other three competitors; (3) training more learning samples is a simple and effective approach to enhance computing capability of any validated predictor, and under this circumstance the light gradient boosting machine cored predictor still performs relatively better; (4) the light gradient boosting machine cored predictor is characterized with a better robustness as its predicting outcomes, even derived from a sparse‐condition dataset, are also proved more qualified. Consequently, the new proposed predictor is demonstrated as a high‐efficient and robust solver for the prediction of lithofacies and deserves a widespread application in the geological, geophysical and petrophysical research.

The exploration of kimberlite pipes is difficult. However, information about their deep structure is necessary for tasks such as prospecting, exploration and development of kimberlite pipes. Therefore, it is necessary to increase the efficiency of deep structure research. In addition, pipes in different territories differ in their properties. The latter requires the study of the peculiarities of the manifestation of pipes in various areas. The Verkhnetovskaya pipe is the only kimberlite object currently known in the Chernoozerskaya area of the Arkhangelsk diamondiferous province. This indicates the need for a comprehensive study of this pipe. The main reason it is difficult to study the pipes of the Arkhangelsk diamondiferous province is the large overburden thickness. In order to increase the efficiency of prospecting, we used a set of methods: microseismic sounding method, passive seismic interferometry, H/V method, gamma spectrometry and emanation mapping. The analysis showed that a new field measurement technique was tested in this research, resulting in more stable data. The main improvement in field measurements is the determination of the optimal accumulation time of microseism when implementing the microseismic sounding method. In addition, the location of the stations in the implementation of passive interferometry made it possible to minimize the influence of the azimuthal distribution of ambient noise sources. The studies made it possible to construct the geophysical image of the investigated pipe and the surrounding medium. As a result, it was shown that the northwestern side has a vertical structure, while according to the drilling data, the side has a uniform slope. In particular, the presence of previously unknown lateral channels was shown. The proposed methodology made it possible to obtain important information with minimal time and technical costs, confirming the applicability of the proposed methodology in the Chernoozerskaya area.

Litho‐prediction is a non‐unique process that requires more than integrated techniques to reinforce the final results and harmonize hydrocarbon probability. Consequently, integrated procedures have been started with log data of the Mishrif formation in different site locations and involve petrophysics and rock physics analyses altogether to evaluate the general situation of the formation. In the second stage, we performed seismic inversion for 3D seismic data in the Kumaite and Dhafriyah oil fields. A brand‐new technique called pseudo–post‐stack simultaneous inversion has been used to calculate all the elastic engineering properties of the seismic cube. This method is performed by utilizing a real genetic inversion model that would be used to calibrate a low‐frequency model of simultaneous inversion to enhance the resolution and event isolation of the final inverted cube to identify reservoir characterization. The final output then is used to perform quantitative analysis and determine the productive layers within the area of interest. Density–velocity analysis also plays an extraordinary role in establishing the main relationship between different elastic parameters. Finally, by using the local equation and global density–velocity equations, different velocities of carbonate succession have been calculated.

In oil and gas exploration, it is vital to acquire information about the bottom hole conditions. This is done in the field using wireline logging. The sonic log is one of the most prolific logs as it assists in porosity determination, cement evaluation and identification of lithology and gas‐bearing intervals. However, sonic logging tools are not always a part of the wireline logging arrangement. Still, there are sections where the logging data are missing, and in some cases, these are dependent upon old tools. The tool is incapable of recording shear wave transit times. This study explores the possibility of substituting the sonic log using machine learning and artificial intelligence techniques. These techniques can also predict the sonic log data in sections where these are missing or unreliable. Artificial neural networks, decision tree regression, random forest regression, support vector regression and extreme gradient boosting are the most popular tools available at our disposal for making these estimations. This study has compared these different techniques for their effectiveness and accuracy in making sonic transit time predictions based on other available well logs. The obtained results suggest that despite all the attention on artificial neural networks, eXtreme gradient boosting and the random forest regression outperform it for the given purpose. In the case of missing shear transit time data, random forest regression made predictions with a root mean squared error of 1.03 × 10^–4 and a root mean squared error of 0.97 while eXtreme gradient boosting regression did so with a root mean squared error of 1.36 × 10^–4 and the same regression coefficient (0.97). When no sonic data were available, random forest regression estimated shear transit time with a root mean squared error of 6.41×10^–4 and a regression coefficient of 0.95, and compressional transit time with a root mean squared error of 9.06×10^–5 and a root mean squared error of 0.94. The root mean squared error of shear transit time and compressional transit time predictions made using eXtreme gradient boosting were found to be the same as those of random forest regression. The root mean squared error, however, was observed to be slightly less for the compressional transit time predictions and somewhat more for the shear transit time predictions. As data analysis, in general, is a better method for estimation than the use of empirical correlations, these machine learning‐based predictions can serve as powerful tools in the oil and gas exploration industry.