The in situ reservoir status monitoring plays a critical role in natural gas hydrate resource production. Considering the complexity of the field environment, a simulation framework for monitoring gas hydrates with cross-hole electrical resistivity tomography (CHERT) was developed to monitor the hydrate distribution during hydrate formation and dissociation. The simulation study comprised both numerical and physical experiments. The optimal CHERT array was designed through a numerical experiment. The effect of applying CHERT was verified through a physical experiment (a high-resistivity medium and hydrate formation experiment). The results show that improper electrode layouts will lead to varying degrees of low amplitude and blur boundary. An optimal CHERT array of a 100-mm electrode rod spacing, 8-mm electrode ring spacing and 48 electrode rings was obtained. The inversion results obtained using this CHERT array scheme can easily distinguish the distribution of high-resistivity targets and yield satisfactory results in hydrate formation experiments. These findings guarantee data processing and interpretation for applying CHERT in gas hydrate experiments and fields.

As the most direct and important method to detect the Earth’s deep interior, scientific drilling plays a key role in solving the strategic technological issue of deep Earth exploration. This study briefly reviews the development and current situation of domestic and international continental scientific drilling and deep drilling, and analyzes the characteristics and trends of continental scientific drilling development. We summarized the Chinese Continental Scientific Drilling, its key scientific and technological issues, as well as challenges and opportunities. Accordingly, we proposed the development goals, priorities and approaches of the Chinese Continental Scientific Drilling. Continental scientific deep drilling can provide a unique pathway for investigating global concerns in Earth sciences, such as geodynamic processes, geohazards, mineral and energy resources, and environmental changes. However, its implementation depth is constrained by ultra-high temperature and ultra-high pressure in harsh borehole environments. Modern scientific advances have promoted the development of various technologies in continental scientific drilling, which provide important support for ultra-deep and extra-deep drilling. The Chinese Continental Scientific Drilling should aim for 9,000-15,000 m extra-deep boreholes, focusing on scientific issues in the fields of ultra-deep, deep-time and deep-observation. Priority development directions can include deep Earth tectonics, deep-life, deep-climate, and deep-resources. The priority detection technologies can include those for drilling, logging, and long-term observation in ultra-high temperature-pressure environments. Priority should also be given to the experimental technique development for ultra-deep matter, dynamics processes, and petrophysics. This would promote a breakthrough in the capacity and level of Earth interior detection in China.

Characterizing the electrical properties of hydrate-bearing sediments, especially resistivity, is essential for reservoir identification and saturation evaluation. The variation in electrical properties depends on the evolution of pore habits, which in turn are influenced by the hydrate growth pattern. To analyze the relationship between hydrate morphology and resistivity quantitatively, different micromorphologies of hydrates were simulated at the pore scale. This study was also conducted based on Maxwell’s equations for a constant current field. During numerical simulation, three types of hydrate occurrence patterns (grain-cementing, pore-filling and load-bearing) and five types of distribution morphologies (circle, square, square rotated by 45°, ellipse and ellipse rotated by 90°) in the pore-filling mode were considered. Moreover, the effects of porosity, the conductivity of seawater, the size of the pore-throat and other factors on resistivity are also discussed. The results show that the variation in resistivity with hydrate saturation can be broadly divided into three stages (basically no effect, slow change and rapid growth). Compared with the grain-cementing and pore-filling modes, the resistivity of the load-bearing mode was relatively high even when hydrate saturation was low. For high hydrate-saturated sediments (Sh > 0.4), the saturation exponent n in Archie equation was taken as 2.42 ± 0.2. The size of the throat is furthermore the most critical factor affecting resistivity. This work shows the potential application prospects of the fine reservoir characterization and evaluation of hydrate-bearing sediments.

The electrical resistivity tomography (ERT) method plays an essential role in researching electrical properties at the core scale, and the resistivity image inversion reconstruction technique is the key to constraining measurement accuracy. In recent years, with the improvement of computer operation capabilities and the acquisition of a large amount of geophysical data, the inversion algorithms represented by machine learning (ML) have made it possible to reconstruct ERT images automatically. However, the solution of ERT image inversion has large uncertainty due to high nonlinearity. Meanwhile, traditional ML methods are not initially developed for evaluating the uncertainty of such reconstruction results. In this article, we propose a novel Monte Carlo (MC)-Net ML scheme to quantitatively estimate the uncertainty of the ERT image reconstruction by introducing the MC dropout strategy with a multiple stochastic method to approximate Bayesian inference. As one important evaluation index, the correlation coefficient (CC) between target image and reconstructed image is used to compare the effects of six types of simulated data carried by different ML schemes. The inversion results show that MC-Net increases the lower limit of CC on the premise of ensuring average CC. Moreover, the accuracy of regression and classification results defined in this article also obtained the highest value by MC-Net. With the proposed scheme, we can further provide a quantification of the uncertainty to determine whether reconstructed images are reliable after reconstructing the ERT images, which is of great significance for actual intelligent inversion tasks.

In marine pore-filling natural gas hydrate reservoirs, hydrates are mainly produced in four occurrence modes: contact-cemented, grain-coated, matrix-supported, pore-suspended. Discrepancies in reservoir elastic-electrical properties are caused by different hydrate occurrence modes, and integrated acoustic and resistivity log data processing can effectively identify the hydrate occurrence mode. We used petrophysical models to simulate the reservoir acoustic and electrical responses, based on common log data（incl. resistivity and longitudinal velocity）, and identified hydrate occurrence modes and calculated hydrate saturation. We also quantitatively characterized the occurrence modes, via calculating the relative proportion of the four hydrate occurrence modes. Actual drilling data from three typical marine hydrate reservoirs（i.e., Shenhu area of the South China Sea, Blake Ridge in North America, Hikurangi margin in New Zealand） were used as examples to quantify the hydrate reservoir occurrence mode:（1） In the hydrate reservoir at Shenhu Site SH2, hydrates are predominantly matrix-supported, accounting for ～64% of the total;（2） In the reservoir at Blake Ridge Site 994C, the hydrates are mainly of contact-cemented（27%） and grain-coated（51%） modes;（3） In the hydrate reservoir at Hikurangi Site U1518B, the hydrates are mainly of grain-coated（32%） and matrix-supported（47%） modes. Previous experimental studies on hydrate formation and occurrence mode show that the hydrates are more likely to be stored as contact-cemented, grain-coated and matrix-supported mode, which supports the analytical reliability. The integrated acoustic and electrical log data processing here enables the quantitative evaluation of hydrate reservoir occurrence mode in marine pore-filling gas hydrate reservoirs.

现阶段适用于天然气水合物资源开采过程获得其空间分布变化的现场监测技术仍不完善.以电阻率层析成像(ERT)技术为基础,研发一套新的ERT阵列(由两组平行的垂向电极组合而成,每组阵列有24个环形电极)应用于水合物储层动态变化监测模拟实验.通过开展物理模拟实验,采用新的E RT方法对水合物生成过程进行了动态监测,并分析了动态监测方法的应用效果.通过对实验数据统计分析,92％的数据标准偏差小于5％,证实该装置获取的电学数据资料具有较高质量.通过对高阻介质模拟实验,发现温度压力条件的变化对E RT监测结果影响较小,成像结果显示饱含3.5％NaCl溶液的沉积物电阻率约为1Ω·m与阿尔奇公式计算所得电阻率基本一致,验证了其具有较高的适用性.新的ERT方法有效地监测了水合物生成过程中电阻率变化,平均电阻率随着水合物的生成从0.95Ω·m增大至1.95Ω·m并观察到"爬壁效应",证明其在水合物监测中具有良好的应用效果,并有助于实现对水合物饱和度空间分布的动态监测.研究成果为开展实验室内水合物生成和分解过程演化研究提供了方法和技术支持,也有助于研发现场天然气水合物储层动态变化监测技术和设备.

As a new energy source, gas hydrates have attracted worldwide attention, but their exploration and development face enormous challenges. Thus, it has become increasingly crucial to identify hydrate distribution accurately. Electrical resistivity tomography (ERT) can be used to detect the distribution of hydrate deposits. An ERT inversion network (ERTInvNet) based on a deep neural network (DNN) is proposed, with strong learning and memory capabilities to solve the ERT nonlinear inversion problem. 160,000 samples about hydrate distribution are generated by numerical simulation, of which 10% are used for testing. The impact of different deep learning parameters (such as loss function, activation function, and optimizer) on the performance of ERT inversion is investigated to obtain a more accurate hydrate distribution. When the Logcosh loss function is enabled in ERTInvNet, the average correlation coefficient (CC) and relative error (RE) of all samples in the test sets are 0.9511 and 0.1098. The results generated by Logcosh are better than MSE, MAE, and Huber. ERTInvNet with Selu activation function can better learn the nonlinear relationship between voltage and resistivity. Its average CC and RE of all samples in the test set are 0.9449 and 0.2301, the best choices for Relu, Selu, Leaky_Relu, and Softplus. Compared with Adadelta, Adagrad, and Aadmax, Adam has the best performance in ERTInvNet with the optimizer. Its average CC and RE of all samples in the test set are 0.9449 and 0.2301, respectively. By optimizing the critical parameters of deep learning, the accuracy of ERT in identifying hydrate distribution is improved.

Seafloor polymetallic sulfides mainly occur in mid-ocean ridges are increasingly regarded as strategic and promising metal resources for the future. One of the most difficult to detect in situ is disseminated seafloor polymetallic sulfide. It is necessary and urgent to search effective exploration and evaluation technologies. Relaxation time spectra (RTS) originated from induced polarization logging in petroleum exploration, the recognition and research on RTS of metal ore especially seafloor polymetallic sulfide is obviously insufficient. In this study, the theory of intrinsic semiconductor polarization was introduced into the traditional RTS based on membrane polarization theory and a physical property evaluation method for disseminated seafloor polymetallic sulfide rock was proposed. Synthesized and rock physics experiment data were calculated and analyzed to verify the validity of this method. The result shows that this method can effectively evaluate the induced polarization characteristics of disseminated seafloor polymetallic sulfide rocks, including the evaluation of conventional physical properties such as the relaxation time constant and chargeability, as well as the identification of possible polarization sources inside the rock. This method also can be used to evaluate the volume content and grain size of sulfide particles. Based on the rock physics experiment, the RTS of seafloor polymetallic sulfide rocks was divided into four typical types, which can indicate the induced polarization characteristics of rocks containing metal sulfides with multiple distribution. Delay duration is an important factor that affects the evaluation of induced polarization characteristics. Slightly larger delay duration can cause the induced polarization effect of disseminated metal sulfides to be greatly underestimated. This method has a directive function for the induced polarization logging interpretation of seafloor polymetallic sulfide rocks.

The induced polarization (IP) method plays an important role in the detection of seafloor polymetallic sulfide deposits. Numerical simulations based on the Poisson–Nernst–Planck equation and the Maxwell equation were performed. The effects of mineralized structures on the IP and electrical conductivity properties of seafloor sulfide-bearing rocks were investigated. The results show that total chargeability increases linearly as the volume content of disseminated metal sulfides increases when the volume content is below 20%. However, total chargeability increases nonlinearly with increasing volume content in vein and massive metal sulfides when the volume content is below 30%. The electrical resistivity of disseminated metal sulfides mainly depends on the conductivity of pore water. The electrical resistivity of vein and massive sulfides mainly depends on the volume content and the length of sulfides. Increase in the aspect ratio (0.36 to 0.93) of seafloor massive sulfides causes relaxation time constants and total chargeability to decrease. Relaxation time constants and total chargeability also decrease with increase in the tortuosity of seafloor vein sulfides from 1.0 to 1.38. This study is of great value for the electrical survey of seafloor polymetallic sulfide deposits.

The induced polarization (IP) method can play an important role in the exploration of seafloor polymetallic sulfide deposits. Compared to frequency-domain IP, time-domain IP (TDIP) requires a simpler apparatus configuration and can be more widely and economically deployed for operations in seafloor environments. To investigate the effect of the seafloor environment on the TDIP measurement and find suitable parameters to characterize metallic bodies, laboratory experiments on synthetic samples were carried out based on a special electrical experimental system. The time-domain Cole–Cole model and relaxation time distribution (RTD) method were combined to process and interpret the measured data. The results show that the volume content of metallic minerals in ore-bearing rocks can be directly quantified by the total chargeability. The sizes of metallic particles can be approximately determined by the relaxation time defined from the peak of the RTD. The RTD method was used to distinguish multiple polarizable sources, such as sulfide and basalt. TDIP surveying in the marine environment is more efficient than surveying in the terrestrial environment. A short time delay used before starting secondary voltage measurements is more suitable for a successful TDIP survey in a high-salinity environment. In addition, the chargeability is shown to be more sensitive to the variation in the volume content of metallic minerals than the direct current resistivity.