Search

Objective Borehole Televiewer (BHTV) imaging serves as an effective tool for analyzing deep rock formations in geological engineering, offering crucial insights into surface discontinuity extensions and the distribution of buried discontinuities, especially within complex geological environments associated with hydraulic engineering projects. Recognizing the current issues of subjectivity and low efficiency, prevalent in the manual calculation of rock discontinuity morphological characteristics within rock masses, we propose an intelligent approach for segmenting geometric data from digital borehole images. Leveraging the segmentation outcomes, an image thinning technique is employed to facilitate precise quantitative analysis of borehole data. Methods In this research, we employ deep learning models to intelligently identify fractures within BHTV images, utilizing various network structures such as Unet, SegNet, and DeepLabV3. The recognition results are compared with traditional image processing methods, demonstrating the advantages of deep models in accurately segmenting complex geological images. Furthermore, we enhance the model's performance by incorporating an attention mechanism into the encoder-decoder process.Once precise segmentation of rock discontinuities is achieved, the fracture skeleton is extracted using the image thinning method, representing fractures as one-pixel-width curves. Ultimately, automated calculations is completed for dip strike, dip angle, and fracture thickness. Results This method is applied to segment and calculate borehole televiewer images in hydraulic engineering. Comparing the results of manual extraction and automatic extraction of fracture information, the error of dip strike and dip angle is less than 3°, and the fracture thickness error is less than 0.65 mm. Conclusion The results verify that the intelligent calculation method of fracture information proposed in this paper. The proposed method has wide application prospects in hydraulic engineering.

[Introduction] Under the existing converter station reactive power compensation scheme, the synchronous condenser only provides transient reactive power support, and the AC (Alternating Current) filter undertakes steady-state reactive power compensation and adjustment tasks, with a lack of coordination scheme between the two and insufficient economy. [Method] On the basis of ensuring the dynamic characteristics of the synchronous condenser, based on the operating characteristics of the DC (Direct Current) sending and receiving end systems, firstly, a reactive power optimization scheme for the converter station was proposed, in which the regulator was used to replace a part of the part of the AC filter capacity with the regulator, and both participated in steady state reactive power compensation. Ubsequently, a two-layer reactive power adjustment strategy for the converter station was proposed according to the sequence of actions. The first layer of reactive power adjustment was undertaken by the synchronous condenser, which was realized by adjusting the output within the range of steady state operation. The second layer of reactive power adjustment was started after the first layer of reactive power adjustment, and the AC filters were switched on group by group for reactive power adjustment until the reactive power adjustment requirement of the converter station was reached. Finally, a LCC-HVDC (Line-Commutated Converter High Voltage Direct Current) system simulation model was built in PSCAD for simulation verification. [Result] The simulation results show that the reactive power compensation scheme and the two-layer reactive power adjustment strategy proposed can optimize the number of AC filter configuration groups in the converter station, reduce the frequency of AC filter switching in the converter station. [Conclusion] The proposed scheme achieves optimization of the existing LCC-HVDC converter station reactive power compensation scheme and reactive power adjustment strategy.

Series of experiments were performed to simulate the invasion of formation sand into and the plugging process of gravel-pack at different viscosities and flowing rates of fluid. Two types of formation sands with the medium size of 0.10 mm and 0.16 mm and the quartz sand and ceramsite of 0.6—1.2 mm were used in the experiments. A new viscosity-velocity index (the product of fluid viscosity and velocity) was put forward to characterize the influencing mechanism and law of physical property and flow condition of formation fluid on gravel-pack plugging, and a new method to optimize the production rate of wells controlling sand production with gravel-packing was proposed. The results show that the permeability of formation sand invaded zone and final permeability of plugged gravel-pack have negative correlations with viscosity and flow velocity of fluid, the higher the flow velocity and viscosity, the lower the permeability of formation sand invaded zone and final permeability of plugged gravel-pack will be. The flow velocity and viscosity of fluid are key factors affecting plugging degree of the gravel zone. The viscosity-velocity index (v-v index) can reflect the flow characteristics of fluid very well and make it easier to analyze the plugging mechanism of gravel zone. For different combinations of fluid viscosity and flow velocity, if the v-v index is the same or close, their impact on the final gravel permeability would be the same or close. With the increase of the v-v index, the permeability of plugged gravel zone decreases first, then the reduction rate slows down till the permeability stabilizes. By optimizing production and increasing production step by step, the optimal working scheme for sand-control well can reduce the damage to gravel-pack zone permeability caused by sand-carrying fluid effectively, and increase well productivity and extend the sand control life. Key words: gravel-pack, sand retention experiment, plugging law, viscosity-velocity index, working scheme, optimization method, plugging simulation

To deal with sand production problems during the process of producing natural gas from hydrate-bearing sediments (HBS) using reservoir-fluid extraction method, a new gravel sizing method for sand control packing named “Hold coarse while eliminate fine particle (HC & EF method)” was developed for the clayey hydrate-bearing formations. Site X, in Shenhu area, South China Sea was taken as an example to describe detailed gravel sizing procedure. On the basis of analyzing basic particle size distribution (PSD) characteristics of HBS at Site X, the formation sand was divided into two components, which are coarse component and fine component. The gravel sizes for retaining coarse component and eliminate fine component were calculated, respectively. Finally, intersection of these two gravel sizes was taken as the proper gravel size for Site X. The research results show that the formation at Site X is clayey sand with poor sorting and uniformity, proper gravel size for upper segment packing is 143−215 μm, while that for lower segment packing is 240−360 μm. In consideration of the difficulty of layered sand control operation on offshore platform, proper gravel packing size for Site X is recommended as 215−360 μm. Key words: gas hydrate, production test, sand management, gravel sizing, South China Sea, Shenhu area

In order to find out the human factors leading to coal mine general accidents, the paper counts the coal mine accidents occurred in 2019. Based on the investigation reports, HFACS is used to carry out the statistics and analysis of 43 coal mine general accidents. Then the relationships between different human factors, which are involved in the HFACS-PEFE framework, are identified by conducting chi-square test and odds ratio analysis. Four routes of coal mine general accidents are obtained. Among them, the route comprised of organizational process loopholes → invalid correction → individual mental state → habitual violation is the easiest way to lead to general accidents in coal mines. Finally, according to the analysis results, the measures to prevent coal mine general accidents are proposed.