Assessment of wellbore integrity failure risk and hazardous zones in depleted reservoirs underground gas storage during the operation processes.

Depleted reservoirs are widely used for underground gas storage due to their advantages of low construction cost and easy development. Under the influence of complex geological conditions and frequent operations, the integrity of the wells in depleted reservoirs is prone to failure, which would potentially lead to gas leakage. In this study, by using a finite element-based computational fluid dynamics model, we have developed evaluation criteria for assessing the severity of the occurred wellbore integrity failure and the risk of the un-occurred wellbore integrity failures respectively to identify hazardous zones potentially prone to wellbore integrity failure. The study results indicate that the gas storage wellbore integrity failure is prone to occur inside the wellbore structure in the direction of the minimum ground stress near the lower boundary of the formation interlayer. The wellbore integrity failure hazardous zones are mainly concentrated at the formation interlayer boundaries. The practical guidelines and solutions derived from current research results can provide an accurate direction for monitoring and protecting work of wellbore integrity and avoid environment pollution problems caused by natural gas leakage.

Pore-throat structure characterization of carbon fiber reinforced resin matrix composites: Employing Micro-CT and Avizo technique.

Micro-CT technique poses significant applications in characterizing the microstructure of materials. Based on the CT three-dimensional(3D) reconstruction technology and "Avizo" 3D visualization software, the microscopic pore-throat structure of porous media can be quantitatively characterized. This paper takes the carbon fiber reinforced resin matrix composites as an example to introduce the operation process of "Avizo" in details, which mainly covers the following modules: Volume Edit, Interactive Thresholding, Fill Holes, Mask, Separate Objects and Generate Pore Network Model, then further discuss the difficult problems when the "Avizo" is employed to analyze. The microstructures of carbon fiber reinforced resin matrix composites illustrate that pores in the upper part of sample are dramatically dispersed, and mainly concentrated in the lower part of sample. The porosity of adopted cuboid is 3.6%, accordingly the numbers of pores and throats reach 268 and 7, respectively. The equivalent radius of pores seems mainly distributed in the range of 0.7-0.8µm, accounting for 28.73% of the total pore number. The surface area of pore ranges from 5 to 10µm2, accounting for 14.16% of the total pore number. The pore volume concentrates in the range of 1-20µm3, accounting for 57.46% of the total pore number. In addition, the equivalent radius of throat mainly concentrates in the range of 1-5µm, the overall length of throat is distributed in the range of 37-60µm, and the equivalent area of throat is distributed non-uniformly in the range of 5-75µm2. This work provides a basis for the further investigation of fluid migration mechanism and law in the composite materials by the numerical simulation methodology.

Predicting ultrasound tongue image from lip images using sequence to sequence learning.

Understanding the dynamic system that produces speech is essential to advancing speech science, and several simultaneous sensory streams can be leveraged to describe the process. As the tongue functional deformation correlates with the lip's shapes of the speaker, this paper aims to explore the association between them. The problem is formulated as a sequence to sequence learning task and a deep neural network is trained using unlabeled lip videos to predict an upcoming ultrasound tongue image sequence. Experimental results show that the machine learning model can predict the tongue's motion with satisfactory performance, which demonstrates that the learned neural network can build the association between two imaging modalities.