Layer | Fill | Outline |
---|
Map layers
Theme | Visible | Selectable | Appearance | Zoom Range (now: 0) |
---|
Fill | Stroke |
---|---|
Collaborating Authors
Liu, Xiaohua
Application of CNN Deep Learning to Well Pump Troubleshooting via Power Cards
Zhou, Xiangguang (PetroChina Research Institute of Petroleum Exploration & Development) | Zhao, Chuanfeng (China University of Petroleum-Beijing) | Liu, Xiaohua (PetroChina Research Institute of Petroleum Exploration & Development)
Abstract Recent years have seen extensive applications of deep learning, especially in identification and analysis of images, audios and texts, but incipient applications in petroleum industry. Shapes of loops in power cards of a pumping unit are valuable indicators for pump troubles. These troubles may cause engineering accidents, increase operation costs and reduce operation efficiency. This paper applies image recognition technique based on Convolution Neural Network (CNN) to well pump troubleshooting via power cards. Recent years have seen extensive applications of deep learning, especially in identification and analysis of images, audios and texts, but incipient applications in petroleum industry. Shapes of loops in power cards of a pumping unit are valuable indicators for pump troubles. These troubles may cause engineering accidents, increase operation costs and reduce operation efficiency. This paper applies image recognition technique based on Convolution Neural Network (CNN) to well pump troubleshooting via power cards. Firstly, we establish mathematical models both for displacements of the polished rod clamp of a pump and for loads of the polished rod during a reciprocating movement, and preset input parameters corresponding to pump trouble types and severity levels. Ideal benchmarking power cards as the media for pump troubleshooting are generated by simulating complete pumping processes via running the mathematical models with the preset pumping parameters. Secondly, we establish a power card classification model with the AlexNet method. Then we train it with the ideal benchmarking power cards to develop its function of pump troubleshooting and increase the classification accuracy. This model gains robustness and universality from manually presetting parameters for and full coverage of trouble types and severity levels. Thirdly, we train the classification model with real power cards and obtain the preliminary classification results. A further training makes it more practical and applicable to local operations of pump troubleshooting. In the further training, we localize the ideal benchmarking power cards via manual inspection and local expertiseby adjusting the preliminary classification results honoring field expertise. Finally, we randomly divide the localized benchmarking power cards into one training set and one testing set, and then train the classification model with the training set and then apply it to the testing set. The final classification results revealthe high accuracy and practicability of the classification model. It is recommended that GPU should be used for calculation with the classification model to satisfy clients' requirements for higher speeds and efficiency. It provides a feasible method to exploit the potential value of oilfield data assets. The work in this paper will function as a stepping stone in applying ideas, algorithms and models of artificial intelligence to more extensive and thorough aims.
Evaluation of Dynamic Reserves in Ultra-Deep Naturally Fractured Tight Sandstone Gas Reservoirs
Luo, Ruilan (RIPED, PetroChina) | Yu, Jichen (RIPED, PetroChina) | Wan, Yujin (RIPED, PetroChina) | Liu, Xiaohua (RIPED, PetroChina) | Zhang, Lin (RIPED, PetroChina) | Mei, Qingyan (PetroChina Southwest Oil& Gas Company) | Zhao, Yi (PetroChina Southwest Oil& Gas Company) | Chen, Yingli (PetroChina Southwest Oil& Gas Company)
Abstract Ultra-deep naturally fractured tight sandstone gas reservoirs have the characteristics of tight matrix, natural fractures development, strong heterogeneity and complex gas-water relations. There is strong uncertainty of gas reserves estimation in the early stage for such reservoirs, which brings big challenge to the development design of gas fields. Taking Keshen gas field in Tarim basin as example, during the early development stage, the dynamic reserves were much less than those of proven geologic reserves. As results, the actual production performances are obviously different from those of conceptual design. What are the reasons? How to adjust the development program of gas field? Based on special core analysis, production performance analysis, gas reservoir engineering method, and numerical simulations, influencing factors on evaluation of dynamic reserves for ultra-deep fractured tight sanstone gas reservoirs are analyzed. The results show that rock pore compressibility, recovery percent of gas reserves, gas supply capacity of matrix rock, water invasion are the major factors affecting the evaluation of dynamic reserves. On the basis of above analysis, some suggestions are given for the evaluation of dynamic reserves in Ultra-deep fractured tight sandstone gas reservoirs. For this kind of reservoirs, it is reasonable to determine the gas production scale based on dynamic reserves instead of proven geological reserves.
Diagnosis and Analysis of Well Abnormal Flowing Data in Dina HPHT Gas Field, West China
Liu, Xiaohua (Research Institute of Petroleum Exploration & Development-Langfang, CNPC) | Zou, Chunmei (Research Institute of Petroleum Exploration & Development-Langfang, CNPC) | Liu, Hualin (Research Institute of Petroleum Exploration & Development-Langfang, CNPC)
Abstract This paper presents an integrated approach for diagnosis and synthesis of production data for gas wells in Dina HPHT reservoir, where most of the production wells exhibit everlasting abnormal flow with limited analyzable data For production data diagnostics, we prove that for gas wells at high pressure stage (both reservoir pressure and wellbore flowing pressure are above 50MPa), the qg/(Ppi - Ppwf) versus tca plot for gas system can be substituted rigorously by qg/(Pi - Pwf) versus Gp/qg plot (similar form as liquid system) in Palacio-Blasingame decline type curve analysis; and for gas Flowing Material Balance (FMB) analysis, the qg/(Ppi - Ppwf) versus (qgtca)/[(Ppi - Ppwf)Cti] plot can be substituted by qg/(Pi - Pwf) versus Gg/(Pi - Pwf) plot. These reduced functions are used to generate simplified Blasingame diagnostic plot and FMB diagnostic plot for high pressure gas system. These simplified diagnostic plots, with no iterations or integrations when calculating, provide a quick and straightforward solution for instant detection of abnormal flowing data in field. In well production profile synthesis, the dependence of well OGIP on individual-well-flow-rate/reservoir-flow-rate ratio is given based on "pseudo-steady-state" flow theory in a multiwell reservoir. Interference effects are coupled into the analytical model by assigning varied OGIP in production profile synthesis, which reduce the risk of erroneous results in modeling long-term production profile for wells with limited analyzable history data and make the "generated" well behavior consistent with the multi-well-reservoir behavior.
- Asia (1.00)
- North America > United States > Virginia > Virginia County (0.40)
- North America > United States > Texas > Yoakum County (0.40)
- North America > United States > Texas > Wichita County (0.40)
- Asia > China > Xinjiang Uyghur Autonomous Region > Tarim Basin > Dina Field (0.99)
- North America > Canada > Alberta > Dina Field > Acl Dina 7-36-44-1 Well (0.97)