节点文献

气举井嵌入式测试系统研制与应用

【作者】 喻欣

【导师】 李颖川;

【作者基本信息】 西南石油大学 , 油气田开发, 2007, 硕士

【摘要】 气举采油法是应用较早也较为广泛的一种机械采油方法,在采油工程中起着举足轻重的作用,已在我国中原、吐哈、塔里木、四川、南海等油气田得到了推广应用。随着油气田开发进入中后期,地层能量、产能逐渐降低,其注气量不断增加,气举效率随之降低,同时易受井口回压波动影响,气举井稳定性变差,管理难度增大,导致气举井井数正逐渐减少。因此,如何实时跟踪气举井工作状态参数,及时调整注气参数,确保气举井高效运行已成为目前国内连续气举采油的关键问题。本文在气举井动态测试仪器研制及其分析方法研究方面主要完成了以下工作:(1)以嵌入式系统构建动态测试的硬件平台,结合无线数传技术设计了测试系统的分布式通讯网络,研制了气举井动态测试仪器。并以uClinux系统为基础,应用分层软件设计方法,综合设计系统层、中间层和应用层,编制了测试仪专用软件。软、硬件结合,形成了一套气举井动态测试系统,(2)自行设计研制了气举可视化实验架,实验架总高为18.2m,能模拟5(1/2)″套管内采用API1.9″油管球塞气举或常规连续气举排水采气两相流举升过程,开展了一系列常规连续气举和球塞气举实验研究。(3)将气举井嵌入式动态测试系统应用于气举实验架上,对注气量、注气压力、举升液量、投球频率等实验参数进行实时监测。经近两年的实验研究表明:动态测试系统运行稳定,测试数据准确,同时获取了大量的实验测试数据。(4)优选差分法对差压和压力实验测试数据进行去噪处理,从数据时域、频域内提取了标准偏差、非对称系数、峭度函数等10个特征量,建立了举升段塞特征样本库。以特征样本为输入参数,实际段塞类型为预期输出参数,训练了举升段塞神经网络分析模型。在此基础上编制了分析软件,能够准确判断球塞段塞和无球段塞。为球塞气举现场应用提供了必要的技术支持。(5)在实验架应用的基础上,将嵌入式气举井动态测试系统拓展到现场应用的无线远程测试系统。应用该系统采集的中原油田连续气举井生产数据,分析了油压、套压波动对气举井生产的影响,初步建立了气举井典型工况特征样本库,编写了气举井油压、套压神经网络分析软件,可判断正常生产,油套连通,间歇出油三种工况。将上述研究成果集成,初步完成气举井动态测试综合配套技术,成功地从物理模拟实验架测试走向现场无线远程监测,为气举井动态参数监测提供了重要的技术手段,分析了气举井工作状态,为及时有针对性地调整气举井工作参数,使其高效稳定工作提供了技术支持,表明本文研究成果对连续气举采油工艺具有重大的现实意义和理论价值。

【Abstract】 Gas lift is a kind of mechanical oil production method which is used widely and early. It plays a significant role in the oil production engineering. Because of its great advantages, gas lift has been successfully used in many oil fields such as Zhongyuan, Tuha, Talimu, Sichuan, Nanhai etc. However, when the exploitation of oil-gas field goes to the middle or last phase, the stratum energy and the production capacity are shrinking, the gas injection pressure and the gas injection are increasing, correspondingly, the efficiency of the gas lift is also decreasing. Besides, the task of harmonizing the parameter of formation feed flow and gas injection rating becomes much more difficult, which is tended to be effected by the fluctuation of the back pressure of the mouth of the well. The stability of the gas lift well becomes worse and the well is hard to be controlled, which lead to the number of the gas lift well dropping off. Therefore, how to track the real-time work behavior parameter of the gas lift well, adjust the gas injection parameter, and ensure the high efficient function of the gas lift well are the main problems of continuous oil extraction in China. The following research work on the development of the dynamic test apparatus of gas lift well and the correlative theory has been finished in this paper:1)In the paper, the dynamic test hardware platform which is integrated into the embedded system is constructed, the distributed communication network of the testing system with the wireless data translation technology is designed, and the dynamic checkout apparatus of gas lift well is developed. On the basis of the uClinux system, the system layer, the intermediate layer and the application layer are scientifically designed with the method of layered software. An integrative program of dynamic test system is developed, which provide important instrument and platform for dynamic monitoring parameter of gas lift well.2)Gas lift simulative testing equipment with a height of 18.2 meters is designed and built up, which can take a series experimental study of conventional continuous gas lift mainly used for simulative study of water drainage of gas well and pig lift simulation with 51/2" casing and API1.9" oil slug.3) Dynamic testing and control system which is integrated into the embedded system is applied to gas lift experimental equipment to make real-time test and control on several important experimental parameters such as the quantity of gas injection, the gas injection pressure, lifting liquid production, ball injection frequency, etc. The experiment results in recent two years illustrate that the dynamic testing and control system run well, testing data can properly reflect the behavior of gas lift system. Meanwhile, an abundant of testing data is acquired.4)Well comparing several kinds of methods of processing the testing data, the paper options the calculus of differences to analyze the experimental data of pressure and differential pressure. It picks out ten characteristic quantities such as standard deviation, non-symmetry coefficient and kurtosis function from time-frequency domain and establishes the characteristic sample database of the lifting slug. The paper uses the method of BP neural network system, it takes the characteristic samples as the input parameter and the actual type of the slug as the expecting output parameter. It practices the analytical neural network model of lifting slug. On the basis of the model, it designs the software which can make a correct estimate on the pig slug and normal slug and gives an important technical support to the application on the normal slug lift in the fieldwork.5) Based on the application of experiment equipment, the embedded dynamic testing control system is extended to the application of wireless remote testing system in the fieldwork. Applying the production data of the continuous gas lift well in Zhongyuan oil field, the paper analyzes the influence of the fluctuation of oil pressure and casing pressure on the production of the gas lift well, makes a study on the diagnoses of the working behaviors of gas lift well, establishes the characteristic sample database of gas lift well and designs the software about the neural network of the oil pressure and casing pressure of gas lift well. With the software, it can differentiate three kinds of working behaviors which are the normal production, the connection of oil casing and the batch production.Integrating the research work above can complete the dynamic testing synthesis compound technology of gas lift well, make the physical simulative experimental equipment test successfully transformed to the wireless remote monitoring in the fieldwork. It efficiently collects the dynamic parameter of gas lift well and analyzes the work behaviors of gas lift well. The research work made by the paper has both great practical significance and theoretical worth to the technics of continuous gas lift oil extraction. It also makes the parameter of gas lift well correctly adjusted to ensure the high efficiency.

【关键词】 气举嵌入式系统动态测试神经网络
【Key words】 gas liftembedded systemdynamic testneural network
  • 【分类号】TE937
  • 【被引频次】1
  • 【下载频次】218
节点文献中: 

本文链接的文献网络图示:

本文的引文网络