Multiple fault diagnosis of down-hole conditions of sucker-rod pumping wells based on Freeman chain code and DCA

It is important to achieve continuous, stable and efficient pumping well operation in actual oilfield operation. Down-hole pumping well working conditions can be monitored in real-time and a reasonable production scheme can be designed when computer diagnosis is used. However, it is difficult to make a comprehensive analysis to supply efficient technical guidance for operation of the pumping well with multiple faults of down-hole conditions, which cannot be effectively dealt with by the common methods. To solve this problem, a method based on designated component analysis （DCA） is used in this paper. Freeman chain code is used to represent the down-hole dynamometer card whose important characteristics are extracted to construct a designated mode set. A control chart is used as a basis for fault detection. The upper and lower control lines on the control chart are determined from standard samples in normal working conditions. In an incompletely orthogonal mode, the designated mode set could be divided into some subsets in which the modes are completely orthogonal. The observed data is projected into each designated mode to realize fault detection according to the upper and lower control lines. The examples show that the proposed method can effectively diagnose multiple faults of down-hole conditions.

A Method for Forecasting Failures of Sucker Rod-pumped Wells

An exact forecast of the failures of a sucker rod-pumped well in a production area means much for an oilfield’s operation budget, operational arrangement and production plan. In this paper, according to the characteristics of failed sucker rod-pumped well randomness and strong outburst, with the gray GM (1,1) forecast model and the Markov forecast model combined, gray GM (1,1) forecast model is utilized to handle the primary data of an oilfield, and Markov forecast model is utilized to calculate the state transfer probability of forecast value. Then, the gray Markov forecast model considering the influence of randomness factors is formed. Field results prove that the calculation precision of this method is higher and the practicability is greater.

Numerical Prediction and Performance Experiment in a Deep-well Centrifugal Pump with Different Impeller Outlet Width

The existing research of the deep-well centrifugal pump mainly focuses on reduce the manufacturing cost and improve the pump performance, and how to combine above two aspects together is the most difficult and important topic. In this study, the performances of the deep-well centrifugal pump with four different impeller outlet widths are studied by the numerical, theoretical and experimental methods in this paper. Two stages deep-well centrifugal pump equipped with different impellers are simulated employing the commercial CFD software to solve the Navier-Stokes equations for three-dimensional incompressible steady flow. The sensitivity analyses of the grid size and turbulence model have been performed to improve numerical accuracy. The flow field distributions are acquired and compared under the design operating conditions, including the static pressure, turbulence kinetic energy and velocity. The prototype is manufactured and tested to certify the numerical predicted performance. The numerical results of pump performance are higher than the test results, but their change trends have an acceptable agreement with each other. The performance results indicted that the oversize impeller outlet width leads to poor pump performances and increasing shaft power. Changing the performance of deep-well centrifugal pump by alter impeller outlet width is practicable and convenient, which is worth popularizing in the engineering application. The proposed research enhances the theoretical basis of pump design to improve the performance and reduce the manufacturing cost of deep-well centrifugal pump.

Advantages of employing multilevel monitoring wells for design of tunnels subjected to multi-aquifer alluvial

For tunnels being excavated through multiple knowledge of the aquifers’hydraulic head becomes essential for determining groundwater inflow into the tunnel and analyzing its stability,specifically using multilevel monitoring systems.In the multi-aquifer alluvial section of the Glas tunnel(Iran),since the hydraulic head calculations were based on the data obtained from single-piezometer boreholes,the excavation risk was assessed to be at high level and the tunnel seemed to be unstable,thus an incorrect conclusion was derived from the misleading data.To take cost mitigation measures into account,it was necessary to calculate the hydraulic head at tunnel level accurately.By installing nested and clustered wells the mean hydraulic head was measured to be 70 m,significantly different from the 90 m previously determined by boreholes.Considering the updated value,the groundwater inflow and bulkhead load,formerly calculated as 0.65 m^(3)/s and 9.5 bars,were determined to be 0.49 m^(3)/s and 7.5 bars,respectively.

Intra-acceptor hole relaxation in Be δ-doped GaAs/AlAs multiple quantum wells

This paper studies the dynamics of intra-acceptor hole relaxation in Be δ-doped GaAs/AlAs multiple quantum wells （MQW） with doping at the centre by time-resolved pump-probe spectroscopy using a picosecond free electron laser for infrared experiments. Low temperature far-infrared absorption measurements clearly show three principal absorption lines due to transitions of the Be acceptor from the ground state to the first three odd-parity excited states respectively. The pump-probe experiments are performed at different temperatures and different pump pulse wavelengths. The hole relaxation time from 2p excited state to ls ground state in MQW is found to be much shorter than that in bulk GaAs, and shown to be independent of temperature but strongly dependent on wavelength. The zone-folded acoustic phonon emission and slower decay of the wavefunctions of impurity states are suggested to account for the reduction of the 2p excited state lifetime in MQW. The wavelength dependence of the 2p lifetime is attributed to the diffusion of the Be atom δ-layer in quantum wells.

Wellsnet油井监控系统在鄯善油田的应用

Wellsnet油井监测系统是江苏华盛信息技术有限公司根据油田生产的实际需求,结合吐哈油田的生产实际,研制的新一代应用于石油开采行业的抽油井自动化监控系统,它集油井数据采油、生产、控制、监测、报警和管理、地面向地下延伸于一体;融合了多项高新技术。为了提高油井的日常管理水平,及时发现和掌握抽油机的工作状态和泵运行情况,及时采取相应措施,保证抽油机平稳高效运行,从而达到有效提高采油时率的目的,为了达到这个目的,鄯善油田在130口抽油井上安装了油井监控系统,使抽油井的管理迈向信息化时代。

T2Well单井地下水源热泵水-热耦合数值模拟研究

以北京地区实际井参数刻画的单井地下水源热泵为研究对象,利用T2Well模拟器对单井系统地源热泵运行过程进行全面模拟,并对主要参数进行敏感性分析,探究对单井系统生产井温度、热突破时间的影响。距离井筒相同位置不同深度的点,埋深越大受回灌水温度的影响越小。每个运行周期结束时,温度场分布都能得到一定程度的恢复(同一周期内初始温度场相比)。水平方向渗透率同垂向渗透率的比值越大,越有利于回灌水在含水层中的运移,可有效保证热泵系统的取热温差。初步预计一个单井水源热泵系统能满足北京地区约1.5万m^2面积建筑的供暖需求,约1.6万m^2面积建筑制冷需求。

基于停泵压降数据的页岩气井单段裂缝参数反演——以长宁N209井区页岩气井为例

水平井多级分段多簇压裂是页岩气藏有效开发的关键技术,准确评价缝网参数是开展压裂效果评估、压裂工艺优化和压后产能预测等工作的重要前提.传统页岩气井压裂裂缝参数表征主要采用裂缝扩展模拟与现场监测等手段,对压裂施工停泵压降曲线的信息挖掘与利用不足.采用数字滤波方法消除停泵压降数据的水锤波动,利用分形裂缝表征页岩气井压裂缝网,结合数值试井分析实现了单段裂缝参数反演,建立了一套基于停泵压降数据的缝网参数反演工作流程.该成果在长宁公司N209井区压裂井成功应用,并对裂缝参数评价结果与一体化压裂模拟和产气剖面测试等反演结果进行了验证.结果表明,基于停泵阶段压降数据的单段裂缝参数反演结果显著提升了参数解释精度,有助于揭示页岩气井段间产气能力差异的原因;气井产能与裂缝总长和裂缝表面积的相关性明显高于利用广域电磁法确定的压裂改造体积,其中,早期日产气量与总缝长相关性最高;预计最终回收率与裂缝表面积相关性最高;缝网总长与单簇压裂液规模强线性相关,反映了长宁区块页岩气井压裂工艺2.0的技术先进性.获取页岩气井单段裂缝参数可为产能主控因素分析和压裂效果评价提供更加有效的分析指标.

地浸采铀注液井小直径潜水泵抽水洗井试验

某地浸采铀试验采区注液井周边含矿层堵塞严重。注液井套管内径仅有80 mm,在常规空压机抽水洗井结束后,在3口注液井中开展了小直径潜水泵抽水洗井试验。初始抽水量均在4.5 m_(3)/h左右,期间最小抽水量为3.4 m_(3)/h,其中2口井的第1股最浑浊洗井水的含砂量分别为4.36、5.19 kg/m_(3)。粒度、电镜扫描和化学组成分析显示,洗井水沉降干渣中的矿物碎屑主要为粉砂、细砂碎屑,少量黏土碎屑;化学沉淀主要是CaCO_(3)、MgCO_(3)和FeCO_(3);潜水泵抽水洗井累计运行约27.5 h,未出现任何故障。试验表明,小直径潜水泵抽水洗井可行且投资少、成本低、安全性高、实施方便。

On measures to improveeffectiveness of pump suc-tionreverse circulation drillin,

ＯＮＭＥＡＳＵＲＥＳＴＯＩＭＰＲＯＶＥＥＦＦＥＣＴＩＶＥＮＥＳＳＯＦＰＵＭＰＳＵＣＴＩＯＮＲＥＶＥＲＳＥＣＩＲＣＵＬＡＴＩＯＮＤＲＩＬＬＩＮＧＷｕＧｕａｎｇｌｉｎＣｈｅｎｇｄｕＩｎｓｔｉｔｕｔｅｏｆＴｅｃｈｎｏｌｏｇｙ，Ｃｈｅｎｇｄｕ６１００５９，Ｃ...

井下旋流分离管柱的振动特性研究

为研究旋流分离管柱在注入泵和采出泵两螺杆泵抽吸条件下的振动特性,对其进行模态分析和瞬态动力响应分析。结果表明,不同工况下,旋流分离管柱的固有频率有所不同;固定注入泵一端和固定采出泵一端相比,固定采出泵一端各阶固有频率较小;螺杆泵转动引起的激振力和内部流量变化都会导致管柱变形,其中螺杆泵转动对旋流分离管柱变形影响较大;相对于固定采出泵,固定注入泵条件下旋流分离管柱产生的振动位移更大,是旋流分离管柱自身结构和注入泵转速产生更大激振力综合作用导致的结果;控制两个螺杆泵工作状态的时间差,理想情况下可以控制旋流分离管柱的总变形。

地热井用潜水泵冲蚀磨损特性

为研究颗粒属性对地热井用潜水泵外特性和冲蚀磨损的影响,采用离散相模型和RNG k-ε湍流模型对两级潜水泵内流场进行了全流道数值模拟.结果表明:随着固体颗粒体积分数的增大,潜水泵各级扬程和效率均呈下降趋势,单级扬程最大下降1.68 m,单级效率最大下降5.18%;叶轮的冲蚀磨损主要出现在叶片工作面进口,导叶的冲蚀磨损主要出现在导叶工作面出口和导叶前盖板;随着固体颗粒密度的增大,各部件的平均冲蚀率增加明显;叶轮和导叶的最大冲蚀率均随固体颗粒粒径的增大呈增大趋势,其中导叶的最大冲蚀率随颗粒粒径变化相比叶轮更加明显,当颗粒粒径超过0.650 mm时,过流部件最大冲蚀磨损率急剧上升.

“两抽两灌”水源热泵系统应用效果分析及优化布置

为探究丹东市某工程地下水源热泵系统的实际应用效果及布井方式的合理性,采用COMSOL Multiphysics软件,建立三维水热耦合数值模型,对“两抽两灌”水源热泵系统进行为时一年的运行水温监测,并对实际应用中抽灌井的布置方式以及抽灌井间距是否存在含水层“热贯通”问题进行分析,将存在的问题与工程场地条件相结合。结果表明,采用“抽灌井平行布置”时,且抽灌井间距离为55 m时,发生“热贯通”的程度最轻,为丹东市水源热泵规划提供科学设计方法及参考依据。

基于无监督学习的抽油机井示功图自动聚类与批量标注方法

为充分利用大量未标注样本、节约人力与时间,提出了基于无监督学习的抽油机井示功图自动聚类与批量标注方法。首先,将抽油机驴头往复运动产生的位移、载荷数据转化为示功图图片样本,其中,示功图的横坐标为位移,纵坐标为载荷;其次,加载在ImageNet上训练过的带有一系列权重参数、具有强特征提取能力的卷积神经网络模型;然后,去除该网络模型的全连接层,利用该网络模型提取示功图图片样本的特征;最后,利用k-means聚类算法对提取到的特征进行聚类分析,将具有相似特征的示功图聚到同一文件夹中。批量的对示功图聚类结果进行快速标注,从而形成抽油机井故障诊断的示功图样本集。实验随机搜集了100口抽油机井的20 000条示功图数据,结果表明,基于无监督学习的抽油机井示功图自动聚类与批量标注方法耗时短、准确率高,为示功图样本集标注提供了一种高效方法,对于充分挖掘油田大数据的应用价值具有示范意义。

煤层气L型水平井电潜螺杆泵井清防砂工艺技术创新及应用

套管完井压裂投产的煤层气L型水平井出砂严重,倒置式电潜螺杆泵举升工艺受压裂砂影响故障频繁,影响了连续稳定排采,因此确定了倒置式电潜螺杆泵排采过程中由于压裂砂导致的故障点,并针对性创新应用了井筒内、导流罩及泵内清砂工艺技术,同时优化防砂管柱设计,确保了连续稳定排采。

基于改进ResNet的示功图分类算法研究

示功图是反映抽油机井工作状态的重要图示,通过分析示功图的闭合曲线形状,可以得出抽油机井的具体工作状态,从而可以判断出抽油机井是否发生故障以及具体的故障类型。随着深度学习的发展,基于深度神经网络的示功图分类也逐渐应用到了抽油机井工况检测当中。该文提出了基于改进ResNet的示功图分类算法,通过优化残差结构和引入SE子结构等措施,提高了分类准确性和鲁棒性。改进的残差结构嵌入了SE子结构,对输入特征进行降维的同时也减小了参数的数量,在降低计算量的同时也添加了更多非线性因素,通过不断增加有效特征的权重,不断减小无效特征的权重,进而完成了特征重标定,不仅起到加速网络收敛的作用,也使模型更加轻量化,从而提高了模型的性能。相较于其它模型,改进的ResNet模型可以更好地适应示功图分类任务,分类效果更好。实验结果表明,基于改进ResNet的示功图分类算法在精确率、召回率和F1值上均优于其它示功图分类算法。该研究为抽油机井工况检测系统提供了更好的理论支持。

基于机理仿真的抽油机有杆泵举升系统生产参数优化设计

基于传统经验法和预测模型法的抽油机有杆泵举升系统设计方法在实际应用中要么过于简单要么过于复杂,设计方案的可靠性还有待提升。因此,应用大数据驱动技术从海量抽油机井举升系统设计方案中探索规律,推荐以提高产量和节能降耗为目标的抽汲参数优化设计最优方案,将提升抽油机井举升系统设计方法的科学性和可操作性。文章分析了抽油机井产量主控因素,将软件仿真和现场获取的抽油机井举升系统设计样本进行聚类分析融合得到新的样本库,建立了抽油机井举升系统设计方案的深度学习模型,研究了基于匹配算法的方案相似度模型,开发了抽油机井有杆泵举升系统生产参数设计推荐系统。根据实例计算验证,该系统能够根据待设计井的地质特征和开发特征,从海量样本中匹配到与地质油藏条件相似度高且运行效果良好的举升设计方案,推荐方案的举升参数及举升效果与原有方案相比有较大的提升,证明了计算模型的实际可行性,为抽油机举升系统优化设计提供了新思路及方法。

地热井用两级潜水泵内流熵产诊断及水力性能优化

为了提升地热井用潜水泵的运行效率,根据地热井用潜水泵的流动参数和实际运行工况,完成了两级潜水泵过流部件水力设计,并进行了全流道流动数值模拟和试验验证;采用熵产分析法定位了潜水泵内流场的高能耗区域,主要集中在级间交汇和动静部件干涉处;采用正交设计和结构参数敏感性分析方法,确定了结构参数对潜水泵性能的影响规律,以叶轮叶片出口角、出口宽度、叶片包角和导叶轴向长度为优化变量,以扬程和效率为优化目标,通过遗传算法、最优拉丁超立方采样和熵权TOPSIS法对两级潜水泵进行水力优化。结果表明,额定工况下,优化泵的扬程增加了24.9%,效率增加了5.7%,总水力损失减小了1024.9 W,高能耗区域明显降低,水力性能显著提升,分离流、二次回流和冲击流动等不良流动得到显著抑制;大流量工况处理能力大幅度提升,高效区间范围扩大了70%,运行稳定性得到显著提高。研究内容为潜水泵的设计及内流场研究提供理论指导。

基于有效控制抽水量和地面沉降的长江Ⅰ级阶地深基坑深井降水参数取值优化

针对目前深基坑深井降水研究中仅关注目标降深及相应地面沉降,未考虑满足目标降深条件下的有效控制抽水量问题,以武汉长江隧道明挖段深基坑深井降水工程为例,通过抽水试验分析了井管直径、进水段的相对位置b/M以及井的不完整程度l/M对基坑涌水量的影响,确定了止水帷幕长度、井管直径、过滤管有效长度和合理位置等降水参数,根据试验数据拟合出该地区非完整井涌水量经验公式;建立了控制抽水量和地面沉降的深基坑深井降水参数优化设计模型,求解Maxf(l)得出的过滤管长度,与抽水试验结果相近。结果表明:止水帷幕进入强透水层厚度2/3处时,其对基坑坑内抽水量和坑内外水位有控制作用;选用Φ250井管、过滤管设置在强透水层上部(b=0)、l/M值较小及附加阻力系数差(ξ_(1)-ξ_(2))最大时,基坑涌水量较小;q_(0)-l、Δq_(0)/Δl-l曲线中曲率最大点对应的过滤管长度范围为14.0~16.0 m,其对应的基坑涌水量q0最小;考虑井群干涉的影响,干涉井涌水量在基坑中心处最小、角点处最大。