海底管道清管器运行时间预测  

Prediction on running time of pigs in submarine pipelines

作　　者：曹杨 王红红 林超 李忠涛 CAO Yang;WANG Honghong;LIN Chao;LI Zhongtao(CNOOC Research Institute Co.Ltd.;CNOOC China Limited,Tianjin Branch)

机构地区：[1]中海油研究总院有限责任公司 [2]中海石油(中国)有限公司天津分公司

出　　处：《油气储运》2022年第4期451-457,共7页Oil & Gas Storage and Transportation

基　　金：工业与信息化部高技术船舶科研项目“水下油气生产系统全生命周期可靠性保障技术体系研究”,MC-201709-S01;中海石油(中国)有限公司科技项目“在役油气生产平台上部设施安全评估技术研究”,YXKY-2018-ZY-06。

摘　　要：海底管道清管器运行时间对于清管收球作业具有重要的参考作用。为了明确清管器运行时间的影响因素,运用灰色关联分析法对海管里程、海管进出口压力差、输气量、海管进出口温度差、过盈量、清管器质量、输油量、含水率进行关联度计算,分析各影响因素与清管器运行时间的相关性,并以此构建了网络结构为(8,11,1)的清管器运行时间BP神经网络预测模型。以渤海某油田海底管道的清管作业数据记录为样本,对清管器运行时间进行预测及验证,结果表明:该模型预测的清管器运行时间的平均误差为8 min,平均相对误差为12.5%。研究结果可为制订合理的海底管道清管收球作业方案提供参考。The running time of pigs in submarine pipelines is an important reference for pig receiving operation.In order to identify the factors that influence the running time of pigs,the correlation degree of pipeline mileage,pipeline inlet/outlet pressure difference,gas transport capacity,pipeline inlet/outlet temperature difference,amount of interference,pig quality,oil transport capacity and water cut was calculated with the grey incidence analysis method.Meanwhile,the incidence between the influencing factors and the running time of pigs was also analyzed,and on this basis,a BP neural network prediction model of pig running time in the network structure of(8,11,1)was established.Then,the running time of pig in submarine pipeline was predicted and verified based on the data record of pigging operation in a submarine pipeline of the Bohai Oilfield.It is indicated that the running time of pigs predicted by the BP prediction model has an average error of 8 min and an average relative error of 12.5%.In general,the research results could provide reference for the preparation of a reasonable plan for pig receiving operation in the submarine pipelines.

关 键 词：海底管道 清管器 运行时间 影响因素 灰色关联度 BP模型 

分 类 号：TE832[石油与天然气工程—油气储运工程]