摘要
煤层气排采开发阶段,井眼轨迹与煤层的有效接触面积对优化采气速度和提高采收率有着重要意义。而在煤层气水平井钻井过程中实时、精确地识别煤层,可以明显地提高井眼轨迹在煤层中延伸长度,增加有效接触面积。应用LWD数据开展煤层识别已在工程现场得到广泛应用,而整合综合录井数据实时识别煤层的方法还处于研究阶段。基于煤层与围岩的地层岩性差异在综合录井数据上表现的特征,采用BP神经网络算法,以综合录井数据为依托,提出了实时识别煤层的录井解释方法。研究显示,通过煤层识别录井解释方法,应用综合录井数据不但可以实现煤层实时识别,而且获得的分析结果还可以指导水平段轨迹在煤层中的延伸,为水平井钻进过程中煤层识别提供了新的思路。结论认为,该方法性能稳定,数据来源广泛,响应时间短,准确度高,并拓宽了综合录井数据的应用领域,可以在煤层气水平井导向钻井的研究工作中发挥更大的作用。
At the stage of out-flow and extraction of coalbed methane wells,the effectual contact area between the hole trajectory and the coal bed is of great significance to optimize production rates and recovery.If coal beds are identified in accuracy and immediately in the process of horizontal drilling in coalbed methane gas wells,the length of well trajectory will be more extended in coal beds,thus the effectual contact area will be widened.Although the method of coal bed identification by use of logging while drilling(LWD)data has been widely applied in fields,the real-time identification of coal beds by use of the integrated logging data is still under research.In view of this,based on the features representing the lithologic differences between coal beds and their periphery rocks from the integrated logging data,a new method of logging interpretation for real-time identification of coal beds is presented on the integrated logging data by use of the BP neural network algorithm.Our research shows that with this new method the coal beds can be easily and immediately identified,and the achieved results can also be regarded as guidance for monitoring the extension of horizontal well section's track in the coal beds.This new method provides a new idea for the coal bed identification during the process of horizontal drilling.Field practices show that this new method has a reliably good performance with many advantages like wide data sources,short response time,and a high accuracy,and the integrated logging data can thus be taken good advantage of and will play more important role in the studies of horizontal drilling in coalbed methane gas wells.
出处
《天然气工业》
EI
CAS
CSCD
北大核心
2010年第10期60-63,共4页
Natural Gas Industry
基金
"十一五"国家科技重大专项"大型油气田及煤层气开发--山西沁水盆地煤层气水平井开发示范工程"(编号:2008ZX05061)"煤层气水平井综合地层判识技术"的部分研究成果
关键词
煤层气
水平井
综合录井
神经网络
煤层识别
实时
水平段轨迹
coalbed methane gas,horizontal well,integrated logging,neural network,coal bed identification
