子洲地区山西组2段气藏储集层流体分布特征研究
|
摘要
储集层作为油气储集的场所和油气开发的对象,历来是油气藏开发所关注的重点。储集层空间分布与性质好坏,决定了储集流体的空间分布特征,最终直接影响气井产能和气藏开发。子洲地区山西组2段气藏砂体物性变化较大,砂体储集岩性较多,是典型的低孔低渗气藏,受构造-岩性控制。在有效识别储集层基础上再有效识别气藏流体,摸清砂体空间分布特征和流体分布规律,能够合理制定气井产能和气藏开发策略,为气藏的高效合理开发提供保证,延长气藏生命周期,提高最终采收率。
结合子洲气藏山2气藏地质特征,论文以常规测井资料为基础,结合现场多口井岩芯描述结果,提取对应的砂、泥和煤岩性的相关测井属性特征,综合应用多元判别分析和BP神经网络方法科学有效的识别了山2岩性;在综合识别各单井砂体基础之上,分析了研究区山2砂体的平面展布规律;通过测井属性和岩芯分析结果对比拟合,建立了有效的孔、渗、饱测井参数解释模型;在定义不同产层标准基础上,采用流体声阻抗、多元判别分析和支持向量机等多种方法研究不同产层流体测井响应特征及判识,并对全区单井储层流体进行了有效识别,筛选出了流体判别的一些较好方法;最后结合水化学分析数据和生产井动态数据,建立起工区不同水型的判别标准,并指出富水区和气水界面所在,讨论了富水区生产井的生产管理。
结果表明,子洲山2气藏砂体分布较复杂,山_2~3小层是气藏主力产层;多元判别方法和BP神经网络方法识别岩性效果较好,可以应用于工区的岩性识别中;论文建立的储层孔渗饱解释模型较好;在讨论的流体识别方法中,推荐使用多元判别分析、流体声阻抗法和支持向量机进行子洲气田山2段流体识别;多角度全面分析了全区53口投产井的生产动态特征和全区不同水类的离子化学特征,按照水气比变化特点将投产井划分为水气比上升、水气比稳定和水气比下降三大类来讨论;全面系统总结出工区所有投产井的出水机理,即透镜体水、致密地层水和边底水;结合水样离子分析资料进一步分析了不同地层水的水化学特征和离子参数分布规律,计算了不同地层水的特征系数,制定了工区不同水型的识别标准;同时,论文指出工区存在三大块富水区,且富水区地层水性质存在较大差异;提出了工区透镜体水和致密地层水与边底水矿化度差别受凝析水稀释影响的观点。根据渗流力学理论,对比气井产能方程,结合相对渗透率曲线,从含水饱和度角度讨论了产水对产能影响;建立了边底水的气藏理想模型,计算了气井排水采气合理井径、合理产量和合理生产压差;以气水剖面为手段剖析了研究区气水空间分布特征,指出气藏不存在统一的气水界面,且至西向东气水界面相差较大,进而指出了无水区及潜在含水区所在;最后系统总结了研究区的四类气水分布模式,提出工区内地层水型可能受煤层水影响的观点。
通过研究,对子洲山2气藏砂体展布及流体空间分布特征有了更明确的认识,也为现场应用特别是合理制订气藏开发方式、延长气藏无水采气期,延长气藏稳产时间、提高气藏采收率等提供可靠依据。
Reservoir is defined as a place to accumulate oil and gas or a subject of hydrocarbon bearing pool empoldering, which is focused by oil & gas development engineers. Distribution and quality of the reservoir determine the fluid reservoir spatial distribution. Ultimately they make a direct impact on gas production and gas reservoir development. Physical property of gas pool in Shanxi-2 formation in Zizhou area has a great change, and it has more lithologic character. it is a typical low-porosity & low-permeability formation, which is controlled by structure and lithology. Based on the effective identification of the reservoir, It is useful for high efficient and reasonable development through effective liquid identification and confirming liquid distribution.
Considering the geologic characteristics, based on traditional logging data and getting logging properties of sand, shale, and coal from description of rock core, this paper has scientifically and effectively identified lithology with multiple discriminant analysis and BP neural network, and analyzed the distribution of sand body after identify the sand of single well. This paper has established log interpretation modeling through fitting the logging properties and rock core analytic data, such as porosity, permeability and water saturation. At the same time, the author tries to study pay formation liquid log response and discrimination via some different methods, such as fluid acoustic impedance, multiple discriminant analysis and support vector machine, based on different production formation criterion and selects some applicable methods. At last, according to water chemistry and production data, the paper establishes the identification criterion for different water, and further points out the rich water area, the altitude of water contact, and the management of production wells where are nearby the rich water area.
The results have shown that Shan-23 is the major productive formation. Multiple discriminant analysis and BP neural network are useful for the identification of the lithology. And the log interpretation modeling is effective. Multiple discriminant analysis, fluid acoustic impedance and support vector machine are the recommended practicable methods for Zi-Zhou reservoir during all the discussed methods. In addition, this paper sums up three water output mechanisms of lenticular water, tight water and fringe/basal water combined production performance data of 53 brought-in wells that are divided to raised, sustained, and descendant water/gas ratio, and global water chemical analysis data, and works out the recognition criteria of different water. In addition, the paper points that there are three rich water areas during the reach area and there is different characteristic for different water of the three rich water areas. Besides, the paper is also points that the total salinity of fringe/basal water was not different from lenticular water and tight water because of the condensed water. Calling on water saturation, this paper has discussed the relationship between production water and production gas. Considering the relationship, this paper has built the gas field modeling containing fringe/basal water, calculated the feasible well diameter, output and drawdown pressure. According to the prior study, this paper researches the liquid distribution based on gas-water section, and points that the gas reservoir does not have united gas/water surface, and the surface differed from west to east. Simultaneously, the paper points the location of no-water and potential-water. At last, the paper generalizes four gas/water distribution patterns and proposes the concept that the formation water salinity is affected by coal water.
Sand body and liquid distribution of Shan-2 gas reservoir have become clearer than ever before through this research, which also provide a reliable basis for gas reservoir exploitation, prolonging water free gas recovery and stable production period of gas well, enhancing gas recovery, etc, in further research.
结合子洲气藏山2气藏地质特征,论文以常规测井资料为基础,结合现场多口井岩芯描述结果,提取对应的砂、泥和煤岩性的相关测井属性特征,综合应用多元判别分析和BP神经网络方法科学有效的识别了山2岩性;在综合识别各单井砂体基础之上,分析了研究区山2砂体的平面展布规律;通过测井属性和岩芯分析结果对比拟合,建立了有效的孔、渗、饱测井参数解释模型;在定义不同产层标准基础上,采用流体声阻抗、多元判别分析和支持向量机等多种方法研究不同产层流体测井响应特征及判识,并对全区单井储层流体进行了有效识别,筛选出了流体判别的一些较好方法;最后结合水化学分析数据和生产井动态数据,建立起工区不同水型的判别标准,并指出富水区和气水界面所在,讨论了富水区生产井的生产管理。
结果表明,子洲山2气藏砂体分布较复杂,山_2~3小层是气藏主力产层;多元判别方法和BP神经网络方法识别岩性效果较好,可以应用于工区的岩性识别中;论文建立的储层孔渗饱解释模型较好;在讨论的流体识别方法中,推荐使用多元判别分析、流体声阻抗法和支持向量机进行子洲气田山2段流体识别;多角度全面分析了全区53口投产井的生产动态特征和全区不同水类的离子化学特征,按照水气比变化特点将投产井划分为水气比上升、水气比稳定和水气比下降三大类来讨论;全面系统总结出工区所有投产井的出水机理,即透镜体水、致密地层水和边底水;结合水样离子分析资料进一步分析了不同地层水的水化学特征和离子参数分布规律,计算了不同地层水的特征系数,制定了工区不同水型的识别标准;同时,论文指出工区存在三大块富水区,且富水区地层水性质存在较大差异;提出了工区透镜体水和致密地层水与边底水矿化度差别受凝析水稀释影响的观点。根据渗流力学理论,对比气井产能方程,结合相对渗透率曲线,从含水饱和度角度讨论了产水对产能影响;建立了边底水的气藏理想模型,计算了气井排水采气合理井径、合理产量和合理生产压差;以气水剖面为手段剖析了研究区气水空间分布特征,指出气藏不存在统一的气水界面,且至西向东气水界面相差较大,进而指出了无水区及潜在含水区所在;最后系统总结了研究区的四类气水分布模式,提出工区内地层水型可能受煤层水影响的观点。
通过研究,对子洲山2气藏砂体展布及流体空间分布特征有了更明确的认识,也为现场应用特别是合理制订气藏开发方式、延长气藏无水采气期,延长气藏稳产时间、提高气藏采收率等提供可靠依据。
Reservoir is defined as a place to accumulate oil and gas or a subject of hydrocarbon bearing pool empoldering, which is focused by oil & gas development engineers. Distribution and quality of the reservoir determine the fluid reservoir spatial distribution. Ultimately they make a direct impact on gas production and gas reservoir development. Physical property of gas pool in Shanxi-2 formation in Zizhou area has a great change, and it has more lithologic character. it is a typical low-porosity & low-permeability formation, which is controlled by structure and lithology. Based on the effective identification of the reservoir, It is useful for high efficient and reasonable development through effective liquid identification and confirming liquid distribution.
Considering the geologic characteristics, based on traditional logging data and getting logging properties of sand, shale, and coal from description of rock core, this paper has scientifically and effectively identified lithology with multiple discriminant analysis and BP neural network, and analyzed the distribution of sand body after identify the sand of single well. This paper has established log interpretation modeling through fitting the logging properties and rock core analytic data, such as porosity, permeability and water saturation. At the same time, the author tries to study pay formation liquid log response and discrimination via some different methods, such as fluid acoustic impedance, multiple discriminant analysis and support vector machine, based on different production formation criterion and selects some applicable methods. At last, according to water chemistry and production data, the paper establishes the identification criterion for different water, and further points out the rich water area, the altitude of water contact, and the management of production wells where are nearby the rich water area.
The results have shown that Shan-23 is the major productive formation. Multiple discriminant analysis and BP neural network are useful for the identification of the lithology. And the log interpretation modeling is effective. Multiple discriminant analysis, fluid acoustic impedance and support vector machine are the recommended practicable methods for Zi-Zhou reservoir during all the discussed methods. In addition, this paper sums up three water output mechanisms of lenticular water, tight water and fringe/basal water combined production performance data of 53 brought-in wells that are divided to raised, sustained, and descendant water/gas ratio, and global water chemical analysis data, and works out the recognition criteria of different water. In addition, the paper points that there are three rich water areas during the reach area and there is different characteristic for different water of the three rich water areas. Besides, the paper is also points that the total salinity of fringe/basal water was not different from lenticular water and tight water because of the condensed water. Calling on water saturation, this paper has discussed the relationship between production water and production gas. Considering the relationship, this paper has built the gas field modeling containing fringe/basal water, calculated the feasible well diameter, output and drawdown pressure. According to the prior study, this paper researches the liquid distribution based on gas-water section, and points that the gas reservoir does not have united gas/water surface, and the surface differed from west to east. Simultaneously, the paper points the location of no-water and potential-water. At last, the paper generalizes four gas/water distribution patterns and proposes the concept that the formation water salinity is affected by coal water.
Sand body and liquid distribution of Shan-2 gas reservoir have become clearer than ever before through this research, which also provide a reliable basis for gas reservoir exploitation, prolonging water free gas recovery and stable production period of gas well, enhancing gas recovery, etc, in further research.
引文
[1]贺自新等.鄂尔多斯盆地演化与油气[M].北京:石油工业出版社.2003
[2]戴金星,邹才能等.中国大气田形成条件和主控因素[J].天然气地球化学,2007,18(4):473-484.
[3]李贤庆等著.鄂尔多斯盆地中部气田地层流体特征与天然气成藏[M].北京:地质出版社,1997.
[4]刘方魁等编著.油气田水文地质学原理[M].北京:石油工业出版社,1991.
[5]Collins A G.1975.Geochemistry of oilfield waters,chap.1,New York:W.H.Freeman and Company,267-269.
[6]高锡兴.中国含油气盆地油田水[M].北京:石油工业出版社,1994.
[7]刘崇禧.我国陆相盆地油田水文地球化学特征[J].地球化学,1982,11(2):190-197
[8]杨忠辉等,油气田的水文地球化学标志及其应用[J].石油天然气地质,1982,3(4):327-334.
[9]张金来.我国油田水的基本特征及其分类讨论[J].地质论评,1979(2)64-69.
[10]张金来.中国东部中、新生代陆相广盆与槽盆油田水化学场比较[J].石油与天然气地质,1983,4(2):222-228.
[11]汪义先.泌阳凹陷油田水地球化学特征及其与油气的关系[J].石油实验地质,1983,5(4):298-303.
[12]秦成明,应丹琳.四川盆地二叠系、上三叠系水文地质垂直分带及预测模型[J].石油勘探与开发,23(1):30-32.
[13]汪蕴璞,林锦璇等.论含油气盆地含水系统和水文地质期的划分[J].地球科学,1995,20(4):393-398.
[14]楼章华,金爱民等.海相地下水文地球化学与油气保存条件评价[J].浙江大学学报(工学版),2006,40(3):501-505.
[15]宋子齐,袁福文.一种评价储层的多参数优先相似选择法[J].西安石油学院学报,1988,3(3):9-17.
[16]仲自勉,陈宝树等.用模糊聚类分析技术识别储层[J].测井技术,1995,19(1):28-34.
[17]李亚林,雷晓等.用神经网络方法识别碳酸盐裂缝系统中的气与水[J].天然气工业,2000,20(1):32-36.
[18]Nestor A Rivera,Shubhankar Ray.Well-log Feature Extraction Using Wavelets and Genetic Algorithms[J].AAPG Annual Meeting,2002.
[19]卢颖忠,李保华等.测井综合特征参数在碳酸盐岩储层识别中的应用[J].中国西部油气地质,2006,2(1):109-113.
[20]张松扬.致密砂岩气层测井特征参数提取方法研究及识别评价技术[D].北京:中国石油大学工学博士学位论文,2007.
[21]龙建勋.鄂尔多斯盆地子洲地区下二叠统山西组二段地层水及气水分布规律研究[D].成都:成都理工大学能源学院,2008.
[22]付新华,魏新善等.鄂尔多斯盆地天然气勘探形势与发展前景[J].石油学 报,2006,27(6):1-4,13.
[23]赵贤正,李景明等.中国天然气勘探快速发展的十年[M].北京:石油工业出版社,2002,9-16.
[24]马新华.鄂尔多斯盆地天然气勘探开发形势分析[J].石油勘探与开发,2005,32(4):50-53.
[25]戴金星,邹才能等.中国大气田形成条件和主控因素[J].天然气地球科学,2007,18(4):473-484.
[26]周文等.鄂尔多斯盆地靖边气田北二区马五_1气藏产水机理研究[R].成都:成都理工大学能源学院,2003.
[27]李仲冬等.榆林子洲地区山2段地层水成因及分布规律研究[R].成都:成都理工大学能源学院,2008.
[28](美)P.A.迪基.石油开发地质学[M].北京:石油工业出版社,1982.
[29]Dickey,P.A,and Carlos Soto R,1974,Chemical composition of deep subsurface waters of the western Anadarko Basin:Soc Petrol.Eng.Pap.SPE5178,18P.
[30](美)柯林斯著.油田水地球化学[M].北京:石油工业出版社,1984.
[31]徐国盛.含油气系统中的古水文地质分析—以陕甘宁盆地和四川盆地为例[D].成都:成都理工学院博士学位论文,1998.
[32]库丽曼·木沙太.川中地区上三叠统须家河组四段气藏成藏机理研究[D].成都:成都理工大学博士学位论文,2007.
[33](美)J.W.哈博,D.F.梅里亚姆.电子计算机在地层分析中的应用[M].科学出版社,1978.
[34]陈力.D函数规律在油气综合化探中的应用.物探与化探[J],1988,12(2):114-121.
[35]Wayne A.Van Voast(美).与煤层气有关的地层水的地球化学特征[J].国外油气地质信息,2004,3:24-31.
[36]雍世和,张超谟.测井数据处理与综合解释[M].北京:石油大学出版社,1996.
[37]周文,童孝华.地球物理测井原理及应用[M].成都理工大学石油工程内部教材.
[38]杜奉屏.油矿地球物理测井[M].地质出版社,1984.
[39]Vapnik V.统计学习理论[M].许建华,张学工译.北京:电子工业出版社,2004.274-420.
[40]Cristiannini N,Shawe-Taylor J.支持向量机导论[M].李国正,王猛,曾华军译.北京:电子工业出版社,2004.24-108.
[41]Burges C.A tutorial on support vector machines for pattern recongnition[J].Data Mining and Knowledge Discovery,1998,2(2):121-167.
[42]朱国强,刘士荣,俞金寿.支持向量机及其在函数逼近中的应用[J].华东理工大学学报,2002,28(5):555-559.
[43]乐友善,袁全杜.支持向量机方法在储层预测中的应用[J].石油物探,2005,44(4):388-392.
[44]刘得军,冉群英,王斌.支持向量机在大庆齐家凹陷测井解释中的应用[J].石油物探,2007,46(2):156-161.
[45]彭涛,张翔.支持向量机及其在石油勘探开发中的应用综述[J].勘探地球物理进展,2007,30(2):91-95.
[46]Vapnik V.The nature of statistical learning theory[M].New York:Springer,1995.54-96.
[47]Weston J，Watkins C．Multi-class support vector machines[D]．Royal Holloway College，1998．
[48]黎华继，张晟等．新场气田须二气藏气水分布探讨[J]．天然气技术，2008，2(2)：28-31．
[49]庄惠农．气藏动态描述与试井[M]．北京：石油工业出版社，2004．
[50]尹晓贺．川中地区上三叠统须家河组埋藏史恢复及重点构造气水分布控制因素探讨[D]．成都：西南石油大学硕士学位论文，2006．
[51]符勇．论油气成藏的水动力作用[J]．新疆石油地质，2005，26(5)：517-519．
[52]符勇，李世英．石油水文地质学简介[J]．石油知识，1991，(2)：28-29．
[53]杨川东．优选管柱排水采气工艺技术的发展及成效[J]．钻采工艺，1997，20(3)：33-37．
[54]杨桦，杨川东．优选管柱排水采气工艺的理论研究[J]．西南石油学院学报，1994，16(4)：56-65．
[55]杨川东．采气工程[M]．北京：石油工业出版社，2004．
[56]黄炳光，刘蜀知等．气藏工程与动态分析方法[M]．北京：石油工业出版社，2004．
[57]李晓平，王会强．边水气藏气井合理产量生产压差及产量的确定[J]．天然气工业，2008，28(7)：85-86．
[58]陈元千．油气藏工程计算方法[M]．北京：石油工业出版社，1990．
[59]李传亮，宋洪才等．带隔板底水气藏油井临界产量计算公式[J]．大庆石油地质与开发，1993，12(4)：43-46．
[60]李传亮．修正DUPUIT临界产量公式[J]．石油勘探与开发，1993，20(4)：91-95．
[61]R．D．Oden and Jennings，Modification of the Cullender and Smith Equation More Accurate Bottomhole Pressure Calulations in Gas Wells[J]．SPE 17306．
[62]姜亦栋，杜殿发等．底水油藏采水注销机理研究[J]．胜利油田职工大学学报，2008，22(5)：42-44．
[63]杨秉乾译．用气水同采工艺提高水驱气藏采收率的可行性研究[M]．北京：油气田开发工程译丛，1989．
[64]何顺利等．子洲气田主力气层气藏描述[R]．西安：中国石油长庆油田分公司(内部资料)，2008．
[2]戴金星,邹才能等.中国大气田形成条件和主控因素[J].天然气地球化学,2007,18(4):473-484.
[3]李贤庆等著.鄂尔多斯盆地中部气田地层流体特征与天然气成藏[M].北京:地质出版社,1997.
[4]刘方魁等编著.油气田水文地质学原理[M].北京:石油工业出版社,1991.
[5]Collins A G.1975.Geochemistry of oilfield waters,chap.1,New York:W.H.Freeman and Company,267-269.
[6]高锡兴.中国含油气盆地油田水[M].北京:石油工业出版社,1994.
[7]刘崇禧.我国陆相盆地油田水文地球化学特征[J].地球化学,1982,11(2):190-197
[8]杨忠辉等,油气田的水文地球化学标志及其应用[J].石油天然气地质,1982,3(4):327-334.
[9]张金来.我国油田水的基本特征及其分类讨论[J].地质论评,1979(2)64-69.
[10]张金来.中国东部中、新生代陆相广盆与槽盆油田水化学场比较[J].石油与天然气地质,1983,4(2):222-228.
[11]汪义先.泌阳凹陷油田水地球化学特征及其与油气的关系[J].石油实验地质,1983,5(4):298-303.
[12]秦成明,应丹琳.四川盆地二叠系、上三叠系水文地质垂直分带及预测模型[J].石油勘探与开发,23(1):30-32.
[13]汪蕴璞,林锦璇等.论含油气盆地含水系统和水文地质期的划分[J].地球科学,1995,20(4):393-398.
[14]楼章华,金爱民等.海相地下水文地球化学与油气保存条件评价[J].浙江大学学报(工学版),2006,40(3):501-505.
[15]宋子齐,袁福文.一种评价储层的多参数优先相似选择法[J].西安石油学院学报,1988,3(3):9-17.
[16]仲自勉,陈宝树等.用模糊聚类分析技术识别储层[J].测井技术,1995,19(1):28-34.
[17]李亚林,雷晓等.用神经网络方法识别碳酸盐裂缝系统中的气与水[J].天然气工业,2000,20(1):32-36.
[18]Nestor A Rivera,Shubhankar Ray.Well-log Feature Extraction Using Wavelets and Genetic Algorithms[J].AAPG Annual Meeting,2002.
[19]卢颖忠,李保华等.测井综合特征参数在碳酸盐岩储层识别中的应用[J].中国西部油气地质,2006,2(1):109-113.
[20]张松扬.致密砂岩气层测井特征参数提取方法研究及识别评价技术[D].北京:中国石油大学工学博士学位论文,2007.
[21]龙建勋.鄂尔多斯盆地子洲地区下二叠统山西组二段地层水及气水分布规律研究[D].成都:成都理工大学能源学院,2008.
[22]付新华,魏新善等.鄂尔多斯盆地天然气勘探形势与发展前景[J].石油学 报,2006,27(6):1-4,13.
[23]赵贤正,李景明等.中国天然气勘探快速发展的十年[M].北京:石油工业出版社,2002,9-16.
[24]马新华.鄂尔多斯盆地天然气勘探开发形势分析[J].石油勘探与开发,2005,32(4):50-53.
[25]戴金星,邹才能等.中国大气田形成条件和主控因素[J].天然气地球科学,2007,18(4):473-484.
[26]周文等.鄂尔多斯盆地靖边气田北二区马五_1气藏产水机理研究[R].成都:成都理工大学能源学院,2003.
[27]李仲冬等.榆林子洲地区山2段地层水成因及分布规律研究[R].成都:成都理工大学能源学院,2008.
[28](美)P.A.迪基.石油开发地质学[M].北京:石油工业出版社,1982.
[29]Dickey,P.A,and Carlos Soto R,1974,Chemical composition of deep subsurface waters of the western Anadarko Basin:Soc Petrol.Eng.Pap.SPE5178,18P.
[30](美)柯林斯著.油田水地球化学[M].北京:石油工业出版社,1984.
[31]徐国盛.含油气系统中的古水文地质分析—以陕甘宁盆地和四川盆地为例[D].成都:成都理工学院博士学位论文,1998.
[32]库丽曼·木沙太.川中地区上三叠统须家河组四段气藏成藏机理研究[D].成都:成都理工大学博士学位论文,2007.
[33](美)J.W.哈博,D.F.梅里亚姆.电子计算机在地层分析中的应用[M].科学出版社,1978.
[34]陈力.D函数规律在油气综合化探中的应用.物探与化探[J],1988,12(2):114-121.
[35]Wayne A.Van Voast(美).与煤层气有关的地层水的地球化学特征[J].国外油气地质信息,2004,3:24-31.
[36]雍世和,张超谟.测井数据处理与综合解释[M].北京:石油大学出版社,1996.
[37]周文,童孝华.地球物理测井原理及应用[M].成都理工大学石油工程内部教材.
[38]杜奉屏.油矿地球物理测井[M].地质出版社,1984.
[39]Vapnik V.统计学习理论[M].许建华,张学工译.北京:电子工业出版社,2004.274-420.
[40]Cristiannini N,Shawe-Taylor J.支持向量机导论[M].李国正,王猛,曾华军译.北京:电子工业出版社,2004.24-108.
[41]Burges C.A tutorial on support vector machines for pattern recongnition[J].Data Mining and Knowledge Discovery,1998,2(2):121-167.
[42]朱国强,刘士荣,俞金寿.支持向量机及其在函数逼近中的应用[J].华东理工大学学报,2002,28(5):555-559.
[43]乐友善,袁全杜.支持向量机方法在储层预测中的应用[J].石油物探,2005,44(4):388-392.
[44]刘得军,冉群英,王斌.支持向量机在大庆齐家凹陷测井解释中的应用[J].石油物探,2007,46(2):156-161.
[45]彭涛,张翔.支持向量机及其在石油勘探开发中的应用综述[J].勘探地球物理进展,2007,30(2):91-95.
[46]Vapnik V.The nature of statistical learning theory[M].New York:Springer,1995.54-96.
[47]Weston J，Watkins C．Multi-class support vector machines[D]．Royal Holloway College，1998．
[48]黎华继，张晟等．新场气田须二气藏气水分布探讨[J]．天然气技术，2008，2(2)：28-31．
[49]庄惠农．气藏动态描述与试井[M]．北京：石油工业出版社，2004．
[50]尹晓贺．川中地区上三叠统须家河组埋藏史恢复及重点构造气水分布控制因素探讨[D]．成都：西南石油大学硕士学位论文，2006．
[51]符勇．论油气成藏的水动力作用[J]．新疆石油地质，2005，26(5)：517-519．
[52]符勇，李世英．石油水文地质学简介[J]．石油知识，1991，(2)：28-29．
[53]杨川东．优选管柱排水采气工艺技术的发展及成效[J]．钻采工艺，1997，20(3)：33-37．
[54]杨桦，杨川东．优选管柱排水采气工艺的理论研究[J]．西南石油学院学报，1994，16(4)：56-65．
[55]杨川东．采气工程[M]．北京：石油工业出版社，2004．
[56]黄炳光，刘蜀知等．气藏工程与动态分析方法[M]．北京：石油工业出版社，2004．
[57]李晓平，王会强．边水气藏气井合理产量生产压差及产量的确定[J]．天然气工业，2008，28(7)：85-86．
[58]陈元千．油气藏工程计算方法[M]．北京：石油工业出版社，1990．
[59]李传亮，宋洪才等．带隔板底水气藏油井临界产量计算公式[J]．大庆石油地质与开发，1993，12(4)：43-46．
[60]李传亮．修正DUPUIT临界产量公式[J]．石油勘探与开发，1993，20(4)：91-95．
[61]R．D．Oden and Jennings，Modification of the Cullender and Smith Equation More Accurate Bottomhole Pressure Calulations in Gas Wells[J]．SPE 17306．
[62]姜亦栋，杜殿发等．底水油藏采水注销机理研究[J]．胜利油田职工大学学报，2008，22(5)：42-44．
[63]杨秉乾译．用气水同采工艺提高水驱气藏采收率的可行性研究[M]．北京：油气田开发工程译丛，1989．
[64]何顺利等．子洲气田主力气层气藏描述[R]．西安：中国石油长庆油田分公司(内部资料)，2008．