油田混层开采产能动态监测技术
|
摘要
油田混层开采产能动态监测技术研究手段非常丰富，包括气相色谱指纹法、液相色谱法、紫外光谱法、同位素法、色谱-质谱法等。由于油田混层开采技术对样品采集的要求低，不影响原油的生产，具有快速和低成本的特点，可与常规测井手段配合使用，其应用前景十分广阔。
色谱指纹技术是油藏地球化学研究的重要技术之一，对于研究油田横向和垂向流体连通性、流体单元划分、合采油井分层产量贡献确定、定量监测剩余油空间分布等具有重要意义。随着油田勘探开发的深入，油田生产面临诸如产能下降、含水升高、层间层内矛盾突出、剩余油分布状况不清以及多层合采产能贡献计算等问题，尤其多层合采产能贡献研究更是油藏管理者一直关心的核心问题之一。这就需要深入研究色谱指纹技术在油田开发动态监测中的应用。
本论文在运用色谱分析研究原油全烃技术理论的基础上，通过大量的单层原油和混合原油的模拟配比实验，运用色谱烃指纹峰高比、内标法绝对定量分析等方法建立了化学模型，利用非线性神经网络方法和支持向量机(Support VectorRegression SVR)建立了多层混采原油分层产能贡献数学模型，为实现油田混层开采产能动态监测提供了方法。
本论文重点对文昌13-1／2油田的混采油井产能贡献动态监测进行了研究。首先，注重了解油田的地质现状，对一个区块进行油井分层产能分配计算前，必须根据精细地质、油藏地球化学、油田开发设计等调研工作，以掌握各油井油层的连通状况；掌握油田各油层受储层本身厚度、孔隙度、渗透率等非均质性的影响因素，避免因为这些因素的误判所带来的计算误差。有机相和源岩成熟度的变化决定了石油组成差异，导致了油藏单元之间色谱指纹不同，因此原油色谱指纹的变化是原油非均质性的具体表现。其次，研究使用了先进的分析仪器，由于色谱检测系统运行的稳定性对于分析试验结果的可靠性至关重要，从而影响到分层原油产能配比的准确性，因此，在改进的色谱指纹分析方法中，选用了世界先进的气相色谱仪和化学工作站系统，并采用了自动进样器进样，重新摸索了仪器分析实验条件，使仪器检测精度有了新的提高，保证了分析数据的准确性。再次，研究针对不同混采条件建立了本区多层混采油井分层产量贡献的判别模型。对于两层合采问题的数学模拟计算，一般用线性拟合的方法，即选取相关性好的峰高比特征指纹参数建立化学和数学模型，其数学模型问题基本解决。对于三层及以上多层合采问题，分层原油与合采原油烃指纹浓度之间已不是简单的线性关系，而是一种复杂的曲线性关系。因而，必须采用指纹绝对定量和非线性模型来实现对多层合采原油分层贡献的计算。
本研究应用神经网络和支持向量机分别建立了非线性数学模型及软件，应用地球化学指纹绝对定量方法对文昌13-1/2油田混采油井产能分配进行动态监测，计算了5口混采油井的分层产能贡献，其解释结果与停产测试结果基本一致，为该油田制定生产调整方案提供理论依据。
There are a variety of techniques employed to geochemical research of oil production allocation, including gas chromatographic fingerprinting, liquid chromatographic, ultraviolet spectrum, stable isotope analysis, GC-MS method, etc. The geochemical techniques for research of oil production allocation are generally characterized by quick, low cost and low sampling requirements, which can be easily combined with conventional logging. Therefore, the geochemical techniques have a very bright prospect in terms of its application.
Gas chromatographic fingerprinting is one of important techniques for oil reservoir geochemistry, which have been proved to be significant for the horizontal and vertical connectivity of fluids, division of fluid units, determination of oil production allocation, monitoring of the remnant oil distribution and so on. With the exploration and development of an oilfield, many issues will be encountered in oil production, such as decrease in production potential, water-contained oil increasing, contradictions of production allocation between/within producing zones, the unknown distribution of the remanent oil, as well as the contributions of the oil-producing zones in hybrid production well. In particular, the research for the production allocation in multiple-zone production well has been one of the core issues and has paid widespread attention from the managers of oil reservoirs, consequently an in-depth study is required on the application of GC fingerprinting technique in dynamic monitor of oilfield development.
In the thesis the mathematic models to compute the production allocation of hybrid producing wells are established and the dynamic monitoring technique is provided for multiple-zone production in an oilfield by using non-linear artificial neural networks and support vector regression, which are made based on both technical and theoretical study of total hydrocarbons GC fingerprint, by means of compound-specific peak height ratio, absolute quantitative analysis with internal standards, and a great number of matching experiments of crude oils from individual zone.
The models and technique are applied to Wenchang 13-1/2 oilfield to monitor the dynamic production allocation of hybrid recovery wells and calculate contribution of individual producing zone in hybrid recovery wells.
First of all, the study was focused on geological investigation of the current status of the oilfield. Before calculation of production allocation of different producing zones in a block was performed, a detail investigation must be made, including the status of connectivity between zones in the oilfield, the geology and geochemistry of the oil reservoirs and the design/development history of the oilfield in order to well understand such factors as thickness, porosity and permeability of the reservoirs that may affect the oil-producing zones in the oilfield. In addition, GC fingerprint features of crude oil are affected by the organic facies and maturity of source rocks, which will result in the difference in crude oil compositions and GC fingerprint features. Therefore, GC fingerprint features represent the inhomogeneity of the crude oil from individual producing zone.
Secondly, that modern instrument is employed is very important for the stability and the reliability of the analytical results. This is a key for the accuracy calculation of production allocation from different oil-producing zones in the hybrid co-recovery wells. Thus, a state-of-the-art gas chromatograph, chemical workstation system and an automatic sample-feeding machine are employed. In this way the precision for detection of the instruments has been greatly improved to enhance the accuracy and reliability of the analytical data.
Finally, different fitting models and geochemical fingerprint features are used to predict the contribution of individual producing zone according to hybrid recovery conditions. For the recovery from two producing zones, the linear fitting and the peak height ratios of geochemical fingerprints are employed and the satisfied results can be obtained. For the hybrid recovery of multiple (3 and more than 3) producing zones, however, the fingerprints of the crude oils from individual producing zone show a complex, non-linear relationships, it can't be fit by a simple, linear function. Thus, absolute quantification fingerprints and non-linear fitting have to use for calculation of the production allocation from multiple producing zones.
The BP nueral networks, SVR and software are here developed. Together with the absolute quantification fingerprints, BP and SVR are applied to monitoring the dynamic production allocation of the hybrid recovery wells in Wenchang 13-1/2 oilfield. The production allocations of 5 hybrid recovery wells are computed, the results showing that the predicted values are basically consistent with those measured by stop production. This provides a theoretical support and applied technique for adjusting production allocation in this oilfield.
色谱指纹技术是油藏地球化学研究的重要技术之一，对于研究油田横向和垂向流体连通性、流体单元划分、合采油井分层产量贡献确定、定量监测剩余油空间分布等具有重要意义。随着油田勘探开发的深入，油田生产面临诸如产能下降、含水升高、层间层内矛盾突出、剩余油分布状况不清以及多层合采产能贡献计算等问题，尤其多层合采产能贡献研究更是油藏管理者一直关心的核心问题之一。这就需要深入研究色谱指纹技术在油田开发动态监测中的应用。
本论文在运用色谱分析研究原油全烃技术理论的基础上，通过大量的单层原油和混合原油的模拟配比实验，运用色谱烃指纹峰高比、内标法绝对定量分析等方法建立了化学模型，利用非线性神经网络方法和支持向量机(Support VectorRegression SVR)建立了多层混采原油分层产能贡献数学模型，为实现油田混层开采产能动态监测提供了方法。
本论文重点对文昌13-1／2油田的混采油井产能贡献动态监测进行了研究。首先，注重了解油田的地质现状，对一个区块进行油井分层产能分配计算前，必须根据精细地质、油藏地球化学、油田开发设计等调研工作，以掌握各油井油层的连通状况；掌握油田各油层受储层本身厚度、孔隙度、渗透率等非均质性的影响因素，避免因为这些因素的误判所带来的计算误差。有机相和源岩成熟度的变化决定了石油组成差异，导致了油藏单元之间色谱指纹不同，因此原油色谱指纹的变化是原油非均质性的具体表现。其次，研究使用了先进的分析仪器，由于色谱检测系统运行的稳定性对于分析试验结果的可靠性至关重要，从而影响到分层原油产能配比的准确性，因此，在改进的色谱指纹分析方法中，选用了世界先进的气相色谱仪和化学工作站系统，并采用了自动进样器进样，重新摸索了仪器分析实验条件，使仪器检测精度有了新的提高，保证了分析数据的准确性。再次，研究针对不同混采条件建立了本区多层混采油井分层产量贡献的判别模型。对于两层合采问题的数学模拟计算，一般用线性拟合的方法，即选取相关性好的峰高比特征指纹参数建立化学和数学模型，其数学模型问题基本解决。对于三层及以上多层合采问题，分层原油与合采原油烃指纹浓度之间已不是简单的线性关系，而是一种复杂的曲线性关系。因而，必须采用指纹绝对定量和非线性模型来实现对多层合采原油分层贡献的计算。
本研究应用神经网络和支持向量机分别建立了非线性数学模型及软件，应用地球化学指纹绝对定量方法对文昌13-1/2油田混采油井产能分配进行动态监测，计算了5口混采油井的分层产能贡献，其解释结果与停产测试结果基本一致，为该油田制定生产调整方案提供理论依据。
There are a variety of techniques employed to geochemical research of oil production allocation, including gas chromatographic fingerprinting, liquid chromatographic, ultraviolet spectrum, stable isotope analysis, GC-MS method, etc. The geochemical techniques for research of oil production allocation are generally characterized by quick, low cost and low sampling requirements, which can be easily combined with conventional logging. Therefore, the geochemical techniques have a very bright prospect in terms of its application.
Gas chromatographic fingerprinting is one of important techniques for oil reservoir geochemistry, which have been proved to be significant for the horizontal and vertical connectivity of fluids, division of fluid units, determination of oil production allocation, monitoring of the remnant oil distribution and so on. With the exploration and development of an oilfield, many issues will be encountered in oil production, such as decrease in production potential, water-contained oil increasing, contradictions of production allocation between/within producing zones, the unknown distribution of the remanent oil, as well as the contributions of the oil-producing zones in hybrid production well. In particular, the research for the production allocation in multiple-zone production well has been one of the core issues and has paid widespread attention from the managers of oil reservoirs, consequently an in-depth study is required on the application of GC fingerprinting technique in dynamic monitor of oilfield development.
In the thesis the mathematic models to compute the production allocation of hybrid producing wells are established and the dynamic monitoring technique is provided for multiple-zone production in an oilfield by using non-linear artificial neural networks and support vector regression, which are made based on both technical and theoretical study of total hydrocarbons GC fingerprint, by means of compound-specific peak height ratio, absolute quantitative analysis with internal standards, and a great number of matching experiments of crude oils from individual zone.
The models and technique are applied to Wenchang 13-1/2 oilfield to monitor the dynamic production allocation of hybrid recovery wells and calculate contribution of individual producing zone in hybrid recovery wells.
First of all, the study was focused on geological investigation of the current status of the oilfield. Before calculation of production allocation of different producing zones in a block was performed, a detail investigation must be made, including the status of connectivity between zones in the oilfield, the geology and geochemistry of the oil reservoirs and the design/development history of the oilfield in order to well understand such factors as thickness, porosity and permeability of the reservoirs that may affect the oil-producing zones in the oilfield. In addition, GC fingerprint features of crude oil are affected by the organic facies and maturity of source rocks, which will result in the difference in crude oil compositions and GC fingerprint features. Therefore, GC fingerprint features represent the inhomogeneity of the crude oil from individual producing zone.
Secondly, that modern instrument is employed is very important for the stability and the reliability of the analytical results. This is a key for the accuracy calculation of production allocation from different oil-producing zones in the hybrid co-recovery wells. Thus, a state-of-the-art gas chromatograph, chemical workstation system and an automatic sample-feeding machine are employed. In this way the precision for detection of the instruments has been greatly improved to enhance the accuracy and reliability of the analytical data.
Finally, different fitting models and geochemical fingerprint features are used to predict the contribution of individual producing zone according to hybrid recovery conditions. For the recovery from two producing zones, the linear fitting and the peak height ratios of geochemical fingerprints are employed and the satisfied results can be obtained. For the hybrid recovery of multiple (3 and more than 3) producing zones, however, the fingerprints of the crude oils from individual producing zone show a complex, non-linear relationships, it can't be fit by a simple, linear function. Thus, absolute quantification fingerprints and non-linear fitting have to use for calculation of the production allocation from multiple producing zones.
The BP nueral networks, SVR and software are here developed. Together with the absolute quantification fingerprints, BP and SVR are applied to monitoring the dynamic production allocation of the hybrid recovery wells in Wenchang 13-1/2 oilfield. The production allocations of 5 hybrid recovery wells are computed, the results showing that the predicted values are basically consistent with those measured by stop production. This provides a theoretical support and applied technique for adjusting production allocation in this oilfield.
引文
侯读杰，张敏，赵红静．油藏及开发地球化学导论[M]．北京：石油工业出版社，2001．
尹太举，白全明．混采井产量贡献估算[J]．新疆石油天然气质，2005，1(3)：47-62．
魏福军，张云霞．油藏地球化学新技术及其在西达里亚油田开发中的应用前景[J]．石油实验地质，1998，20(3)：257-260．
金晓辉，朱丹，林壬子．油藏色谱指纹非均质性形成机理及其稳定性实验模拟[J]．地质评论．2003，49(2)：217-221
常象春．张忮焕．王铁冠．混采油井产能贡献的油藏地球化学研究[J]．西安石油学报(自然科学版)．2000，15(4)：35-38
何文祥．王培荣．刘逸．单层原油配比实验图版及其在混采油井中的应用[J]．石油勘探与开发．2001，28(6)：82-83
黄保家．李旭红．陈飞雄．地球化学指纹技术在油气藏连通性及配产研究中的应用[J]．中国海上油气(地质)，2002，12(5)：302-308
朱俊章，侯读杰，黄保家．文昌13-1油田部分混采油井产能分配计算[J]．天然气地球科学．2004，15(2)：187-190
林青，孙福街，黄保家．由多甲基萘参数计算混采井单层产能贡献[J]．中国海上油气．2005，17(5)：316-331
金晓辉．朱丹．林壬子．油田开发动态色谱指纹监测技术的数学模拟研究[J]．石油勘探与开发．2001，28(3)：86-89
金晓辉．朱丹．林壬子．原油气相色谱指纹可配比性实验研究[J]．石油实验地质．2003，25(13)：53-92
冯子辉，张居和，关秋华．用族组成和全烃色谱确定原油单井产能配比[J]．石油仪器．1999，13(4)40-42
邹宇峥．蔡元明．马亭等．如何消除应用于混采井原油产量分配计算的气相色谱指纹技术存在的缺陷[J]．石油实验地质．2001，23(2)：212 220
陈世加，陈小凡．混采井分层产能确定的新方法[J]．西南石油学院学报．2001，23(6)：31-33
刘玉海，王城镇，李国军．色谱指纹技术在确定混采井单层产能中的应用[J]．内蒙古石油化工．2005，(1)：92-93
张居和．方伟．冯子辉等．多层混采原油分层产能贡献监测色谱技术[J]．石油学报．2004，25(4)：75-79
王跃文．卢双舫．方伟．多层混采产能配比的算法研究及应用[J]．石油实验地质．2005，27(6)：630-634
陈永红．地球化学指纹技术在油藏动态监测中的应用[J]．山东科技大学学报(自然科学版)．2005，24(1)：49-52
邹宇峥，马亭，梅博文．混采井原油产量分配计算的新技术——紫外光谱法[J]．石油与天然气地质．2000，21(6)：168-175
王振平，付晓泰，卢双舫．混采油井产量贡献液相色谱监测法[J]．地球化学．2000，29(3)：297-301
王强，付晓文，徐志明．稳定碳同位素在油气地球化学中的应用及存在的问题[J]．天然气地球科学．2005，16(2)：233-237
黄海平，许晓宏．天然气同位素特征及作用[J]．石油与天然气地质．1997，18(2)：136-139
薛磊等．一种基于支持向量回归的多层分层产能贡献预测模型[J]．石油化工高等学报 2006
李俊良等《文昌13-1／2油田油藏地质综合研究》1999
南海西部研究院《原油性质分析报告》1999
苏崇华等《文昌13-1／2油田油藏地质综合研究》1997
苏崇华等《文昌13-1／2油田油藏地质综合研究》2005
林壬子，张敏．油藏地球化学进展[M]．西安：陕西科学技术出版社，1996．245～276
陈世加，林峰，陈小凡．利用原油色谱指纹技术确定混采油井分层产能贡献[J]．石油勘探与开发，1999，26(6)：63-71．
苏秀芳，黄光辉，张敏，朱俊章．有机地球化学分析方法及其应用[D]．塔里木石油勘探开发指挥部，江汉石油学院地球化学研究中心．1999
付晓文，吴刚．原油指纹技术在陆梁油田产能监测中的应用[J]．西南石油学院学报，2002，24(5)：4～6
阎平凡，张长水．人工神经网络与模拟进化计算[M]．北京：清华大学出版社，2003．
Vapnik V．统计学习理论的本质[M]．北京：清华大学出版社，1999．
邓乃扬，田英杰．数据挖掘中的新方法——支持向量机[M]．北京：科学出版社，2004．
杨晓敏，吴炜，黎涛，．基于Gabor变换和支持向量机的车牌字符识别算法[J]．四川大学学报，2005，37(5)：130-134．
付晓文，陈世加，吴刚．原油指纹技术在陆梁油田产能监测中的应用[J]．西南石油学院学报，2002，24(5)：4-6．
Nello Cristianini，John Shawe-Taylor．支持向量机导论[M]．北京：电子工业出版社，2004．
刘广利，杨志民．一种新的支持向量回归预测模型[J]．吉首大学学报，2002，23(3)：28-31．
王延江，杨培杰，史清江．一种基于支撑向量机学习预测井眼轨迹的新方法[J]．石油学报．2005，26(5)：98-101．
代为钢，孔枫等，《文昌13-1／2油田地球化学研究》海西部研究院内部资料2003
S Gunn. Support Vector Machines for Classification and Regression[R].Technical Report. University of Southamptom, 1998.
Kaufman R L, Ahmed A S, Moldowan J M, A new technique for the analysis of commingled oil and its application to production allocation calculations[A]. Proceedings of the 16th Annual convention of Indonesian Petroleum Assoc[C]. Indonesion Petroleum Association, 1987.247-268
Kaufman R L, Ahmed A S, Gas chromatography as a development and production tool for fingerprinting oils from Individual Reservoirs: Applications in the Gulf of Mexico[A]. In: GCS/SEPM Foundation Ninth Annual Research Conference Proceedings, 1990. 263-282
Kaufman R L, Dashti H, Kabir C S, et al. Characterizing the greater Burgan Field: Use of geochemistry and oil fingerprinting [A]. In: SPE Paper No. 37803, 1997.385-394
Mark A. McCaffrey, Henry A. Legarre, and Scott J. Johnson. Using biomarkers to improve heavy oil reservoir management: an example from the Cymic Field, Kern country, California. AAPG Bulletin, v.80, No.6
R.J. Hwang, D.K. Baksin, S.C. Teerman. Allocation of commingled pipeline oils to field production. Organic Geochemistry 31(2000): 1463-1474
Kaufman R L. Gas chromatography as adevelopment and production tool for fingerprinting oils from individual reservoirs: application in the Gulf of Mexico[A]. Gulf Coast Section of the society of Economic pala eontologists and Mineralogists Foundation Ninth Annual Research Conference Proceedings[C]. 1990
Hwang R J, Ahmed A S.oil composition Variation and reservoir continuity: Unity field, Sudan [J]. Organic Geochemistry, 1994,21(2):
England W.A., The organic geochemistry of petroleum reservoirs Org. Geochem. 1990,116(1～3):
Kaufman R L, Ahmed A S, Moldowan J M. A new technique for the analysis of commingled oils and its application to production allocation calculations[A]. Procedings of the Sixteenth Annual Conention of Indonesian Petroleum Association. [C]Indonesian Petroleum Association, 1987,247-268.
Larter S Rand Aplin A C. Reservoir geochemistry :methods,application and opportunities, In:The Cubitt J N and England W A eds.Geochemistry of Reservoir[C] England,Geological Society Special Publication,1995,No.86:5～32.
尹太举，白全明．混采井产量贡献估算[J]．新疆石油天然气质，2005，1(3)：47-62．
魏福军，张云霞．油藏地球化学新技术及其在西达里亚油田开发中的应用前景[J]．石油实验地质，1998，20(3)：257-260．
金晓辉，朱丹，林壬子．油藏色谱指纹非均质性形成机理及其稳定性实验模拟[J]．地质评论．2003，49(2)：217-221
常象春．张忮焕．王铁冠．混采油井产能贡献的油藏地球化学研究[J]．西安石油学报(自然科学版)．2000，15(4)：35-38
何文祥．王培荣．刘逸．单层原油配比实验图版及其在混采油井中的应用[J]．石油勘探与开发．2001，28(6)：82-83
黄保家．李旭红．陈飞雄．地球化学指纹技术在油气藏连通性及配产研究中的应用[J]．中国海上油气(地质)，2002，12(5)：302-308
朱俊章，侯读杰，黄保家．文昌13-1油田部分混采油井产能分配计算[J]．天然气地球科学．2004，15(2)：187-190
林青，孙福街，黄保家．由多甲基萘参数计算混采井单层产能贡献[J]．中国海上油气．2005，17(5)：316-331
金晓辉．朱丹．林壬子．油田开发动态色谱指纹监测技术的数学模拟研究[J]．石油勘探与开发．2001，28(3)：86-89
金晓辉．朱丹．林壬子．原油气相色谱指纹可配比性实验研究[J]．石油实验地质．2003，25(13)：53-92
冯子辉，张居和，关秋华．用族组成和全烃色谱确定原油单井产能配比[J]．石油仪器．1999，13(4)40-42
邹宇峥．蔡元明．马亭等．如何消除应用于混采井原油产量分配计算的气相色谱指纹技术存在的缺陷[J]．石油实验地质．2001，23(2)：212 220
陈世加，陈小凡．混采井分层产能确定的新方法[J]．西南石油学院学报．2001，23(6)：31-33
刘玉海，王城镇，李国军．色谱指纹技术在确定混采井单层产能中的应用[J]．内蒙古石油化工．2005，(1)：92-93
张居和．方伟．冯子辉等．多层混采原油分层产能贡献监测色谱技术[J]．石油学报．2004，25(4)：75-79
王跃文．卢双舫．方伟．多层混采产能配比的算法研究及应用[J]．石油实验地质．2005，27(6)：630-634
陈永红．地球化学指纹技术在油藏动态监测中的应用[J]．山东科技大学学报(自然科学版)．2005，24(1)：49-52
邹宇峥，马亭，梅博文．混采井原油产量分配计算的新技术——紫外光谱法[J]．石油与天然气地质．2000，21(6)：168-175
王振平，付晓泰，卢双舫．混采油井产量贡献液相色谱监测法[J]．地球化学．2000，29(3)：297-301
王强，付晓文，徐志明．稳定碳同位素在油气地球化学中的应用及存在的问题[J]．天然气地球科学．2005，16(2)：233-237
黄海平，许晓宏．天然气同位素特征及作用[J]．石油与天然气地质．1997，18(2)：136-139
薛磊等．一种基于支持向量回归的多层分层产能贡献预测模型[J]．石油化工高等学报 2006
李俊良等《文昌13-1／2油田油藏地质综合研究》1999
南海西部研究院《原油性质分析报告》1999
苏崇华等《文昌13-1／2油田油藏地质综合研究》1997
苏崇华等《文昌13-1／2油田油藏地质综合研究》2005
林壬子，张敏．油藏地球化学进展[M]．西安：陕西科学技术出版社，1996．245～276
陈世加，林峰，陈小凡．利用原油色谱指纹技术确定混采油井分层产能贡献[J]．石油勘探与开发，1999，26(6)：63-71．
苏秀芳，黄光辉，张敏，朱俊章．有机地球化学分析方法及其应用[D]．塔里木石油勘探开发指挥部，江汉石油学院地球化学研究中心．1999
付晓文，吴刚．原油指纹技术在陆梁油田产能监测中的应用[J]．西南石油学院学报，2002，24(5)：4～6
阎平凡，张长水．人工神经网络与模拟进化计算[M]．北京：清华大学出版社，2003．
Vapnik V．统计学习理论的本质[M]．北京：清华大学出版社，1999．
邓乃扬，田英杰．数据挖掘中的新方法——支持向量机[M]．北京：科学出版社，2004．
杨晓敏，吴炜，黎涛，．基于Gabor变换和支持向量机的车牌字符识别算法[J]．四川大学学报，2005，37(5)：130-134．
付晓文，陈世加，吴刚．原油指纹技术在陆梁油田产能监测中的应用[J]．西南石油学院学报，2002，24(5)：4-6．
Nello Cristianini，John Shawe-Taylor．支持向量机导论[M]．北京：电子工业出版社，2004．
刘广利，杨志民．一种新的支持向量回归预测模型[J]．吉首大学学报，2002，23(3)：28-31．
王延江，杨培杰，史清江．一种基于支撑向量机学习预测井眼轨迹的新方法[J]．石油学报．2005，26(5)：98-101．
代为钢，孔枫等，《文昌13-1／2油田地球化学研究》海西部研究院内部资料2003
S Gunn. Support Vector Machines for Classification and Regression[R].Technical Report. University of Southamptom, 1998.
Kaufman R L, Ahmed A S, Moldowan J M, A new technique for the analysis of commingled oil and its application to production allocation calculations[A]. Proceedings of the 16th Annual convention of Indonesian Petroleum Assoc[C]. Indonesion Petroleum Association, 1987.247-268
Kaufman R L, Ahmed A S, Gas chromatography as a development and production tool for fingerprinting oils from Individual Reservoirs: Applications in the Gulf of Mexico[A]. In: GCS/SEPM Foundation Ninth Annual Research Conference Proceedings, 1990. 263-282
Kaufman R L, Dashti H, Kabir C S, et al. Characterizing the greater Burgan Field: Use of geochemistry and oil fingerprinting [A]. In: SPE Paper No. 37803, 1997.385-394
Mark A. McCaffrey, Henry A. Legarre, and Scott J. Johnson. Using biomarkers to improve heavy oil reservoir management: an example from the Cymic Field, Kern country, California. AAPG Bulletin, v.80, No.6
R.J. Hwang, D.K. Baksin, S.C. Teerman. Allocation of commingled pipeline oils to field production. Organic Geochemistry 31(2000): 1463-1474
Kaufman R L. Gas chromatography as adevelopment and production tool for fingerprinting oils from individual reservoirs: application in the Gulf of Mexico[A]. Gulf Coast Section of the society of Economic pala eontologists and Mineralogists Foundation Ninth Annual Research Conference Proceedings[C]. 1990
Hwang R J, Ahmed A S.oil composition Variation and reservoir continuity: Unity field, Sudan [J]. Organic Geochemistry, 1994,21(2):
England W.A., The organic geochemistry of petroleum reservoirs Org. Geochem. 1990,116(1～3):
Kaufman R L, Ahmed A S, Moldowan J M. A new technique for the analysis of commingled oils and its application to production allocation calculations[A]. Procedings of the Sixteenth Annual Conention of Indonesian Petroleum Association. [C]Indonesian Petroleum Association, 1987,247-268.
Larter S Rand Aplin A C. Reservoir geochemistry :methods,application and opportunities, In:The Cubitt J N and England W A eds.Geochemistry of Reservoir[C] England,Geological Society Special Publication,1995,No.86:5～32.