论文标题:原油指纹技术在油井生产管理中的应用研究
Research on the Application of Oil Fingerprints Technique in Oil Wells Management
论文作者
论文导师 李桃；李彦平,论文学位 硕士,论文专业 管理科学与工程
论文单位 沈阳大学,点击次数 88,论文页数 92页File Size1466K
2007-01-26论文网 http://www.lw23.com/lunwen_2471147/
Oil fingerprint; Commingled oil well; Multi-oil production analysis software; Back propagation network;LiaoHe oilfield
在油田勘探开发过程中,生产井多为多层合采,确定各层产油贡献率对于及时了解产油剖面的产油能力,监测各油层随时间的变化情况以及提高勘探区块的储量估算方面都起到至关重要的作用。原油色谱指纹技术的成功研究,可以节约大量勘探试油经费,在不关井、不下井下工具的情况下即可完成产油剖面测试,可真正实现勘探开发一体化,可连续多次进行快速、准确分析、重现性和可靠性好。本文运用高分辨率全烃色谱分析技术对原油进行色谱检测,原油色谱指纹主要指原油经气相色谱仪检测得到的保留时间、峰高、峰面积等信息,利用这些信息建立原油色谱指纹数据库,并根据单层原油的指纹差异及混合原油的指纹变化特征确定分层产能贡献率;研究了标准指纹图谱的特征峰与待测指纹图谱特征峰之间的相似性指标与评价方法,原油指纹经相似性评价后可识别出不同层原油所含相同物质成分的特征峰的保留时间漂移量;在此基础上进行原油指纹特征峰识别,并提出谱峰分形识别方法,用来配对原油色谱指纹峰和剔除歧义数据;最后针对二层合采井和二层以上的合采井分别提出了原油指纹产能分析模型和BP网络模型,并用分形维数对二层产能分析模型进行了改进。在理论研究的基础上,开发了原油产能分析软件,在实验室进行了二层与三层原油的配方试验。计算结果与实际配方对比表明:二层实际配方与软件分析的误差在7.5%以内,三层误差在16%以内。与生产测井的测试结果对比,误差在允许的范围内。说明采用原油指纹技术研制的产能分析软件具有合理性,在油田现场可以用来计算合采井各层的产能贡献率,为油田的生产管理提供了可靠的决策依据。
In the process of prospect for oil, it is important to commingled oil well that makes sure the contribution of the certain single zone in commingled oil, from which we can analysis the ability of oil production, the variation of oil zone and the total oil production in oilfield. Oil fingerprint technique has the advantages of the low cost and evaluating the oil zone without stopping production. It analysis the contribution of the certain single zone in commingled oil quickly, repeatedly and exactly. This thesis uses the technology of whole-oil to analysis the oil fingerprint which including the height of peak, area of peak and the symmetry of peak etc, the oil fingerprint database is built on the base of oil fingerprint data. The single zone contribution ratio is made, according to the oil fingerprint differences of individual zone and fingerprint variation characteristics of commingled oil. A new algorithm of similarity between the standard and unknown chromatographic fingerprint is provided, and an estimate index is also provided, combined with those two methods, we can analysis the same material’s variety of reserve time in difference oil zone. Based on the above research, an oil fingerprint distinguishing method is provided to deal with and eliminate the oil fingerprints data, and what is more fractal distinguishing method is provided. Finally, this thesis builds up an oil fingerprints calculation model and a Back Propagation Network model to analysis the two and above pay zones in commingled oil well, besides using the fractal theory to improve the oil fingerprints calculation model. Based on the theory research, the multi-oil production analysis software is developed, analyzes the two and three pay zones. The results show that the errors between production logging result and fingerprints calculation result are all below 7.5% for two zones, and 16% for three zones commingled oil well, the errors are all under the allowed range. Therefore that software developed based on oil fingerprints technique can be used to calculate the contribution of individual zone in commingled oil well, and provides exactly data to manage the oilfields.
|
