| 可变点约束叠前流体因子直接提取方法 |
| |
| 引用本文： | 杨培杰,王长江,毕俊凤,刘书会.可变点约束叠前流体因子直接提取方法[J].地球物理学报,2015,58(6):2188-2200. |
| |
| 作者姓名： | 杨培杰 王长江 毕俊凤 刘书会 |
| |
| 作者单位： | 1. 中石化胜利油田分公司勘探开发研究院, 东营 257015;2. 中石化胜利油田分公司博士后工作站, 东营 257015 |
| |
| 基金项目： | 国家科技重大专项"渤海湾盆地精细勘探关键技术"(2011ZX05006) 资助. |
| |
| 摘 要： | 以Gassmann流体因子(Gassmann Fluid Item,GFI)为目标,提出了一种流体因子直接提取的新方法.首先,以贝叶斯反演框架为基础,将似然函数、先验信息以及Gassmann流体因子近似方程相结合,得到初始的目标函数;其次,进一步在初始目标函数中加入可变数量的点约束信息,并得到最终的目标函数;最后,通过求解该目标函数,就直接提取出了Gassmann流体因子.该方法的主要特点是不需要初始模型的参与,而是通过一个约束模型来控制提取结果的稳定性和准确性,并且可以从约束模型中选定不同数量的约束点进行约束,称为可变点约束.给出并讨论了三种常用的不同点约束模式和原则,并用模型说明了它们不同的约束效果.模型验证和实际应用结果皆以表明,该方法即使在叠前数据信噪比很低的情况下也能较好地提取出Gassmann流体因子,流体因子提取结果客观性高、稳定性好,并且能够与已知的流体解释结果很好地匹配,益于进一步推广应用.
|
| 关 键 词： | 流体因子 叠前反演 可变点约束 直接提取 客观性 稳定性 |
| 收稿时间： | 2014-02-14 |
Direct extraction of the fluid factor based on variable point-constraint |
| |
| YANG Pei-Jie,WANG Chang-Jiang,BI Jun-Feng,LIU Shu-Hui.Direct extraction of the fluid factor based on variable point-constraint[J].Chinese Journal of Geophysics,2015,58(6):2188-2200. |
| |
| Authors: | YANG Pei-Jie WANG Chang-Jiang BI Jun-Feng LIU Shu-Hui |
| |
| Institution: | 1. GeoScience Research Institute of Shengli Oilfield, SINOPEC, Dongying 257015, China;2. Postdoctoral Workstation of Shengli Oilfield, SINOPEC, Dongying 257015, China |
| |
| Abstract: | Fluid factor extraction plays an increasing important role in fluid discrimination. The conventional way of such extraction through prestack inversion is to calculate fluid factors indirectly from P-wave velocity, S-wave velocity and density data which can be derived from inversion of seismic data. However, this method has two disadvantages. One is that the density data imbedded in fluid factors is more contaminated by noise than the inverted P-wave and S-wave reflectivity even with large incident angles. The other is that the indirect way of fluid factor estimation can create more uncertainties caused by the indirect calculation. This article focuses on the direct extraction of Gassmann fluid item (GFI), which is the real factor that reflects the influence of fluid in porous rock as Russell et al. discussed. The objective is to improve accuracy and stability of fluid factor extraction compared with the conventional way.#br#A novel method for direct extraction of fluid factors, named variable point-constraint fluid factor direct extraction (VPC-FFDE), is developed that uses variable point-constraint strategy to extract GFI the Gassmann fluid item from prestack data directly. The initial objective function is build combining likelihood function, priori information and GFI approximate equation. The final objective function is yielded by adding a variable number of constraint points to the initial objective function. Three different point-constraint patterns are examined, and different constraint effects are illustrated using synthetic data. Instead of the initial model, this method uses a constraint model to improve the accuracy and stability of the extraction results. The core of the proposed approach is to control the extraction results by adding a variable number of constraint points into the extraction process. Either accurate constraint points or the extremely low frequency model can be used, and different numbers of constraint points can be chosen during the constraint process. It does not need to obtain P-wave velocity, S-wave velocity and density first, and therefore can avoid accumulation of errors that often appear with indirect approach.#br#We applied the proposed method to the Chengdao area of the Shengli Oilfield, Sinopec. The area is about 150 km2. We chose prestack angle stacks in this case. Neither the structural high nor the bright spot is unambiguous for the prediction of gas or oil sands in this area, and the SNR is also a little low. We used a constant value constraint model to constrain the extraction process and pattern 3 was chosen in this example. As can be seen from the extraction results, the direct GFI extraction profile in the three layers are all characterized by low values, but the GFI value of the oil layer is relatively lower compared to the water layer, which has already been verified in the fluid substitution model in the previous section. This result gives a clear indication of the lateral extent and vertical extent of the oil layer and water layer. The extraction results are consistent with current oil production and joint interpretation results with only well information. However, the indirect GFI extraction result is somewhat more ambiguous, and the resolution is also lower. The actual application results show that compared with the indirect GFI extraction results, the direct ones have higher resolution and accuracy, and can match the well logging interpretation results perfectly. We implemented the procedure to distinguish different fluids. The proposed method is accurate and reliable. We implemented the purpose of more accurate fluid discrimination through fluid factor direct extraction.#br#We proposed a novel approach to extract GFI directly based on GFI linearized approximation, Bayesian inversion framework and variable points-constrain strategy. The likelihood function and priori information contribute to the high extraction resolution. The strategy of variable points-constraint renders the extraction process more stable and not sensitive to the constraint model. Model validation and actual application results show that the proposed method can produce good application effects even if the SNR of prestack data is low, and therefore is beneficial for further popularization and application. |
| |
| Keywords: | Fluid Factor Prestack Inversion Variable Point Constraint Direct Extraction Objectivity Stability |
| 本文献已被 CNKI 等数据库收录！ |
| 点击此处可从《地球物理学报》浏览原始摘要信息 |
| 点击此处可从《地球物理学报》下载免费的PDF全文 |
|
