在西伯利亚一个复杂油藏的开发中，通过数据集成方法实现了储层性质的精准预测。 编译 | 惊蛰 在油田开发过程中，能否成功钻探至预测的甜点，其关键在于降低作业风险、减少不确定性。钻探阶段的一项重要任务是提高流体界面与储层物性的准确性。为了提高Chonskaya油田勘探过程中地球物理预测的准确性，作业人员开展了地质勘探工作，包括在单一观测网上进行的三维地震勘探与高密度电测研究（时域电磁）。 Chonskaya油田利用时域电磁（TDEM）与地震勘探技术，研究了****平方公里的区域。因为TDEM信号可以提供陆相储层中饱和流体的相关信息，在建立与更新储层的复杂地质模型时，需依靠TDEM数据。 如何将地震数据与电测数据相结合，如何基于TDEM测量来做出决策，这些都需要正演建模分析。根据复杂地质模型的构建经验以及模拟结果，综合利用地震数据与电测数据，可实现纯油区、水/油区与纯水区的划分；预测勘探目标中可能存在的流体类型；并勾勒出Nepa-Botuoba背斜构造中盐下复合体的陆相岩石中储层物性差的区域。 基于目标储层的地质与地球物理先验信息，包括地震数据的运动与动态解释、测井数据解释、岩石力学模型、储层地质模型（二维或三维），建立符合TDEM数据的地质模型。随后，将综合模型与TDEM反演数据进行对比分析。在定性与定量对比的基础上，对地质模型中的假设进行修正。修正后的假设可以缩小地质不确定性，进而调整地质勘探方案。 区块*的开发是该项目的驱动因素之一，四口位于B**构造穹顶位置的井成功实现了原油开采，而位于构造更低位置的另一口井却采出了水。考虑到储层的缓倾斜构造，油水界面(OWC)的不确定性(主要不确定性之一)，对含油面积影响显著，进而影响了初始储量。 在最初阶段，针对不同类型的流体(含油的W-*井与含水的W-*井)，建立了TDEM信号的综合模型。在这种情况下，将横向测井数据平均到地电层的范围，模拟信号响应。接下来，进行数据反演(利用噪声信号重建地电层的地电特征)。 通过对W-*井与W-*井的重建信号分布进行分析，可知当B**储层处于油、水饱和条件下时，目的层的不同地电属性。在初步模拟的基础上，选择*区块作为开发与测试数据集成技术（地震与TDEM）的试验区。 地质建模的数据集成 在两种地球物理方法（TDEM与地震）的基础上，若要建立统一的地质模型与后续的复杂预测，则需要在建立地质模型的过程中，利用TDEM数据进行修正。本文采用的技术是电测、地震数据的稳定集成，旨在通过建立满足地震、电测数据的地质模型，来减少流体界面(油/水界面与气/水界面)、含水饱和度以及储层物性相关的不确定性，从而提高地质储层模型的精度。 建立综合地质模型，进行后续模型运行，包括以下几个阶段：获得储层的初始地电特征。这涉及TDEM数据的反演，与建立基岩背景电导率模型，以确定储层的横向电导率。通常，目的层的地电层厚度超过*** m。在Chonskaya项目中，地电层以Vendian陆相沉积的目的层为代表，产层由B**、B**储层，以及上覆与底层泥岩组成。 建立一个综合地电储层模型。在地质认识的基础上，建立储层的综合地电模型。利用地质模型的含水饱和度与孔隙度参数，通过电测饱和模型确定储层电阻率，再基于净厚度确定储层的横向电导率。根据输入参数估算综合模型的地电特性。如果目标复合体包含两个或两个以上的储层，则分别估算每个储层的地电储层模型，然后将每个储层的电导率图相加，确定横向电导率的整体特征。将储层横向电导率的综合模型与TDEM电导率进行对比分析。 地质模型的排序与选择。这些地质模型对应的综合地电模型，应根据它们与初始地电模型的相似性进行排序。该阶段包括，将模拟的电阻率参数与综合/初始地电模型的横向电导率进行定性与定量的比较，从而确保在地质模型正确的基础上得出结论。模型的选择标准包括相关系数与均方根偏差。然后，选择一组最符合标准的综合模型。 利用TDEM数据对地质模型进行校正。为了使建立的地质模型，能够和基于TDEM的初始地电模型之间，形成较高的关联度，从而开发出一种算法。该算法可不断修正地质参数的预测图，直到横向电导率的测量值与模拟值之间的偏差达到最小。在这种情况下，变量参数被指定为地质与地球物理先验信息调节的变化范围。若地质认识与初始地电特征具有最高贴合度，即可建立地质模型，以符合两种地球物理法的测量结果。地质认识的持续发展（考虑各种油/水界面与储层物性的更新）可实现地质模型的建立，且综合模型与地电测量值之间的误差最小。 在构建Chonskaya油田*区块的综合地质模型时，已对该技术进行了测试。在井数据、概念建模、地震与电测勘探的基础上，建立了目的层的地质模型。油/水界面的位置是该区块地质模型的关键不确定性。 界面不确定性的初始范围为**米。地质模型中油/水界面的变化，结合综合模型与测得的地电参数之间的定量对比，可将预计油/水界面的不确定范围降低至*米。储量不确定性的范围减少了**%，P**油藏的储量增加了****万吨。基于该分析，可调整探井井位，使之能够与油/水界面相交，从而完全消除该区块内油/水界面的不确定性。 结果表明，在建立地质模型，更新东西伯利亚的流体类型与储层物性时，可能会使用TDEM数据。在复杂的地质与物理条件、研究程度低、地质不确定性高的情况下，本文推荐集成描述不同地层物性的地震与电测数据，并在合理集成的情况下，相互查漏补缺。这种方法允许在建立模型的过程中使用地球物理数据，以便在现场评估与减少地质不确定性方面做出有效决策。在开发油田的勘探阶段，该项能力极为重要。The terrigenous sediments of the subsalt complex within the Nepa-Botuoba anteclise are characterized by complex geology: nonanticlinal traps, complex tectonics, lateral reservoir heterogeneity, and digenesis that controls reservoir distribution. The block structure of the Chonskaya group of fields and their poor coverage with exploration wells leave a number of uncertainties with regard to water/oil and gas/water contacts in blocks with proved oil content, as well as the main fluid types in exploration blocks. The success of predicting prospective sites for exploration drilling is influenced by minimal risks and reduced uncertainty during field development. An important task at the stage of exploration drilling is to increase the reliability of fluid contacts and reservoir properties. To improve the quality of geophysical predictions during exploration of the Chonskaya group of fields, geological exploration operations were conducted, including *D seismic surveys and high-density electrical studies [time-domain electromagnetic and magnetic (TDEM)] on a single observation network. The area studied by TDEM and seismic surveys on the observation network in the Chonskaya group covers ****?km*, more than **% of the total territory of the license areas. In this regard, the methods to honor the TDEM data when building and updating complex geological models of reservoirs are being actively developed, because the TDEM signal can provide information about fluids saturating terrigenous reservoirs. To understand the capabilities of the technology for integrating seismic and electrical data and making a decision on the basis of the TDEM survey, forward modeling is necessary. On the basis of the modeling results and experience in building complex geological models, the combined use of electrical and seismic data has the potential to be used to separate pure-oil, water/oil, and pure-water zones; to predict probable fluid types in the exploration targets; and to outline zones with poor reservoir properties in terrigenous rocks of the subsalt complex within the Nepa-Botuoba anteclise. A geological model honors the TDEM data through building a geoelectrical reservoir model using a priori geological and geophysical information about target reservoirs—the kinematic and dynamic interpretation of seismic data, well-log--interpretation data, a petrophysical model, and a geological model of the reservoir (*D or *D). Next, a comparative analysis of the synthetic geoelectrical model with the TDEM inversion data is conducted. The geological model assumptions are revised on the basis of qualitative and quantitative comparison. The revised assumptions allow reduction of the range of geological uncertainties and adjustment of the geological exploration program. Preconditions for Use of TDEM To Determine Reservoir-Fluid Types Block * is one of the project drivers. Four wells drilled in the dome part of the B** structure produced commercial oil inflows, and another well, located much lower in the structure, produced an inflow of water. Given the gently sloping structure of the reservoir, the oil/water-contact (OWC) uncertainty (one of the key uncertainties) affects the oil-bearing area significantly and, as a consequence, the volume of initial in-place reserves. At the initial stage, for the wells with different types of fluid (W-* with oil and W-* with water), a synthetic model of the TDEM signal was built. In this case, the lateral logging data were averaged to the scale of geoelectrical layers and the signal response was simulated. Next, data inversion was performed (a reconstruction of geoelectrical characteristics of the geoelectrical layer from a noisy signal). The analysis of the reconstructed signal distribution in Wells W-* and W-* shows different geoelectrical properties of the target interval when the B** reservoir is saturated with oil and with water. On the basis of the preliminary simulation, Block * was selected as a pilot area for the development and testing of the data--integration technology (seismic and TDEM). Data Integration in the Geological Modeling Process Building a unified geological model and subsequent complex prediction on the basis of results of two geophysical methods (TDEM and seismic) requires a methodology for honoring the TDEM data in the process of building a geological model. The technology used here is robust integration of electrical and seismic data, aimed at improving the accuracy of the geological reservoir model by reducing uncertainties related to fluid contacts (OWC and gas/water contact), water saturation, and reservoir properties by building geological models that satisfy seismic and electrical survey data. Building an integrated geological model and conducting subsequent model runs includes the following stages. Obtaining Initial Geoelectrical Characteristics of a Reservoir.?This involves inversion of TDEM data and the building of a background conductivity model of host rocks to determine lateral reservoir conductivity. As a rule, the geoelectrical layer of the target interval is more than *** m thick. In the Chonskaya project, the geoelectrical layer is represented by the target interval of Vendian terrigenous deposits, the productive part of which consists of two reservoirs, B** and B**, and over- and underlying mudstone units. The complete paper provides equations by which geoelectrical reservoir characteristics can be obtained. Building a Synthetic Geoelectrical Reservoir Model.? A synthetic geoelectrical model of the reservoir is built on the basis of a geological realization. The water saturation and porosity parameters of the geological model are used, through an electrical saturation model, to determine reservoir resistivity, and the net thickness is used to determine lateral conductivity of the reservoir. The geoelectrical characteristics of the synthetic model are estimated on the basis of the input parameters. If the target complex contains two or more reservoirs, the geoelectrical reservoir models are estimated for each reservoir separately and then the integral characteristic of the lateral conductivity is determined by summing the conductivity maps of each reservoir. The synthetic model of the reservoir lateral conductivity is compared with the TDEM conductivity. Ranking and Selecting the Geological Models.?The corresponding synthetic geoelectrical models of these geological models should be ranked on the basis of their similarity to the initial geoelectrical model. This stage includes a qualitative and quantitative comparison of the simulated resistivity parameters with the lateral conductivity of the synthetic and initial geoelectrical models, which allows conclusions to be reached on the basis of the geological model correctness. The model selection criteria include a correlation coefficient and a root-mean-square deviation. Next, a set of synthetic models with the best criteria is selected. Correction of the Geological Model With TDEM Data.?To build a geological model with a high degree of correlation that satisfies the initial TDEM-based geoelectrical model, an algorithm was developed. The algorithm corrects the prediction maps for the geological realization of parameters until a minimum discrepancy is achieved between the observed and the synthetic lateral conductivity of the model. In this case, the variable parameters are assigned the variation range regulated by a priori geological and geophysical information. The geological realizations that have the highest degree of convergence with the initial geoelectrical characteristics allow building a geological model that honors the results of two geophysical methods. Consistent development of geological realizations (consideration of various OWC levels and update of reservoir properties) allows building a geological model with minimal mis-ties of the synthetic and observed geoelectrical field. This technique has been tested when building a comprehensive geological model of Block * of the Chonskaya group of fields. The geological model of target horizons is built on the basis of well data, concept modeling, and seismic and electrical exploration. OWC position is a key uncertainty in the geological model of this block. The initial range of contact uncertainty was ** m. The OWC variations in the geological model and a consistent quantitative comparison of the maps of synthetic and observed geoelectrical parameters allowed reduction of the estimated OWC to * m. The range of reserves uncertainty was reduced by **%, and P** reserves increased by ** million tons. The analysis allowed relocation of the exploration well to intersect the OWC most probably and to remove OWC uncertainty completely within the block. The results demonstrate the potential of using TDEM data when building a geological model for updating fluid type and reservoir properties in Eastern Siberia. In the context of complex geological and physical conditions, a low degree of study, and high geological uncertainties, it is recommended to integrate the seismic and electrical data that describe various physical parameters of the subsurface and, in the case of reasonable integration, complement each other. This approach allows use of geophysical data in the process of building a model for making effective decisions on field appraisal and reducing geological uncertainties, an ability of particular importance at the exploration stages of field development.