New seismic wave model for tight reservoirs: Incorporating non-Darcy flow and fractional viscoelasticity

We present a workflow for the quantitative integration of 4D seismic and production data into reservoir simulation models. Reservoir models are probabilistically history matched to 4D seismic and production historical data simultaneously. The timely incorporation of 4D seismic into reservoir models, coupled with the associated reduction of production forecast uncertainty, has a significant impact in the modern reservoir management of oil and gas fields. One of the outstanding features of the work presented is the use of the 4D seismic data at the seismic trace level (i.e. seismic wiggle) compared to the impedance level. In the joint inversion approach, the reservoir models are jointly calibrated to the production data and the 4D seismic data simultaneously through a single inversion process. There is no need for separate seismic inversion for impedance or for pressure/saturation. The calibrated reservoir models honor both the historical production data and the 4D seismic traces. The joint inversion ensures consistency among the observed 4D seismic traces, the observed production data, and the geological, seismic, petrophysical and flow models. This consistency, while it is guaranteed in the joint inversion, is difficult to attain through the traditional approach of separate inversion processes for 4D seismic and production data. The joint inversion process seeks: a) to shorten the time between 4D seismic acquisition and its incorporation into reservoir models, b) to use 4D seismic in a quantitative way, and c) to reduce uncertainty in reservoir models and thus reduce uncertainty in the production forecasts. This paper presents and discusses results from application of the joint inversion workflow to a synthetic field case based on a real field. The example covers the areas of a) history matching of flow simulation models to 4D seismic and production history, and b) estimation of uncertainty. The joint inversion process requires a multidisciplinary effort. It involves the joint modeling of geology, flow simulation, rock physics, and seismic modeling under a common earth model, and thus it naturally fosters the efficient interaction among geologists, geophysicists, and reservoir engineers.

A new high seismic resolution ocean bottom node (OBN) was acquired and evaluated in complex sandstone reservoirs, from shallow to deep reservoirs. The integrated reservoir characterization with new high seismic data and model are used for reservoir model in complex reservoirs. The field has overpressure, multi fluid contacts, multi-reservoir subunits, structural and stratigraphic sand discontinuities. Therefore reliable updated reservoir characterization in deep reservoirs and estimation of fluids in place requires an integrated subsurface data approach. The advanced reservoir characterization with new seismic velocity have been implemented in the workflow of reservoir characterization. The available wells formation pressures, PVT, image log, core-logs saturation height model (SHM) and production data have been used to calibrate in the compartment and reservoir discontinuity evaluation. The new velocity high resolution of 3D-seismic has been utilized for recent pore pressure evaluation. There are variation rock properties from shallow to deep reservoirs due to reservoir compaction, overpressure reservoirs have been observed in deep reservoirs, it tends to preserve porosity and has sufficient permeability for oil productivity from the deep reservoirs. The updated reservoir model is intended to provide the oil and condensate sweet spots for further potential wells of development and appraisal in the matured oil field area. The integrated reservoir data with the latest seismic data have been used to provide reliable and good of reservoir properties and results for updated reservoir model. The high resolution seismic has been integrated with open and cased-hole logs data such as image resistivity/density, chromatography, pulse-neutron capture and production logs have been used to verify fault, fluid contacts, contribution, water saturation changes and production optimization. And for every reservoir subunit, the formation pressure has been used to identify an initial oil water contact and reservoir compartment/sand discontinuity evaluation, thus it has provided high resolution and reliability results for complex reservoir modeling. The saturation height model (SHM) has been used, a quasi-SHM for unavailable core data in deep reservoir and PVT data have been used in the reservoir evaluation/modeling, for the development of multi-layers, over pressure and complex sandstone reservoirs. The potential oil sweet spots for further appraisal and development can be modeled in the reservoirs and fields area. The integrated approach reservoir characterization in this paper shows the value of advanced reservoir characterization with the latest data in the complex reservoirs. It has utilized the integration updated high seismic resolution and previous reservoir fluid-rock data. The evaluation also has provided the reliable for reservoir modeling, oil in place volume, recent oil sweat spot data, field appraisal and development purposes. Classification: Public

A 1000 bopd was achieved from successful integrated study of Kurau Field in 2007 year. All best practices of the study describe in this paper. The objective of this study is to build a complete reservoir model from an integrated study involving all aspects of engineering & G&G, such as data review, G&G modeling, reservoir characterization, reservoir modeling and prediction, in order to optimize the development of Kurau Field. The milestone starts from interpreting the 3D seismic data, structural or zonation model, petrophysics analysis, reservoir characterization, static modeling, dynamic modeling, and proceeds until it reaches the development and performance prediction of Kurau Field. More than just a simple model, Kurau reservoir model resulted from all data and results from G&G such as interpretation of 3D seismic, G&G modeling (static model), environmental correction for well logs from Petrophysical Software analysis together with reservoir engineering works such as rock and fluid properties analysis, rock characterization and also production engineering aspects. Geomodeling software was used to build 3D seismic to reservoir modeling, with additional Simulation Software to complete the simulation. The history matching analysis shows a satisfactory level of production profile comparison results of model performances with actual production data. Several infill wells and even EOR potential can be identified from this reservoir simulation model. Using an integrated study workflow from G&G, Petrophysic and Reservoir Engineering work had greatly improved the reliability of this reservoir model. Recently updated, those infill wells proved able to yield an additional 1000 bopd to Kurau field, as previously predicted. An integrated study from Engineering & G&G Department was required when we attempted to build a reservoir model. Such a model can be used to determine infill well potential, work over result estimation and also the EOR project. Achieving an increment of the oil recovery factor is the most beneficial result from this reservoir simulation.

The finite-difference (FD) injection method by Robertsson and Chapman5,6, can be used to efficiently synthesize the seismic response after local model alterations. By integrating the FDinjection method with a reservoir simulator and petrophysical model, a highly efficient and accurate tool for predicting the seismic response of reservoir fluid flow has been developed. In a case study, six complete marine seismic surveys over a reservoir at different production stages were simulated with computational savings greater than a factor of 50. Introduction In time-lapse seismics, the difference between seismic data sets acquired at different times during the production process is used to infer changes in the distribution of fluids and pressure due to production. To decide whether a repeat seismic survey is able to detect fluid changes predicted by reservoir simulation, and to check if an existing seismic data set is consistent with the reservoir model, a seismic forward model is required. The predicted distributions of water saturation, pressure and temperature from a reservoir simulator can be combined with rock properties, via a petrophysical model, to provide the elastic moduli for the seismic modeling (e.g., Huang et al. 4 and Biondi et al.2). We combine the realism of such a mechanism for generating seismic models with the advantages of FD modeling. FD methods allow for the full response to be synthesized as the wavefield interacts with a seismic model. This includes wave propagation in heterogeneous anisotropic and anelastic media, scattering, mode conversions, etc. Although accurate, FD calculations can be highly computationally expensive. The FD-injection technique5,6 provides a means to efficiently compute the response from a seismic model subject to changes within subvolumes and is thus ideally suited for time-lapse studies. We demonstrate our approach through a case study based on the Gullfaks Field in the North Sea, where a successful real time-lapse study was carried out7. The model is not designed to represent the Gullfaks Field, but to demonstrate the application of the finite-difference injection technique in modeling time-lapse seismic data. FD-Injection Initially, the FD response from a full reference model is calculated and the wavefield is recorded at receivers and along a closed surface around a sub-volume. This surface is referred to as the injection surface Si. As changes to the model occur inside the sub-volume, the recorded seismograms can be updated by simulating the response on a small FD sub-mesh encompassing the neighborhood of the sub-volume. In the subsequent simulations on the sub-mesh, the wavefield recorded in the first simulation is injected into the sub-mesh along Si. For the case when no model alterations have taken place, the field inside Si will be identical to the field inside Si in the original FD simulation on the full model. The field outside Si will be as close to zero as machine precision allows, since on Si, the injection procedure exactly generates the ingoing waves and cancels the out-going waves.

The public perception for human-induced and triggered seismicity is often very high and public safety is a major issue for approving authorities. Even weak non-damaging events felt by population have led to major debates and in some cases fostered the closing down of geotechnical projects. Often, we cannot convincingly answer the question of the likelihood that damaging earthquakes may occur or not before the start or during the course of the engineering operations. Passive seismic techniques and advances in theoretical seismicity models are needed to improve the current situation. Seismological methods towards the characterisation of seismic sources are well advanced in mining applications (e.g. Grigoli et al., 2013, Sen et al., 2014, Maghsoudi et al., 2014). Micro-seismicity is also recognized as a valuable tool for completion, optimization, characterization and modelling of reservoirs or storage facilities (e.g. Cesca et al., 2014a). However, although the probabilistic description of seismicity is established in seismology since decades, a probabilistic approach is still not common practice for induced seismicity and reservoir studies in industry. A possible reason may be that seismicity models in seismology are often based on steady state or quasi static loading rate conditions, a situation rarely valid for engineering activities. Therefore, seismicity models considering the stress and pressure loading conditions of engineering activities need to be further developed andvalidated. The discrimination between natural and human related earthquakes is important for both issues. For instance, the nature of induced seismicity implies nearby geotechnical operations and engineering activity. However, the occurrence of a close-by earthquake does not always mean that the earthquake is human related, especially in regions with high tectonic activity. We need to establish community accepted methods for the discrimination of events. These should consider (geo-)physical and structural parameter from the natural background processes and the human related activities (e.g.Dahm et al., 2010b). Beside the discrimination by means of source parameter estimations (e.g. Cesca et al., 2013b, 2014b), probabilistic methods based on seismicity parameters are important (e.g. Dahm et al., 2012). A database of case studies, including both seismicity and production parameters, is important to validate such models and to establish common accepted procedures. Unfortunately, such databases are still not available or not accessible to a wider scientific community. The presentation reviews and summarizes the basics of seismicity models and their current role in natural earthquake and triggered / induced earthquake studies. Specific problems and questions for the different cases of induced / triggered seismicity and applications will be clarified. The behaviour and different aspects of a rate and state seismicity model are demonstrated. We discuss implementations of the rate and state model (Dietrich, 1994) for the characterization of triggered and induced seismicity, comprising the characterization of seismicity clouds related to hydro-fracture formation and the evaluation and discrimination of isolated significant earthquakes. Knowledge gaps and unsolved questions will be discussed.

Three-dimensional (3D) reservoir models are best created with a combination of well logs and 3D seismic data. However, the effective integration of those results in the reservoir modeling was not easy due to limited seismic resolution. With the increasing quality of seismic data and widely application of new methods, High-resolution seismic stochastic inversion volume was used as a direct input to reduce the uncertainty of the reservoir model in this paper. Through the past years study, workflows were developed which use the high resolution seismic stochastic inversion as a direct input for reservoir modeling. The workflows mainly include three steps. The first step is target processing. Wavelet transform was applied to achieve noise elimination and resolution improvement. Base on the High-resolution seismic data, the second step is seismic stochastic inversion. After the process of time-depth conversion, the high resolution 3D data from seismic stochastic seismic inversion and well logs data were used to reservoir modeling as a direct input. Experiment results show that noise elimination and resolution improvement achieved by using wavelet transform brings desirable convenience, high efficiency and good fidelity. The seismic response of sand reservoir in the high resolution seismic section becomes clearer than the seismic data before target processing, and matched with the well log interpretation. As the seismic stochastic inversion process is controlled not only by the acoustic impedance features, variographic model, and histogram, but also by high-resolution seismic data, the possible number of solutions is reduced, thus decreasing the nonuniqueness of the solution. The seismic stochastic inversion results were multiple 3D volumes with the same horizontal resolution of the seismic and with the vertical resolution, which is matched with the well data. The reservoir model base on the high-resolution 3D data from seismic stochastic inversion and well logs data will reduce the uncertainty greatly, especially in the area Where reservoir exhibits strong heterogeneity. A postmortem is presented showing a successful well that were better explained by this model result based on data existing before the well were drilled. With the increasing quality of seismic data, and the progress of target processing, seismic stochastic inversion and reservoir modeling technology, the high-resolution seismic data will play more and more important role in reservoir modeling. High-resolution seismic stochastic inversion as a direct input for reservoir modeling will reduce the uncertainty of model greatly.

Two independently derived reservoir models of a stratigraphically complex oil field in Alberta, Canada, were used as the bases for reservoir simulation and history-matching. This study is an attempt to compare the applicability of each method for obtaining a reliable description of the reservoir. As discussed in paper SPE 15505, the first reservoir model was based on log and core data from six producing and two dry wells drilled within a portion of the field on which a 3-D seismic survey had been acquired; the second reservoir model was developed by supplementing well data with petrophysical parameters derived from seismic lithologic modeling of the 3-D seismic reflection data. Using structure and net porosity-thickness models developed by these two approaches at two different stages of the field development drilling, history-match calculations of water cut were made on each model using the limited reservoir production and pressure history available to date. The results suggest that the seismically enhanced reservoir model produces a more accurate simulation of the observed performance than that achieved with a "conventional" reservoir model based solely on well control. This observation is particularly true at earlier stages of development drilling when inadequate reservoir descriptions can be produced based on limited well control. This study supports the potential technical benefit of using 3-D seismic data to obtain an improved reservoir description and thereby, lead to improved design of field development and operational programs.

Influence of variable uncertainties in seismic tomography models on constraining mantle viscosity from geoid observations

H041 Seismic Forward Modeling for More Representative Reservoir Model P. Breton* (Total SA) D. Modin (Total SA) O. Duplantier (Total SA) P. Thore (Total SA) & M. Bredel (EOPGS) SUMMARY Seismic forward modeling process from reservoir grid appears to be critical to check the good coherency between the reservoir model and the seismic. In-house seismic forward modeling tool chain has been optimized and directly implemented in the geosciences workstation. Two main benefits are illustrated: - use of 3D lithoseismic cubes enhances the horizontal resolution of the geological heterogeneity organization inside the reservoir model. But an horizontal upscaling of lithoseismic cubes

The seismic data set is a fundamental requirement for producing oil and gas fields. It provides understanding of the structure and stratigraphy of the reservoirs, and it is now routinely employed in reservoir modeling for advancing insights into how fields are being produced. Modern 3D seismic data were first developed in the 1960s, but it wasn't until the 1980s that seismic interpretation software enabled the building of gridded reservoir models from seismic interpretations. Reservoir modeling utilizing seismic interpretations drove insights into reservoir quality and performance, helping to understand the communication between reservoir units and wells, particularly in fields with many wells. But key challenges such as the cost of building or updating reservoir models and scale variance created barriers for early industry-wide adoption. Data integration required calibration to correct and account for the difference in measurement scales of seismic data and well data, as well as to create robust relationships between seismic properties and petrophysical properties in the model. Over time, technological advancements led to a reduction in the cost of reservoir modeling, while increased acquisition, processing, and utility of seismic data provided the means to drive innovation toward incorporating seismic. Today, 3D and 4D seismic data play pivotal roles in defining and updating reservoir models where hundreds to thousands of simulations can be realized in a reservoir model to explore history matching and model uncertainties.

Modern three-dimensional (3D) seismic data assist not only in delineating reservoir geometry, but also in predicting porosity and lithology variations away from well control. This case study of an oil-producing channel sand in the Taber/Turin area, Alta., Canada illustrates the improvement in reservoir characterization achieved with an integrated approach incorporating both well and seismic information.

To estimate strong ground motions caused by future earthquakes in Japan and to more accurately predict seismic hazards and tsunamis, it is necessary to accurately model the geometry of the subducting plate and the seismic velocity structure around Japan, particularly in offshore areas. Although various seismic velocity structure and plate boundary models have been proposed around Japan, they are all managed individually and differ in extent, data type, and format. Ensuring consistency among those models requires knowledge of their spatial distribution around the subduction zones of Japan. Here, we describe a database system to store and serve various velocity structure and plate geometry datasets from around Japan. Seismic structure models in this database include 3D seismic velocity models obtained by seismic tomography, 3D plate geometry models, 2D seismic velocity structure models, 2D plate geometry models obtained by offshore seismic surveys, and hypocenter distributions determined by offshore observations and the Japan Meteorological Agency. Using this database (currently available only in Japanese), users can obtain data from several structural models at once in the form of the original model data, equal-interval gridded data in a text file, and Keyhole Markup Language (KML) data. Users can grasp the distributions of all available seismic models and hypocenters using a web-based interface, simultaneously view various models and hypocenters as KML output files in Google Earth, and easily and freely handle the structural models in a selected area of interest using the gridded text-file output data. This system will be useful in creating more accurate models of the geometries of the subducting plate and the seismic velocity structure around Japan.

Full-field reservoir models provide key input to annual business plans and reserve booking. They support the long-term field development plan by enabling well target optimization, identification of infill opportunities, water-flood management, and well-surveillance and intervention strategies. It is crucial to constrain the model with all available static and dynamic data to improve its predictive power for confident decision making. Across Shell's global deepwater portfolio, a model-based probabilistic seismic amplitude-variation-with-offset (AVO) inversion methodology is used to constrain reservoir properties as part of a comprehensive quantitative seismic reservoir modeling workflow. Promise, a proprietary probabilistic inversion tool, estimates values of reservoir properties and quantifies their uncertainties through repeated forward modeling and automated quality checking of synthetic against recorded seismic data. During workflow execution, available geologic, petrophysical, and geophysical data are incorporated. As a consequence, the reservoir models are consistent with all relevant subsurface data following their update through inversion. Model-based inversion establishes a direct link between static model properties and elastic impedances. Probabilistic inversion output is an ensemble of posterior static models. The inversion process automatically sorts through the ensemble. It can directly provide low, mid, and high cases of the inverted models that are ready to be used in hydrocarbon volume estimation and multiscenario dynamic modeling for history matching and production forecasting. For successful and efficient delivery of full-field reservoir models with uncertainty assessment using model-based probabilistic AVO inversion, early integration of interdisciplinary subsurface data and cross-business collaboration are key.

Accurate anisotropic seismic velocity model building is the key to the success of seismic depth imaging projects in complex geological settings. Tomography has been an industry standard velocity model building tool for decades, but simultaneously solving for P-wave velocity, epsilon, and delta with surface seismic data only is an underdetermined inverse problem and unstable. The ambiguity in seismic migration velocity model leads to structural uncertainty in seismic image and is carried over to uncertainty in reservoir modeling. In this paper, we introduce a new method using rock physics compaction modeling of sandy shales to constrain the anisotropic tomography. An effective-media rock model was calibrated with well data for sedimentary basin and was used to build initial vertical transverse isotropy (VTI) velocity models. By running a stochastic simulation of the rock physics model, covariance functions were extracted from possible combination of to P-wave velocity, epsilon, and delta as a priori information to constrain the following anisotropic tomography updates and uncertainty analysis. The case study area is in the Green Canyon in the Gulf of Mexico. The results show that we can successfully constrain three parameters of tomography with the prior information from rock physics. We also performed seismic uncertainty analysis to assess the non-uniqueness of the tomography solutions. 500 velocity models with equivalent residual move out were generated and used to map migrate the reservoir structures. The gross rock volume P10, P50 and P90 were calculated from these 500 realizations to demonstrate the reduction of uncertainty from the rock physics constraints. Introduction The correctness of seismic depth imaging plays important role in E&P for Oil and Gas industry, since we are exploring in very complex areas with more imaging challenges. For instance, in the Gulf of Mexico, the presence of salt and anisotropy requires careful earth model building and advanced imaging tool such as reverse time migration. The key for success relies on building an accurate and geological plausible migration velocity model. Reflection tomography in post-migrated depth domain has been an industry standard velocity model building tool for a decade (Stork, 1992, Woodward et al., 2008). However, tomography with surface seismic data only is an ill-posed and underdetermined problem. The surface seismic acquisition geometry decides that we can solve earth model parameterizes better in one direction than the other, because of the lack of horizontal rays. In area with complex geology like salt and carbonates, the subsurface illumination can be very poor and make the problem more underdetermined. Better seismic survey designs (Moldoveanu et al., 2009) with more azimuthal coverage and longer offset can significantly improve subsurface illumination, and thus, reduce model building uncertainty. Borehole measurements, such as check shot transit times, are widely used to provide vertical velocity information, and combined with surface seismic data at well location to derive anisotropic parameters. Traditionally the process is manually done by fixing vertical velocity and picking anisotropic parameters to flat common imaging gathers. Recent examples of localized tomography (Bakulin et al., 2009) demonstrated that anisotropic parameters can be solved simultaneously at well locations with check shot transit times.

The objectives of this work were to achieve realistic reservoir modeling using seismic inversion volumes, advance petrophysical & rock typing analysis for predictive modeling of reservoir quality sands of Lam Formation. Generation of predictive scenarios for the constructed reservoir model for reduction in uncertainty. Identification of new infill well locations based on the predictive reservoir sand distribution. Reevaluation of in place volumetrics for Lam Formation in Prospect D Field. The methodology adopted to achieve the above results. The methodology for this work contained the following steps; In the first step seismic inversion was performed on Prospect D Field 3D seismic dataset to obtain volume of clay seismic volumes and facies volume. Petrophysical rock typing combined with the core data analysis was sued to calibrate the inverted volumes by identification of clay typing. The seismic inversion volumes were integrated with the depositional settings of Lam Formation to qualitatively interpret the inverted volumes. Reservoir modeling was performed using seismic inversion facies volumes and petrophysical rock type model to predict and distribute the different depositional facies being controlled by inversion trends. Generation of In place volumes and predictive scenarios for reduction in uncertainty and attributing more predictive strength to the reservoir model. On the basis of reservoir modeling two (2) prospective areas were observed to show good quality sand bodies which were non-tight and reflected good reservoir properties. The seismic inversion volumes captured the depositional trends within the Lam Formation showing the variation between channel complexes to reservoir quality delta sand bodies. Two (2) to Four (4) infill well locations were identified along with forecasted results which showed positive results based on the delineated prospective areas. The seismic inversion volume results, petrophysical rock typing combined with core data completely changed the field development plan by identifying new prospective areas which were not identified or interpreted previously.