Layer | Fill | Outline |
---|
Map layers
Theme | Visible | Selectable | Appearance | Zoom Range (now: 0) |
---|
Fill | Stroke |
---|---|
Collaborating Authors
Leeuwenburgh, Olwijn
A Hybrid Data-Physics Framework for Reservoir Performance Prediction with Application to H2S Production
Leeuwenburgh, Olwijn (TNO (Corresponding author)) | Egberts, Paul J. P. (TNO) | Barros, Eduardo G. D. (TNO) | Turchan, Lukasz P. (Bluware) | Dilib, Fahad (Equinor ASA) | Lødøen, Ole-Petter (Equinor ASA) | de Bruin, Wouter J. (Equinor ASA)
Summary Model-based reservoir management workflows rely on the ability to generate predictions for large numbers of model and decision scenarios. When suitable simulators or models are not available or cannot be evaluated in a sufficiently short time frame, surrogate modeling techniques can be used instead. In the first part of this paper, we describe extensions of a recently developed open-source framework for creating and training flow network surrogate models, called FlowNet. In particular, we discuss functionality to reproduce historical well rates for wells with arbitrary trajectories, multiple perforated sections, and changing well type or injection phase, as one may encounter in large and complex fields with a long history. Furthermore, we discuss strategies for the placement of additional network nodes in the presence of flow barriers. Despite their flexibility and speed, the applicability of flow network models is limited to phenomena that can be simulated with available numerical simulators. Prediction of poorly understood physics, such as reservoir souring, may require a more data-driven approach. We discuss an extension of the FlowNet framework with a machine learning (ML) proxy for the purpose of generating predictions of H2S production rates. The combined data-physics proxy is trained on historical liquid volume rates, seawater fractions, and H2S production data from a real North Sea oil and gas field, and is then used to generate predictions of H2S production. Several experiments are presented in which the data source, data type, and length of the history are varied. Results indicate that, given a sufficient number of training data, FlowNet is able to produce reliable predictions of conventional oilfield quantities. An experiment performed with the ML proxy suggests that, at least for some production wells, useful predictions of H2S production can be obtained much faster and at much lower computational cost and complexity than would be possible with high-fidelity models. Finally, we discuss some of the current limitations of the approach and options to address them.
- North America > United States > Texas (0.68)
- Europe > Norway > Norwegian Sea (0.46)
- Europe > Norway > North Sea > Northern North Sea (0.29)
- Europe > Norway > Norwegian Sea > Halten Terrace > PL 128 > Block 6608/10 > Norne Field > Tofte Formation (0.99)
- Europe > Norway > Norwegian Sea > Halten Terrace > PL 128 > Block 6608/10 > Norne Field > Not Formation (0.99)
- Europe > Norway > Norwegian Sea > Halten Terrace > PL 128 > Block 6608/10 > Norne Field > Ile Formation (0.99)
- (10 more...)
- Information Technology > Modeling & Simulation (1.00)
- Information Technology > Artificial Intelligence > Representation & Reasoning (1.00)
- Information Technology > Artificial Intelligence > Machine Learning > Statistical Learning (0.46)
- Information Technology > Artificial Intelligence > Machine Learning > Neural Networks (0.46)
An Ensemble-Based Method for Constrained Reservoir Life-Cycle Optimization
Leeuwenburgh, Olwijn (TNO) | Egberts, Paul J. (TNO) | Chitu, Alin G. (TNO) | Alim, Marcel (Delft University of Technology)
Abstract We consider the problem of finding optimal long-term (life-cycle) recovery strategies for hydrocarbon reservoirs by use of simulation models. In such problems the presence of operating constraints, such as for example a maximum rate limit for a group of wells, may strongly influence the range of possible solutions. Commercial simulators often offer the possibility to use heuristic rules to deal with constraints during simulation. It is first shown that such an approach imposes serious limitations for important cases of interest, such as finding optimal operating strategies for smart wells. We subsequently consider the use of gradient-based numerical optimization approaches, which are considered the most efficient for realistically complex reservoir cases. Formal treatment of especially output constraints in such approaches is challenging and is usually associated with high computational costs or strong simplifications. We propose and demonstrate a method that addresses these challenges by efficient use of an ensemble to approximate constraint gradients with respect to the well controls. The gradients are then used by an optimization algorithm to maximize an objective function, typically net present value, while avoiding violation of the constraints. The method was implemented in an in-house optimization framework and tested on two example cases. The first example case is a relatively simple homogeneous 2D reservoir with four injection wells and one producing well, all equipped with adjustable valves and operating at fixed pressures. We consider the presence of a maximum field water injection rate constraint, as well as individual well rate constraints. The method is also applied to a realistically complex 3D case with 30 smart wells equipped with ICVs, modified from the original Brugge benchmark model. The results show that in both example cases the method is able to find improved recovery strategies that do not violate the constraints. It is also shown that proper choices for constraint grouping, balancing of objective function and constraints, and of the initial control strategy may positively impact the efficiency and effectiveness of the optimization process. The proposed method is an improvement to constraint handling by simulators for smart well optimization and enables application of optimization workflows to field cases with realistic operating constraints.
- Europe (1.00)
- North America > United States (0.68)
- Research Report > New Finding (0.34)
- Research Report > Experimental Study (0.34)
- Energy > Oil & Gas > Upstream (1.00)
- Water & Waste Management > Water Management > Lifecycle > Disposal/Injection (0.54)
Seismic History Matching of Fluid Fronts Using the Ensemble Kalman Filter
Trani, Mario (Delft University of Technology) | Arts, Rob (TNO) | Leeuwenburgh, Olwijn (TNO)
Summary Time-lapse seismic data provide information on the dynamics of multiphase reservoir fluid flow in places where no production data from wells are available. This information, in principle, could be used to estimate unknown reservoir properties. However, the amount, resolution, and character of the data have long posed significant challenges for quantitative use in assisted-history-matching workflows. Previous studies, therefore, have generally investigated methods for updating single models with reduced parameter-uncertainty space. Recent developments in ensemble-based history-matching methods have shown the feasibility of multimodel history and matching of production data while maintaining a full uncertainty description. Here, we introduce a robust and flexible reparameterization for interpreted fluid fronts or seismic attribute isolines that extends these developments to seismic history matching. The seismic data set is reparameterized, in terms of arrival times, at observed front positions, thereby significantly reducing the number of data while retaining essential information. A simple 1D example is used to introduce the concepts of the approach. A synthetic 3D example, with spatial complexity that is typical for many waterfloods, is examined in detail. History-matching cases based on both separate and combined use of production and seismic data are examined. It is shown that consistent multimodel history matches can be obtained without the need for reduction of the parameter space or for localization of the impact of observations. The quality of forecasts based on the history-matched models is evaluated by simulating both expected production and saturation changes throughout the field for a fixed operating strategy. It is shown that bias and uncertainty in the forecasts of production both at existing wells and in the flooded area are reduced considerably when both production and seismic data are incorporated. The proposed workflow, therefore, enables better decisions on field developments that require optimal placement of infill wells.
- Europe (1.00)
- North America > United States > Texas (0.46)
- Asia > Middle East > Israel > Mediterranean Sea (0.34)
- Asia > Middle East > UAE (0.28)
- Europe > Norway > Norwegian Sea > Halten Terrace > Block 6407/9 > Draugen Field > Rogn Formation (0.99)
- Europe > Norway > Norwegian Sea > Halten Terrace > Block 6407/9 > Draugen Field > Garn Formation (0.99)
- Europe > Norway > Norwegian Sea > Halten Terrace > Block 6407/12 > Draugen Field > Rogn Formation (0.99)
- (2 more...)
Integrated Workflow for Computer Assisted History Matching on a Channelized Reservoir
Peters, Elisabeth (TNO) | Wilschut, Frank (TNO) | Leeuwenburgh, Olwijn (TNO) | van Hooff, Peter (TNO) | Abbink, Oscar (TNO)
Abstract Increasingly computer assisted techniques are used for history matching reservoir models. Such methods will become indispensable in view of the increasing amount of information generated by intelligent wells, in which case manual interpretation becomes too time consuming. Also, with the increasing possibilities for controlling a reservoir, a better prediction of the reservoir behavior is very important. A technique that has received considerable attention lately is the Ensemble Kalman Filter (EnKF), which is a sequential updating technique based on the Bayesian notion of updating prior information with observations (measurements). The EnKF has already been proven to be useful for history matching of real fields. Some of the advantages of the EnKF are its sequential nature, the large number of parameters that can be estimated (10000+) and the fact that an uncertainty estimate is generated. The most important and time consuming step for anyone wanting to do a history match with the EnKF is creating the initial ensemble including defining all the proper initial uncertainties. This paper presents an integrated workflow for creating an initial ensemble for a channelized reservoir and updating the ensemble sequentially using the EnKF. The resulting ensemble of history matched models is then used to predict the expected production and associated uncertainty for a new production strategy. The quality of the prediction is compared to the predictions from a model which was history matched manually. The improved predictive capabilities of the models which were history matched with the EnKF allows for better optimization of the production strategy for the complex channel geometry. Moreover, the updated uncertainty estimate shows the risks involved with some of the newly proposed wells, which allows for better decision making.
- Europe > Netherlands (0.69)
- North America > United States > Texas (0.68)
- Information Technology > Software Engineering (0.70)
- Information Technology > Modeling & Simulation (0.46)
- Information Technology > Data Science > Data Mining (0.34)
A reliable estimate of reservoir pressure and fluid saturation changes from time-lapse seismic data is difficult to obtain. Existing methods generally suffer from leakage between the estimated parameters. We propose a new method using different combinations of time-lapse seismic attributes based on four equations: two expressing changes in prestack AVO attributes (zero-offset and gradient reflectivities), and two expressing poststack time-shifts of compressional and shear waves as functions of production-induced changes in fluid properties. The effect of using different approximations of these equations was tested on a realistic, synthetic reservoir, where seismic data have been simulated during the 30-year lifetime of a water-flooded oil reservoir. Results found the importance of the porosity in the inversion with a clear attenuation of the porosity imprint on the final estimates in case the porosity field or the vertically averaged porosity field is known a priori. The use of a first-order approximation of the gradient reflectivity equation leads to severely biased estimates of changes in saturation and leakage between the two different parameters. Both the bias and the leakage can be reduced, if not eliminated, by including higher-order terms in the description of the gradient, or by replacing the gradient equation with P- and/or S-wave time-shift data. The final estimates are relatively robust to random noise, as they present fairly high accuracy in the presence of white noise with a standard deviation of 15%. The introduction of systematic noise decreases the inversion accuracy more severely.
- Europe > Netherlands (0.47)
- Asia > Middle East > Israel > Mediterranean Sea (0.24)
- Research Report > New Finding (0.48)
- Research Report > Experimental Study (0.34)
Ensemble Methods for Reservoir Life-Cycle Optimization and Well Placement
Leeuwenburgh, Olwijn (TNO) | Egberts, Paul J. (TNO) | Abbink, Oscar A. (TNO)
Abstract Several simple examples are presented that demonstrate the application of an ensemble-based method to production optimization. In particular, some practical aspects of the method such as ensemble size, perturbation, regularization and smoothing, and robust gradient estimation are discussed by comparison with an adjoint approach. The controls in the presented examples are inflow control valve settings for fixed time intervals or well position. We find that the performance of the method is clearly affected by the correlation time scale that is assumed for the controls, as reflected both by the quality of the gradient estimation and the subsequent optimization. The well placement optimization problem is studied for two cases: one is a homogenous reservoir with a sealing fault, and the other a non-homogeneous case. The ensemble optimization method is found to work well for both these simple well placement problems.
The Importance of Localization In the Assimilation of 4D Seismic Data In the Data Assimilation Process Using the EnKF
Trani, Mario (Delft University of technology) | Arts, Rob (Delft University of technology) | Leeuwenburgh, Olwijn (TNO) | Brouwer, Jan (TNO) | Douma, Sippe (Shell International E&P B.V.)
ABSTRACT The Ensemble Kaiman Filter (EnKF) is considered a fast and efficient algorithm in the data assimilation process to estimate reservoir properties from measured data. 4D seismics is an important source of information for the reservoir monitoring and the improvement of the geological model. The use of low frequencies for deep surface seismic makes it very complicated to discriminate and estimate properties for fine-grid reservoir models. In this paper it is demonstrated that using vertically averaged seismic data, inverted as time-lapse differences in pore pressure and saturation, greatly improves the quality of the history match and the estimation of the reservoir state. The EnKF may present some problems when assimilating large amounts of data (frequent 4D seismic), as the flexibility of the model solution is strongly reduced. The conditioning of the covariance matrix in the Kaiman gain is a key to avoid the filter divergence. In this study the localization criterion is based on the mere distance or on the streamlines trajectories. Results from 2D and 3D synthetic examples show the importance of localization to ensure the correct functioning of the filter.
- Geophysics > Time-Lapse Surveying > Time-Lapse Seismic Surveying (1.00)
- Geophysics > Seismic Surveying (1.00)