Identification of a prospect is normally done based on seismic interpretation and geological understanding of the area. However, due to the inherent uncertainties of the data we still observe in many cases that all key petroleum system elements are present, but still the drilled prospect is dry. Such failures are mostly attributed to a lack of understanding of seal capacity, reservoir heterogeneity, source rock presence and maturation, hydrocarbon migration, and relative timing of these processes. The workflow described in this paper aims to improve discovery success rates by deploying a more rigorous and structured approach. It is guided by the play-based exploration risk assessment process. The starting point is always that the process is guided by the the basic understanding of a mature kitchen should always be based on a regional scale petroleum systems model. However, while evaluating prospects, the migration and entrapment component of a prospect should always be investigated by means of a locally refined grid-based petroleum system model. The uniquepart of this approach is the construction of a high-resolution static model covering the prospects, which is built by using available well data, seismo-geological trends and attributes to capture reservoir potential. Additional inputs such as fault seal analysis also helps to understand prospect scale migration and associated geological risks. In the regional play and local prospect-scale petroleum system models, geological and geophysical inputs are utilized to create the uncertainty distribution for each input parameter which is required for assessing the success case volume of identified prospects. The evaluated risk is combined with the volumetric uncertainty in a probabilistic way to derive the risked volumetrics. It is further translated into an economic evaluation of the prospect by integrating inputs like estimated production profiles, appropriate fiscal models, HC price decks, etc. This enables the economic viability of the prospects to be assessed, resulting in a portfolio with proper ranking to build a decision-tree leading to execution and operations in ensuing drilling campaigns.
Multiple point statistical (MPS) simulation is a modern pattern-based geostatistical approach for describing and stochastically simulating geologic formations with complex connectivity patterns. In MPS geostatistical simulation, a template containing data patterns around each simulation cell is used to extract and store the local conditional probabilities from a training image (TI). To generate a simulated sample, a random path is generated to sequentially visit all unsampled grid cells and draw conditional samples from the corresponding stored conditional probabilities. The grid-based implementation of MPS simulation offers several advantages for integration of hard and soft data. In the Single Normal Equation SIMulation (SNESIM) implementation of MPS for facies simulation, it has been observed that the integration of soft data can result in many facies realizations that do not provide consistent patterns with the incorporated probability map. This is partly explained by the Markov property that only considers probabilities that are co-located with the simulation node, and hence ignoring spatial information from neighboring cells. In addition to this effect, we show another important mechanism is in play in the SNESIM algorithm that explains the observed behavior. Specifically, at the early stage of the simulation when the first few percentage of the simulation nodes on the random path are visited the local conditioning data are limited and the resulting conditional probabilities that are obtained from the TI are not strictly constrained. Hence the conditional probabilities cover a wide range of values in the range [0,1]. However, after this initial stage, as the simulated data populate more cells in the model grid, they tend to severely constrain the conditional probabilities to assume extreme values of 0 or 1. With these extreme values at the later stages of the simulation the probability values that are included in the soft data (as secondary source of information) tend to be disregarded and the facies types are predominantly determined by the TI. We demonstrate and discuss this behavior of the SNESIM algorithm through several examples and present strategies that can be adopted to compensate for this effect. The presented examples are related to indirect integration of the flow data by first inferring probabilistic information about facies types and using the results as soft data for integration into SNESIM algorithm.
We present a novel sampling algorithm for characterization and uncertainty quantification of heterogeneous multiple facies reservoirs. The method implements a Bayesian inversion framework to estimate physically plausible porosity distributions. This inversion process incorporates data matching at the well locations and constrains the model space by adding
The proposed workflow uses an ensemble-based Markov Chain Monte Carlo approach combined with sampling probability distributions that are physically meaningful. Moreover, the method targets geostatistical modeling to specific zones in the reservoir. Accordingly, it improves fulfilling the inherent stationarity assumption in geostatistical simulation techniques. Parameter sampling and geostatistical simulations are calculated through an inversion process. In other words, the models fit the known porosity field at the well locations and are structurally consistent within main reservoir compartments, zones, and layers obtained from the seismic impedance volume. The new sampling algorithm ensures that the automated history matching algorithm maintains diversity among ensemble members avoiding underestimation of the uncertainty in the posterior probability distribution.
We evaluate the efficiency of the sampling methodology on a synthetic model of a waterflooding field. The predictive capability of the assimilated ensemble is assessed by using production data and dynamic measurements. Also, the qualities of the results are examined by comparing the geological realism of the assimilated ensemble with the reference probability distribution of the model parameters and computing the predicted dynamic data mismatch. Our numerical examples show that incorporating the seismically constrained models as prior information results in an efficient model update scheme and favorable history matching.
Uncertainty is present at every stage of the subsurface modelling workflow and understanding it is an ongoing challenge for the petroleum industry. Quantifying this uncertainty is a rapidly growing field of study as increasingly available high-performance computing enables the application of traditional statistical methods to this problem. However, the extension of these methods to spatial data remains a challenge for which there is no immediate solution. This paper describes the use of data analytics techniques to incorporate spatial uncertainty in reservoir surfaces into subsurface modelling. A metric usually applied in image analytics, the Modified Hausdorff Distance, is adapted for this purpose. The workflow involves sampling the domain of possible surface realisations, characterising them using this metric and determining the most efficient subset to represent the entire data set. The value of this process is that the selected subset captures spatial uncertainty in the surface rather than only gross rock volume. The proposed technique proved to be a simple process that was able to easily select these surfaces from a stochastically generated set and has been successfully applied to the top reservoir surfaces in two fields.
Zhang, Tuanfeng (Schlumberger-Doll Research) | Tilke, Peter (Schlumberger-Doll Research) | Dupont, Emilien (Schlumberger-Doll Research) | Zhu, Lingchen (Schlumberger-Doll Research) | Liang, Lin (Schlumberger-Doll Research) | Bailey, William (Schlumberger-Doll Research)
This paper proposes a novel approach for generating 3-dimensional complex geological facies models based on deep generative models. It can reproduce a wide range of conceptual geological models while possessing the flexibility necessary to honor constraints such as well data. Compared with existing geostatistics-based modeling methods, our approach produces realistic subsurface facies architecture in 3D using a state-of-the-art deep learning method called Generative Adversarial Networks (GANs). GANs couple a generator with a discriminator and each uses a deep Convolutional Neural Network (CNN). The networks are trained in an adversarial manner until the generator can create "fake" images that the discriminator cannot distinguish from "real" images. We extend the original GAN approach to 3D geological modeling at the reservoir scale. The GANs are trained using a library of 3D facies models. Once the GANs have been trained, they can generate a variety of geologically realistic facies models constrained by well data interpretations. This geomodelling approach using GANs has been tested on models of both complex fluvial depositional systems and carbonate reservoirs that exhibit progradational and aggradational trends. The results demonstrate that this deep learning-driven modeling approach can capture more realistic facies architectures and associations than existing geostatistical modeling methods, which often fail to reproduce heterogeneous nonstationary sedimentary facies with apparent depositional trend.
Shi, Hongfu (China National Offshore Oil Company) | Yue, Baolin (China National Offshore Oil Company) | Luo, Xianbo (China National Offshore Oil Company) | Shi, Fei (China National Offshore Oil Company) | Xiao, Bo (China National Offshore Oil Company)
The exploration and development of offshore oilfield facing unprecedented challenges include the decline in the quality of oil reserves, increase of invest and strict environmental protection policies. Usually, low permeability reservoir, heavy oil reservoir complex fault block and small reservoir located far from an existing facility are classified into marginal oilfield. More and more marginal oilfield is put on the schedule of development. In the view of economic, The internal rate of marginal oilfield return is lower than the benchmark rate of return of the industry, but higher than the cost discount rate of the industry. An integrated work flow is presented to improve the tap the potential and mitigate the risk of marginal oilfield involved in dependent development of small oilfields, unit exploitation of small oilfield group, simple platform, extended reach well and phased development. The LD oil field is taken as an example to state the strategy of marginal oilfield.
Ismail, Hasnol Hady (PETRONAS Research Sdn Bhd) | Lew, Chean Lin (PETRONAS Research Sdn Bhd) | Mohamad Som, Muhd Rapi (PETRONAS Research Sdn Bhd) | Abdul Kadir, Mohd Fauzi (Energy Quest Sdn Bhd) | Ahmad Tajuddin, Mohamad Raisuddin (Energy Quest Sdn Bhd)
Modelling of meandering fluvial reservoirs with point bars and crevasse splays is very challenging. The conventional modelling approaches, especially for meandering fluvial reservoirs, are mainly controlled by wells, which have contributed to uncertainties in lateral variations between and away from well control. Integration of the improved sedimentology, geophysics and 3D reservoir geomodelling techniques of fluvial reservoir system are proposed in the study. In stratigraphic and structural framework building, the improved methodologies included 3D seismic geobody extraction, stratal slicing and high order architectural elements interpretation. 3D geobody extraction and stratal slicing technique enhanced interpreter ability to visualize fluvial features at specific time-equivalent stratigraphic surface. In lithofacies modelling, more refined high-order architectural elements were modelled using methodologies included 3D facies seismic probability, local varying azimuth and dip angle to capture lateral accretion of point bars inside the channels for better facies distributions following point bar architectures. In property modelling, porosity was populated using Gaussian Random Function Simulation constraint to lithofacies trend to control the distribution of porosity away from wells. This methodology resulted in the porosity distributions being well controlled following the lithofacies trend. The proposed workflows and methodologies enable geomodeller to produce a more geological realistic meandering fluvial reservoir model with internal lithofacies and property distribution honouring well data and input distribution.
Xu, Wei (CNOOC Research Institute Co., Ltd.) | Zhang, Yingchun (CNOOC Research Institute Co., Ltd.) | Fang, Lei (CNOOC Research Institute Co., Ltd.) | Jing, Zhiyi (CNOOC Research Institute Co., Ltd.) | Zou, Jingyun (CNOOC Research Institute Co., Ltd.) | Liu, Jun (CNOOC Research Institute Co., Ltd.)
The Albert Basin of Uganda is located at the northern end of the western branch of the East African Rift System. It is a graben rich in oil and gas with a shallow research degree. In the south of the basin, a fan delta system controlled by the boundary fault is developed in the Miocene formation. Due to the few wells and poor quality of seismic data in this area, it is difficult to predict the spatial distribution of sedimentary reservoir sands. In this paper, sedimentary forward modeling coupled with 3D geological modeling is used to provide new ideas for reservoir prediction.
Sedimentary facies analysis is based on core description, well logs, paleontology, heavy mineral content and grain size data. Quantitative analysis of accommodation space, source supply, and sediment transport parameters can help explain the main factors that controlled the sedimentation. Milankovitch cycle method was used to establish the time scale of the basin. The simulation results were combined with 3D geological modeling to quantify the characteristics of the sand body distributions.
Sedimentary facies analysis shows that the Miocene formation in the south of Albert Basin deposited in a shallow lacustrine environment. A proximal fan delta deposition with subaqueous distributary channels was controlled by the east boundary faults. Firstly, the accommodation space was estimated according to the thickness of the stratum and the change of the ancient water depth. The source supply was estimated by the area of the project and formation thickness, and the transportation parameters were estimated according to the nonlinear transportation model based on the traction flow with a little gravity flow. Secondly, an astronomical stratigraphic framework of the Miocene strata in the south of Albert basin was established through the Milankovitch cycle stratigraphy, and it was used to restrain the process of stratigraphic forward modeling and to reproduce the sedimentary evolution process in the geological historical period. Thirdly, the stratigraphic forward modeling results were resampled into the geological model, a 3D reservoir probability distribution model is established from trend modeling to quantitatively characterize the spatial distribution of sand bodies. Finally, the sandstone distribution simulation results were transformed into quantitative control constraints for 3D geological facies modeling. Thus, the new approach significantly promotes the facies model quality and provides robust results for petrophysical property models.
Integration of stratigraphic forward modeling with 3D geological modeling can effectively solve the problem of reservoir characterization in an early stage of oilfield development through the interaction of the dual model coupling. This method has unique advantages in the reservoir research in the area with fewer data and great variation of sand.
Dhote, Prashant (Kuwait Oil Company) | Al-Adwani, Talal (Kuwait Oil Company) | Al-Bahar, Mohammad (Kuwait Oil Company) | Al-Otaibi, Ahmad (Kuwait Oil Company) | Chakraborty, Subrata (Schlumberger) | Stojic, Slobodan (Schlumberger)
Subsurface petroleum industry is burdened with uncertainties in every aspect from exploration to production due to limitations of accessibility to reservoir and technology. The most important tools used to understand, quantify and mitigate the uncertainties are geostatistical static modeling and numerical dynamic simulation geomodels. Geomodels are widely used in the industry for characterizing the reservoir and planning favorable development strategy. It is vital instrument for maximizing asset value and optimize project economics.
Static geomodels are foundation for all the advanced numerical and analytical solutions to solve the intricacies of reservoir performance. At the same time, it is where all the static and dynamic geological and engineering observations get integrated to develop common understanding of the reservoir for future studies. Understanding of the above observations and imaging of reservoir framework by individual is the basis for building static geomodels. Hence, at time, the process is highly subjective and proper QC'ing of the models to achieve the general and specific modeling objectives becomes imperative. Simple Questionaries’ based QC'ing and ranking methodologies are also controlled by subjectivity and individual preferences.
In the present endeavor, quantitative ‘Key Performance Indicators (KPIs)’ based standard static geomodeling practices and QC'ing methodologies at corporate level are developed in specially designed "Process Implementation Project (PIP) – Hydrocarbon resource and Uncertainty Management"’ under the aegis of ‘Kuwait Oil Company (KOC) - Reservoir Management Best Practices Steering Committee'.
The main objectives are to establish a practical modeling process, workflows and criteria to standardize modeling processes. A structured self-guidling modeling document has been developed with self-assemment guidelines and questionary. Finally, for each individual process a set of KPIs are specified as minimum standard to meet to obtain the approval of static model.
The present efforts are important for any geologists, geomodelers and reservoir engineers dealing with geostatistical and numerical reservoir modeling and will provide the KPI's based general practices for quality assurance (QA) and QC'ing of the models.