Static Reservoir Models Research Articles

Improved lithology prediction in channelized reservoirs by integrating stratigraphic forward modelling: Towards improved model calibration in a case study of the Holocene Rhine-Meuse fluvio-deltaic system

Stratigraphic forward modelling (SFM) provides the means to produce geologically coherent and realistic models. In this paper, we demonstrate the possibility of matching lithological variability simulated with a basin-scale advection-diffusion SFM to a data-rich real-world setting, i.e. the Holocene Rhine-Meuse fluvio-deltaic system in the Netherlands. SFM model calibration to real-world data in general has proven non-trivial. This study focuses on a novel inversion process constrained by the top surface and the sand proportion observed at specific pseudo-wells in the study area. Goodness-of-fit expressed by a new fitness function gives the error calculated as the average of two calibration constraints. Computational efficiency was increased significantly by implementing a new optimization process in two hierarchical steps: a) optimization in terms of sediment load and discharge, which are the most influential parameters having the largest uncertainty and b) optimization with respect to the remaining uncertain parameters, these being sediment transport parameters. The calibration process described allows for the most optimal combination of achieving acceptable levels of goodness-of-fit, feasible runtimes and multiple (non-unique) solutions to obtain synthetic stratigraphic output best matching real-world datasets.By removing model realizations which are geologically unrealistic, calibrated SFM models provide a multiscale stratigraphic framework for reconstructing static models of reservoirs which are consistent with the palaeogeographic layout, basin-fill history and external drivers (e.g. sea level, sediment supply). The static reservoir models that are matched with highest certainty therefore contain the highest geological realism and may be used to improve deep subsurface reservoir or aquifer property prediction.The new methodology was applied to the well-established Holocene Rhine-Meuse dataset, which allows a rigorous testing of the optimization; the calibrated SFM allows investigation of controls of the Holocene development on the sedimentary system.

Hydrocarbon Reservoir Characterization of ‘UDI’ Field, Western Niger Delta

The increasing demand for petroleum products has posed a challenge to the search for oil and gas. This search for hydrocarbon has developed due to advances in computational techniques to evaluate the probability of hydrocarbon proneness of a basin, thereby limiting the risk factor associated with hydrocarbon. This study was therefore designed to assess the hydrocarbon potential and generate a static reservoir model of UDI Field, Onshore Niger Delta.
 Well, the correlation was carried out to establish stratigraphic continuity of the reservoir sand bodies. The identified potential reservoir intervals were tied to the seismic data using available check shot survey data. With a good match achieved, seismic events were interpreted through paying attention to reflection continuity, amplitude and frequency. Interpreted horizons were converted to surfaces using a convergent interpolation algorithm. Faults within the Field showed a dominant East-West trend with two (2) major faults and five (5) minor ones. A Pixel-based facies model was built based on the normal distribution of the upscaled lithofacies log using the Sequential Indicator Simulation algorithm. Petrophysical models were built by constraining the petrophysical logs to the facies models using Sequential Gaussian simulation algorithm. 
 Four potential reservoir intervals, A100, A125, A150 and A200 were delineated. Average petrophysical parameters were computed for all the four intervals and the results revealed the reservoir intervals to be of good quality. Sand A100 has the highest average porosity value of 29.4%, while Sand A200 has the lowest value of 25.3%. Net-to-gross ratio also follows the pattern of decreasing value with depth. Sand A150 has the highest average gross thickness value, 170.4 m, while Sand A200 has the least thickness of 80.5 m. The net-to-gross ratio preserved the pattern of gross thickness and this resulted in Sand A150 still having the highest Net thickness and Sand A200 having the least Net sand thickness. The relatively large net sand thicknesses, high net-to-gross ratio values and the high porosity values all support the reservoir intervals within UDI Field to be of good quality.
 Extrapolations of reservoir properties away from good control honored the geological interpretation of reservoir Sand A125 thereby reducing the subsurface reservoir uncertainties.
 The availability of pressure data of the reservoir will help in establishing whether the reservoir is compartmentalized and hence the model can be updated to accommodate the effect of compartmentalization.

Integrated reservoir and basin modeling in understanding the petroleum system and evaluating prospects: The Cenomanian reservoir, Bahariya Formation, at Falak Field, Shushan Basin, Western Desert, Egypt

Robust inferences on understanding the reservoir heterogeneity and processes of the respective petroleum system are essential to reduce risks throughout the petroleum exploration and development phases. In this paper we present integrated static reservoir modeling and basin modeling to better characterise the reservoir rock; hydrocarbon-bearing sandstones of the Cenomanian Bahariya Formation. This reservoir is an element of the petroleum system; whose times of hydrocarbon generation-migration-accumulation, and entrapment, are also to be addressed; in Falak Field, Meleiha Area, Shushan Basin, Western Desert, Egypt. We utilized dataset including seismic profiles and well logs of four wells. Stratigraphic and structural modeling was performed to delineate the reservoir framework, then facies and petrophysical modeling, using stochastic simulation algorithms, was performed to delineate the reservoir architecture. 1D basin modeling was performed to retrieve the geohistory and maturation models of the area. Combining both reservoir and basin models was aimed, in this study, at identifying new petroleum prospects and estimating the volume of hydrocarbons within the reservoir intervals of the study area. According to the constructed reservoir and basin models, we identified three new petroleum prospect areas; to the NE, NW, and southern parts of Falak Field; that lead to reliable consistent inferences reasonably have sound impact on the oilfield development phases of the area.

A time-lapse case study in West Africa: Integrating disciplines for a complete reservoir study and field management

Development and management of deepwater oil fields requires material investment and must be supported by application of quality technological solutions and tools. To maximize the advantages that can derive from these, it is essential to integrate technologies and disciplines throughout the field cycle. This integration makes it possible to identify and fix issues early and seize business opportunities quickly. We describe the activity carried out on a field characterized by a high-quality oil-bearing clastic reservoir located in a deepwater belt along a West African continental passive margin. We highlight how geoscience specialists from both geologic and geophysical branches have contributed to construction of the static and dynamic reservoir model, field monitoring, model updating, and operations optimization. The wide geologic and geophysical experience in the area made it possible to build a petroelastic model and a sedimentological model that were integrated in the population of a geocellular model. Seismic data also were used to support reservoir monitoring of the producing field. A dedicated 4D seismic project helped our understanding of fluid movement and fault behaviors in the field. The 4D information was interpreted and shared within the integrated team with the goal of identifying further business opportunities and driving decision making over our area of interest.

Synthesis and Experimental-Modelling Evaluation of Nanoparticles Movements by Novel Surfactant on Water Injection: An Approach on Mechanical Formation Damage Control and Pore Size Distribution

Water injection is used as a widespread IOR/EOR method and promising formation damages (especially mechanical ones) is a crucial challenge in the near-wellbore of injection wells. The magnesium oxide (MgO) NanoParticles (NPs) considered in the article underwater flooding experiment tests to monitor the promising mechanical formation damage (size exclusion) in lab mechanistic scale include micro-scale classical deep bed filtration model, permeability, and pore size distribution. The averaged upper-scale equations were constructed on the water injection basis on the presence of NPs. The model validation to adjust the equation of state was obtained based on fluid samples from the laboratory and simulation tests. The permeability decline (up to 50% initial permeability) was important when the optimum value of capturing the probability coefficient (pa) is 0.7 mismatched on the conventional simulation results. Pore size distribution in each simulation time step based on retention concentrations determined in the sandstone samples. Formation damage analyses on the presence of NPs showed that modification of the static reservoir models has excellent potential regarding porosity and permeability maps, in large-scale simulation. This study displays an improved approach to NPs’ movement through a porous medium which will act as a benchmark for future waterflooding EOR projects in sandstone oil reservoirs or similar basins all over the world.

Composite rock types as part of a workflow for the integration of mm-to cm-scale lithological heterogeneity in static reservoir models

Abstract Lithological heterogeneity in the form of sedimentary structures such as cross and planar and as interfaces between massive beddings is known to impact fluid flow and fluid-rock reactions in the subsurface. Such heterogeneity exists at mm-to cm-scale and is typically not accounted for in conventional geological models due to technical restrictions posed by the wireline log resolution and the inefficiency of reservoir modelling softwares to handle such fine scale information. Hence, multiphase flow and reactive transport simulations based on conventional models fail to appropriately account for the impact of sedimentary heterogeneity. We present a new workflow which allows building reservoir models where such fine-scale lithological heterogeneity can be integrated, thereby overcoming the limitations associated with log resolution and modelling softwares. First, the rock types are coupled as composites in a systematic manner at a discretization level that can be handled by the modelling softwares. The mm-scale rock properties quantified from core samples are accounted for in the composite rock type characterization. Lithotype logs of the composites are derived by modifying the color-coded depth record based on a combination of different wireline logs. The lithotype logs form the basis for property population in the reservoir models. The coupling of core and wireline data allows to account for the mm-to cm-scale heterogeneity in coarsely discretised reservoir models. The application of the new workflow is demonstrated for the development of a 2-D reservoir model at the CO2CRC's Otway Research site with a vertical resolution of 5 cm while incorporating sedimentary heterogeneity at 5 mm scale. The composite rock type method would allow for the impact of fine scale heterogeneity to be directly accounted for in static reservoir models.

Time-lapse seismic history matching with an iterative ensemble smoother and deep convolutional autoencoder

We have developed a time-lapse seismic history matching framework to assimilate production data and time-lapse seismic data for the prediction of static reservoir models. An iterative data assimilation method, the ensemble smoother with multiple data assimilation is adopted to iteratively update an ensemble of reservoir models until their predicted observations match the actual production and seismic measurements and to quantify the model uncertainty of the posterior reservoir models. To address computational and numerical challenges when applying ensemble-based optimization methods on large seismic data volumes, we develop a deep representation learning method, namely, the deep convolutional autoencoder. Such a method is used to reduce the data dimensionality by sparsely and approximately representing the seismic data with a set of hidden features to capture the nonlinear and spatial correlations in the data space. Instead of using the entire seismic data set, which would require an extremely large number of models, the ensemble of reservoir models is iteratively updated by conditioning the reservoir realizations on the production data and the low-dimensional hidden features extracted from the seismic measurements. We test our methodology on two synthetic data sets: a simplified 2D reservoir used for method validation and a 3D application with multiple channelized reservoirs. The results indicate that the deep convolutional autoencoder is extremely efficient in sparsely representing the seismic data and that the reservoir models can be accurately updated according to production data and the reparameterized time-lapse seismic data.

Reservoir modelling notional CO 2 injection into the Precipice Sandstone and Evergreen Formation in the Surat Basin, Australia

The Lower Jurassic Precipice Sandstone and Evergreen Formation are an important prospective reservoir–seal pair for CO 2 storage in the Surat Basin, Australia. However, there is little seismic and well data to constrain reservoir modelling in the best notional injection area. To test the likely storage performance, three contrasting sector-scale static reservoir models were built to capture the range of geological uncertainty in facies distribution and reservoir properties. These considered sectors of the Surat Basin with different palaeogeographical arrangements. The models were focused on capturing detail at the interface between the top of the Precipice Sandstone (Blocky Sandstone Reservoir: BSR) and the overlying basal portions of the Evergreen Formation (Transition Zone: TZ), a critical area for understanding CO 2 injection. Object modelling was used for the BSR and lower TZ. Stochastic modelling was implemented for the upper TZ and the Ultimate Seal because these zones were less sensitive to facies distributions. Porosity was modelled stochastically, and permeability calculated using porosity–permeability transformation functions. Dynamic simulation showed the TZ has the capacity to arrest CO 2 flow out of the BSR given appropriate CO 2 injection conditions. This study shows a method of capturing uncertainty in geological heterogeneity when data are sparse or absent. The promising initial modelling results of CO 2 injection into the Surat Basin suggests that it presents a real option for carbon storage at a climate mitigation scale. Further investigation should focus on assessing other major risks associated with carbon storage such as fault seals, reactive fluid transport and the impact of legacy wells. This article is part of the Energy Geoscience Series available at https://www.lyellcollection.org/cc/energy-geoscience-series

Quantifying the impact of the structural uncertainty on the gross rock volume in the Lubina and Montanazo oil fields (Western Mediterranean)

Abstract. Structural uncertainty is a key parameter affecting the accuracy of the information contained in static and dynamic reservoir models. However, quantifying and assessing its real impact on reservoir property distribution, in-place volume estimates and dynamic simulation has always been a challenge. Due to the limitation of the existing workflows and time constraints, the exploration of all potential geological configurations matching the interpreted data has been limited to a small number of scenarios, making the future field development decisions uncertain. We present a case study in the Lubina and Montanazo mature oil fields (Western Mediterranean) in which the structural uncertainty in the seismic interpretation of faults and horizons has been captured using modern reservoir modeling workflows. We model the fault and horizon uncertainty by means of two workflows: the manually interpreted and the constant uncertainty cases. In the manually interpreted case, the zones of ambiguity in the position of horizons and faults are defined as locally varying envelopes around the best interpretation, whose dimensions mainly vary according to the frequency content of the seismic data, lateral variations of amplitudes along reflectors, and how the reflectors terminate around faults when fault reflections are not present in the seismic image. In the constant case, the envelope dimensions are kept constant for each horizon and each fault. Both faults and horizons are simulated within their respective uncertainty envelopes as provided to the user. In all simulations, conditioning to available well data is ensured. Stochastic simulation was used to obtain 200 realizations for each uncertainty modeling workflow. The realizations were compared in terms of gross rock volumes above the oil–water contact considering three scenarios at the depths of the contact. The results show that capturing the structural uncertainty in the picking of horizons and faults in seismic data has a relevant impact on the volume estimation. The models predict percentage differences in the mean gross rock volume with respect to best-estimate interpretation up to 7 % higher and 12 % lower (P10 and P90). The manually interpreted uncertainty workflow reports narrower gross rock volume predictions and more consistent results from the simulated structural models than the constant case. This work has also revealed that, for the Lubina and Montanazo fields, the fault uncertainty associated with the major faults that bound the reservoir laterally strongly affects the gross rock volume predicted. The multiple realizations obtained are geologically consistent with the available data, and their differences in geometry and dimensions of the reservoir allow us to improve the understanding of the reservoir structure. The uncertainty modeling workflows applied are easy to design and allow us to update the models when required. This work demonstrates that knowledge of the data and the sources of uncertainty is important to set up the workflows correctly. Further studies can combine other sources of uncertainty in the modeling process to improve the risk assessment.

Guest Editorial: Digital Transformation: Quest for Operational Efficiency

Guest editorial In recent times, we have been inundated with articles concerning the oil and gas industry’s race toward digitalization and automation. The pillars of such a transformation are the rapidly evolving Industrial Internet of Things (IIoT), secure cloud computing, data analytics, artificial intelligence (AI), and machine learning. In addition, advances in sensor technology have enabled important breakthroughs in near-continuous real-time measurements, potentially transfiguring our ability to derive pertinent information from real-time measurements of processes that change over time. Sensors that are able to continuously report current downhole conditions in producing wells and surface facilities have become powerful tools for managing oilfields. The skillful combination of these rapidly emerging digital and permanent sensor technologies is moving the industry toward advances in automation to optimize asset performance across the full life cycle of a reservoir with minimum human intervention. Seeing the Unseen We strive to see what is beneath the surface. Our ability to construct a digital twin image of the subsurface is solely dependent upon the data we measure. Oilfield data are classified into two categories: data at rest (archived) and data in motion (live streaming data). The huge amounts of data gathered over time pose a formidable challenge to automation. Secrets to E&P problems may lie hidden within the massive amount of accumulated data. This quandary leads me to something I read in the MIT Sloan Management Review in the early 1990, called metaknowledge, an appreciation of what we know and what we do not know. Normally, we define knowledge as all the facts we have accumulated over time. Metaknowledge is a measure of our sense of awareness of the nature, scope, limits, and uncertainties associated with our primary knowledge. Often metaknowledge is more important than primary knowledge. Metaphorically, the authors, J. Edward Russo and Paul J. H. Schoemaker, described the difference by stating that knowing when to see a doctor (metaknowledge) is more important than how much we know about medicine (primary knowledge). To image the subsurface, we must extract and interpret pertinent information (metaknowledge) buried inside the massive amount of archived data (primary knowledge). It requires Big Data solutions and the development of complex mathematical algorithms that at times can mimic the reasoning processes of the human brain through AI. The integration of 3D seismic data, through geology and petrophysics is at the core of defining a static reservoir model, which can predict reservoir performance. The static reservoir model provides a 3D database for storing 3D reservoir properties referred to as data at rest. The data at rest is updated periodically as new data are acquired (well log and time lapse 3D seismic). Streaming live data measured by permanent sensors are referred to as data in motion. Data in motion include pressure and rate measured at the wellhead or bottomhole and temperature from distributed temperature sensing (DTS) fiber-optic cables at various measurement points (nodes) in the wellbore, on the seabed, and in surface facilities.

Widespread nearshore and shallow marine deposition within the Lower Jurassic Precipice Sandstone and Evergreen Formation in the Surat Basin, Australia

Abstract In the Surat Basin of eastern Australia, the Lower Jurassic Precipice Sandstone and Evergreen Formation are a highly prospective reservoir-seal pair for notional future carbon capture and storage. However, the succession remains poorly constrained from a paleodepositional standpoint and this has impacted the capacity to construct predictive reservoir models. Here we integrate sedimentological, ichnological, and palynological data from ten cores located across a large region of the northern and central basin to produce conceptual depositional models for the interval. Our analysis shows that the Lower Jurassic Series consists of fifteen recurring sedimentary facies that are arranged into six facies associations – braidplain, lower delta plain, subaqueous delta, delta-influenced shoreface, tidally-influenced shoreline, and restricted marine shoals. The facies associations occur in the context of large scale fluvio-deltaic systems that occupied the basin. These results are supported by ichnological indications of marine and brackish-water, a coastal suite of palynomorphs, and marine palynomorphs including rare dinocysts, acritarchs, and copepod fragments. The very low abundance of marine palynomorphs are confined to the upper portion of the Evergreen Formation, and in combination with the sedimentology and ichnology suggest that marine influence increased through time as the basin was filled. The elucidation of marine-influenced deposition contravenes all but the most recent facies interpretations of the Precipice Sandstone and Evergreen Formation, and suggests that the paleogeography of the Mesozoic of eastern Australia needs to be reconsidered. Importantly, the nearshore and shallow marine depositional affinity has important implications for the size, orientation, and distribution of geobodies when building geologically-realistic static reservoir models for dynamic flow simulation.

Static reservoir modeling of El Wastani formation, for justifying development plans, using 2D seismic and well log data in Scarab field, offshore Nile Delta, Egypt

Static reservoir modeling adequately and precisely defines the reservoir framework (geometry) and architecture (property). So as to substantially reduce the expected risk while defining the hydrocarbon-bearing reservoir boundaries, static geocellular model was constructed, in this study, to define lithology and fluid stochastic distribution, and volume of hydrocarbons as well, within the Late Pliocene hydrocarbon-bearing El Wastani Formation, (sandstone channel-1 and channel-2), in Scarab Field (offshore Nile Delta, Egypt). Twenty 2D seismic lines and well log data, of four wells, were used in this study. Stratigraphic, structural, facies, and petrophysical models were applied not only to estimate the volume of hydrocarbons within the reservoir but also to identify the best location for proposing new development wells that are of crucial value for oil industry development. In accordance with the constructed reservoir geocellular model, we found that the northwestern part of Scarab field is the best location for running further oilfield development plans.

Application of geostatistical seismic inversion in reservoir characterization of Sapphire gas field, offshore Nile Delta, Egypt

Seismic reservoir characterization becomes challenging when reservoir thickness goes beyond the limits of seismic resolution. Geostatistical inversion techniques are being considered to overcome the resolution limitations of conventional inversion methods and to provide an intuitive understanding of subsurface uncertainty. Geostatistical inversion was applied on a highly compartmentalized area of Sapphire gas field, offshore Nile Delta, Egypt, with the aim of understanding the distribution of thin sands and their impact on reservoir connectivity. The integration of high-resolution well data with seismic partial-angle-stack volumes into geostatistical inversion has resulted in multiple elastic property realizations at the desired resolution. The multitude of inverted elastic properties are analyzed to improve reservoir characterization and reflect the inversion nonuniqueness. These property realizations are then classified into facies probability cubes and ranked based on pay sand volumes to quantify the volumetric uncertainty in static reservoir modeling. Stochastic connectivity analysis was also applied on facies models to assess the possible connected volumes. Sand connectivity analysis showed that the connected pay sand volume derived from the posterior mean of property realizations, which is analogous to deterministic inversion, is much smaller than the volumes generated by any high-frequency realization. This observation supports the role of thin interbed reservoirs in facilitating connectivity between the main sand units.

Geologically-Realistic Facies Distribution in Static Reservoir Models for CO2 Sequestration: From Core to Wireline Logs to Grid Cells

The Precipice Sandstone and Evergreen Formation are an important target for potential carbon sequestration in the Surat Basin, Australia. However, the geological data necessary to characterize flow units and build reliable static reservoir models is mostly absent from the basin centre where prospective storage is the highest. To deal with this limitation, detailed facies analysis was undertaken on the relatively limited core dataset, comprising 8 wells and over 2000 m of cumulative section. From this, neural networks were trained to recognize facies using wireline logs as a means to predict the distribution of geological units across the basin. Facies from 186 wells were used to produce basin wide gross depositional environment maps and the maps allowed type locations to be identified to showcase the range of possible scenarios at the basin centre where injection potential is the highest. Facies predictions were upscaled to grid cells and these then helped guide object modelling of geobodies. The result was geologically-realistic facies distributions that controlled parameterization of three alternative static models. This work is an important contribution to the world of reservoir modelling, as it shows an important method that can be used to predict a range of possible geobody distributions and corresponding flow units in areas where data is limited.

History Matching of a Channelized Reservoir Using a Serial Denoising Autoencoder Integrated with ES-MDA

For an ensemble-based history matching of a channelized reservoir, loss of geological plausibility is challenging because of pixel-based manipulation of channel shape and connectivity despite sufficient conditioning to dynamic observations. Regarding the loss as artificial noise, this study designs a serial denoising autoencoder (SDAE) composed of two neural network filters, utilizes this machine learning algorithm for relieving noise effects in the process of ensemble smoother with multiple data assimilation (ES-MDA), and improves the overall history matching performance. As a training dataset of the SDAE, the static reservoir models are realized based on multipoint geostatistics and contaminated with two types of noise: salt and pepper noise and Gaussian noise. The SDAE learns how to eliminate the noise and restore the clean reservoir models. It does this through encoding and decoding processes using the noise realizations as inputs and the original realizations as outputs of the SDAE. The trained SDAE is embedded in the ES-MDA. The posterior reservoir models updated using Kalman gain are imported to the SDAE which then exports the purified prior models of the next assimilation. In this manner, a clear contrast among rock facies parameters during multiple data assimilations is maintained. A case study at a gas reservoir indicates that ES-MDA coupled with the noise remover outperforms a conventional ES-MDA. Improvement in the history matching performance resulting from denoising is also observed for ES-MDA algorithms combined with dimension reduction approaches such as discrete cosine transform, K-singular vector decomposition, and a stacked autoencoder. The results of this study imply that a well-trained SDAE has the potential to be a reliable auxiliary method for enhancing the performance of data assimilation algorithms if the computational cost required for machine learning is affordable.

Combined Study of Static and Dynamic Reservoir Modelling for the CO2 Storage Project in Deep Saline Aquifer in Zhundong, Xinjiang, China

Geological storage of CO2 has been identified as an effective technology to reduce greenhouse gas emissions and mitigate global climate change. Deep saline aquifers are recognised as having the highest CO2 storage potential. The Junggar Basin is located in the northern Xinjiang and has extensive distributed deep saline aquifers, which could be the effective sites for CO2 storage. CO2 injectivity and storage capacity were investigated through both static and dynamic modelling on the Cretaceous Donggou Formation aquifer in Zhundong area, Junggar Basin. A static reservoir model was constructed by integrating well data and seismic attributes, and the best estimate of storage capacity (P50) was estimated to be approximately 72 million tonnes using a storage coefficient of 2.4% (P50). Dynamic simulation provided a comprehensive understanding of injectivity, storage capacity and explanation of the different storage mechanisms after CO2 injection. The total injection of CO2 was 31.4 million tonnes with five injection wells. Simulations suggest that at year 300 after injection, 28% of the injected CO2 was stored by residual trapping and 26% of the injected CO2 was dissolved into formation water. The modelling results suggest that there is good potential for large scale CO2 aquifer storage in the Junggar Basin.

Intelligent Completion in Water-Injector Well Improves Field Development

This article, written by JPT Technology Editor Chris Carpenter, contains highlights of paper SPE 192850, “Improving Field Development Through Successful Installation of Intelligent Completion in a Water-Injection Well,” by Eglier Yanez, Mattheus Uijttenhout, Maher Zidan, Rail Salimov, Salem Al-Jaberi, Al Anoud Al-Shamsi, Amnah Al-Sereidi, Mohamed Mostafa Amer, Yousef Al-Hammadi, Abdullah Abdul-Halim, Giovani Caletti, Mustapha Adli, Yousif Hasan Al-Hammadi, and Fahad Mustafa Al-Hosani, ADNOC, prepared for the 2018 Abu Dhabi International Petroleum Exhibition and Conference, Abu Dhabi, 12–15 November. The paper has not been peer reviewed. The operator piloted a new well-completion design combining inflow-control valves (ICVs) in the shallow reservoir and inflow-control devices (ICDs) in the deeper reservoir, both deployed in a water-injector well for the first time in the company’s experience. The objectives were to improve reservoir management, reduce well-construction complexity, and achieve cost optimization. This paper covers the overall aspects of well performance in commingled injection mode. Introduction The offshore field was discovered in 1971 in the Arabian Gulf. On the basis of results from exploratory wells, three main reservoirs were proposed for further development (A, B, and C). The commercial phase of the field started successfully, with first oil in 2015 provided by the two first wellhead towers as part of the Phase I development plan. The completion design described in this paper is aligned with an optimized field-development plan (FDP) resulting from a study initiated in 2017. The main objectives of the study were integrating new acquired data to upgrade static and dynamic reservoir models to identify any opportunity for extra production and cost optimization. Early water breakthrough caused by the initial elected well pattern and the associated oil-producer-to-water-injector spacing is a significant risk. Therefore, accessibility and selectivity to control inflow are strongly recommended for the upper reservoir (A) to mitigate the breakthrough risk and secure long well life and maximize reserves. The initial dual completion, with two horizontal laterals targeting Reservoirs A and B, and a nonaccessible upper drain was removed, and a single controlled completion was placed. Field-Development Plan Pattern. The FDP consisted of two phases. Phase 1 consisted of an early production scheme from two wellhead towers and export of untreated crude to nearby facilities. This phase contained an intensive data-gathering program to reduce reservoir uncertainties and identify future risks and opportunities clearly. Phase 2 is the full field development, containing the bulk of the oil producers and water-injector (WI) wells drilled from new wellhead towers and newly installed offshore treatment facilities. The overall reservoir performance of Reservoir A was significantly better compared with the original basis of assumption. Consequently, the original line-drive development scheme between oil producers and WIs had to be revised in order to avoid early water breakthrough. As a response, the injector wells were placed outside the oil pool. The behavior of Reservoir B met expectations; for that reason, the line-drive well pattern was maintained in this reservoir.

Bayesian Stochastic Inversion (A case study from an Iranian oil field)

We have implemented an estimation procedure whereby wireline data can be extrapolated away from existing well using geostatistical inversion of post-stack 3D seismic data. This procedure works directly in a fine-scale stratigraphic grid, and is conditioned by well and seismic data. It uses a Bayesian framework and a linearized, weak contrast approximation of the Zoeppritz equation to construct a joint log-Gaussian posterior distribution for Pwave impedances. Variograms are also estimated from well-log, seismic data and acoustic inversion results that define the expected degree of lateral smoothness away from the well. A sequential Gaussian Simulation algorithm is applied to sample the posterior PDF and generates multiple, high-resolution realizations of the acoustic impedance which can be utilized to generate stochastic realizations of petrophysical variables that not only honor the well-log data, but most importantly, that fully honor the 3D seismic. Sensitivity analysis was also performed to find the optimum values for S/N ratio, Variogram ranges and prior standard deviation. We had just one well in the area and preferred to keep it to perform quality control of the inversion results. Without constraining with well, geostatistical inversion still can be used to estimate acceptable static reservoir model for the subsequent simulation and planning of in-fill drilling and/or enhanced-oil-recovery operations.

Using neural networks and the Markov Chain approach for facies analysis and prediction from well logs in the Precipice Sandstone and Evergreen Formation, Surat Basin, Australia

Abstract Facies analysis is crucial for reservoir evaluation because the distribution of facies has significant impact on reservoir properties. Artificial Neural Networks (ANN) are a powerful way to use facies interpretations from core to determine equivalent facies from wireline logs in uncored wells. However, ANN do not incorporate information that relates to facies successions. This has limited the ability to effectively use facies information derived from logs alone in reservoir modelling, especially at the regional scale where data is often sparse, clustered, or incomplete. In this study, based on observations of 8 cored wells with a total thickness of ∼2000 m, 20 core facies were defined that range from 0.22 m to 11.56 m thick. Facies were based on grain size, sedimentary structures, and ichnological characteristics; Each facies corresponds to a distinct depositional sub-environment within the broader context of a large nearshore to shallow marine system. It was essential that these facies were incorporated into reservoir models to accurately map the distribution of reservoir and seal geobodies for CO2 storage assessment in the Surat Basin, Australia. However, core data are few and far between in the Surat Basin. To use core-defined facies in the absence of core, six wireline log parameters – gamma ray, density, sonic, neutron, photoelectric factor, and deep resistivity were plotted in multidimensional space and examined using Linear Discriminator Analysis. Combined with model recognition and Fisher Canonical Discriminance, the 20 core facies were simplified into 10 representative wireline log facies (WLF) with unique petrophysical parameters. We then used the Markov Chains Approach (MCA) to determine the significance of vertical facies transitions, which supported the interpretation that facies group into 5 distinct associations: (1) channel-levee complex; (2) lower delta plain; (3) subaqueous delta; (4) shoreface and; (5) tidal flats and channels. Based on the facies analysis and statistical classification, Multilayer Perceptron Classifier, a type of neural network method was applied using a training set of three cored wells that had all 6 wireline log data and using the facies successions determined from the MCA. Results show that the accuracy of WLF prediction ranges from 66% to 99% (ca. 83%). The accuracy of facies recognition decreased step wise with a decreasing number of logs as input data, such that when only gamma ray, density, deep resistivity, and sonic were used to train neural networks the accuracy dropped to between 45 and 98% (ca. 67%), depending upon the facies. This was considered the lowest acceptable threshold of accuracy for facies determination for input into reservoir models for carbon capture and storage. The results of this study show that sedimentary facies can be accurately predicted for uncored intervals in the Precipice Sandstone and Evergreen Formation to improve facies mapping and static reservoir modelling. Additionally, wireline log facies are helpful for interpreting Lower Jurassic stratigraphy, depositional setting, and basin evolution in the Mesozoic of Eastern Australia.

Geostatistical 3D geological model construction to estimate the capacity of commercial scale injection and storage of CO2 in Jacksonburg-Stringtown oil field, West Virginia, USA

As concerns around global warming increase, carbon capture, utilization and geological storage (CCUS) is a promising way to reduce the emissions of anthropogenic CO2 into the atmosphere. Sequestering the CO2 into depleted hydrocarbon reservoirs with associated enhanced oil recovery (EOR) is the most achievable approach under current economic constraints since it increases recovery of existing oil reserves, and bridges the gap between regional-scale CO2 capture and geologic sequestration. However, one of the challenges is the paucity of modern subsurface data required for adequate reservoir characterization and storage estimates. The Upper Devonian fluvial sandstone reservoirs in the Jacksonburg-Stringtown oil field in West Virginia, which have produced over 22 million barrels of oil since 1895, are ideal candidates for CO2 sequestration coupled with EOR. Reservoir storage capacity and oil recovery factors are keys for the evaluation of coupled CO2 storage and CO2-EOR process. Regression relationships between wireline logs and core measured data for porosity, permeability are constructed by artificial neural network and support vector machine in core-scale; then extended from core-scale to well-scale, where wells do not have porosity and permeability wireline logs; finally a 3D static geological model is generated based on the Random Gaussian Function simulation method and well-established variogram models generated by detailed data analysis. A static 3D reservoir model, which integrates detailed geological knowledge and existing legacy geological data from Jacksonburg-Stringtown oil field, is constructed to estimate theoretical CO2 storage capacity. Depending on the proposed 3D geological model, the best regions for coupled CCUS-EOR are located in southern portions of the field, and the estimated CO2 theoretical storage capacity for Jacksonburg-Stringtown oil field varies from 24 to 383 million metric tons. The estimated results of CO2 sequestration indicate that the Jacksonburg-Stringtown oilfield has significant potential for CO2 storage and value-added EOR.