Understanding the impact of fractures on fluid flow is fundamental for developing geoenergy reservoirs. Pressure transient analysis could play a key role for fracture characterization purposes if better links can be established between the pressure derivative responses (p′) and the fracture properties. However, pressure transient analysis is particularly challenging in the presence of fractures because they can manifest themselves in many different p′ curves. In this work, we aim to provide a proof-of-concept machine learning approach that allows us to effectively handle the diversity in fracture-related p′ curves by automatically classifying them and identifying the characteristic fracture patterns. We created a synthetic dataset from numerical simulation that comprised 2560 p′ curves that represent a wide range of fracture network properties. We developed an unsupervised machine learning approach that can distinguish the temporal variations in the p′ curves by combining dynamic time warping with k-medoids clustering. Our results suggest that the approach is effective at recognizing similar shapes in the p′ curves if the second pressure derivatives are used as the classification variable. Our analysis indicated that 12 clusters were appropriate to describe the full collection of p′ curves in this particular dataset. The classification exercise also allowed us to identify the key geological features that influence the p′ curves in this particular dataset, namely (1) the distance from the wellbore to the closest fracture(s), (2) the local/global fracture connectivity, and (3) the local/global fracture intensity. With additional training data to account for a broader range of fracture network properties, the proposed classification method could be expanded to other naturally fractured reservoirs and eventually serve as an interpretation framework for understanding how complex fracture network properties impact pressure transient behaviour.@en

We demonstrate how modelling decisions for a giant carbonate reservoir with a thick transition zone in the Middle East, most notably the approach to reservoir rock typing and modelling the initial fluid saturations, impact the hydrocarbon distributions and oil-in-place estimates in the reservoir. Rather than anchoring our model around a single base case with an upside and downside, we apply a comprehensive 3D multiple deterministic scenario workflow to compare-and-contrast how modelling decisions and geological uncertainties influence the volumetric estimates. We carry out a detailed analysis which shows that the variations in STOIIP estimates can be as high as 28% depending on the preferred modelling decision, which could potentially mask the impact of other geological uncertainties. These models were validated through repeated and randomised blind tests. We hence present a quantitative approach that helps us to assess if the static models are consistent in terms of the integration of geological and petrophysical data. Ultimately, the decision which of the different modelling options should be applied does not only influence STOIIP estimates, but also subsequent history matching & forecasts.@en

This paper successfully applied the geoengineering workflow for integrated well-test analysis to characterize fluid flow in a newly discovered fractured reservoir in the Barents Sea. A reservoir model containing fractures and matrix was built and calibrated using this workflow to match complex pressure transients measured in the field. We outline different geological scenarios that could potentially reproduce the pressure response observed in the field, highlighting the challenge of non-uniqueness when analysing well-test data. However, integrating other field data into the analysis allowed us to narrow the range of uncertainty, enabling the most plausible geological scenario to be taken forward for more detailed reservoir characterization and history matching. The results provide new insights into the reservoir geology and the key flow processes that generate the pressure response observed in the field. This paper demonstrates that the geoengineering workflow used here can be applied to better characterize naturally fractured reservoirs. We also provide reference solutions for interpreting well tests in fractured reservoirs where troughs in the pressure derivative are recognizable in the data.@en

In water driven gas reservoirs, aquifer encroachment has major implications on development decisions and production management. Water encroachment patterns depend on several static and dynamic parameters spanning numerous spatial scales, from pore to basin scale. This research focuses on improving the characterization of aquifer encroachment patterns, utilizing a large offshore, water driven, gas reservoir as a case study. An experimental design workflow applied to a numerical simulation sector-model is used to evaluate and quantify key parameters. Results of this research provide insight to improved understanding of water encroachment patterns and methods that improve the ability to model aquifer dynamics.

@en

The estimation of permeabilities is a key objective of a well test. In Naturally Fractured Reservoirs (NFRs), this is particularly important because the permeability derived from pressure transient analysis can yield quantitatve information about the characteristics of both, the fractured and matrix system. In this study, we present new insights that enable us to understand the nature of the permeability values obtained from well testing in NFRs. We considered cases that are little explored in the literature like the impact of varying apertures along the fracture planes or disconnected fractures on the effective permeability of the fracture-matrix system. We showed that if only 5% of the fracture connections are closed, the fracture network no longer percolates, that is the fractures begin to act as isolated features and the effective total permeability of the fracture-matrix system is reduced significantly. We propose a way forward for distinguishing between connected and unconnected fracture networks, which uses the ratio of well test-derived permeability and the average core permeability. We suggest that if this ratio is larger than 10, the fracture network is likely to be connected while values below 10 likely correspond to either disconnected, poorly connected or very low-conductivity fracture networks.

@en

Geological reservoirs can be extensively fractured but the well-test signatures observed in the wells may not show a pressure transient response that is representative of naturally fractured reservoirs (NFRs): for example, one that indicates two distinct pore systems (i.e. the mobile fractures and immobile matrix). Yet, the production behaviour may still be influenced by these fractures. To improve the exploitation of hydrocarbons from NFRs, we therefore need to improve our understanding of fluid-flow behaviour in fractures.

Multiple techniques are used to detect the presence and extent of fractures in a reservoir. Of particular interest to this work is the analysis of well-test data in order to interpret the flow behaviour in an NFR. An important concept for interpreting well-test data from an NFR is the theory of dual-porosity model. However, several studies pointed out that the dual-porosity model may not be appropriate for interpreting well tests from all fractured reservoirs.

This paper therefore uses geological well-testing insights to explore the limitations of the characteristic flow behaviour inherent to the dual-porosity model in interpreting well-test data from Type II and III NFRs of Nelson's classification. To achieve this, we apply a geoengineering workflow with discrete fracture matrix (DFM) modelling techniques and unstructured-grid reservoir simulations to generate synthetic pressure transient data in both idealized fracture geometries and real fracture networks mapped in an outcrop of the Jandaira Formation. We also present key reservoir features that account for the classic V-shape pressure derivative response in NFRs. These include effects of fracture skin, a very tight matrix permeability and wells intersecting a minor, unconnected fracture close to a large fracture or fracture network. Our findings apply to both connected and disconnected fracture networks.@en

Fractures can be first-order controls on fluid flow in hydrocarbon reservoirs. Understanding the characteristics of fractures such as their aperture, density, distribution, conductivity, connectivity, etc, is key for reservoir engineering and production analysis. Well testing plays a key role in the the characterisation of fractured reservoirs, especially. New advances in the Pressure Transient Analysis (PTA) have enabled the interpretation of production data in a way where the resulting geological scenarios are in better agreement with fracture patterns observed in outcrop analogues. Traditionally, Drill Stem Test (DST) data have been the primay source of information for well testing. However, we hypothesise that wireline conveyed tools designed for Interval Pressure Transient Testing (IPTT) could yield a more throrough description of the near-wellbore heterogeneities, including fractures. Hence, we investigate the applicability of IPTT for characterising fractured reservoirs using detailed numerical simulations models with accurate wellbore representation to generate synthetic IPTT responses that can obtained through a next-generation wireline testing tool called SATURN. We particularly focus on cases where fractures are present in the near-wellbore region but do not intersect the wellbore. The study included parameters such as fracture densities and conductivities, distance between fractures and wellbore and the vertical extension of the fractures across geological beds. The impact of the different fracture scenarios on the pressure transient tests was recorded as characteristic signatures on diagnostic plots (pressure derivative curves). We have called these curves "IPTT-Geotypes"; they can be used to assist the interpretation process of IPTT responses. To the best of our knowledge, this is the first time pressure derivative type curves for IPTT in fractured reservoirs are presented in the literature. A field example of an IPTT case was analysed using the concept of geological well testing. We integrated the information from petrophysical logs and the IPTT-Geotypes to assist the calibration of a reservoir model developed to represent the geological setting of the tested reservoir interval. The results provided a sound interpretation of the reservoir geology and quantitative estimation of the matrix and fracture parameters.

@en

Fractures often influence production in hydrocarbon reservoirs, yet the pressure transients observed in the wells might not show the conventional well-test signatures. In this case, the effect of fractures on production would be misinterpreted or even completely missed. The heterogeneous nature of fractured reservoirs makes them difficult to characterize and develop. In addition, the location of a producer within the fracture network also affects the pressure response; however, conventional well-test analysis assumes that the producer is located in symmetrical fracture networks. In this paper we investigate the effects of variations in fracture conductivity and location of the producer in the fracture network on the pressure-transient responses. To overcome the limitations of the dual-porosity (DP) model, this study uses a discrete fracture/matrix (DFM) modeling technique and an unstructured-grid reservoir simulator to generate pressure transients in all analyzed fracture networks. Our rigorous and systematic geoengineering work flow enables us to correlate the pressure transients to the known geological features of the simulated reservoir model. We observed that the simulated pressure transients vary significantly depending on the location of the producer in the fracture network and the properties of the fractures that the producer intercepts. Our findings enable us to interpret some unconventional features of intersecting fractures with variable conductivity. We observed that the behavior of two intersecting fractures, in which the well asymmetrically intercepts a finite-conductivity fracture, can be similar to that of a well intercepting a fracture in a connected fracture network with uniform fracture conductivity. Furthermore, a well intercepting a finite-conductivity fracture in naturally fractured reservoirs (NFRs) with both finite- and infinite-conductivity fractures would yield a DP response (V-shape) that might otherwise be absent if the fracture network is assumed to have uniform conductivity.

@en

An innovative multi-deterministic scenario workflow was applied to one of the giant and complex carbonate reservoirs in the Middle East. The application of this workflow had the objective to quantify how geological uncertainties and different modelling decisions impact the stock tank oil-initially-in-place (STOIIP) estimates and flow behaviour in this reservoir. In particular, we focused on the uncertainties related to the presence of fractures, reservoir rock typing, and modelling the initial hydrocarbon distribution. Based on the available static and dynamic data we considered two key scenarios, the absence of fractures and the presence of sparse, fault-controlled fractures. In the first scenario, we investigated how different reservoir rock typing methods impact permeability distributions. We further quantified changes in hydrocarbon distribution and analysed how a novel approach that combines capillary pressure and log-derived J-function affects the saturation models. In the second scenario, we used the effective medium theory to calculate permeability multipliers for the regions where fractures are expected. This enabled us to effectively represent fractures in a single-porosity reservoir model. The representativeness of the different models was analysed through blind tests using static data as well as history matching using dynamic data. The most significant findings of our work are that subtle changes in modelling decisions and reservoir rock typing have major consequences for the saturation model, leading to up to 20% change in STOIIP estimates. Such uncertainties must be carried forward in future reservoir management decisions and when estimating reserves. The blind tests showed that a saturation model based on the combination of core- and log-derived J-functions gave the most robust STOIIP estimates. These particular saturation models further led to a much-improved history match, especially for wells located in the transition zone of the reservoir. The best history matches were obtained once sparse, fault-controlled fractures were included in the reservoir model using effective medium theory. The presence of fractures specifically improved the history matching quality for wells located close to the faults; these wells were very difficult to match in the past. Our work clearly demonstrates that a multi-deterministic scenario workflow is key to explore the appropriate range of geological uncertainties, and that, equally important, the impact of different modelling decisions must be accounted for when quantifying uncertainty during reservoir modelling. This is particularly applicable to giant carbonate reservoirs where relatively minor changes in the workflow and data interpretation can have major consequences on STOIIP estimates, dynamic behaviours, and reserve estimates. Multi-stochastic modelling workflows which anchor the reservoir to a single base case are not capable of achieving this.

@en

A geological well testing model containing fractures and matrix was built to match an observed complex well test response in a reservoir that was known to be fractured. The resulting match of a well intersecting minor fractures near a major fracture was considered a good match in consideration of alternative geological scenarios. This match leads to a better understanding of the key reservoir features to reliably support development decisions

@en

We apply, evaluate and compare several reservoir rock typing (RRT) approaches for a giant carbonate reservoir in the Middle East that is characterized by poor rock quality and a thick transition zone. These workflows are used to define RRTs by comparing several approaches, namely pseudo clouds from porosity and permeability data for each subzone, global hydraulic elements, petrophysical grouping using MICP data and Lucia classes. Then, the permeability models obtained from each RRT approach are used as input data for building the water saturation models using two different approaches, Leverett J-functions and the Geo2Flow software. The application of this workflow leads to a significantly change of the saturation distribution and subsequently revised the oil-in-place estimation for the field, especially because the reservoir has a thick transition zone. The variations in saturation distributions and oil-in-place arising from the different RRT approaches are likely larger than the uncertainty in the geological data itself. The saturation distributions do not only impact oil-in-place estimates but also predictions for the subsequent reservoir simulation models.@en

We apply a new and geologically consistent workflow to re-evaluate the saturation distribution in a giant and complex carbonate reservoir from the Middle East that comprises a thick transition zone. The workflow makes use of all the available capillary pressure, porosity, and permeability, data, irreducible water saturations, Archie exponents and cementation exponents. By using these data, the new workflow enables us to model hydrocarbon distributions more robustly. Hence the resulting 3D saturation model honours saturation logs at the wells exactly and allowed us to identify new reservoir compartments, leading to improved estimates of hydrocarbon reserves, and an estimated change in reserves of over 20%. The workflow is embedded in the Geo2Flow software and can be run conveniently from Petrel. The new saturation distribution needs to be considered as a new model scenario for any subsequent dynamic reservoir simulations.

@en

Naturally fractured reservoirs (NFRs) account for a significant amount of the world conventional reserves but suffer from low recovery factors. Multiple techniques are, often in combination, used to detect the presence and extent of fractures in a reservoir. Of particular interest to this work is the use of dual-porosity model for analysis of well test data in order to identify and interpret the behaviour of fluid flow in NFR. This model, originally developed by Warren & Root (1963), has since been the industry standard for modelling NFRs and interpreting well-test data from NFRs. However, several studies have shown that the dual-porosity responses expected for naturally fractured reservoirs are not always observed where the wellbore intersect fractures, even for heavily and well-connected fractured network reservoirs. Our research aims to examine the reservoir features that cause the dual-porosity response to absent in some naturally fractured reservoirs and to be present in others. We demonstrate when dual porosity models are valid for well-test interpretation and can capture the key reservoirs features that characterise flow behaviours in NFRs. These features include the effect of fracture skin, network connectivity, and network size. The findings allow us to interpret well-tests in NFR more reliably.

@en

Geological well testing is a valuable tool that allows us to improve understanding of pressure transient behaviour in a fractured reservoir. However, not all wells in a fractured reservoir will show pressure transients that are expected for NFRs. Our findings demonstrate that high resolution models with proper grid refinement around the wells and fractures are required to model pressure transient behaviour adequately and produce a physically meaningful wellbore response for a fractured reservoir. The key concept for interpreting well test data from fractured reservoirs is the dual-porosity model. This model, originally developed by Warren and Root (1963) has been the industry standard for modelling NFRs and interpreting well-test data from NFRs for more than 50 years. Although there are a number of factors impacting the exact shape of the pressure transients, our results suggest that observing the classical “V-shape” in the pressure derivative, as expected from a dual-porosity model may be an exception, rather than a rule in NFR, even for well-connected fracture networks. Our work quantifies when and why the assumptions inherent to the dual-porosity model break down when interpreting well-test data from NFR. @en

Field X comprises a giant Palaeogene limestone reservoir with a long production history. An original geomodel used for history matching employed a permeability transform derived directly from core data. However, the resulting permeability model required major modifications, such as horizontal and vertical permeability multipliers, in order to match the historic data. The rationale behind these multipliers is not well understood and not based on geological constraints. Our study employs an integrated near-wellbore upscaling workflow to identify and evaluate the geological heterogeneities that enhanced reservoir permeability. Key among these heterogeneities are mechanically weak zones of solution-enhanced porosity, leached stylolites and associated tension-gashes, which were developed during late-stage diagenetic corrosion. The results of this investigation confirmed the key role of diagenetic corrosion in enhancing the permeability of the reservoir. Insights gained from the available production history, in conjunction with petrophysical data analysis, substantiated the characterization of this solution-enhanced permeability. This study provided valuable insights into the means by which a satisfactory field-level history match for a giant carbonate reservoir can be achieved. Instead of applying artificial permeability multipliers that do not necessarily capture the impacts of geological heterogeneities, our method incorporates representations of fine-scale heterogeneities. Improving the characterization of permeability distribution in the field provided an updated and geologically consistent permeability model that could contribute to the ongoing development plans to maximize incremental oil recovery.@en

Obtaining a fit-for-purpose rock-type classification that adequately incorporates the key depositional and diagenetic heterogeneities is a prime challenge for carbonate reservoirs. Another prevailing issue is to integrate the static and dynamic data consistently with the rock-typing scheme in order to correctly initialise the reservoir flow simulation model. This paper describes a novel near-wellbore rock-typing and upscaling approach adopted to address the crucial challenges of integrating reservoir rock-typing and simulation in carbonate reservoirs. We demonstrate this workflow through a case study for a highly heterogeneous Eocene-Oligocene limestone reservoir, Field X. Geological studies carried out in Field X suggested that the key permeability pathways are strongly related to the mechanism of reservoir porosity and permeability evolution during late-burial corrosion. The rock-typing and upscaling methodology described in this paper involves the geological-petrophysical classification of the key reservoir heterogeneities through systematic evaluation of the main paragenetic events. Associations between the depositional and late-burial corrosion features, and their impact on reservoir flow properties, were accounted for in our workflow. Employing near-wellbore rock-typing and upscaling workflow yielded consistent initialisation of the Field X reservoir simulation model and therefore improved the accuracy of fluids-in-place calculation. Subsequently, the cumulative production curves computed by the reservoir simulation model of Field X showed closer agreement to the historic production data. The revised Field X simulation model is now much better constrained to the reservoir geology and provides an improved geological-prior for history matching.

@en

This study illustrates a novel near wellbore rock typing and upscaling approach to obtain fit-for-purpose rock typing scheme to improve the characterisation and simulation of a highly heterogeneous offshore carbonate field, Field X. The key geological heterogeneities present in Field X were caused by late burial (mesogenetic) dissolution, such as chalky micro-porosity, macro-porosity including vuggy and moldic pores, leached stylolites and associated tension gashes. The prime challenges addressed are to incorporate the influence of diagenesis on reservoir petrophysical properties, adequately represent the multi-scale and multi-modal pore types and integrate dynamic data through rock typing. Another major issue is to upscale the petrophysical properties of the rock types to the reservoir model using appropriate geostatistical tools. The rock typing and upscaling methodology we adopted involves the geological-petrophysical classification of these heterogeneities through systematic evaluation of the key paragenetic events, and considers the crucial aspects of near wellbore modelling and upscaling. The outcome of this study has significantly improved characterization of porosity and permeability distributions and dynamic calibration in Field X. As a result, a new alternative geomodel scenario was obtained, which is now much better constrained to the reservoir geology and shows improved match to historic production data.@en

New reservoir characterization methods are needed to integrate multiscale exploration and development data, particularly at the interface between well and field models. In this paper, we illustrate a novel workflow involving high-resolution near-wellbore modeling (NWM), which allows us to accurately include seismic, wireline data, image logs, and well core logs from highly heterogeneous reservoirs in field-scale reservoir simulations. We demonstrate that an NWM-enhanced geoengineering workflow has the potential to improve reservoir characterization by applying it to a realistic clastic reservoir with high variance at small scale. We have performed a number of sensitivities comparing conventional local grid refinement (LGR) in the near-wellbore region with that involving NWM, and we obtained a significant increase in the accuracy of reservoir characterization and the calibration of dynamic models. Centimeter-scale models, containing several million cells, representing the fine geological details of the nearwellbore region, were constructed with available data from core and openhole well-log suits. The resulting well models were upscaled into regular grids with the highest resolution possible through the NWM software and incorporated into a field-scale simulation model to evaluate the dynamic behavior of the reservoir with a static-model transient test. Our results show that the use of NWM tools for reservoir modeling yields more precise flow calculations and improves our fundamental understanding of the interactions between the reservoir and the wellbore.@en

We have investigated the role of high-frequency cycles (HFCs) in carbonates during water flooding. These are normally sub-seismic and therefore below grid resolution and must be upscaled, both in terms of single- and multiphase flow behaviour, to predict time to water breakthrough, recovery factors, and the habitat of remaining oil properly. Particularly the latter could be important when designing appropriate enhanced oil recovery schemes that target the remaining oil. The importance of selecting a high vertical resolution grid has been demonstrated for HFCs with continuous petrophysical log gradients and discontinuities, and the consequence these high permeability ranges have on waterflood velocity has been presented: capturing all of a continuous gradient requires increasing the vertical resolution so as to predict a fast enough water front, and a realistic distribution of the residual oil within the HFCs. In addition, the importance of selective diagenesis has been investigated by considering the effect of early-stage diagenesis at the top of HFCs in end-member Greenhouse and Icehouse climates. HFCs are very common in carbonates, even if they are nested within larger-scale heterogeneous geobodies, and the value of this investigation is in its application to real reservoir models.@en

Carbonate reservoirs host a major portion of the world's remaining conventional and unconventional hydrocarbon reserves, typically containing multi-scale geological heterogeneities varying over many orders of magnitude in size. Characterizing and representing them robustly in reservoir models is a prime challenge in carbonate reservoir simulation. One of the key aims of this paper is, hence, to present a novel near wellbore upscaling (NWU) workflow that addresses the challenges associated with conventional carbonate modelling workflows. The NWU workflow provides a systematic geostatistical approach to obtain more realistic representation of multi-scale geological-petrophysical heterogeneities in complex carbonate reservoir simulation models. Using well log and core data, near wellbore regions were recreated to represent the core scale heterogeneities via high resolution geostatistical models. These core/centimeter scale permeability models were then upscaled into wireline/decametre scale using flow-based upscaling. The results, coupled with wireline data were used to generate global porosity-permeability and vertical-horizontal permeability relationships for reservoir simulation. Importantly, the workflow mitigates sample bias, which is frequently observed in the core data for carbonate reservoirs. We have applied our approach to a mature carbonate field, to model and upscale crucial multi-scale heterogeneities ubiquitous in the reservoir. These heterogeneities, such as mechanically weak zones of enhanced micro- and macro-porosity, leached stylolites and associated tension gashes, were caused by diagenetic corrosion. Core plugs representivity is always an issue in carbonates and these highly corroded features were very difficult, if not impossible, to sample due to their fragility. As a result, the field suffers from inherent sample biasing and insufficiency of Routine Core Analysis (RCA) data, consequently underestimating the permeability in the simulation model. The workflow presented here has enabled the authors to re-evaluate the reservoir permeability model by accounting for as yet under-sampled geological heterogeneities. The paper represents a focused individual study addressing this specific issue and doesn't necessarily reflect the operator's full understanding of this multifaceted field. Our new permeability model has addressed the need for artificial permeability multipliers and provided insight on the potential causes of the original mismatch. As a result, a new alternative model scenario has been built to help guide the on-going development plans and forecasting incremental oil recovery.@en