For every barrel of oil, about three to four barrels of water is produced. Water is part of every operation in upstream oil and gas: we produce it, we process it, we inject it. It affects our reserves because it may drive or sweep the oil out of the pores. It is a source of corrosion and scaling in pipe and in the reservoir. Measuring formation water resistivity (R w ) goes beyond using it as the basis of petrophysical well log interpretation. It is the key to telling different waters apart for taking the most representative samples. We introduce a calibrated induction-based water resistivity measurement sensor, which is configured to accurately measure R w in the flowline of a formation testing tool. The induction-based operating principle of the sensor eliminates the use of electrodes and the associated fouling of the measurement due to coating or accumulation of particles on the electrodes. Instead, the sensor induces an electric current through a nonconductive, neutrally wetting flowline tube that is proportional to the conductivity of the fluid column within the tube. The resulting current at the receiver coil is then converted into resistivity. A case study presents data from a focused water-sampling station in a transition zone in a well drilled with water-based mud (WBM). The resistivity contrast between the mud filtrate and the formation water is low and mobile oil mixes with the formation water and mud filtrate. Despite these difficult conditions, the downhole measurement clearly shows the cleanup progress in real time and compares well with the surface measurements of the water samples. The ability to differentiate formation water from WBM filtrate with low resistivity contrast in the presence of oil places the station depth in the transition zone and enables accurate interpretation of contacts, saturation, and ultimately hydrocarbon in place. The sensor package is suitable for use up to 200-degC temperature and 35,000-psi pressure. The sensor can measure a wide range of resistivity, from 0.01 to 65 ohm.m. Measurements performed on known fluids prove its high accuracy of ±5% or less for resistivity below 10 ohm.m at a resolution of 0.001 ohm.m. The design eliminates any dead volume and all flowline fluid passes through the sensor. The sensor tube is smoothly flushable for fast dynamic response in multiphase slug flow. This paper also discusses optimal sensor placement and operational techniques to achieve best results in multiphase flow environments. The accuracy and resolution of the resistivity measurement enables direct comparison of guard and sample flowlines during focused sampling and provides differentiation even when the contrast between filtrate and formation water is low. The results can serve as a direct R w measurement, for example in an exploration scenario, as successfully shown in another PDO trial, or can be compared to other sources of R w measurement or used to improve the accuracy of alternatives to the Archie equation, such as dielectric dispersion.

Wireline formation testing has evolved from discrete pressure measurements, introduced in the 1950s to measuring pressure gradients and fluid contacts since the 1970s. Technology introduced in the late 1980s and onwards added interval pressure transient testing, focused sampling, and downhole fluid analysis. Thirty years later, this paper shows data examples of a recently developed formation testing platform in a wide range of environments, and applications, that change how we plan, acquire, and use formation testing. The dual-flow-line architecture of the formation testing platform is designed to systematically address shortcomings of legacy technology, enabling focused sampling in the tightest conventional formations, as well as transient testing in high mobility environments. Specialized pre-job planning software evaluates conveyance options to minimize friction and borehole contact, estimates the available flow rate, compares cleanup performance of the different inlets, and simulates transient testing responses. During the operation, the platform uses hardware embedded automation algorithms that execute routine tasks in a consistent and highly efficient manner, leaving more time for the user to focus on data quality and value of the measurements. Case studies from Mexico, Norway, and the US demonstrate specific improvements in capability and performance. Field data from Mexico shows focused sampling of gas condensate from a heterogeneous submillidarcy carbonate formation in an HP/HT well drilled with oil-based mud. Controlled downhole decompression of crude oil in the flowline at a sampling station in Norway enabled real-time measurement of its bubblepoint pressure to within 6 psi of that measured in the laboratory. Another case study integrates accurate relative asphaltene gradients into an existing reservoir fluid study to prove reservoir connectivity across a large lateral distance in a producing field. Application of the dual packer subsystem demonstrates inflation within four minutes and pure oil samples within 90 minutes on station in a 1.5-md/cp fractured basement formation. The fine pump control at a low rate enabled sampling just below reservoir pressure in Alaska and a case from the Gulf of Mexico demonstrates the real-time impact of fluid properties on the understanding of reservoir architecture and completion design. The presented examples highlight the impact of downhole automation, define the new operating envelope for formation testing in the most challenging environments, and highlight how the technology development leads to decision making on a broad reservoir scale by providing contextual answers rather than an accumulation of facts and figures.

A major concern in offshore production is the well intervention cost and the time taken to perform the operations. These issues relate to the need for a rig or a similar unit to perform those operations. This paper presents the current stage of the development of an autonomous robot for rigless well interventions. The robot enters the well through the Christmas Tree using a special tool. This proposal will allow significant cost reduction by avoiding the need for a rig to do light workover operations in offshore locations. This strategy reduces intervention time since a simpler unit may deploy the robot. Examples of operation performed by the robot are production logging, gas lift valve replacement, sliding sleeve opening and closing and plug insertion or removal. A Simplified Robot allowed evaluating the technical feasibility of this proposal, both theoretically and experimentally. This Simplified Robot allowed evaluating the Traction System, the Energy Storage, and the Positioning Algorithm. After that, the design of the Real Robot started. This system will be modular and adjustable according to the needed operation.

Wireline formation testing has been routinely used for reservoir characterization in the Gulf of Mexico. The significant impact that fluid composition has on designing and operating production and processing facilities for these complex reservoirs requires comprehensive fluid sampling during exploration and appraisal. Vertical and lateral connectivity, fluid compatibility when commingling to meet production targets, or asphaltene onset potential along the production flow path are examples of the need for thorough understanding of fluid property variations. This case study combines the use of the newest generation of formation testing platform with innovative workflows to acquire critical fluid data and refine the geological reservoir model in near real time. Conveyance difficulties foreseen in the S-shaped well suggested that the use of a shorter, lighter tool could increase the chances of success in a field where past keyseating episodes left few reliable alternatives for conveyance apart from drillpipe-conveyed logging. The application of the smart sampling platform capabilities introduced important changes to the standard protocols used in these cases. The increased cleanup efficiency of the new formation testing technology enabled fast investigation of reservoir fluids at multiple depths with robust metrology for specific properties. The outstanding sample purity formed the foundation of advanced interpretation workflows. With its high-quality fluid data, this case was used to perform a "proof of concept" of advanced interpretation workflows aimed at updating geological and fluid models in near real time once the downhole fluid analysis data become available at the end of each sampling station. The formation testing data and the near real-time answers expedited tailoring the completion of the well, which involved a nontrivial commingle decision. The fluid information also provided clarity for the geological model in a section of the field where the current seismic interpretation had shown uncertainty for the reservoir architecture in the downthrown block of the field. The combination of new formation testing hardware and short-cycle workflows makes this case study the first of its kind. The proof of hardware performance and data processing/interpretation capability cements the case for higher-density data streams and a shortened data life cycle. It serves as a preview of how digitally enabled formation testing operations will be conducted in the future to provide new insights to reservoir understanding.

Well integrity is very important for well operation in oil and gas industry. Advanced technologies and analyses have been developed to ensure well integrity. One of the most important analysis to be done is the evaluation of the cement bond which measures the presence and bonding of cement between casing and the formation in a particular depth or interval. This evaluation is critical for hydraulic isolation to withstand subsequent completion and production operation. Usually, Cement bond evaluation is done from wireline logging, with a tool that transmits acoustic waves and compute the acoustic energy propagating through the casing, the cement and the formation. The amplitude and the attenuation of the returned wave can be correlated with Bond Index (BI), due to the difference in acoustic impedance between casing and cement. Over years, different wireline technologies have been developed and matured for cement bond evaluation. Nowadays, it can be found scenarios where wireline operation is not feasible, such as in highly deviated wells or when a quick turnaround is needed. Acoustic technology in Logging While Drilling (LWD) has been developed to measure the compressional and shear slowness of the formation. These devices are able to transmit acoustic wave at different frequencies and receive them after propagation through a medium, applicable in cased hole condition. This paper presents a case study where the LWD Acoustic technology was successfully used to evaluate the cement bond in a high deviated well in Brazil and the results show the capabilities of the technology to generate a qualitative cement bond data quality. These results were successfully validated using sonic and ultrasonic wireline technologies. This comparison, between LWD acoustic and wireline results, is shown in this paper, including details of the methodology applied to the test, parameters considered and the data processing.

The Operator and the license partnership have set an extremely high ambition for recovery from the Johan Sverdrup field, even before a barrel of oil has been produced. How is this possible? This paper describes the characteristics of the reservoir, as well as early assessments and investments for improved oil recovery (IOR) to ensure flexibility. In addition, data acquisition, reservoir monitoring, new technologies and digitalisation, as well as new ways of working are addressed. This will be the key enablers for a recovery of more than 70% of the field’s oil resources. Johan Sverdrup is the third largest oil field on the Norwegian Continental Shelf (NCS) with a recoverable volume range of 2,2 to 3,2 billion b.o.e. The reservoir is characterized by excellent reservoir properties with a strongly undersaturated oil. The primary drainage strategy is water flooding, including re-injection of all produced water, supplemented by water-alternating-gas (WAG) injection at the end of the oil production plateau. The field came on stream in October 2019. Going back to the early stages of the Johan Sverdrup field development, it was obvious from the start that this would be an independent development solution with a long lifetime. Given the excellent reservoir, this was considered as a unique opportunity to plan for a high resource exploitation, and make sure that future business opportunities in this context could be utilized in a technical and economically attractive way. A very early screening was conducted to investigate which IOR measures should be further matured. With subsurface evaluations as the base, this maturation also included assessments on technical feasibility and potential implications for development solutions. The objective was to ensure sufficient flexibility in early field design. It also implied that the Johan Sverdrup license had to consider pre-investments prior to any implementation decision. Data acquisition and reservoir monitoring strategies were also started early on, which e.g. led to a full field Permanent Reservoir Monitoring (PRM) decision, with installation starting summer 2019. This gives a baseline for parts of the field before production start, and when completed in 2020 it will be the world’s largest fiber based PRM system. Fiber optics are also installed in the wells. In addition, a dedicated observation well is part of the development plan. The idea is that PRM and fiber data results, in addition to repeated logging in the observation well, will be key information to evaluate business cases for future IOR or new technology measures. Digitalisation has also been a key aspect of this, and several subsurface-focused digitalisation initiatives have been implemented during the field development, giving the operator the opportunity to implement new ways of working and enabling new ways of cooperation in the partnership as data and applications are shared within the owner group in a digital setting. The overall objective of digitalisation in this context is to further optimize the analysis and management of the Johan Sverdrup reservoir – and hence value of the Johan Sverdrup field – for the license owners.

This research advances the use of microwave radar to assess geomaterial properties by the novel application of frequency modulated continuous wave (FMCW) radar to uniaxially loaded sandstones in the laboratory. Here contrasts in reflection coefficient are interpreted to represent microfracture formation and cross-linking under compressive loading prior to through-going sample failure. The Darney and Red St. Bees sandstones are characterized via optical microscope, X-ray radiography and tomography and by SEM methods and are tested for FMCW response under uniaxial compressive loads, with results showing that FMCW interrogation in the K-band is highly sensitive to the progression of bulk axial shortening and material bulk composition. The results reported here present an external, non-contact radio-frequency sensing modality using microwave radar to evaluate the samples as they are deformed. These samples are sandstones chosen for their distinct petrophysical properties and their significant differences in textural heterogeneity and anisotropy. This novel sensing application yields repeatable and consistent results, which are currently verified for four laboratory-deformed sandstone samples, where characteristic spikes in reflection coefficient, Γ, are observed. These spikes, which have not been observed in any other sensing modality, and occurring approximately 20 seconds prior to macroscopic yield, were observed in all samples tested. This research offers new insights into the evolution of deformation and failure in a porous geomaterial, here sandstone, and represents a departure from the current practices of post failure analysis. The effects of these new measurands (and how they change as a function of load) can be used to recognize more clearly the step-wise evolution of the sample progression to bulk failure and potentially be developed to provide a precursor warning of imminent sample failure. By increasing understanding of the damage development this can help refine the associated geomechanical theory. Better theory underpinning better calculations and simulations of a geomechanical reservoir issue can be critical in managing key seal and trap properties of a reservoir, enabling the operator to make improved operational decisions, increasing field productivity and production efficiency.

PNPR cluster consists of three fields, namely PX, NX and PR (combined STOIIP ~200 MMstb), located ~300 km offshore of Peninsular Malaysia. Throughout its journey of monetizing marginal waxy crude, many challenges and hurdles have arisen, including sustaining oil production rate above economic threshold, pipeline clogging, and FPSO fuel uncertainties, which requires collaboration between surface and subsurface team to develop unique solutions in managing these downturns. Critically, PNPR cluster is expected to reach economic limit within few years’ time. This paper will elaborate on how IOR is achieved in PNPR cluster, historically and in near future. Ever since first production by PX and NX in 2004, infill drilling campaigns have been needed to sustain production above the economic limit of 5,000 bopd. Later in 2009, approximately a year after PR kicked off its first oil; the14 km pipeline to FPSO was plugged due to wax accumulation as a result of prolonged shutdown. A pipeline restoration project was embarked on involving installation of pipe in pipe (PiP), which utilizes hot water circulation as pipeline heating element. Another complexity has been consistently supplying gas to the FPSO for fuel, which involves a cement packer and adding perforation jobs in gas wells. Additionally, the waxy crude in these fields requires gas lift to be produced, particularly after water production started to escalate. This gives an opportunity to introduce through tubing electrical submersible pump (TTESP) to the field, while reducing dependence on gas lift. Financial wise, cost optimization initiatives are necessary to maintain the operability of the fields. To date, five infill projects have been successfully completed, contributing to IOR by bouncing back PNPR oil production rate. Additionally, a gas cap blow down (GCBD) from NX J80 reservoir also managed to improve reservoir recovery factor (RF) while supplying additional gas for fuel. Meanwhile, the PiP system, an enabler for IOR, has successfully ensures smooth crude oil delivery above pour point temperature from PR Platform to FPSO. In terms of gas fuel supply forecast, proper gas wells production phasing is planned to secure steady supply until 2023. IOR through artificial lift, TTESP is planned to be executed soon in one idle production well with potential gain of 500 bopd, hence eliminating option to workover the well, which is costly. Viewing IOR from economic standpoint, operating expenditure (OPEX) reduction through new philosophies were implemented, including reduction of FPSO charting rate, proactive maintenance and low-cost chemical bull heading, resulting in better cash flow for PNPR. It is expected that existing PNPR wells can recover 2 MMstb of oil through extension of economic life via incoming infill drilling in 2021, translating into 1-2% increase from current RF. Moreover, PX and NX already produced ~80% more reserves than originally booked in the first FDP.

In this paper, we examine the evaluation of a low resistivity pay siliciclastic reservoir in Bohai Bay, China. The reservoir consists of fine to silt sized sand grains with dispersed illite and smectite minerals filling the pore throats. Due to the large surface area associated with the small grains and the clay minerals, a significant amount of irreducible water is bound to the rock surface, dramatically lowering the resistivity of the pay zone. This phenomenon makes open hole log interpretation quite challenging because pay zone resistivity can sometimes read very close to that in water zone. Wireline formation tester shows oil gradient in sands where open hole logs indicate as wet and shaly. However, these sands were excluded from the initial reserves estimation so that uncertainty is minimized. In order to confirm hydrocarbon production potential and grow reserves in the low-resistivity sands, the operator perforated these sand intervals and acquired production logs. The borehole fluid density and water holdup from production logs clearly indicate that there's hydrocarbon production coming from the low resistivity zones. Detailed analysis revealed that these sands contribute 46 cubic meters of oil per day with no water production. This result confirms our theory that the low resistivity is caused by bound water trapped in clay minerals. These rocks are actually saturated with oil with almost no free water. Based on this finding, we made a few adjustments to our petrophysics model. We modified the amount of bound water associated with clay so that there's no free water in the zone. Also, we adjusted the parameters used in the shaly sand saturation equation to better account for the conductivity contributed by clay. With this improved interpretation model, we made production predictions for a few offset wells based on their open hole logs. Additional perforations were added in the low resistivity sands of these wells according to log interpretations. Production history of these wells are highly consistent with our predictions and shows that the low resistivity sands produce good quality oil with almost no water. These newly discovered pay zones increased proven reserves by about 430 thousand tons. Production logging is typically a production diagnosis tool to improve the understanding of reservoir petrophysics. The novelty of our idea is that we use production logging to provide the ground truth of reservoir fluids in the low resistivity pay and use this information to improve our petrophysics model. The improve model allows us to accurately predict the production from low-resistivity pay sands in other wells, without the need for further production logging.

The largest risks in deepwater well construction often occur in the shallow, surface hole sections where seismic uncertainty is high. This seismic uncertainty translates into increased error when estimating the position and extent of shallow hazards such as gas charged sediments, high pressure zones, wellbore stability concerns, faults, etc. Should these hazards be encountered unexpectedly, or without adequate preparation, then the associated operational costs and risks to safety can be extremely high. It is therefore critical that these risks are managed appropriately to ensure drilling success. Minimizing the uncertainty in the surface seismic requires adequate knowledge of both the compressional and shear velocities of these shallow sediments. Being able to turn this information into real-time drilling solutions to mitigate risk requires the acquisition of these measurements using logging while drilling (LWD) sonic tools in a wide variety formation types, often in large, and potentially washed out, surface holes and in a wide variety of mud types. This paper will discuss the physics of acquiring quadrupole shear measurements in large boreholes and slow formations. The limits of measurement will be explored, using both modeled data and real well examples, in a variety of formations, with different mud types, and in different borehole sizes. The effect on measurement quality as a result of the interplay among these three key parameters is revealed. The results show that, although the acquisition of reliable shear data in these conditions is challenging, it is not impossible, provided that the LWD tool is correctly designed, the physics are understood, and suitable processing applied. The latest generation of LWD sonic tools can therefore offer unique solutions to managing drilling risk. As LWD sonic tools have become more advanced, they have become more routine in use and are now commonly used in multiple hole sections from surface to total depth. Utilizing the compressional and shear measurements from these tools as inputs into analysis to address concerns such as wellbore stability or seismic remigration requires a thorough understanding of the physics of the measurement, the effects of the borehole conditions and the limits of their capability. The challenges inherent in acquiring compressional velocity in large holes and slow formations are well documented, but the effect of these formations on measurements of LWD quadrupole shear has been much less discussed. The newly conducted intensive modeling study and field data analysis of the LWD quadrupole modes reveal that the effect of the drilling mud properties (slowness and density) is also important as well as the formation properties and the borehole size for deriving robust shear slowness.

The global oil industry has a large installed base of mature offshore platforms and facilities, often with declining production. A trend is to use this infrastructure to produce and process new discoveries, reducing time to production and recovery costs while extending infrastructure operating life. However, production of new oil fields using existing infrastructure presents unique technical challenges. Accurate, repeatable and reliable measurement of produced oil and water is vital for fiscal, allocation or royalty purposes, as well as automation and control. This challenging measurement requires high-energy mixing with low power consumption and negligible pressure loss in a very small footprint. The measurement uncertainty of net oil (oil minus water) for fiscal and allocation purposes is defined by international standards and contracts. With the increase in the use of declining facilities, a new technical challenge has emerged. To be able to accurately allocate the new fluids to the field, partner or different tax regimes they must be measured to fiscal accuracy prior to being mingled with other fluids for processing. Oil and water measurements are often performed at the output of first stage separation. This creates a challenge, as the fluids are close to the critical pressure, meaning that any pressure drop potentially results in cavitation. In addition, production pipelines are generally less than 8 inches, so any obstruction causes an undesired pressure loss or restriction. These requirements highlighted the need for an effective, nonintrusive mixing device to enable operators to accurately measure and control the new fields being developed with today's greater capital constraints. Research was conducted, as part of a Joint Industry Project, at Imperial College in London followed by extensive computation fluid dynamics modeling to develop a conceptual design. The design was independently tested at the National Engineering Laboratory at East Kilbride, UK. The test independently verified that the design met the uncertainty performance criteria of the international API and ISO sampling standards across the wide range of water cut seen in both mature and new production. Once verified, the design was scaled for operating line sizes seen in the targeted application, and early deployment sites were identified. The paper discusses the application envelope identified by offshore operators and the technical challenges they were seeking to solve. It follows the design process, highlighting the choices made and the results and methodology used in the independent testing to verify performance.

Intelligent well completion (IWC) or smart wells have been used in the industry for more than two decades. Numerous benefits from this technology have been reported in the literature. These benefits are associated with improved field development economics and better reservoir management. The field economics are mainly impacted by reduction of the well count and the increase of the estimated-ultimate recovery (EUR) and net-present value (NPV) by commingling different reservoirs. The reservoir management benefits are less tangible and more difficult to quantify. They have to rely on certain assumptions that may be controversial among the team members and management. Therefore, the IWC business case is usually supported only by the improved economics. A key input on the economics is the production forecast. The industry lacks a standard workflow to predict the production from an IWC. Reservoir engineers have to be creative using their available tools to model the performance of this complex well operation. The methods regularly used are simple inflow-performance relationships (IPR), nodal analysis, well modeling with pipeline-network simulators, and 3D reservoir simulators with multi-segment well capabilities. The latter is the most appropriate to capture the value for new developments, but still extensive customization is required to optimize the production forecast while honoring all the constraints in the field operational philosophy; the others are more suitable for routine production optimization. In our case, deepwater Gulf of Mexico (GOM), the operating philosophy is (1) use primary well-head choke to control well production, (2) limit the pressure drop across the inflow-control valves (ICV), and (3) limit the producing zone drawdown. In addition, the productivity index (PI) degradation mechanism has to be included in the forecast. None of the commercial reservoir simulators have these requirements built-in. This paper presents an approach that integrates detailed wellbore modeling, reservoir simulation, and the field operating philosophy. The workflow is based on a pipeline network simulator (GAP™) coupled to 3D reservoir simulation models in INTERSECT™. The IWC optimization under the operational philosophy is executed through a Python script—INTERSECT custom functionality—that manages the GAP network information obtained from OpenServer™ and the reservoir simulation data to select the ICV settings that optimize the well production while honoring all constraints. This workflow provides an accurate production forecast assuring all critical operating constraints are incorporated.

The interpretation of well logs is crucial for understanding subsurface geological formations and reservoir characterization. Categorizing subsurface into several geological facies types assists in identifying potential locations of hydrocarbon resources. Geologists interpret each well log feature such as gamma ray, sonic velocity, neutron porosity and resistivity, based on logging characteristics and their petrophysical domain knowledge. By integrating with core analysis result, well logs interpretation result becomes accurate, and geologists finally determine rock types of subsurface for the target well. Recently, various studies propose to leverage the power of machine learning algorithms for facies classification to reduce laborious interpretation work. Most works utilize supervised learning algorithms that require input/label paired data for training. In case of facies classification, however, the accurate core analysis for determining facies is an expensive process, and the other methods require hard work and often result in noisy data. In this paper, we develop an unsupervised facies identification model based on deep neural networks. Our model is purely unsupervised, which do not require facies label for input well logs. Because well logs response from the same rock type would have high similarity, we apply an unsupervised clustering algorithm, where data points with similar traits are assigned to the same group, while dissimilar data points are assigned to the different groups. We demonstrate that our unsupervised facies identification model outperforms other unsupervised clustering algorithms for facies identification. Particularly, our clustered result shows highly similar distribution to the true facies distribution, which suggests future development of automated facies identification.

Integrating discrete facies classification into the estimation of formation permeability is a crucial step to improve reservoir characterization and to preserve heterogeneity quantification. Therefore, it is essential to obtain the most accurate estimation of permeability in non-cored intervals in order to attain realistic reservoir characterization and modeling. In our most recent paper [OTC-30906-MS], the electrofacies classifications have been conduced for a well from a carbonate reservoir in a Giant Southern Iraqi oil field. The same predicted discrete electrofacies distribution was included in this paper along with well logging interpretations to model and predict the reservoir core permeability for all wells. The well logging interpretations that were included in permeability modelling are neutron porosity, shale volume, and water saturation as a function of depth. The regression and machine learning approaches adopted for permeability modelling are multiple linear regression (MLR), smooth generalized additive Modeling (SGAM) and Random Forest (RF) Algorithm. The classified electrofacies were considered as a discrete independent variable in the core permeability modelling to provide different model fits given each electrofacies type in order to capture the different permeability variances. The matching visualization between the observed and predicted core permeability, the computed root mean square prediction error and adjusted squared R were considered as validation and accuracy tools to compare between the three modelling approaches. Since there are too many intervals with missing core permeability measurements, the modelling was first adopted on the intervals that have permeability readings (known subset). The prediction was then conducted given the same known permeability intervals in addition to the entire dataset (full dataset). The root mean square prediction error and adjusted squared R for the Random Forest were significantly better than in both MLR and SGAM for the known subset and full dataset. It can be concluded that combining the electrofacies in one permeability model has accurate, fast and an automation procedure of prediction for other wells. The two machine learning algorithms were implemented by R software, the most powerful statistical programming language.

The plug-and-perf (PnP) method is widely used globally for multistage fracturing operations. With only a few jobs performed worldwide, coiled tubing (CT) assisted PnP operations in high-pressure/high-temperature (HP/HT) wells are largely uncharted and most challenging. The "A" field in India has HP/HT formations, with bottomhole temperature (BHT) of 310°F and reservoir pressure of 9,000 psi. Whereas the PnP method is widely used globally, there are few examples in wells with completion restrictions or whose downhole conditions dramatically increase depth inaccuracies and equipment damage. This study describes how to address challenges such as depth correlation (which affects plug-setting depth accuracy), low injectivity, completion restrictions, and heavy brines (which damage CT). To gain further understanding of operations, simulations are sensitized to identify solutions for pumping rates, HP/HT conditions, well kill fluid, milling, and cleanouts where obstructions hindered BHA penetration. The proposed best practices presented here are for primary CT operations involved in the complete PnP cycle, such as wellbore displacement, well dummy run-drift, setting the isolation plug and milling, acidizing using jetting tools, sand cleanout using gels having best performance in HP/HT environment and motor-mill runs with durable resistance in harsh environment, well kill (using 13.65-ppg calcium bromide), and nitrogen lift. The featured case studies describe operations including seven bridge plugs being set at accurate depths and milled after fracturing; cleanout of a 340-m sand column is also featured, as well as well kills with heavy brines. Optimized operational parameters such as CT speed, pumping rates, and the use of smaller outer diameter bottomhole assemblies doubled operational efficiency during those operations.

Downhole conditions of oil and gas reservoirs change with time. Monitoring this change is critical to enhance hydrocarbon recovery from the reservoir. Conventionally, downhole measurements of physical and chemical properties of downhole formation fluids are taken using wireline logging or using permanent downhole sensors. Wireline logging is a complex operation that requires several miles of wireline cable, a winch, a crane and a specialized crew that knows how to operate this equipment [ 1 ]. In addition, a blowout preventer, a lubricator and a specialized crew to install and operate are needed. The complexity and cost of wireline operations makes it difficult to acquire reservoir data frequently. The other alternative for gathering downhole data more frequently is installing permanent sensors or optical fibers in the well [ 2 - 3 ]. However, due to the harsh downhole conditions, these sensors need to be extremely reliable and continuously maintained. Moreover, surface data acquisition systems for these sensors increase their cost significantly and reduces their applicability to every well. Both methods have their own limitations and there is a need for more practical and less expensive oil field instruments for well logging. Engineering small, inexpensive and robust logging instruments has its own challenges. Solving these challenges has been the main focus area of Sensors Development Team at Aramco Research Center in Houston. This paper describes one of the concept ideas that was successfully engineered and tested in our facility and successfully deployed and retrieved in the field in Saudi Arabia in collaboration with our EXPEC ARC and Northern Area Production Engineering and Well Services Divisions in Saudi Arabia.

Exploration wells in India are being drilled to target tight sandstone reservoirs of Mesozoic age with potential hydrocarbon accumulation in the Gulf of Kutch (GK) & Gulf of Saurashtra (GS). Typical permeability ranges from 0.1 md to 2.5 md with effective porosity in range of 8% to 13% in good sandstone layers. These boreholes are seen to have substantial risks in drilling and completion, as evident by the severe wellbore instability events in the offset wells. Presence of wide breakouts and stress rotations along the wellbore axis with variation in dips along fractures are often observed, indicating complicated wellbore failure modes. Considering the increasing exploration activity and field development plans for these basins located in western offshore India, maintenance of hole integrity till open hole logging and optimizing perforation depths for hydro-fracturing jobs are the main challenge in upcoming wells. An advanced acoustic geomechanics workflow for the estimation of formation horizontal stresses from the inversion of four velocity radial profiles was implemented in shallower formations of offset wells with the help of advanced borehole 3D acoustic profiling measurements. Based on the estimation of stress parameters, stable mud weight windows and rock mechanical properties were estimated for a planned well. Reported over-gauged hole condition, caving, and stuck tool incidents are found at the depth intervals where the estimated breakout mud limit exceeds the mud weight used during drilling of offset wells. Pore pressure value is in range of 13.7 ppg to 14.1 ppg EMW with sudden change in regime during transition from Trap formation to Mesozoic formation. Stable mud weights to avoid breakouts are in range of 14.2 ppg to 14.7 ppg EMW for vertical well. The local normal fault regime is identified with few interbedded stress barriers. Simulated ECD value ranges 15.6 ppg to 16.1 ppg for an 8.5-in section with 600 gpm. The rock mechanical properties constrained by inverted stresses and wellbore failure were used to optimize bit and BHA design in the planned well. Real-time monitoring has been performed in the planned well using logging while drilling data, borehole images, and cutting analysis. Intermediate advanced acoustic profiling logs and drilling parameters were used in the planned well to directly estimate stress gradients and proactively keep mud weight within the stable mud weight window. By monitoring magnitude of breakouts present in LWD images and caving type, further remedial actions were taken. Few high angle features with gas peaks were encountered as seen in offset wells. Hole cleaning was performed regularly through pumping high viscous pills and cutting load was monitored at each connection through ECD and ESD variations. Near balance equilibrium was maintained to minimize hoop stress related skin around the sandstone layers. The modifications resulted in significant reduction of nonproductive time in the planned well as compared to the offset wells by 40%. Successful casing and cementing of 7-in liner to planned depth with good cement bond was also achieved, unlike offset wells where wellbore instabilities were severe. Upon designing an appropriate stimulation strategy in a HPHT and high-stress environment, closure pressure and breakdown pressure ranges were 12,300 psi to 13,500 psi and 13,000 psi to 14,800 psi, respectively, in sandstone with 8% to 13% effective porosity. To optimize expensive hydraulic fracturing stimulation treatments, the open hole advanced 3D acoustic logs were analyzed to identify and test gas bearing intervals. Intervals with suspected critically stressed natural fractures as indicated by 3D shear moduli changes were identified. The optimized intervals flowed gas to surface at good rates with minimal stimulation, proving commercial viability of the field.

Field A is located offshore of Peninsular Malaysia, consists of multi stacked reservoirs where previously the vertical communication between different units was considered to be well understood. Separate reservoir models were built independently where fair to good history matching were achieved during the dynamic modelling study. Subsequently two drilling campaigns were executed based on the simulation results. On top of that, regular reservoir surveillance and frequent simulation model updates have assisted in assessing vertical connectivity. North of Field A, a recent drilling campaign in early 2019 indicated possible communication between infill wells in Unit 35U and water injectors in Unit 35L. One of the well is experiencing severe watercut production than initially forecasted. It is well understood that the seismic amplitude for Unit 35L is being overlapped by the stronger response of Unit 35U in the northern area. Due to that, the reservoir modelling and understanding for Unit 35L in this region is mainly driven by actual well controls and the geological model, rather than from seismic. Meanwhile in the Southern area, a stable reservoir pressure trend is observed in Unit 27 despite reduction of cumulative voidage replacement ratio (VRR) from 0.6 to 0.4. Additionally well C02ST1 was producing oil at an increasing rate, a trend which the current reservoir model failed to match. The model also failed to explain the unreasonably high recovery factor (RF) of 50% for Unit 27 given the delayed water injection and small gas cap support. Back to the North, well interference testing was conducted by shutting in water injectors in Unit 35L and observing the pressure response of a new well in Unit 35U using a permanent downhole gauge (PDG). Observations indicated a decrease in pressure response once injection was turned off, confirming the communication between the two units. Moving forward, two injectors for Unit 35L will be temporarily suspended to relax the watercut trend in the new wells, thus improving the oil rate. In the South, a material balance study and reservoir 3D model suggest additional volume needs to be introduced for Unit 27, which indicates that it is probably draining Unit 35U oil due to a higher level of vertical communication and overlapping sand bodies between both units. This is also supported by produced water salinity readings of well C02ST1 (Unit 27) having similar values as Unit 35U wells rather than other Unit 27 wells. Communication between different layers often requires longer time and a significant amount of production to be well established and observed. This case study proves that reservoir understanding can still be a mystery even when the field is reaching 20 years of production.

This paper presents the lessons learnt during an intervention to remove a gas hydrate blockage and reinstate oil production from an oil well in Field D. It also uses economics to justify facilities projects for hydrate prevention and flow assurance. Field D is a deep water oil field with 10 subsea production and water injection wells tied back to a floating, production, storage, and offloading (FPSO) vessel. Field D was shut down for turnaround maintenance work during the summer of 2016. After the field was brought back online, one of the production wells (D2) failed to flow. An evaluation of the pressure and temperature data suggested that the well had a tubing restriction. This was attributed to hydrate formation and blockage caused by limited methanol injection capacity. A number of attempts were made to induce the well with no success. A subsea intervention vessel was then hired to execute a clean out intervention operation, and this restored oil production from the D2 well. To minimise hydrate blockage and oil production losses, the asset team completed a feasibility study to evaluate the viability of installing a second methanol umbilical and a test separator. The hydrate clean-out intervention reinstated oil production from the D2 well, and the feasibility study suggested that installing a second methanol umbilical and a test separator are economically attractive as standalone or joint projects. It is recommended that flow assurance strategies are regularly updated as production fluids change over the lives of petroleum fields, from dry oil production to high water cut production. This paper presents hydrate remediation steps in a producing oil field, and outlines practical methods to justify methanol umbilical capacity enhancement projects.

Brazilian Pre-Salt reservoir characterization and monitoring is a challenge for surface geophysical methods due to the inherently minute observable signals from the reservoirs, located below the massive salt, in depth ranges > 5000 m. The status of a program to develop a wireline deployed 3-axis gravity sensor with a target sensitivity of ~ 5 μGal is firstly introduced. This is followed by a feasibility study for the potential deployment as a time-lapse gravity survey (4D gravity monitoring) within the pre-salt. A conceptual pre-salt reservoir model for the Libra field, Santos Basin, the offshore of Brazil, is developed, built on available pre-salt knowledge provided from seismic imaging and well log data, including reservoir and production characteristics. The model is based on water substituting oil over 6 months within and through the reservoir, with adoption of an oil production rate for the field published in 2017 by the ANP. These data are used to forward model gravity to understand the potential for a detectable signal and thereby establish a baseline for a time-lapse gravity survey (4D gravity) that could be used to monitor the Libra Field. A clear gravity response >80 μGal is observed over a six month period in the reservoir, due to water replacing oil, at the defined oil production rate. In a three-axis measurement the vertical axis is directly related to the magnitude of the fluid substitution and the two horizontal axes are sensitive to the fluid movement directions. Together, these suggest that an annual survey with a limited well stock could be effective in monitoring this type of reservoir and that a wireline deployed 3-axis gravity tool is likely to provide significant additional surveillance to constrain a reservoir production strategy through better appreciation of the direction of water movement. The follow-on step would be to model further field data and run a baseline survey to develop a novel reservoir surveillance method within the Pre-Salt.