Intelligent multilateral well completions provide downhole flow rate, pressure, and temperature measurements at multiple well segments which allows for a continuous spatiotemporal data stream. Such an extensive data input poses a challenging task to decide on the optimal strategy of manipulating the inflow control valve (ICV) settings over time for best performance. This study investigated the use of machine learning to analyze and predict well performance given different ICV settings to ultimately maximize the well output.
A commercial reservoir simulator was used to generate two synthetic reservoir models: homogeneous (Case A) and heterogenous (Case B). These synthetic data were used to train, validate, and test machine learning models. The reservoir cases were generated based on a segmented, trilateral producer completed with three ICV devices installed at tie-in segments. The data used were measurements of wellhead and downhole flow rates across ICV segments over a period of 4,000 days. A total of 1,330 experiments were conducted with an eight-day timestep, generating a total of 667,660 sample data points for each of Case A and Case B. Fully connected neural networks were used to fit the data while model generalizability was enhanced using regularization techniques, namely L2 regularization and early stopping.
Both random sampling and Latin Hypercube Sampling (LHS) methods were evaluated in constructing the training, validation, and testing splits. Trained with different sample sizes drawn from the 1,330 simulated data histories for the two reservoir models, the proposed neural network showed excellent results. Given only ten simulated choices of ICV settings for training, the network proved capable of predicting oil and water production profiles at surface for both homogeneous and heterogeneous reservoir models with over 0.95 coefficient of determination (R2) when evaluated at unseen, test ICV settings. Extending the problem to downhole flow performance prediction, about 40 training simulated settings were necessary to achieve 0.95 R2. We observed that LHS was superior to random sampling in both R2 average and confidence interval. We also found that increasing the training and validation sample sizes increased the test R2 when testing against unseen cases. Study results suggest the applicability of machine reinforcement learning to estimate the well output at different ICV settings, where the neural network model depends fully on the real-time well feedback and production measurements.
By using a machine learning approach during the operation of a well with multiple ICV settings, it would be feasible to estimate the lateral-by-lateral output at unseen scenarios. Hence, it becomes possible to maximize the well output by using an optimization algorithm to determine the optimal ICV settings.
This seminar will teach participants how to identify, evaluate, and quantify risk and uncertainty in everyday oil and gas economic situations. It reviews the development of pragmatic tools, methods, and understandings for professionals that are applicable to companies of all sizes. The seminar also briefly reviews statistics, the relationship between risk and return, and hedging and future markets. Strategic thinking and planning are key elements in an organisation’s journey to maximise value to shareholders, customers, and employees. Through this workshop, attendees will go through the different processes involved in strategic planning including the elements of organisational SWOT, business scenario and options development, elaboration of strategic options and communication to stakeholders.
Green fields today mostly can be regarded as marginal fields and successfully developed. It covers the complete assessment of the oil and gas recovery potential from reservoir structure and formation evaluation, oil and gas reserve mapping, their uncertainties and risks management, feasible reservoir fluid depletion approaches, and to the construction of integrated production systems for cost effective development of the green fields. Depth conversion of time interpretations is a basic skill set for interpreters. There is no single methodology that is optimal for all cases. Next, appropriate depth methods will be presented. Depth imaging should be considered an integral component of interpretation. If the results derived from depth imaging are intended to mitigate risk, the interpreter must actively guide the process.
Decisions in E&P ventures are affected by Bias, Blindness, and Illusions (BBI) which permeate our analyses, interpretations and decisions. This one-day course examines the influence of these cognitive pitfalls and presents techniques that can be used to mitigate their impact. Bias refers to errors in thinking whereby interpretations and judgments are drawn in an illogical fashion. Blindness is the condition where we fail to see an unexpected event in plain sight. Illusions refer to misleading beliefs based on a false impression of reality.
Padhy, Girija Shankar (Kuwait Oil Company) | Kasaraneni, Pruthvi Raj (Kuwait Oil Company) | Al-Rashidi, Tahani (Kuwait Oil Company) | Tagarieva, Larisa (Weatherford Oil Tool Middle East Ltd) | Abba, Abdessalem (Weatherford Oil Tool Middle East Ltd)
Carbonate Reservoir characteristics and fluid properties can vary among multiple layers within the same stratigraphic unit. The objective of this case study is to emphasize the added values of integrating the data from a newly introduced formation testing technology along with open hole logs and core data to enhance the understanding of the Minagish Ooilte reservoir permeability distribution and fluid typing.
The methodology implies the first time application of the newly introduced formation testing techology external mounted quartz pressure gauge and fluid typing sensors (density, viscosity, resistivity, capacitance, pressure and temperature), which could minimize reservoir fluid samples contamination and later validated by comparison to laboratory analysis results. The fluid sampling operation was conducted in different reservoir units with varying mobility values where the tested zones were selected based on the pressure pretests done prior to the sampling deployment. The success criteria to evaluate the pressure measurements capability of the new techgnology was met as set by the operator to have accuracy within 0.1psi range for two build-up in pretest at the same point. The data was integrated with open hole logs and laboratory measurements to provide a comprehensive formation evaluation and conclusive reservoir characterization after validation of the permeability.
Heterogeniety in permeability measured/captured through RFT-tool was helpful to understand the reservoir flow capacity at the well location and subsequently select the right perforation intervals. Multiple fluid samples collected during this job aided in understanding the compositional variation with depth in the reservoir. Conjoining fluid variation with flow capacity of the reservoir was immensely useful to understand the true oil potential of the well and eventually select right production allowables. Production performance and productivity of the resulting well obtained after completing in the appropriate interval is better than other wells in the near vicinity.
The high well performance and productivity reflect the value of the information provided by the novel formation testing technology sonde helped, as it achieve the well objectives, design the appropriate completion and most importantly resolve many Minagish Oolite reservoir characterization uncertainties in a timely efficient operation.
Ben Amor, Faical (Schlumberger Oilfield Eastern Limited) | Madhavan, Sethu (Kuwait Oil Company) | Edwards, Keith (Kuwait Oil Company) | Kalyanbrata, Datta (Kuwait Oil Company) | Filak, Jean Michel (Kuwait Oil Company)
A Cretaceous carbonate reservoir, deposited in a shoal complex environment, produced only 10% of the estimated STOIIP, yet currently suffering from a rapid reservoir pressure decline. Recently acquired geoscience and engineering data revealed a lot of subsurface uncertainties. To boost the production and support a pressure maintenance project, a reviewed reservoir evaluation was critical to narrow down the uncertainties on reservoir structure, Tarmat prediction, rock quality, oil distribution and connectivity of aquifer with neighboring fields.
Owing to lateral variation of depositional environments field-wise and a complex diagenetic processes history, mutli-scale heterogeneities are seen both vertically and areally. Capitalizing on limited dataset, collected from early development wells and previously overseen deep exploration wells, a fully integrated approach was required to address these heterogeneities. Fault data, integrated with log & pressure data analysis and seismically mappable Tarmat related flat spots enabled to decipher reservoir compartmentalization. Detailed property modeling used a hybrid hierarchical workflow of deterministic, statistical and stochastic techniques, and allowed capturing depositional/diagenetic rock quality variations, including seismically mappable Tarmats' overprint, calibrated with well and geochemical data.
At borehole-scale, a cascaded PCA-NNs methodology in a hierarchical order yielded the best results in predicting lithofacies at un-cored intervals using wireline logs, thus enabling more favorably the comparison with the benchmark charts than the clusters generated by directly using NN with the same original logs.
Starting from a statistically proven tight relationship between borehole lithofacies, reservoir rock types and porosity, well-calibrated inverted seismic porosity maps have been used in combination with their corresponding kriged lithofacies proportion maps, together with well/seismic based variography analysis and sedimentological/stratigraphical concepts, to generate lithofacies trend maps. Thesemaps will be the main input we used for 3D facies distribution at the field scale.
The quantification of lithofacies statistical correspondances between well and seismic inversion data, enabled to segregate between reservoir shoal facies (porous limestone), non-reservoir facies (tight limestone), and intermediate-quality facies (fine-grained packestone). Seismic-scale sedimentary/diagenetic bodies were explicitly integrated into the facies model, however high-resolution borehole facies were stochastically populated, through constraining them to pre-established lithofacies trend maps. This served to directly constrain the 3D porosity distribution and, in turn, reservoir rock types - integrating lithology, petrophysics and reservoir behavior - all closely linked to each other.
The objective of this paper is to describe the production enhancement by the application of Propellant stimulation perforation during testing. During testing of a very tight carbonate reservoir (5% average porosity) the well productivity before and after propellant stimulation was studied for it's effectiveness.
Propellant stimulation is achieved by burning the propellant material chemically and generating the gas by combustion. Gases generate a peak pressure that exceeds the fracture gradient of formation. High pressure gases injects at extremely high rates for a few milliseconds, resulting in creating micro fractures in the reservoir near the wellbore area which may result in good reservoir connectivity.
Based on the subsurface information from offset wells, a vertical exploratory well drilled to delineate the potential of the target reservoir. Three sets of intervals were perforated in the target carbonate reservoir, in underbalanced condition and all intervals were tested together with Drill Stem Test (DST) tools. Matrix stimulation carried out using emulsified acid. During cleanup period, the flowing pressure continuously declined and finally only gas return was observed at surface. Flow period lasted for 30 hours. Production logging results showed that only the top perforation interval was contributing to the well flow.
After detailed review of PLT results and open hole logs the middle perforations was selected for the propellant stimulation. Well was filled with 2% KCL brine and the middle section was stimulated thru’ tubing using 15ft. of 2″ propellant stimulation tool on wireline. Matrix stimulation repeated with diverter and emulsified acid for all the perforated interval.
Flowed the well for cleanup followed by rate measurement for 15 hours showed improved flowing pressure and increased liquid rate. Second production logging results showed that both the top and middle perforation interval is contributing to the total flow. Middle perforation contributing to flow after use of matrix propellant stimulation.
Propellant stimulation was successfully applied in tight carbonate reservoir. The production logs recorded pre & post of the propellant stimulation clearly indicates gain in oil production rate & improvement of flowing pressure in tight carbonate reservoir. During Shut-in survey, no cross flow was observed between the perforations and no flow behind casing. The data acquired using production logging will provide procedures for testing new exploration wells in similar reservoirs. Propellant stimulation is economical and enhances the effectiveness of standard acid stimulation in carbonate reservoirs.
Propellant stimulation executed in tight carbonate reservoir of exploration well in State of Kuwait was remarkable success. This mechanism will aid to produce oil from the tight carbonate reservoirs.
Horizontal wells are considered superior to vertical and deviated wells because they increase reservoir contact; however, they can cone unwanted fluids (gas, water) causing reduced oil recovery and early well abandonment. Inflow Control Devices (ICDs) are typically installed along the completion string to delay coning and restrict water/gas influx. Once the coning occurs, conventional ICDs, such as channels and orifices, were found to be inadequate in choking back the unwanted fluids. Thus, new types of "autonomous" ICDs, or AICDs, were developed that choke back unwanted fluids more than conventional ICDs. Conversely, such AICDs have limitations related to bulkiness, moving parts, wellsite adjustability, flow performance predictability, and erosion.
To overcome these limitations, a new AICD, operating on a principle of a cyclone, was developed by a synergy of the latest numerical technologies, such as Computational Fluid Dynamics (CFD) utilizing a high-fidelity Large Eddy Simulation (LES) turbulence model, and Design of Experiments (DOE) techniques. This CFD-driven design optimization involved utilization of high-performance computing (HPC) coupled with experimental validation. A DOE matrix of CFD analyses runs was performed to identify a geometry that would generate significantly higher pressure drop for water and gas than for oil.
Early multiphase testing on a prototype device validated the concept, and CFD was used to improve the understanding of the operating principle and hence the design. CFD was further used to extrapolate the flow performance to a wider range of operating conditions. An expanded flow performance map and the use of non-dimensional parameters led to the development of a mechanistic AICD performance model which further enhanced our understanding of AICDs and allowed reservoir software programs to evaluate the production performance of wells with AICDs versus wells with conventional ICDs or no inflow control. The overall result is the new cyclonic AICD presented herein which is: 1) relatively compact, 2) without moving parts, 3) erosion resistant, 4) superior in multiphase performance, 5) easily adjustable at the wellsite with many settings, 6) accurately modeled with CFD, and 7) easy to incorporate into state-of-the-art reservoir simulation models.
Dhote, Prashant Dhote (Kuwait Oil Company) | Al-Bahar, Mohammad (Kuwait Oil Company) | Cole, Anthony (DeGolyer and MacNaughton) | Al-Sane, Amal (Kuwait Oil Company) | Bora, Anup (Kuwait Oil Company) | Sreenivasan, Ashique (Kuwait Oil Company)
Residual Oil Zones (ROZs) are an area of incrasing attention from hydrocarbon E&P industry with ever depleting reserves in known oil accumulations and advent of Carbon Dioxide (CO2) Capture and Storage needs and technology. ROZ can serve as viable solution to both the future problems as a possible vast new oil resource and a prospect for reducing carbon emission. ROZs can be defined as thick pile of low-quality reservoir rock below traditional oil-water contact with about residual oil saturations of mainly irreducible oil resulting from the natural flushing of reservoir due to buoying forces and aquifer action in geological past in earlier oil-filled part of reservoir. The production of oil from ROZs from such reservoirs is technically and economicaly feasible through application of enhanced oil recovery techniques - largely through missible CO2 flooding/injection in the zone because of the nature of fluid and reservoir rock. The depostional and tectonic regime in the Kuwait Petroliferous Basins is investigated to demonstrate the occurrence of and independently assess ROZ potential. The understanding of Kuwait Petroliferous Basin indicates that ROZs might be developed by hydrodynamic actions associated with tectonic regime. The degradation of oil by water action and related increase of sulfur content of crude oil can be used as workable proxy for identification ROZ potential of the rerservoir. The regional mapping, understanding of tectionic history and regional systhesis of crude oil composition shows an extensive stratigraphic and lateral existence of ROZ potential across the Kuwait Petroliferous Basin.
This study aims to provide strategic roadmap and detail data acquisition program that will reveal ROZ production potential in Kuwait for Kuwait Oil Company (KOC).
To keep pace with increasing importance of unconventional hydrocarbons and consequent changes in the global energy landscape, the State of Kuwait has embarked on a strategic plan of evaluating and developing these resources. Synergistic interpretation of exploration datasets has brought out the exploration potential of the resources. These resources are prolific and occur at multiple stratigraphic levels in diverse settings in carbonate reservoirs.
Two types of unconventional resources, self-sourced hydrocarbons (shale hydrocarbons) and tight hydrocarbons, have been identified and evaluated with workflows specific to each type. The Cretaceous Makhul and the Jurassic Najmah formations have emerged as important self-sourced hydrocarbon reservoirs. The play existence is demonstrated by the presence of free hydrocarbons contents, substantial thickness, overpressures and positive production tests in both plays. The Makhul play is characterized by total organic carbon (TOC) content between 4 - 7% and lies in the middle to late maturity oil window. It is over-pressured and has thickness in the range of 70 - 300 ft. The Najmah play is characterized by TOC content between 9 - 14% and is in the late maturity oil-condensate window. The play is systematically over-pressured, with pressure reaching the natural fracturing conditions. The Najmah play has shale gas and shale oil resource potential, while the Makhul play has potential for shale oil resource. The other speculative shale gas play is Qusaiba Shale which has not been drilled so far in Kuwait. Fractured carbonate units in the Cretaceous Upper Minagish and Upper Makhul; the Jurassic Hith, Najmah, Upper Sargelu and Marrat; and the Triassic Lower Jilh and Lower Khuff constitute the tight reservoir resources. These reservoirs are characterized by porosity typically less than 5% and permeability mostly less than 0.1 millidarcy (mD). These reservoirs are productive in structurally deformed areas where natural open fractures are well developed and critically stressed.
Resource specific and technology intensive multi-disciplinary workflows from play assessment to commercial production are crucial for effective leverage of the unconventional resources.