The paper shares a 4-years’ experience of "Gazprom Neft" PJSC on Digital Twin Learning Program in training of holistic multidisciplinary petroleum asset management and engineering based on the on-line cloud PetroCup software facility. The objective of the program was to train and test large amounts of managers and engineers with minimum off-work time and motivate self-improvement among the employee. The program includes warm-up videos, immersive master-classes, training courses, discussion clubs and Annual Corporate Championship, with a strong focus on home learning, remote communication, simulation-based exercises and automated testing/certification. The program is divided into Master Development Planning (MDP) and Well & Reservoir Management (WRM) domains which are related to different stages of the petroleum asset lifecycle. The interaction with simulator takes 2-3 days for WRM and 5 days for MDP and engages a multidisciplinary team: asset manager, economist, contract engineer, surface facility engineer, reservoir engineer, geologist, petrophysicist, simulation engineer, well test engineer, well and log analyst and production technologist. The session starts by reading the existing field data and its history and then perform well drilling, completions, workovers, well tests, open-hole and cased-hole logging, manage production and injection targets, build/modify the surface production/injection facilities and receive the fully automated asset response in the form of the field reports, very much in the same way as in real life. Once session is over the simulator generates a detailed debriefing report on team performance in numerous areas: economical, production, injection, reservoir and well performance so that team can understand where it did a good job and where it was not efficient. The current paper shows how this facility has been integrated into the corporate staff capability program, expanded to anchor universities and shed the light to the future perspectives.

The paper shares a practical case of production analysis of mature field in Western Siberia with a large stock of wells (> 1,000) and ongoing waterflood project. The main production complications of this field are the thief water production, thief water injection and non-uniform vertical sweep profile. The objective of the study was to analyse the 30-year history of development using conventional production and surveillance data, identify the suspects of thief water production and thief water injection and check the uniformity of the vertical flow profile. Performing such an analysis on well-by-well basis is a big challenge and requires a systematic approach and substantial automation. The majority of conventional diagnostic metrics fail to identify the origin of production complications. The choice was made in favour of production analysis workflow based on PRIME metrics, which automatically generates numerous conventional production performance metrics (including the reallocated production maps and cross-sections) and additionally generates advanced metrics based on automated 3D micro-modelling. This allowed to zoom on the wells with potential complications and understand their production/recovery potential. The PRIME analysis has also helped to identify the wells and areas which potentially may hold recoverable reserves and may benefit from additional well and cross-well surveillance.

In adaptation of reservoir models a direct gradient backpropagation through the forward model is often intractable or requires enormous computational costs. Thus one have to construct separate models that simulate them implicitly, e.g. via stochastic sampling or solving of adjoint systems. We demonstrate that if the forward model is a neural network, gradient backpropagation becomes naturally involved both in model training and adaptation. In our research we compare 3 adaptation strategies: variation of reservoir model variables, neural network adaptation and latent space adaptation and discuss to what extent they preserve the geological content. We exploit a real-world reservoir model to investigate the problem in practical case. The numerical experiments demonstrate that the latent space adaptation provides the most stable and accurate results.

We present a novel technique for assessing the dynamics of multiphase fluid flow in the oil reservoir. We demonstrate an efficient workflow for handling the 3D reservoir simulation data in a way which is orders of magnitude faster than the conventional routine. The workflow (we call it “Metamodel”) is based on a projection of the system dynamics into a latent variable space, using Variational Autoencoder model, where Recurrent Neural Network predicts the dynamics. We show that being trained on multiple results of the conventional reservoir modelling, the Metamodel does not compromise the accuracy of the reservoir dynamics reconstruction in a significant way. It allows forecasting not only the flow rates from the wells, but also the dynamics of pressure and fluid saturations within the reservoir. The results open a new perspective in the optimization of oilfield development as the scenario screening could be accelerated sufficiently.

BACKGROUND:The clinical course of COVID-19 critically ill patients, during their admission in the intensive care unit (UCI), including medical and infectious complications and support therapies, as well as their association with in-ICU mortality has not been fully reported. OBJECTIVE:This study aimed to describe clinical characteristics and clinical course of ICU COVID-19 patients, and to determine risk factors for ICU mortality of COVID-19 patients. METHODS:Prospective, multicentre, cohort study that enrolled critically ill COVID-19 patients admitted into 30 ICUs from Spain and Andorra. Consecutive patients from March 12th to May 26th, 2020 were enrolled if they had died or were discharged from ICU during the study period. Demographics, symptoms, vital signs, laboratory markers, supportive therapies, pharmacological treatments, medical and infectious complications were reported and compared between deceased and discharged patients. RESULTS:A total of 663 patients were included. Overall ICU mortality was 31% (203 patients). At ICU admission non-survivors were more hypoxemic [SpO2 with non-rebreather mask, 90 (IQR 83 to 93) vs. 91 (IQR 87 to 94); P<.001] and with higher sequential organ failure assessment score [SOFA, 7 (IQR 5 to 9) vs. 4 (IQR 3 to 7); P<.001]. Complications were more frequent in non-survivors: acute respiratory distress syndrome (ARDS) (95% vs. 89%; P=.009), acute kidney injury (AKI) (58% vs. 24%; P<10-16), shock (42% vs. 14%; P<10-13), and arrhythmias (24% vs. 11%; P<10-4). Respiratory super-infection, bloodstream infection and septic shock were higher in non-survivors (33% vs. 25%; P=.03, 33% vs. 23%; P=.01 and 15% vs. 3%, P=10-7), respectively. The multivariable regression model showed that age was associated with mortality, with every year increasing risk-of-death by 1% (95%CI: 1 to 10, P=.014). Each 5-point increase in APACHE II independently predicted mortality [OR: 1.508 (1.081, 2.104), P=.015]. Patients with AKI [OR: 2.468 (1.628, 3.741), P<10-4)], cardiac arrest [OR: 11.099 (3.389, 36.353), P=.0001], and septic shock [OR: 3.224 (1.486, 6.994), P=.002] had an increased risk-of-death. CONCLUSIONS:Older COVID-19 patients with higher APACHE II scores on admission, those who developed AKI grades ii or iii and/or septic shock during ICU stay had an increased risk-of-death. ICU mortality was 31%.

Purpose: To investigate age-related differences in outcomes of critically ill patients with sepsis around the world. Methods: We performed a secondary analysis of data from the prospective ICON audit, in which all adult ( >16 years ) patients admitted to participating ICUs between May 8 and 18, 2012, were included, except admissions for routine postoperative observation. For this sub-analysis, the 10,012 patients with completed age data were included. They were divided into five age groups - <= 50, 51-60, 61-70, 71-80, >80 years. Sepsis was defined as infection plus at least one organ failure. Results: A total of 2963 patients had sepsis, with similar proportions across the age groups (<= 50 = 25.2%: 51-60 = 30.3%; 61-70 = 32.8%; 71-80 = 30.7%; >80 = 30.9%). Hospital mortality increased with age and in patients >80 years was almost twice that of patients <= 50 years (493% vs 25.2%, p < .05). The maximum rate of increase in mortality was about 0.75% per year, occurring between the ages of 71 and 77 years. In multilevel analysis, age > 70 years was independently associated with increased risk of dying. Conclusions: The odds for death in ICU patients with sepsis increased with age with the maximal rate of increase occurring between the ages of 71 and 77 years. (C) 2019 Elsevier Inc. All rights reserved.

Abstract The topic of the paper is an approach to find optimal regimes of miscible gas injection into the reservoir to maximize cumulative oil production using a surrogate model. The sector simulation model of the real reservoir with a gas cap, which is in the first stage of development, was used as a basic model for surrogate model training. As the variable (control) parameters of the surrogate model parameters of gas injection into injection wells and the limitation of the gas factor of production wells were chosen. The target variable is the dynamics of oil production from the reservoir. A set of data has been created to train the surrogate model with various input parameters generated by the Latin hypercube. Several machine learning models were tested on the data set: ARMA, SARIMAX and Random Forest. The Random Forest model showed the best match with simulation results. Based on this model, the task of gas injection optimization was solved in order to achieve maximum oil production for a given period. The optimization issue was solved by Monte Carlo method. The time to find the optimum based on the Random Forest model was 100 times shorter than it took to solve this problem using a simulator. The optimal solution was tested on a commercial simulator and it was found that the results between the surrogate model and the simulator differed by less than 9%.

This paper considers the development of a computationally fast model for simulation of multiphase flow in porous media for a heterogeneous reservoir with the unlimited number of wells characterized by a different type of completion. This fast solution has been obtained by means of replacing the differential equation governing the flow in porous media by approximate governing equations which are parametrized by convolutional neural networks. The matching of the dynamic properties of the original and reduced models is ensured by conservation of spatial invariance property of the equations. The suggested approach is characterized by the minimal number of limitations and shortcomings related to geological-hydrodynamical structure and size of the original model. Also, there is no necessity of additional model training for reservoirs not included in a training dataset. Suggested approach has been evaluated on the synthetic benchmark test model SPE10, where a significant decrease in computational time has been demonstrated comparing to a traditional commercial reservoir simulator. Based on the results of all demonstrated test case scenarios, it could be noted that hybrid hydrodynamic modeling leads to a significant reduction in computational cost (by a factor of few hundreds), maintaining at the same time required accuracy of calculations.

Due to need for optimal management of a large wells stock with a relatively limited amount of conditioned data, the need for a focus rapid detection of implicit complications in the work (mechanical wear of the working bodies of the pump) and an increase in overhaul period of submersible equipment increases. Real-time monitoring of all wells in the debit is limited due to infrastructure problems and high costs of measuring activities. Despite this, the possibility of such monitoring is not excluded due to the availability of field information that correlates with the well flow rate and the mode of its operation. The article presents an algorithm that, based on telemetry ESP to estimate the parameters that affect the performance of the well and submersible equipment, such as the coefficient of degradation of the outlet characteristics of the pump, the actual efficiency (coefficient of performance) of ESP and the thickness of the deposition of paraffin on the inner walls of the tubing. When processing field data, the parameters correlating with the well flow rate were revealed, which allowed to build a model of a virtual flow meter to verify the existing flow rate measurements and restore the missing values. The basis of the physical and mathematical approach is an algorithm that connects the parameters of the ESP system with the flow rate through the power consumption and hydraulic calculations of the gas-liquid mixture flow in the tubing. After preliminary calibration of pressure-flow characteristics of ESP for real mode of operation, daily measurements of fluid flow rate with a periodicity of 1 hour were calculated.

Process of oil-and-gas field development optimization under the conditions of a mineral raw material base deterioration and increase in a share of hard-to-recover reserves is the integral part of commercial production stage, especially in the last stage of development. Decisions regarding the optimization of the development system with contour water flooding under the conditions of a high water-cut of well production need to be made using additional instruments for the decision making, such as 1-D, 2-D and 3-D models. Using of simulation does not exclude a participation of experts in such work and imposes great responsibility on them in making decisions. Searching for optimal decisions under the oil-and-gas field development optimization based on physic-mathematical models together with the participation of recovery and development experts is the basis for managerial decision making in oil-and-gas production companies. This article shows the principles of the oil-and-gas field development optimization based on the existing forecast model and describes an industrial example of such optimization instrument usage together with the participation of the experts.

Summary Currently, many fields in Russia are at the final development stage known not only for a decreasing percentage of active recoverable reserves, but also for a growing share of hard-to-recover reserves, confined, for example, to complex reservoirs featuring primarily highly zonal and layered heterogeneity. In this context, the priority issue is to spot areas containing remaining reserves with a view to subsequently recovering them. Obviously, this issue cannot be tackled without the aid of advanced technologies of crosswell formation evaluation. This publication presents the remaining reserves detection technology involving integrated multi-well diagnostics and reservoir flow model (RFM) [1] calibration techniques focused on refining development targets and shaping an appropriate RFM to come up with an action plan to tap reserves and ramp up production at the Company's current assets.

The paper describes the principal possibility of using machine learning methods for verifying and restoring the quality of oilfield measurements. Basic methods for screening incorrect values have been given and approaches for solving three problems have been recommended: Correctness analysis of well logging data Quality control of physical and chemical fluid properties (PVT-studies) Separation between the base production and effect from well interventions (WI) to predict the performance of hydraulic fracturing (frac). The main deliverable is a set of algorithms based on machine learning methods, which allows to automatically process large volumes of field data. A number of approaches is proposed, including using modern methods of machine learning, to restore the missing values and the quality of algorithms operation.

Abstract Traditional well testing methods applied for production wells for estimation of reservoir porosity and permeability, pressure, and other parameters imply the shutdown of wells to carry out pressure build-up tests. These tests entail significant losses in oil production and, in many cases, do not provide enough information in the event of low-permeability reservoirs because of the insufficient duration of the testing. An alternative to the pressure build-up test is an estimation of well and reservoir parameters based on production data (decline-curve analysis) carried out by means of special software. As compared to the pressure build-up test, the interpretation of decline-curve analysis allows to decrease losses in oil production during shutdowns, assess well interferences and boundary conditions, etc. (Ipatov and Kremenetskiy, 2008). Because of the growing number of wells equipped with high-precision pressure sensors for downhole telemetry and due to ongoing tasks related to monitoring and control of the development process, it is necessary to reduce labor hours required for production data processing. This thesis provides a description of the system designed to introduce automation into the interpetation of production data by means of the machine learning techniques.

We describe an experiment performed at the LULI laser facility using an advanced radiographic technique that allowed obtaining 2D, spatially resolved images of a shocked buried-code-target. The technique is suitable for applications on Fast Ignition as well as Warm Dense Matter research. In our experiment, it allowed to show cone survival up to Mbar pressures and to measure the shock front velocity and the fluid velocity associated to the laser-generated shock. This allowed obtaining one point on the shock polar of porous carbon.

The developed algorithm allows to calculate the depth of hydrate flowing wells according to the well design, reservoir characteristics, mode of operation of the well and the physical and chemical properties, as well as the depth of hydrate mechanized wells into account the characteristics of ESP

The authors present a method of flooding process control as a result of the complex system analysis of carried out specialized well testing. We considered several approaches which allow to determine the optimal mode of operation for any injection well when certain formation reservoir properties are known.