Production Choke Research Articles

Oil and gas flow anomaly detection on offshore naturally flowing wells using deep neural networks

The oil and gas industry is changing. The drive towards cleaner and safer operations is becoming increasingly important. Researchers are looking for more efficient and accurate ways to detect faults that could lead to environmental and sustainability issues. This study aims to enhance the safety and sustainability of the oil and gas industry by improving existing artificial intelligence approaches to automate monitoring and detection of malfunctions. This article explores the application of deep neural networks for anomaly detection in monitoring oil and gas flow in natural flow offshore wells, proposing an innovative approach that takes advantage of the power of Genetic Algorithms and Gated Recurrent Units (GRU). The study aims to enhance the safety and sustainability of the oil and gas industry by leveraging artificial intelligence to automate the monitoring and detection malfunctions. Utilizing a comprehensive dataset from the 3W Petrobras project, which includes real-time data from 21 wells collected between 2012 and 2018, the research focuses on detecting various anomalies such as abrupt increases in basic sediment and water, spurious closures of downhole safety valves, severe slugging, flow instability, rapid productivity loss, quick restrictions in the production choke, scaling, and hydrate formation in production lines. The methodology integrates Long Short-Term Memory (LSTM) networks and GRU backbones with genetic algorithms to optimise model performance. Several hyperparameter optimisation tools were explored innovatively, focusing mainly on Genetic Algorithms, and it was possible to obtain an algorithm with 2 stacked GRU with better comparative performance compared to what is reported in the literature and producing an F1 equal to 0.97. The findings demonstrate the potential of AI to improve real-time anomaly detection, thereby reducing operational risks and contributing to the industry's transition towards greener practices. It also underscores the importance of open data and collaborative efforts in advancing AI applications in the oil and gas sector, aligning with the United Nations' Sustainable Development Goals to mitigate climate impact and promote responsible consumption and production.

Fault Diagnosis in Gas Lift System Using PDF Data

Fault detection and isolation in the gas lift system were implemented assuming the gas lift variables are stochastic. Injection valve coefficient (Civ), production choke coefficient (Cpc), annulus pressure (Pa), and wellhead pressure (Pwh) were observed to show variations with faults presence. By simulating these gas lift variables as stochastic, the probability density function (PDF) data were used to generate decision functions for both the detection and isolation of the gas lift valve faults. The scheme accurately detected and isolated faults in the injection valve coefficient (Civ) and production choke coefficient (Cpc). The result of this diagnosis will aid the proper implementation of fault tolerant control in the gas lift system which will lead to its optimal operation.

Wavelet Transform Processing in Detecting Failures in Offshore Well Production

Brazil has a significant offshore oil production, which dates back to the late 1960s and is currently focused on exploring pre-salt reservoirs. The drilling technology Petrobras uses is considered a world standard: in 2020, it allowed offshore production to reach 97% of the country’s total oil production. During the process, however, unwanted events, and even operational failures may occur, which are capable of significant damage. Thus, failure detection is extremely important to prevent production losses or delays, to reduce costs and to avoid accidents. This study uses a real, public database on offshore production, and proposes using wavelet transforms to detect production failures. With the technique, we pinpointed which time intervals between measurements showed relevant variability, and then clustered the data, according to mobile averages, to shrink the record number. Using wavelet transforms, we analyzed which variables could be used as predictors of production failures and identified the temperature read by the Temperature and Pressure Transducer sensor (T-TPT) and the pressure at the Production Choke sensor (P-PCK) as possible predictor variables. We also observed the creation of a filtered series, averaged from the original data series, which maintained its variability, showing the viability of record regrouping in shorter series.

An agent-based dynamic reliability modeling method for multistate systems considering fault propagation: A case study on subsea Christmas trees

The fault propagation of multistate systems (MSSs) with various operational modes is a critical factor affecting the reliability and safety of systems. However, traditional methods have difficulty describing the influence of function transitions and multiple faults of components on fault propagation, and are inapplicable for accurately evaluating the dynamic reliability of MSSs. To solve these issues, an agent-based method is developed to evaluate the dynamic reliability of MSSs. This study proposes a reliability modeling framework that consists of a management agent and several system agents. Combined with different agents, two types of models are established to describe the logical relationships among components. Moreover, Monte Carlo simulation is used to evaluate the dynamic reliability of MSSs, and a visual failure paths analysis method is proposed to inhibit potential risk. The LH4–1 horizontal Christmas tree is used to verify the effectiveness of proposed method. Under the designed working condition, the dynamic reliability curve of system with three operational modes and the visual failure paths including 15-node are obtained. By comparing the influence of failure rates, it is found that the production choke valve, flowlines, tree cap and MEG chemical control valve have a great impact on the system reliability.

Statistical analysis of offshore production sensors for failure detection applications / Análise estatística dos sensores de produção offshore para aplicações de detecção de falhas

Detecting the early stages of failures is an old concern of petroleum industry. In order to tackle this problem, a novel sensor analysis methodology is proposed. The assessment of production sensors' behavior, individually or in a group, leads to a better understanding of failure modes during oil and gas production. Thus, Principal Components Analysis and Logistic Regression are incorporated as multivariate statistical modeling for studying the impact of different anomalies in production sensors. Therefore, a deep statistical analysis of these sensors can strengthen assumptions for supporting the modeling process of early fault detection systems. Based on a reliable public data set containing data from real wells, the application of the PCA approach combined with a Logistic Regression resulted in better visualization and understanding of some failures that occurred during petroleum production, such as the abrupt increase in BSW (Basic sediment and water), spurious closure of DHSV (Down hole Safety Valve), severe slugging, flow instability, productivity loss, quick restriction in PCK (production choke), scaling in PCK and hydrate formation in production lines. The two statistical approaches were used as a combined method to provide useful information regarding the failure modes in the dataset. Also, the dataset presented two classes that are important for anomaly detection in oil wells: “normal” and “abnormal”, which allow detecting when production is outside its normal condition. Then, using the production sensors analysis with failure data can help to formulate better detection algorithms. By using PCA and Logistic Regression it was possible to identify which set of variables is better for detecting a specific type of problem. The application of these techniques boosts the modeling of early detection systems in oil and gas production. Besides, the assumptions led to conclusions about how to put groups of sensors and abnormalities together and how much time a well stands in a steady normal condition. Other conclusions showed the significance of transient information for fault detection modeling and the need for individual wells analyses. Hence, using PCA for treating and transforming the data brings important contributions for early fault detection modeling, once it allowed insight into how sensors and abnormal events can be related. Consequentially, the present paper has significant novelty contribution: it raises important assumptions that help to build solid knowledge about the anomalies behavior and help researchers to implement a better modeling strategy.

Identifiability and physical interpretability of hybrid, gray-box models - a case study

Model identifiability concerns the uniqueness of uncertain model parameters to be estimated from available process data and is often thought of as a prerequisite for the physical interpretability of a model. Nevertheless, model identifiability may be challenging to obtain in practice due to both stochastic and deterministic uncertainties, e.g. low data variability, noisy measurements, erroneous model structure, and stochasticity and locality of the optimization algorithm. For gray-box, hybrid models, model identifiability is rarely obtainable due to a high number of parameters. We illustrate through an industrial case study - modeling of a production choke valve in a petroleum well - that physical interpretability may be preserved even for non-identifiable models with adequate parameter regularization in the estimation problem. To this end, in a real industrial scenario, it may be beneficial for the model's predictive performance to develop hybrid over mechanistic models, as the model flexibility is higher. Modeling of six petroleum wells on the asset Edvard Grieg using historical production data show a 35\% reduction in the median prediction error across the wells comparing a hybrid to a mechanistic model. On the other hand, both the predictive performance and physical interpretability of the developed models are influenced by the available data. The findings encourage research into online learning and other hybrid model variants to improve the results.

Structural order measure of manufacturing systems based on an information-theoretic approach

A manufacturing structure is the material basis of production operation and management, and its rationality is a necessary prerequisite for ensuring highly efficient operation of manufacturing systems. Therefore, how to determine the rationality of manufacturing structures and accurately screen the optimal scheme in multiple alternatives has become an important production decision-making problem with expert and intelligent systems in recent years and needs to be solved urgently. To solve this problem, we put forward a structure entropy model and a structural order parameter in this paper, respectively. First, combined with the structural characteristics of manufacturing systems, the structure entropy model of manufacturing structures with dynamic characteristics is established and specially treated to make it operable in practical applications. Second, the structural order parameter and the definition of the order of manufacturing structures are developed, which are combined with the structure entropy model to realize the quantitative evaluation of the rationality of manufacturing structures and accurate selection of the optimal structure in the alternatives. Finally, in an empirical study, based on the two parameters of connective paths and connective spans, the structure diagrams of different manufacturing systems are made, which can directly reflect the complexity of manufacturing structures and the importance of equipment utilization, and provide another effective way for the assessment and selection of manufacturing structures. The results of the empirical study demonstrate the effectiveness of the developed approaches, which can solve the production choke point that can only rely on qualitative evaluation or trial operation of all schemes before determining the final one. Therefore, our algorithm not only provides theoretical support and decision-making basis for screening manufacturing structures, but also enriches the evaluation and decision-making methods with expert and intelligent systems.

Developing a Hybrid Data-Driven, Mechanistic Virtual Flow Meter - a Case Study

Virtual flow meters, mathematical models predicting production flow rates in petroleum assets, are useful aids in production monitoring and optimization. Mechanistic models based on first-principles are most common, however, data-driven models exploiting patterns in measurements are gaining popularity. This research investigates a hybrid modeling approach, utilizing techniques from both the aforementioned areas of expertise, to model a well production choke. The choke is represented with a simplified set of first-principle equations and a neural network to estimate the valve flow coefficient. Historical production data from the petroleum platform Edvard Grieg is used for model validation. Additionally, a mechanistic and a data-driven model are constructed for comparison of performance. A practical framework for development of models with varying degree of hybridity and stochastic optimization of its parameters is established. Results of the hybrid model performance are promising albeit with considerable room for improvements.

Optimal tracking control of artificial gas-lift process

Artificial gas-lift (AGL) technique is commonly used to enhance oil production when the reservoir pressure in wells is not enough to sustain acceptable oil flow rate. However, the gas-lift wells are prone to instability, characterized by regular oscillations of pressure and flow. This phenomenon is known as casing-heading instability. It results in production loss and negative impact on downstream equipment, and has been a challenging problem to both industry and academia. In this paper, a novel concept of optimal tracking control is proposed for stabilization and operating mode transition in gas-lift wells when casing-heading phenomenon occurs. The stability of artificial gas-lift process is ensured by manipulating both gas lift choke and oil production choke, where the openings of both choke valves can vary from fully closed to fully open. Through the simulation of the open-loop system, a stability map of AGL process is produced. Then a trajectory optimization algorithm is developed based on this stability map, which is synthesized with a tracking controller to achieve trajectory optimization control. Also, a nonlinear state observer is designed to ensure estimation of unmeasurable variables. Through simulation studies, the effectiveness of proposed trajectory optimization control is demonstrated.

Slugging attenuation using Nonlinear Model Predictive Control in offshore oil production

The increase in oil production in offshore systems can be achieved by active control, as state in the literature. However, the approach based on linear controllers has performance limitations in real applications, because of valve rate of change, as shown in this paper. As an evolution control approach, we show up the advantages of using a Nonlinear Model Predictive Controller (NMPC) by the Local Linearization on the Trajectory (LLT) algorithm, with the Fast Offshore Well Model (FOWM) as internal model, together with a first idea for tuning the controller parameters. Through a monovariable pressure control strategy, manipulating the production choke, we observed a potential production increase around 9.0% using NMPC. In addition, the multivariable advantage of NMPC strategy is stressed including the gas lift flow as a second manipulated variable. With this structure, it is possible to decrease the production choke variation in 76% and the oil production variability by 80% compared to monovariable structure, with similar production gain.

Echo State Networks for data-driven downhole pressure estimation in gas-lift oil wells

Process measurements are of vital importance for monitoring and control of industrial plants. When we consider offshore oil production platforms, wells that require gas-lift technology to yield oil production from low pressure oil reservoirs can become unstable under some conditions. This undesirable phenomenon is usually called slugging flow, and can be identified by an oscillatory behavior of the downhole pressure measurement. Given the importance of this measurement and the unreliability of the related sensor, this work aims at designing data-driven soft-sensors for downhole pressure estimation in two contexts: one for speeding up first-principle model simulation of a vertical riser model; and another for estimating the downhole pressure using real-world data from an oil well from Petrobras based only on topside platform measurements. Both tasks are tackled by employing Echo State Networks (ESN) as an efficient technique for training Recurrent Neural Networks. We show that a single ESN is capable of robustly modeling both the slugging flow behavior and a steady state based only on a square wave input signal representing the production choke opening in the vertical riser. Besides, we compare the performance of a standard network to the performance of a multiple timescale hierarchical architecture in the second task and show that the latter architecture performs better in modeling both large irregular transients and more commonly occurring small oscillations.

System Identification of a Vertical Riser Model with Echo State Networks

System identification of highly nonlinear dynamical systems, important for reducing time complexity in long simulations, is not trivial using more traditional methods such as recurrent neural networks (RNNs) trained with back-propagation through time. The recently introduced Reservoir Computing (RC)∗∗The term reservoir used here is not related to reservoirs in oil and gas industry. approach to training RNNs is a viable and powerful alternative which renders fast training and high performance. In this work, a single Echo State Network (ESN), a flavor of RC, is employed for system identification of a vertical riser model which has stationary and oscillatory signal behaviors depending of the production choke opening input variable. It is shown experimentally that these different behaviors are learned by constraining the high-dimensional reservoir states to attractor subspaces in which the specific behavior is represented. Further experiments show the stability of the identified system.

New Developments in the Control of Fluid Dynamics of wells and risers in oil production systems

A practical control algorithm for stabilizing flow in risers and oil production wells should meet several requirements. i) be simple, ii) able to operate with low-cost measurements and possibly contaminated with noise and iii) stabilize the flow without setting a value for the bottom pressure. An algorithm has been proposed which does not fix any reference for the bottom pressure. It uses as reference a value equal to zero for the derivative of the bottom pressure. This paper presents some changes in the algorithm in order to avoid the difficulties with derivatives and to simplify the tuning of its parameters. It also proposes a control methodology to suppress oscillations in the absence of automated production choke and downhole measurements.

Improved Stable, Optimal Production in Gas Lift Wells: Exploiting Additional Degrees of Freedom

Typical production objectives in lift gas assisted oil wells include stable and optimal production. These objectives are normally daunted by the presence of unstable dynamic behavior resulting from interplay between the energies in the casing head and tubing head of the wells. Moreover, realistic constraints on compressor power and sufficient lift gas availability need to be considered to determine the optimal production from the wells. This paper proposes an alternate optimization formulation to reflect these realistic constraints and exploit the additional degree of freedom associated with the production choke opening. It is demonstrated that a co-ordinated functioning of the choke with the lift gas flow can result in improved and stable production. To overcome the limitations of corrupt measurements in the uncertain downhole environment, we propose the use of a statistical estimator. Validation results involving the simulation model of Jahanshahi et al., (2012) point to the efficacy of the proposed optimization model as well as the soft-sensing approach for estimating downhole pressures.

Closed-loop model identification and PID/PI tuning for robust anti-slug control

Active control of the production choke valve is the recommended solution to prevent severe slugging flow at offshore oilfields. This requires operation in an open-loop unstable operating point. It is possible to use PI or PID controllers which are the preferred choice in the industry, but they need to be tuned appropriately for robustness against plant changes and large inflow disturbances. The focus of this paper is on finding tuning rules based on model identification from a closed-loop step test. We perform an IMC (Internal Model Control) design based on the identified model, and from this we obtain PID and PI tuning parameters. In addition, we find simple PI tuning rules for the whole operation range of the system considering the nonlinearity of the static gain. The proposed model identification and tuning rules show applicability and robustness in experiments on a test rigs as well as in simulations using the OLGA simulator.

Nonlinear model-based control of two-phase flow in risers by feedback linearization

Active control of the production choke valve is the recommended solution to prevent severe slugging flow conditions at offshore oilfields. The slugging flow constitutes an unstable and highly nonlinear system; the gain of the system changes drastically for different operating points. Although PI and PID controllers are the most widely used controllers in the industry, they need to be re-tuned for different operating conditions. The focus of this paper is to design a model-based nonlinear controller in order to counteract nonlinearities of the system. Feedback linearization based on the riser-base pressure and the topside pressure was used for the control design. Stability and convergence of the closed-loop system was verified in theory as well as by experiments on a test rig.

Optimal control strategies with nonlinear optimization for an Electric Submersible Pump lifted oil field

In an Electric Submersible Pump (ESP) lifted oil eld, the ESP of each oil well should be operated inside its operating window. The total power consumed by the ESPs in the oil eld should be minimized. The speed of the ESPs and the production choke valve opening should be optimally chosen for maximizing the total oil produced from the oil eld. At the same time, the capacity of the separator should not be exceeded. In this paper, nonlinear steady state optimization based on Sequential Quadratic Programming (SQP) is developed. Two optimal control structures are proposed in this paper. In the rst case, the optimal pump speed is controlled by a PI controller by varying the electrical excitation signal to the motors. The optimal uid ow rate through each oil well is controlled by another PI controller by varying the production choke valve opening. The paper shows that the production choke valve for each oil well has to be always 100% open to maintain the optimal uid ow rate. In the second case, the production choke valves are considered to be always 100% open as hard constraints. The optimal uid ow rate through each oil well is controlled by a PI controller by varying the pump speed. It is shown that when the optimal uid ow rate is tracked by the controller, the speed of each of the pumps is equal to the optimal pump speed calculated by the optimizer. This basically means that we can achieve the optimization objective with the same optimal results as in the rst case by using only a single PI controller. The limitations of these two optimal control structures for very low values and for very high values of the separator capacity are discussed. For the feasible range of separator capacities, the optimal locus of the uid ow rate and the pump speed are shown in this paper.

Control structure design for stabilizing unstable gas-lift oil wells

Active control of the production choke valve is the recommended solution to prevent casing-heading instability in gas-lifted oil wells. Focus of this work is to find a simple yet robust control structure for stabilization of the system. In order to find suitable control variables, a controllability analysis of the system with different candidate control variables and two alternative manipulated variables was performed. Moreover, to include robustness and performance requirements at the same time, the controllability analysis was extended to a mixed sensitivity H∞ optimization problem. A control structure using only the available topside pressure measurements was found to be effective to stabilize this system.

Controllability analysis of severe slugging in well-pipeline-riser systems

Active control of the production choke valve is the recommended solution to prevent severe slugging flow conditions at offshore oilfields. The focus of this work is to find the structure of a simple, yet robust anti-slug control system. In order to find suitable control variables for stabilization, a controllability analysis of the system with different available measurements or different combinations of them was performed. Moreover, for including robustness and performance requirements at the same time, the controllability analysis was extended to a mixed sensitivity ℋ∞ optimization problem. Two case studies were considered; first, the controllability analysis was performed on a pipeline-rise system using a 4-state model for comparing the results to the previous works. Next, using a 6-state model, the results were extended to a more general well-pipeline-riser system. The controllability results were in accordance with the practical experience in anti-slug control.

Production system analysis for Woollybutt mature oil field using dynamic simulation

This paper describes how multi-phase flow dynamic simulation techniques were applied to support and optimise the day to day production operations for the Woollybutt oil field on the North West Shelf, Australia. Eni Australia is the operator of the: Woollybutt oil project, which consists of two separate fields; Woollybutt North, which has three production wells; and Woollybutt South, which has one production well– Woollybutt–4H (WBT–4H). WBT–4H is located about 7 km from the floating production storage and offloading (FPSO) facilities and shares the production flowline with two other production wells from the Woollybutt North field. Production from WBT-4H accounts for more than 60% of the total production. For this reason, achieving steady production from this well is the highest priority for the operations. To understand the hydrodynamic behaviour and to maximise production from WBT–4H, the production system for the Woollybutt South field was modelled from WBT–4H reservoir inflow points to the FPSO using a commercial dynamic simulation software package. The model successfully matched the pressures measured at the permanent down hole gauge (PDHG) and upstream production subsea choke dynamically, using the arrival pressure at the FPSO and the measured gas lift rate at the FPSO as input data. The model was used to identify the reason for slugging , to locate where the slugging conditions were originating, as well as to investigate optimum gas lift rate and the impact of bringing other wells on production from the Woollybutt South field. Furthermore, the model was used to identify the cause of sudden severe slugging, which forced to reduce the subsea production choke opening. The simulation work concluded that an unstable gas injection at the gas lift orifice valve (GLOV) induced severe slugging. This conclusion was validated operationally in the field soon after. Results of the multi-phase flow dynamic simulation provided good understanding of the hydrodynamic behaviour in the production system and are used to make informed decisions to support and optimise the day to day production operations.