This work presents the modeling and development of a methodology based on Model Predictive Control – MPC that uses a machine learning model, based on Reinforcement Learning, as the method for searching the optimal control policy, and a neural network as a proxy, for modeling the nonlinear plant. The neural network model was developed to predict the following variables: average pressure of the reservoir, the daily production of oil, gas, water and water cut in the production well, for three consecutive values, to perform the predictive control. This model is applied as a strategy to control the oil production in an oil reservoir with existing producer and injector wells. The experiments were carried out on a synthetic oil reservoir model that consists in a reservoir with three layers with different permeability and one producer well and one injector well, both completed in the three layers. There are three valves located into the injector well, one for each completion, which are the handling variables of the model. The oil production of the producer well is the controlled variable. The experiments performed have considered various set points and also the impact of disturbances on the production well. The obtained results indicate that the proposed model is capable of controlling oil production even with disturbances in the producing well, for different reference values for oil production and supporting some features of the petroleum reservoir systems such as: strong non- linearity, long delay in the system response, and multivariate characteristic.
This work presents a system, based on Evolutionary Algorithms, capable of optimizing the controlling process of intelligent wells technology present in Intelligent Fields. The control refers to the opening and shutting operation of valves in these wells. A proactive controlling strategy to find a configuration of opening and shutting valves was assumed. It anticipates and maximizes the oil recuperation, delays the water cut on producer wells, and reduces the quantity of produced water, maximizing the wells life. As a result, the obtained configuration promotes the increasing of NPV (Net Present Value). The use of control strategies to benefit the completion identifies the field as intelligent. The proposed representation can formulate a controlling strategy for all valves, for any desired time interval. To improve the decision making for using or not using smart wells, the fault risk of the control device existing in the intelligent completions was considered into optimization. For this purpose, the system applies Monte Carlo simulation together with some simulation techniques for convergence acceleration and uncertainties representation by probability distributions. Even considering the existence of uncertainties into valves operation, the results obtained in the tests reveal significant gains by using the intelligent completion on the field such as: increasing the recuperation factor of the field, reducing the water inflow and increasing the longevity of the field. For all valves representations, improvements were achieved when compared with the case without valves. The conception and implementation of an intelligent system, capable of supporting the development and management of intelligent petroleum fields, builds up an important advantage for the spreading of intelligent field technology. The results obtained in this work demonstrate that the intelligent control of valves can become a competitive difference in the strategy of hydro-carbon production. Introduction In projects of the petroliferous exploration area [1], the optimization of the exploitation of a field involves the search for production strategies that are more economically attractive. Following this idea, the engineer intervenes in the wells production by performing operations such as: isolating producer intervals, opening of new intervals, acidifications, fracturing, tests of formation for data collection and other restoring operations. The high costs of these operations, however, especially those in offshore fields with wet completion, can make some of these operations unfeasible, and as a consequence, the field management will not be optimal. The concept of wells with intelligent completion arises as a technological alternative. This concept is proposed to reduce the cost of the most common restoring operations, as the isolation or the opening of producer intervals. In addition, the monitoring of the production data in real time - flows, pressures and temperature - allows a better field management. An intelligent completion can be defined as a system capable of collecting, transmitting and analyzing data, which enables the monitoring and the remote drive of flow control devices. As a consequence, the control of reservoir production is made possible. However, generally these technologies are associated with high costs, because they are new and with fewer field information related to reliability and ways of use. This fact makes the assets managers feel a little fearful in approving the implantation of these technologies, especially because there is not a standard methodology to calculate their benefits. Considering the different possible combinations of flow control devices operation, several profiles of production can be generated, suggesting the application of an efficient optimization method that allows the discovery of a profile that optimize the production under some criteria.
This work presents the development and evaluation of a hybrid intelligent system to optimize oil fields development. This system employs the following techniques: evolutionary algorithms to optimize the positioning and characteristics of wells in a reservoir; distributed processing to perform simultaneous reservoir simulations; function approximation models as simulator proxies; and quality maps to use some reservoir information to improve the optimization process. This work represents the first stage in the application of modern methodologies for the analysis of alternatives of oil field development under uncertainties, where no uncertainties are considered. In this sense, the optimization consists in finding wells positioning, type and geometry in a delimited petroleum field, in order to maximize the NPV of alternative, considering some technical constrains as the minimum wells distance and maximum wells trajectory. The problem approached in this work is considered of up most importance and it is recognized as a complex optimization problem, since the benefit of the option to develop an oil field depends on investments which in turn depend on the alternative chosen. The combination with other aspects makes this problem even more complex, yet properly optimized by Evolutionary Algorithms. Introduction The development of a petroleum field can be understood as the needed actions to make the petroleum field productive, these actions can be: drills, injection system, platforms, etc. The way to perform this development defines an alternative. The alternative definition is one of the most important task in petroleum engineering, because this definition impacts the production behavior, future decisions, economic analysis and in consequence, the resultant attractivity of the defined project. This task involves some variables as the number, type and positioning of the petroleum wells; the operation conditions in reservoir, and even the economic scenery. In this work the main activity is determine the wells number and positioning. Most recent optimization systems involve the usage of reservoir simulator; even with a high computational cost, reservoir simulation is still the most reliable way of obtaining forecasts of oil and gas production. In the literature, early works approaching the petroleum fields development use the recovered oil as the optimization criterion to find configurations for oil exploiting that maximize the recovered oil quantity. Nevertheless, more recent works consider the usage of economic optimization criteria. The most used criteria in literature is the Net Present Value (NPV) and will be used in this work. Evolutionary Optimization Evolutionary Optimization basically consists in the application of Evolutionary Algorithms (EAs) 1 to approach complex optimization problems. EAs are global search methods that simulate some of the processes taking place in natural evolution. They maintain a population of potential solutions to a given problem that are transformed, over successive generations, via processes of selection and genetic modification. Even though it seems simplistic from a biologist's viewpoint, these algorithms are robust enough to provide competent adaptive search mechanisms. In the past few years, EAs have been successfully applied to a large number of optimization problems. Some of the most relevant examples belong to the class of combinatorial optimization problems such as the traveling salesperson, scheduling, packing or routing. Additionally, there are many other situations that occur in various industrial, economical, and scientific domains that have also been solved using evolutionary methods. The most know evolutionary algorithms are: genetic algorithms, genetic programming, cultural algorithms, artificial immune systems, co-evolutionary algorithms, differential evolution, among others.
This paper presents a Genetic Algorithm application for selecting the best alternative for oil field development under certainty. The alternatives in this study are related to the arrangement of wells in a known and delimited oil reservoir and serve as a basis for calculating the net present value, which is used to assess the optimization process: the optimal alternative is the one that maximizes the Net Present Value of the field. The results obtained have revealed that the Genetic Algorithm model was able to find good alternatives for the oil field development, achieving good results for the Net Present Value.
This work presents the conception, modeling and development of an E&P projects economical analysis system under uncertainty that integrates the following modules: -an optimization hybrid system for oilfields development based in evolutionary algorithms with distributed evaluation, proxies and use of quality maps to optimize the place and quantity of wells in a delimited petroleum field to maximizing the NPV of the alternative. Also, this system considers some technical constrains as the minimum wells distance and maximum wells trajectory.-a model based on Genetic Algorithms and Monte Carlo simulation designed to find an optimal decision rule for some oil field development alternatives obtained from previous module, considering market uncertainty (oil price), that may help decision-making with regard to: developing a field immediately or waiting until more favorable market conditions. In the economic analysis also is considered, for each alternative under evaluating, the option of investment in information taking in account interactions of different uncertainties types. This analysis considers the option of future production expansion by installing an additional well under reserve volume uncertainties in the in the area to be drained by the additional well. Some computational intelligence techniques were applied in this system as: evolutionary algorithms, neural networks and fuzzy numbers; also, the system uses other techniques as Real Options and Monte Carlo Simulation to treat uncertainties. The obtained outcomes show the benefits to have an integrated decision support system to the decision-making in the economic analysis of oilfields development.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.