Papers by Hamid R. Ghafouri
International Journal of Numerical Methods for Heat & Fluid Flow, Oct 17, 2018
Purpose The purpose of this paper is to present detailed algorithms for simulation of individual ... more Purpose The purpose of this paper is to present detailed algorithms for simulation of individual and group control of production wells in hydrocarbon reservoirs which are implemented in a finite volume-based reservoir simulator. Design/methodology/approach The algorithm for individual control is described for the multi-lateral multi-connection ones based on the multi-segment model considering cross-flow. Moreover, a general group control algorithm is proposed which can be coupled with any well model that can handle a constraint and returns the flow rates. The performance of oil production process based on the group control criteria is investigated and compared for various cases. Findings The proposed algorithm for group control of production wells is a non-optimization iterative scheme converging within a few number of iterations. The numerical results of many computer runs indicate that the nominal power of the production wells, in general, is the best group control criterion for the proposed algorithm. The production well group control with a proper criterion can generally improve the oil recovery process at negligible computational costs when compared with individual control of production wells. Research/limitations/implications Although the group control algorithm is implemented for both production and injection wells in the developed simulator, the numerical algorithm is here described only for production wells to provide more details. Practical/implications The proposed algorithm can be coupled with any well model providing the fluid flow rates and can be efficiently used for group control of production wells. In addition, the calculated flow rates of the production wells based on the group control algorithm can be used as candidate solutions for the optimizer in the simulation-optimization models. It may reduce the total number of iterations and consequently the computational cost of the simulation-optimization models for the well control problem. Originality/value A complete and detailed description of ingredients of an efficient well group control algorithm for the hydrocarbon reservoir is presented. Five group control criteria are extracted from the physical, geometrical and operating conditions of the wells/reservoir. These are the target rate, weighted potential, ultimate rate and introduced nominal power of the production wells. The performance of the group control of production wells with different group control criteria is compared in three different oil production scenarios from a black-oil and highly heterogeneous reservoir.
International Journal of Numerical Methods for Heat & Fluid Flow, Jan 4, 2016
Purpose– The purpose of this paper is to present an efficient improved version of Implicit Pressu... more Purpose– The purpose of this paper is to present an efficient improved version of Implicit Pressure-Explicit Saturation (IMPES) method for the solution of incompressible two-phase flow model based on the discontinuous Galerkin (DG) numerical scheme.Design/methodology/approach– The governing equations, based on the wetting-phase pressure-saturation formulation, are discretized using various primal DG schemes. The authors use H(div) velocity reconstruction in Raviart-Thomas space (RT_0 and RT_1), the weighted average formulation, and the scaled penalties to improve the spatial discretization. It uses a new improved IMPES approach, by using the second-order explicit Total Variation Diminishing Runge-Kutta (TVD-RK) as temporal discretization of the saturation equation. The main purpose of this time stepping technique is to speed up computation without losing accuracy, thus to increase the efficiency of the method.Findings– Utilizing pressure internal interpolation technique in the improved IMPES scheme can reduce CPU time. Combining the TVD property with a strong multi-dimensional slope limiter namely, modified Chavent-Jaffre leads to a non-oscillatory scheme even in coarse grids and highly heterogeneous porous media.Research limitations/implications– The presented locally conservative scheme can be applied only in 2D incompressible two-phase flow modeling in non-deformable porous media. In addition, the capillary pressure discontinuity between two adjacent rock types assumed to be negligible.Practical implications– The proposed numerical scheme can be efficiently used to model the incompressible two-phase flow in secondary recovery of petroleum reservoirs and tracing immiscible contamination in aquifers.Originality/value– The paper describes a novel version of the DG two-phase flow which illustrates the effects of improvements in special discretization. Also the new improved IMPES approach used reduces the computation time. The non-oscillatory scheme is an efficient algorithm as it maintains accuracy and saves computation time.
Water Resources Management, Jan 29, 2022
In Table 1 of this article, the data in columns 4 and 5 were mistakenly listed and in the abbrevi... more In Table 1 of this article, the data in columns 4 and 5 were mistakenly listed and in the abbreviations section, 2 in R2 must be in superscript format. Also, in equation 5, the radi
WIT Transactions on the Built Environment, 1970
Estuaries are main water resources for various drinking, agricultural and industrial purposes in ... more Estuaries are main water resources for various drinking, agricultural and industrial purposes in coastlines. With increasing water demand in these areas, the discharge of industrial and municipal wastewater into rivers and estuaries has increased. On the other hand, seawater intrusion into estuaries has decreased water quality which becomes more severe under the high tides. In order to reduce the undesirable impacts of seawater intrusion, recognition of the influential mechanisms and prediction of their impacts is essential. To achieve this, computer models are usually considered as the main tools. Throughout this paper a novel finite element model has been developed which is particularly applicable in the simulation of the flow hydrodynamics and salt transport in well-mixed estuaries. The presented model consists of two steps, that is, in the first step hydraulic parameters of estuary under tidal wave is computed and in the second step salt concentration at different points during a tide is determined. Continuity and momentum equations are governing equations in the first part while the advection-diffusion equation is used in the second part. Taylor-Galerkin Finite Element technique is applied to the formulation associated with the developed model. According to this approach, governing equations are descritized by Galerkin weighted residual method with respect to space while temporal discritization is implemented by finite difference method. To evaluate the accuracy of the model several analytical and numerical examples have been studied. A comparison between the obtained and expected results indicates the favorable performance of the model and its capability for simulation of seawater intrusion into this type of estuaries.
International Journal of Numerical Methods for Heat & Fluid Flow, Oct 28, 2014
Water Resources Management, Feb 8, 2017
The operation of pumps imposes significant costs on a water distribution system for energy supply... more The operation of pumps imposes significant costs on a water distribution system for energy supply and pumps maintenance. To derive an optimum pumps scheduling program, this study presents a multiobjective optimization problem with the objective functions of 1energy cost and 2-the number of pump switches. The optimization of both objective functions together leads to a multiobjective constrained optimization problem. To solve the problem, the Non-Dominated Sorting Genetic Algorithm, version II, (NSGA-II) is coupled to the EPANET hydraulic simulation model. For constraint handling, some modifications are introduced to the standard NSGA-II to make it self-adaptive through which all constraints of the problem are automatically satisfied. Application of the model to a test example and a real pipe network verifies that the proposed scheme is computationally efficient and reliable. Also, optimization of the real pipe network reveals that by a careful pump scheduling program the total number of pump switches even in optimum operations could be decreased by 69% while the energy cost increases at most by 10%.
International Journal for Numerical and Analytical Methods in Geomechanics, Nov 1, 1996
The mathematical base of the double porosity concept, consisting of the continuity and equilibriu... more The mathematical base of the double porosity concept, consisting of the continuity and equilibrium equation respectively, is briefly reviewed. A quasi-steady-state transfer function, the so-called leakage term, is used. Important aspects of the developed code, based on the double ...
Research Square (Research Square), Jan 24, 2022
This study introduces a two-stage approach for high-resolution leak localization in large-scale p... more This study introduces a two-stage approach for high-resolution leak localization in large-scale pipelines by coupling machine learning to transient hydraulics. The method includes two stages of leak zone identification and in-zone Leak detection. A transient simulation model using the Method of Characteristics (MOC) is developed to generate the learning data for the pipeline under consideration. Afterward, the problem search space is reduced, and the maximum leak detection error is restricted by determining the most likely leaky zone using Support Vector Regression (SVR). Then, the zone dataset is provided by introducing leak candidates to the identified zone. After that, an ensemble classifier consisting of a set of linear discriminant components is trained to reliably detect the exact location of the leak using the majority voting technique. The models are applied to a theoretical pipeline and an experimental Reservoir-Pipe-Valve (RPV) system. The performance of the applied machine learning algorithms is compared to well-known algorithms considering a variety of kernels and hyperparameters. The impacts of different levels of uncertainty in pipe roughness and initial flow on the models' accuracy are also investigated. The results manifest that the proposed model has high accuracy and is stable, and robust against the hydraulic simulation uncertainties.
Computational & Applied Mathematics, Sep 8, 2018
In the present study, a new technique in Discrete Least Squares Meshless method; DLSM; has been a... more In the present study, a new technique in Discrete Least Squares Meshless method; DLSM; has been adopted for stress assessment of two-dimensional elastic domains including a non-growing crack. DLSM is a real meshless method that does not require mesh both in approximation step and in numerical integration step of the solution procedure due to the use of a collection of the un-structured nodal points instead of the mesh and discrete least squares approach to discretize the governing differential equations which can reduce considerably the pre-processing cost of calculations. However, DLSM encounters some difficulties in accommodating the solution procedure in the vicinity of the non-convex boundaries such as cracks. In recent years, researchers utilize some techniques to overcome this problem such as visibility, transparency and diffraction approaches, but these methods require somehow backward corrective operations within each step of the calculation process and suffer from some limitations in application. In this study, however, a new, straightforward and general applicable method has been used for fixing this problem using Moving Least Squares; MLS; shape functions constructed based on the Voronoi tessellation algorithm. In this method, the domain of interest is divided into Voronoi cells and nodes of support domain are selected based on a neighboring criterion. The accuracy and efficiency of the proposed method has been demonstrated by solving some benchmark examples and comparing the obtained results with analytical or valid finite element analyses' results. Keywords Discrete • Least squares • Voronoi • Meshless • Crack Mathematical subject classifications 65 Communicated by Abimael Loula.
Computers & Fluids, Dec 1, 2020
Discontinuous Galerkin (DG) methods due to their robustness properties, e.g. local conservation, ... more Discontinuous Galerkin (DG) methods due to their robustness properties, e.g. local conservation, low numerical dispersion, and well-capturing strong shocks and physical discontinuities, are well-suited for the simulation of Variable Density Flow (VDF) in porous media. This paper aims at introducing, in a unified format, the general class of Interior Penalty DG (IPDG) methods to solve the VDF equations. A combination of symmetric, non-symmetric and incomplete IPDG methods is used to discretize both head and concentration variables. Compatibility analysis is performed to prevent the loss of accuracy of the IPDG methods in simulations of coupled flow and transport equations. An accurate technique is used for time integration, based on a non-iterative procedure and adaptive time stepping with embedded error control. Several benchmarks are investigated to validate the proposed DG scheme and to examine its performance in simulating VDF problems. The new DG scheme reproduces better the experimental data than the conventional SEAWAT model. Its results are in excellent agreement with a recent semianalytical solution of the Henry problem, dealing with seawater intrusion under convection-dominating conditions. The performance of the DG scheme is examined by simulating the challenging problem of natural convection in porous enclosure. The method is compared against a finite element solution obtained with COMSOL multi-physics. The numerical experiments indicate clearly that high-order DG method is much more appropriate than standard conforming Galerkin method in simulating VDF problems while at the same time, guaranteeing a better precision and high-fidelity solutions. The proposed numerical method can be extended to 3D problems.
Hydrological Sciences Journal
Water Resources Management, 2021
This article proposes a methodology to accurately monitor seawater intrusion (SWI) using time-var... more This article proposes a methodology to accurately monitor seawater intrusion (SWI) using time-varied GALDIT vulnerability maps. The properly produced samples are then used as input–output patterns for the approximate SWI simulation. As a novelty, the specific area of high susceptibility to SWI is proposed as the dynamic saltwater wedge position to suitably select the monitoring locations (MLs) from a narrowed area. It is observed that varied initial conditions over time periods have more influence than variable pumping rates on salinity at MLs far from the production wells. Support Vector Regression (SVR), Artificial Neural Network (ANN) and Gaussian Process Regression (GPR) models have been substituted for the numerical model of SWI. Input training patterns of the surrogate models are initial salinity concentrations at selected MLs plus transient pumping values via Latin hypercube sampling. The final salinity at MLs constitutes the output patterns. The paper applies this new methodology to a small study area subject to the SWI problem. The generalization ability of surrogate models for predicting new initial conditions-pumping datasets was evaluated using performance criteria considering the ML locations. All surrogates offered good results for predicting SWI at specified MLs. The SVR model had poor performance compared to ANN and GPR models in MLs near the pumping wells, due to their salinity fluctuations over time.
Journal of Contaminant Hydrology, 2017
A simulation-optimization model is proposed for identifying the characteristics of local immiscib... more A simulation-optimization model is proposed for identifying the characteristics of local immiscible NAPL contaminant sources inside aquifers. This model employs the UTCHEM 9.0 software as its simulator for solving the governing equations associated with the multi-phase flow in porous media. As the optimization model, a novel two-level saturation based Imperialist Competitive Algorithm (ICA) is proposed to estimate the parameters of contaminant sources. The first level consists of three parallel independent ICAs and plays as a pre-conditioner for the second level which is a single modified ICA. The ICA in the second level is modified by dividing each country into a number of provinces (smaller parts). Similar to countries in the classical ICA, these provinces are optimized by the assimilation, competition, and revolution steps in the ICA. To increase the diversity of populations, a new approach named knock the base method is proposed. The performance and accuracy of the simulation-optimization model is assessed by solving a set of two and three-dimensional problems considering the effects of different parameters such as the grid size, rock heterogeneity and designated monitoring networks. The obtained numerical results indicate that using this simulation-optimization model provides accurate results at a less number of iterations when compared with the model employing the classical one-level ICA.
This study introduces a two-stage approach for high-resolution leak localization in large-scale p... more This study introduces a two-stage approach for high-resolution leak localization in large-scale pipelines by coupling machine learning to transient hydraulics. The method includes two stages of leak zone identification and in-zone Leak detection. A transient simulation model using the Method of Characteristics (MOC) is developed to generate the learning data for the pipeline under consideration. Afterward, the problem search space is reduced, and the maximum leak detection error is restricted by determining the most likely leaky zone using Support Vector Regression (SVR). Then, the zone dataset is provided by introducing leak candidates to the identified zone. After that, an ensemble classifier consisting of a set of linear discriminant components is trained to reliably detect the exact location of the leak using the majority voting technique. The models are applied to a theoretical pipeline and an experimental Reservoir-Pipe-Valve (RPV) system. The performance of the applied machin...
The present paper aims to evaluate a class of discontinuous Galerkin methods for modeling of coup... more The present paper aims to evaluate a class of discontinuous Galerkin methods for modeling of coupled flow and mass transport equations in porous medium. Various combinations of primal discontinuous Galerkin methods were used for discretization of the coupled nonlinear system of flow and mass transport equations in a saturated porous medium and a fully implicit backward Euler scheme was applied for temporal discretization. The primal DGs have been developed successfully for density-dependent flows by applying both Cauchy and Dirichlet boundary conditions to the mass transport equation. To avoid the errors arising from non-compatible selection of DG methods for flow and mass transport equations, only compatible combinations were applied. To linearize the resulting nonlinear systems, Picard iterative technique was applied and a slope limiter was used to eliminate the nonphysical oscillations appeared in solution. For the purpose of consistent velocity approximation, Frolkovic-Knabner m...
Journal of American …, 2010
Abstract: A computational model simulating the behavior of the concrete subjected to the high tem... more Abstract: A computational model simulating the behavior of the concrete subjected to the high temperature environment has been presented here by means of micro-planes framework. The constitutive equations using damage formulations developed earlier by (Labibzadeh and ...
Computational and Applied Mathematics, 2018
In the present study, a new technique in Discrete Least Squares Meshless method; DLSM; has been a... more In the present study, a new technique in Discrete Least Squares Meshless method; DLSM; has been adopted for stress assessment of two-dimensional elastic domains including a non-growing crack. DLSM is a real meshless method that does not require mesh both in approximation step and in numerical integration step of the solution procedure due to the use of a collection of the un-structured nodal points instead of the mesh and discrete least squares approach to discretize the governing differential equations which can reduce considerably the pre-processing cost of calculations. However, DLSM encounters some difficulties in accommodating the solution procedure in the vicinity of the non-convex boundaries such as cracks. In recent years, researchers utilize some techniques to overcome this problem such as visibility, transparency and diffraction approaches, but these methods require somehow backward corrective operations within each step of the calculation process and suffer from some limitations in application. In this study, however, a new, straightforward and general applicable method has been used for fixing this problem using Moving Least Squares; MLS; shape functions constructed based on the Voronoi tessellation algorithm. In this method, the domain of interest is divided into Voronoi cells and nodes of support domain are selected based on a neighboring criterion. The accuracy and efficiency of the proposed method has been demonstrated by solving some benchmark examples and comparing the obtained results with analytical or valid finite element analyses' results. Keywords Discrete • Least squares • Voronoi • Meshless • Crack Mathematical subject classifications 65 Communicated by Abimael Loula.
Original Research Paper Received 19 July 2015 Accepted 18 October 2015 Available Online 01 Decemb... more Original Research Paper Received 19 July 2015 Accepted 18 October 2015 Available Online 01 December 2015 In this article, numerical solution of incompressible two-phase flow in isothermal condition, based on wetting pressure-wetting saturation formulation ( using high order primal discontinuous Galerkin (DG) method which can capture the shock fronts of two-phase flow in heterogeneous porous media is considered. In this presented model, the velocity field is reconstructed by H(div) post-process in lowest order of Raviart-Thomas space ( ). Also in this study, the scaled penalty and weighted average (harmonic average) formulation significantly improve the special discretization formulation of governing equations which cause the instabilities in heterogamous media to be reduced. The modified MLP slope limiter is used to remove the non-physical saturation values at the end of each time step. In this study, the slope limiter should be considered as one of the main novelties due to the imp...
Uploads
Papers by Hamid R. Ghafouri