Abstract:Accurate prediction of the flow behaviour of drill cuttings carried by a non-Newtonian fluid in an annular geometry is important for the successful and efficient design, operation, and optimisation of drilling operations. Although it is widely recognised that practical drilling operations hardly involve perfectly spherical cuttings, the relative ease in mathematical description coupled with speedy computation are the main reasons for the prevalence of this simplifying assumption. The possibilities offered by t… Show more
“…The present article adopts a simulation-based optimization method based on the approach proposed by Gunnerud, Foss and coworkers, in which the production network is broken down into several black boxes to formulate explicit relations, rather than treating the entire network as a single black box. 2,6 They argue that although proxy models are required for each section of the entire production network, configuring the optimizer to search using smaller simulators yields faster computations compared to searching over the entire production network. The surrogate modeling approach adopted here also offers a clear opportunity for implementation of a gradient-based optimization solver.…”
Section: Proxy Modeling and The Optimization Frameworkmentioning
confidence: 99%
“…Local searching methods equipped with quadratic proxy models and many reasonable initial guesses are able to exploit smooth features of the objective function, and converge rapidly to a global optimum. The majority of the computational effort involved in solving problems with discrete variables can be attributed to the piecewise linearization procedure and the number of breakpoints required to capture nonlinear-ities in well and pipeline models 6. The absence of these variable types in our case studies significantly reduces the computational effort (as reflected inTable 7) compared to the MILP/MINLP studies.…”
An oil and gas field requires careful operational planning and management via production optimisation for increased recovery and long-term project profitability. This paper addresses the challenge of production optimisation in a field undergoing secondary recovery by water flooding. The field operates with limited processing capacity at the surface separators, pipeline pressure constraints and water injection constraints; an economic indicator (Net Present Value -NPV) is used as the objective function. The formulated optimisation framework adequately integrates slow-paced subsurface dynamics using reservoir simulation and the fast-paced surface dynamics using sophisticated multiphase flow simulation in the upstream facilities. Optimisation of this holistic longterm model is made possible by developing accurate second order polynomial proxy models at each time step. The resulting formulation is solved as a Nonlinear Program (NLP) using commercially available solvers. A comparative analysis is performed using MATLAB's fmincon solver and the IPOPT solver for their robustness, speed and convergence stability in solving the proposed problem.By implementing 2 synthetic case studies, our mathematical programming approach determines the optimal production and injection rates of all wells and further demonstrates considerable improvement to the NPV obtained by simultaneously applying the tools of streamline, reservoir and surface facility simulation for well rate allocation.
“…The present article adopts a simulation-based optimization method based on the approach proposed by Gunnerud, Foss and coworkers, in which the production network is broken down into several black boxes to formulate explicit relations, rather than treating the entire network as a single black box. 2,6 They argue that although proxy models are required for each section of the entire production network, configuring the optimizer to search using smaller simulators yields faster computations compared to searching over the entire production network. The surrogate modeling approach adopted here also offers a clear opportunity for implementation of a gradient-based optimization solver.…”
Section: Proxy Modeling and The Optimization Frameworkmentioning
confidence: 99%
“…Local searching methods equipped with quadratic proxy models and many reasonable initial guesses are able to exploit smooth features of the objective function, and converge rapidly to a global optimum. The majority of the computational effort involved in solving problems with discrete variables can be attributed to the piecewise linearization procedure and the number of breakpoints required to capture nonlinear-ities in well and pipeline models 6. The absence of these variable types in our case studies significantly reduces the computational effort (as reflected inTable 7) compared to the MILP/MINLP studies.…”
An oil and gas field requires careful operational planning and management via production optimisation for increased recovery and long-term project profitability. This paper addresses the challenge of production optimisation in a field undergoing secondary recovery by water flooding. The field operates with limited processing capacity at the surface separators, pipeline pressure constraints and water injection constraints; an economic indicator (Net Present Value -NPV) is used as the objective function. The formulated optimisation framework adequately integrates slow-paced subsurface dynamics using reservoir simulation and the fast-paced surface dynamics using sophisticated multiphase flow simulation in the upstream facilities. Optimisation of this holistic longterm model is made possible by developing accurate second order polynomial proxy models at each time step. The resulting formulation is solved as a Nonlinear Program (NLP) using commercially available solvers. A comparative analysis is performed using MATLAB's fmincon solver and the IPOPT solver for their robustness, speed and convergence stability in solving the proposed problem.By implementing 2 synthetic case studies, our mathematical programming approach determines the optimal production and injection rates of all wells and further demonstrates considerable improvement to the NPV obtained by simultaneously applying the tools of streamline, reservoir and surface facility simulation for well rate allocation.
“…In this paper, we apply a physics-based multidimensional CFD approach for the elucidation of annular geometrical intricacies on the transport phenomena of drill cuttings. The work herein stems from the results of a previous analysis of ours that considers the effect of particle sphericity on cuttings transport (Epelle and Gerogiorgis, 2018b). The obtained results have been postprocessed in a different way (by extracting flow information from strategically positioned lines and planes in the annular domain, as well as considering threshold values) for better understanding of the cuttings transport phenomena.…”
In this study, the two-fluid Eulerian-Eulerian multiphase flow model in ANSYS Fluent (v17.1) is adopted for the simulation of cuttings transport in a deviated annular geometry using two different non-Newtonian drilling fluids (described using the power law and Herschel-Bulkley models). The Syamlal-O'Brien interphase momentum exchange coefficient is implemented to capture non-spherical effects of the drill cuttings. The analysis conducted is based on the hypothesis that a position-dependent profile evaluation is expected to yield more insight into the transport process compared to a volume-averaged analysis over the entire flow domain. This is because the adopted simulation approach takes into account the microscopic particle properties which significantly affect the overall particle transport mechanism. However, this requires the application of robust postprocessing functionalities for data extraction from desired annular regions. Particle sizes considered are in the range of 0.002 m to 0.008 m with a sphericity range of 0.5 to 1.0. A rotational effect is incorporated in our model to describe the drillpipe motion in an annular wellbore with a vertical eccentricity of 0.6. The considered geometry contains a vertical, inclined and horizontal section with interconnecting upper and lower bends. The analysis of the particle velocity and concentration profiles revealed that the mud rheology, particle sphericity and particle sizes play vital roles in determining the cuttings removal process. It is particularly observed that the lower annular region of upper bend, is most susceptible to particle deposition with the lowest transport velocities observed. Our positional variability analysis has shown that the alternating dominance of nonspherical and spherical particles' velocities significantly depends on the nature of the flow (i.e. dense granular flow or dilute annular flow in the upper and lower sections respectively).
“…Such choices dene our method as a PFEM-DEM approach, hence a subcategory of the so-called CFD-DEM methods [67,52,53,55,54,61,62]. The CFD-DEM has been used directly in industrial problems such as uid catalytic cracker (FCC) units [49], pneumatic transport [50] and mud-ow transport problems from the oil industry [51]. But perhaps the most promising use of CFD-DEM procedures is in the derivation of empirical relations and validation examples [47,48], which can be used to improve the models used in coarser but more aordable techniques.…”
This work proposes a fully Lagrangian formulation for the numerical modeling of free-surface particle-laden ows. The uid phase is solved using the Particle Finite Element Method (PFEM), while the solid particles embedded in the uid are modeled with the Discrete Element Method (DEM). The coupling between the implicit PFEM and the explicit DEM is performed through a sub-stepping staggered scheme. This work only considers suspended spherical particles that are assumed not to aect the uid motion. Several tests are presented to validate the formulation. The PFEM-DEM results show very good agreement with analytical solutions, laboratory tests and numerical results from alternative numerical methods.
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.