Electrical Heating of Oil Reservoirs: Numerical Simulation and Field Test Results

Abstract

Summary This work presents the development of a numerical model designed to simulateEOR by in-situ electric heating. The paper includes the results of validationtests vs. analytical solutions, comparisons of oil production and energyconsumption for different electrode schemes, and the results of simulations ofthe Rio Panon, Brazil, pilot test. Panon, Brazil, pilot test. Introduction A number of possibilities exist for the use of electricity to heatoil-bearing formations. These methods can be classified according to themechanism of thermal dissipation that predominates in the process. The linepredominates in the process. The line frequency plays a decisive role in howthe electrical (and electromagnetic) energy is converted to heat. Dielectricheating prevails in the high-frequency range, from radio frequency tomicrowave. The dipoles formed by the molecules tend to align themselves withthe electric field. The alternation of this field induces a rotatory movementon the dipoles with a velocity proportional to the frequency of alternation. The molecular movement can be intense enough to produce considerable heat. Apopular application of this process is encountered in microwave ovens. processis encountered in microwave ovens. Another possibility is inductive heating, where the electric AC flowing through a set of conductors induces a magneticfield in the surrounding medium. The variation of the magnetic field, in turn, induces secondary currents, whose circulation in the medium results in heat. This work is confined to the resistive heating process, which is the majormechanism when DC or low-frequency AC (up to 300 cycles/sec [300 Hz]) is used. Fig. 1 shows a schematic of the process. The electrical heating of reservoirformations was used to enhance oil production as early as 1969, when anexperiment in Little Tom, TX, was reported successful. The production of fourwells had increased from production of four wells had increased from 1 B/D[0.16 m3/d] to an impressive average of 20 B/D [3.18 m3/d] for the experiment, which included wellbore fracturing. The method subsequently attracted theattention of an increasing number of investigators and engineers, and otherfield tests were reported within a few years. The first academic work on theresistive heating process was by El-Feky in 1977. He reported process was byEl-Feky in 1977. He reported on the development and testing of a numericalmodel that was based on an implicitpressure, explicit-saturation formulationover a 2D rectangular grid. Experimental data came from a laboratory modelconsisting of a five-spot waterflood. The electric-heating concept was latercoupled to water injection processes to derive the so-called processes toderive the so-called reservoir-selective-heating method. Until 1986, the fewexisting reservoir simulators for the electrically enhanced process relied onexplicit treatments to determine process relied on explicit treatments todetermine saturation, voltage, temperature, and pressure. Killough and Gonzalespresented a pressure. Killough and Gonzales presented a fully implicit, 3D, multicomponent model in 1986 that was capable of handling water vaporization. The authors focused on the idea of heating waterflood patterns. In 1988, Wattenbarger and McDougal used a 2D simulator to investigate the majorparameters affecting the production response to parameters affecting theproduction response to electric resistive heating. They considered thesteady-state regime to obtain a simple method for estimating the-productionrate. A 1987 pilot test was conducted jointly by Petrobra's, Azevedo Travassos, and ORS Petrobra's, Azevedo Travassos, and ORS Development Co. in the Rio Panonfield. Rio Grande do Norte, Brazil. A single well was connected electrically, and its parameters were thoroughly monitored. The data parameters werethoroughly monitored. The data showed a sudden increase in temperature at thepay zone, accompanied by a remarkable increase in the oil production rate. Thepurposes of the current work were to develop a numerical model for theelectrical heating method and to investigate the response obtained whendifferent electrode schemes are used. Validation of the model includes itsapplication to the Rio Panon example. Mathematical Formulation The problem formulation assumes that electrical heating is applied to adeveloped field containing a high number of producing wells. Therefore, the oilflow part of the study refers to a single drainage area. In most cases, thedrainage area presents a regular shape and can be approximated by an equivalentcircular domain without major effects on the final results. Three furtherassumptions were made in the problem statement and during the numericalsolution. 1. The oil-bearing formation is horizontal and a single layer.2. Theevolution of dissolved gas is not considered separately, and because we areinterested in the effects of viscosity reduction in heavy oils, only two phases(oil and water) are considered. P. 1320