Coupling Reservoir and Well-Completion Simulators for Intelligent Multilateral Wells

Abstract

This article, written by Editorial Manager Adam Wilson, contains highlights of paper SPE 164815, ’Coupling Reservoir and Well-Completion Simulators for Intelligent Multilateral Wells: Part 1,’ by G.A. Carvajal, N. Saldierna, M. Querales, SPE, K. Thornton, SPE, and J. Loiza, Halliburton, prepared for the 2013 SPE Europec/ EAGE Conference and Exhibition, London, 10-13 June. The paper has not been peer reviewed. To maximize well productivity and reservoir oil recovery in mature fields, advanced lateral designs with intelligent completions have been implemented during the last decade. This paper describes an all-in-one system that combines nodal-analysis and numerical-simulation models to calculate the effect of intelligent-completion components—such as swell packers, internal control valves (ICVs), and inflow-control devices (ICDs)—on lateral production profiles. A 3D-grid, multiphase-flow, nonisothermal reservoir numerical simulator was used for advanced well design, and a wellbore nodal-analysis simulator was used to solve the partial-differential equations between the porous medium and the wellbore. Introduction Multilateral wells are often the best choice to maximize reserves recovery and contact with the reservoir and to extend the drainage area. However, several studies have revealed that multilateral wells exhibit a series of natural issues (e.g., high influx at lateral heels because of pressure drops). As a result of such issues, increased water or gas production can decrease oil production over time. Use of intelligent-valve technology can mitigate these problems significantly. Reservoir Complexity and Heterogeneity The reservoir model used here is a synthetic reservoir with an almost constant porosity of 20% and significant areal permeability variation from 1 to 500 md. The main depth is 8,000-ft true vertical depth subsurface. The net pay is 100 ft, and net/gross ratio used for the simulation is 0.8. A 3D sector numerical- simulation model was built with the following dimensions: x=100 cells, y=50 cells, and five layers, for a total of 25,000 cells. The areal permeability distribution is taken from a geostatistical model. The permeability distribution is con-strained by spatial carbonate-facies distribution honoring log and core data of injector and producer wells in the sector model. Reservoir-quality index (RQI) is calculated to screen areas with best reservoir quality; an area where RQI equals 10 is considered the best area to place multilaterals. However, the reservoir exhibits strong permeability contrast, not only among layers but also inside the carbon-ate body, as shown in Fig. 1.