Dynamic Modeling of Invasion Damage and Its Effect on Production in Horizontal Wells

Abstract

This article, written by Assistant Technology Editor Karen Bybee, contains highlights of paper SPE 95861, "Dynamic Modeling of Invasion Damage and Impact on Production in Horizontal Wells," by P.V. Suryanarayana, SPE, and Z. Wu, SPE, Blade Energy Partners; J. Ramalho, SPE, Shell Intl. E&P; and R. Himes, SPE, Stim Lab, prepared for the 2005 SPE Annual Technical Conference and Exhibition, Dallas, 9–12 October. A radially adaptive fine-grid micro-simulator was used to model invasive damage during overbalanced drilling. Near-wellbore pressure and saturation distributions obtained from this model were used as initial conditions for a sector-scale model to study the effects of damage on long-term production performance. Introduction Overbalanced drilling and workover operations often result in invasion of the near-wellbore region by filtrate and solids from the drilling and workover fluids. This invasion causes a reduction in productivity because of degradation of effective permeability. During drilling, mudcake buildup can reduce invasion depth. Mudcake buildup and effectiveness depend on mud formulation, formation characteristics, and overbalance. In horizontal wells, mudcake effectiveness is compromised because of repeated pipe movement against the mudcake, which results in mudcake removal and reformation. Dynamic Damage Modeling A key feature of the model is a fine-grid dynamic microsimulator used to model invasion and obtain a near-wellbore profile of water saturation and reservoir pressure. Saturation-dependent effects such as relative permeability are modeled by use of relative permeability curves based on core tests. The microsimulator has an investigation range of approximately 50 ft from the wellbore. Inputs include the usual reservoir and drilling parameters such as mud properties, overbalance, and drilling time vs. depth. Simulations were performed with and without mudcake to represent the extremes of the invasion boundary. Results from a specially designed core test include undamaged effective permeability, resistance to flow from the wellbore into the reservoir because of the mudcake, absolute permeability damage, and mudcake liftoff pressure. Main output of the invasion model is a near-wellbore saturation and reservoir-pressure profile within the invasion zone that is used with the far-field reservoir properties outside the invasion zone as initial conditions in a sector-scale dynamic simulator with near-wellbore fine grids. Sector dimensions are chosen sufficiently large that no-flow or constant-pressure conditions apply at its boundary. Production is initiated under assumed conditions of rate and pressure constraints, and decline curves are obtained over a desired period of interest (usually 2 to 3 years) for each case of interest. The resulting decline curves for different options can be used to evaluate the effects of invasion on production performance. Fine-Grid Simulator In this work, the near-wellbore fluid-saturation distribution is solved from the two-phase immiscible flow equations as applied to a single-well model. A 2D axisymmetric cylindrical coordinate system is used. Permeability and porosity are assumed to be azimuthally symmetrical along the axis of the horizontal well, and streamlines also are uniformly and symmetrically distributed along the axis. Because the finite-difference method is used to solve the flow equation, extremely fine gridblocks are required to maintain uniform accuracy. A geometric progression scheme is used to generate radially adaptive gridblocks. Finer grid-blocks provide a seamless mechanism to capture the effects of permeability degradation near the wellbore. Typical gridblock size for reservoir-scale models is much larger than the radius of drilling-induced near-wellbore damage. Because a very short time frame (drilling) is being simulated, extremely small timesteps are used to minimize discretization errors.