Phase Redistribution

Key Takeaways

• The pressure derivative "hump" caused by phase redistribution in pressure buildup tests is a characteristic but non-unique signature that can be misidentified as wellbore storage afterflow, a composite reservoir system, or a fault or boundary effect: the hump appears on the log-log diagnostic plot as the pressure derivative rises above the radial flow stabilization level (the "MTR line") at intermediate times during the buildup, typically between 1 and 100 hours after shut-in depending on wellbore geometry and fluid properties; the physical mechanism is the redistribution of gas from a distributed mixture throughout the wellbore to a gas cap at the top of the tubing, which causes the bottomhole pressure (measured by a gauge at or near the perforations) to increase as the gas column above the gauge lightens and then decrease as the gas cap stabilizes; the amplitude of the hump depends on the density contrast between gas and liquid phases, the wellbore volume available for segregation, and the rate of gravitational settling relative to the pressure equilibration rate with the reservoir; distinguishing a phase redistribution hump from reservoir effects requires simulation of the wellbore fluid mechanics or specialized wellbore storage models that account for compressible and segregating multiphase fluids, rather than conventional wellbore storage models that assume a single homogeneous compressible fluid.
• Phase redistribution in horizontal and deviated wellbores creates additional complexity because fluid segregation occurs along the trajectory of the wellbore rather than purely vertically: in a horizontal well with a gas cap in the reservoir, the gas migrates to the high point of the wellbore path during shut-in, which may be at the heel, at a high-angle dogleg, or at any local topographic high in the trajectory; the resulting fluid distribution creates a variable hydrostatic pressure profile along the wellbore that is not captured by simple wellbore storage models; as gas migrates, the water or oil that it displaces flows toward the low points of the wellbore, and the changing hydrostatic gradient along the wellbore causes a complex transient pressure signal at any fixed gauge location; in extended-reach horizontal wells (wellbore lengths of 5,000-10,000 feet), the volume available for phase redistribution is very large, the redistribution process takes many hours to equilibrate, and the resulting pressure signature contaminates a large fraction of the pressure buildup that would otherwise provide useful reservoir data; identification of horizontal well phase redistribution effects requires wellbore simulation models that account for the actual wellbore trajectory and inclination profile, not just the total wellbore volume.
• Phase redistribution effects in gas condensate wells are particularly severe because the wellbore fluid composition changes significantly between flowing and shut-in conditions: during flow, the gas condensate mixture in the wellbore is at or near the flowing wellhead pressure and temperature, and may be above the dew point in the wellbore (single-phase gas); when the well is shut in, the wellbore pressure rises toward the static reservoir pressure and the wellbore cools toward the geothermal gradient, causing retrograde condensation in the wellbore fluid column as conditions pass through the dew point; the condensed liquid is denser than the gas and migrates downward while the gas migrates upward, changing the hydrostatic gradient above the downhole gauge and creating a phase redistribution pressure signature superimposed on the reservoir pressure buildup; the liquid volume that condenses depends on the condensate-gas ratio (CGR) and the phase envelope of the specific fluid, which varies from well to well and requires PVT characterization to quantify; wells with high CGR (greater than 50 STB/MMscf) and high permeability (where the wellbore pressure equilibrates quickly with the reservoir) show the most severe phase redistribution effects because the rapid pressure rise during early shut-in drives rapid condensation and segregation before the pressure derivative has stabilized into radial flow.
• Mitigation of phase redistribution effects in pressure transient tests includes downhole shut-in (using a downhole shut-in tool to close the well at the formation face rather than at the surface, eliminating the wellbore volume available for phase redistribution above the shut-in point), dual-gauge tests with permanent downhole gauges positioned both above and below the likely gas segregation zone (allowing comparison of pressure signals that bracket the redistribution zone), and numerical wellbore simulation that models the multiphase wellbore fluid dynamics during shut-in and strips the redistribution contribution from the measured bottomhole pressure before reservoir analysis; downhole shut-in is the most effective mitigation but adds cost and operational complexity (the shut-in tool must be retrieved or operated downhole, requiring additional wireline or coiled tubing intervention); for wells where downhole shut-in is not practical, the industry standard for handling phase redistribution is to identify the affected portion of the pressure buildup (using the log-log derivative signature) and to analyze only the portion of the data not affected by redistribution, typically the late-time radial flow data that follows the redistribution hump, accepting that some information about near-wellbore conditions is lost from the analysis.
• Phase redistribution in injection wells (water injection, gas injection, and CO2 injection) causes analogous effects when injection is stopped for a falloff test: when water injection into an oil reservoir is stopped, the column of injection water in the wellbore is denser than the oil it displaces near the perforations, and gravity-driven crossflow between the injection water and the reservoir oil creates a transient pressure signal at the downhole gauge that is superimposed on the reservoir falloff response; for gas injection or CO2 injection wells, the low-density injectant in the wellbore migrates to the top of the tubing when injection stops, lightening the hydrostatic column above the gauge and creating an apparent pressure decrease that is not a reservoir pressure signal; the falloff test analysis of injection wells must account for these wellbore effects, either by using downhole injection-point gauges that are positioned below the fluid crossflow zone, by using numerical wellbore models to correct the measured pressure, or by designing the falloff test with sufficient duration that the late-time data (after wellbore phase redistribution has equilibrated) provides a reliable reservoir signal uncontaminated by near-wellbore fluid mechanics.