Reservoir Inflow While Underbalanced Drilling-A Qualitative Perspective

Abstract

Abstract This paper discusses the importance of inflow in underbalanced drilling. The various inflow models are mentioned. A link between dynamic inflow performance model and hydraulic model is emphasized. The potential quantification of inflow on bottomhole pressure changes due to injected blend ratios during underbalanced drilling is highlighted. Using the relationship between inflow and bottomhole pressure changes, optimum wellbore lengths as well as reservoir evaluation could be achieved during underbalanced drilling. Introduction The application of underbalanced drilling is gaining momentum as a new technology capable of providing economic benefit to projects where formation damage and/or problematic drilling events have frequently affected project viability and economics. An unexplored and poorly understood benefit of underbalanced drilling is the ability to quantify reservoir inflow in real time while drilling. Reservoir engineers have not previously been given the opportunity to directly relate inflow to decreased damage due to the underbalanced drilling process. Reservoir modelling and calibration has typical been based on history matching of wells that were drilled "conventionally." In spite of extensive production data and years of knowledge applied in the field, experience has shown that predictions of formation inflow rates for underbalanced wells can be underestimated by as much as 200 to 300%. The methodology used to calculate inflow as applied to wells drilled underbalanced must be revised. Many reservoir inflow computer models exist and are available to the reservoir engineer. Internal calculation engines typically rely on the conservation of mass and/or heat. Basically, these models predict inflow of a fixed well geometry over a specific sandface drawdown in a specified time. Radial, spherical and linear flow regimes ranging from pseudo-steady-state through to steady-state flow conditions are investigated, ultimately providing data regarding total flow to surface. What is not commercially and readily available, is a reservoir model that predicts inflow in a dynamic elemental sense. Simulation of petroleum reservoir performance refers to the construction and operation of a model whose behaviour assumes the appearance of actual reservoir behaviour(1). The model itself is either physical (for example, a laboratory sandpack) or mathematical. A mathematical model is simply a set of equations that, subject to certain assumptions, describes the physical processes active in the reservoir. In calculating the productivity of wells, the common assumption is that inflow into the well is directly proportional to the pressure differential between the reservoir and the wellbore. The proportionality constant PI, is derived from Darcy's law for steadystate radial flow for a single, incompressible fluid. In instances where this relationship holds, a plot of production versus bottomhole pressure yields a straight line. Muskat(2) pointed out that when two-phase (liquid and gas) flow exists in the reservoir, this relationship may not hold; and went on to present theoretical calculations to show that graphs of producing rates versus bottomhole pressures for two-phase flow resulted in curved rather than straight lines. With the existence of a curvature, a well cannot be said to have a single PI, since the value of the slope varies continuously with changes in the drawdown.