Inference of rock pressure-production properties from gas-oil production logs

Abstract

Conventional interpretation of production logs (PL) numerically accounts for fluid productivity (or injectivity) of reservoir rocks by constructing simplified reservoir flow models. They describe the contribution of various fluid-producing rock formations into borehole fluid flow without quantitatively estimating formation dynamic properties. This paper interfaces borehole and formation fluid flow models to link formation petrophysical properties to borehole production logs. A new interpretation method is subsequently developed to diagnose and quantify formation near-borehole permeability and gas saturation from gas-oil production measurements acquired in deviated boreholes. The specific application considered in this study invokes a new coupled flow algorithm to simulate pressure-production behavior of individual rock formations in multilayer reservoirs with/without cross-flow.Simulation of borehole fluid flow is performed by solving separate mass and momentum conservation equations specified for each fluid phase. Drag and buoyant forces are assumed as primary sources for interfacial momentum transfer. Based on a flow-regime map introduced by Hasan and Kabir (1988), we dynamically modify interfacial drag terms to account for fluid-phase slip velocity. Under the assumption of local thermodynamic equilibrium, an equation-of-state (EOS) compositional model is invoked to simulate borehole fluid-phase interfacial mass transfer. Coupling separate borehole and formation fluid flow models is carried out by introducing appropriate source terms into fluid-phase mass conservation equations. Subsequently, a nonlinear inversion-based algorithm estimates near-borehole permeability and gas saturation by minimizing quadratic differences between measurements of borehole fluid velocity, holdup, and pressure, and their corresponding numerical simulations.In a synthetic multilayer reservoir model with a gas cap, PL interpretation recommends a critical bottom-hole pressure to prevent high gas production because of (i) downward advancement of the gas displacement front, and (ii) released gas from oil solution. The estimation error is less than 20% for near-borehole gas saturation, and less than 25% for near-borehole permeability. However, the reliability of estimations is adversely influenced by conditions of immovable gas saturation in rock formations. Additionally, we examine a field example consisting of a gas-oil sand-shale laminated system where PL interpretation is carried out to quantitatively the investigate inflow performance relationship of individual flowing units. The interpretation method integrates well-log-derived and PL-derived permeabilities to quantify the depth distributions of near-borehole skin factor and gas saturation. The estimated near-borehole reservoir model subsequently quantifies more than 60% reduction of unwanted gas production caused by a gas shut-off remedial operation.