COMPARISON OF WIRELINE FORMATION-TESTER SAMPLINGWITH FOCUSED AND CONVENTIONAL PROBES IN THE PRESENCEOF OIL-BASE MUD-FILTRATE INVASION

摘要

Acquisition of fluid samples is challenging in thepresence of oil-base mud (OBM) filtrate invasion dueto its partial or full miscibility with reservoirhydrocarbons. In the course of fluid sampling,varying concentrations of OBM will lead tovariations of fluid properties such as viscosity,density, and gas-oil ratio (GOR). Contamination ofOBM filtrate in the sampled fluid can drasticallyaffect sample quality and lead to non-representativefluid properties. Fluid pumpout time can be extendedto reduce filtrate contamination, although at theexpense of increasing overall rig cost. A focusedprobe can be useful in reducing OBM contaminationby diverting flow into different channels withoutcompromising fluid pumpout time. However, it isimportant to properly quantify the relativeperformance of focused and conventional probes fora wide range of field conditions.The objective of this paper is to appraise theperformance of different probes under the samesimulated field conditions and for a comprehensiveset of petrophysical and fluid properties. We simulatethe process of fluid sampling with a focused probe ina vertical well and compare its performance to that ofa one-inch diameter conventional probe in thepresence of mud-filtrate invasion. A commercialadaptive-implicit compositional reservoir simulator isused to model both invasion and filtrate-cleanupprocesses. Comparison between fluid sampling withfocused and conventional probes identifies cases inwhich focused fluid sampling leads to improvedsample quality in a shorter period of time.Results indicate that sample quality generallyimproves when the flow is split between the guardand sample probes, but the specific amount ofimprovement depends on probe geometry, fluidcomposition, and formation properties. In addition,because the focused probe withdraws fluids through alarger cross-sectional area, the pressure drop causedby the focused probe is smaller than the pressure dropassociated with a conventional probe. Permeabilityanisotropy, presence of a shale boundary, and lack ofmud-filtrate invasion can help to improve samplequality. In addition, fluid cleanup can be acceleratedby altering both the probe design and the flow-rateratio between the sample and guard fluid streamsthereby leading to increased pressure differentialbetween the sample and guard areas. Performance ofthe focused probe improves with increasing fluidfrontdispersion because focusing enhances the“coning” of the mud-filtrate invasion front.Lastly, we perform inversion jointly from transientmeasurements of GOR and probe pressures toestimate formation permeability, anisotropy, and rateof mud-filtrate invasion. It is shown that, in general,focused fluid sampling provides enhanced degrees offreedom in the pressure and GOR transientmeasurements to improve both accuracy andreliability of the estimated formation properties.