RIASSUNTO
ABSTRACT
It is well known that the acquisition of representative formation fluid samples is essential for reservoir management and development. However, because of overbalance pressure in the mud column, mud filtrate invades and contaminates the reservoir fluid during the drilling process but before the mudcake around the wellbore is properly formed. Although water-based mud (WBM) is immiscible with formation fluid, oil-based mud (OBM) is miscible with it. Samples with OBM contamination levels greater than at least 10% for oils and 3% for volatile oils and gas condensates might be considered unusable because the OBM contamination alters the formation fluid properties, preventing an accurate characterization of the reservoir fluid. Despite a large body of research, it is difficult to prevent contamination. Unfortunately, openhole sampling is usually a single opportunity event; by the time the laboratory analysis is complete, it is not possible to acquire additional samples. Consequently, it is important to be able to measure the contamination level of the reservoir fluid as accurately as possible in real time before taking the sample. In addition, by knowing the contamination level in real time, the optimal timing of sampling can easily be determined, which reduces rig time/cost and potential fishing risks. Current techniques to estimate contamination nearly ubiquitously rely on curve fitting of measured properties, such as density, fluid compositions (including gas/oil ratio [GOR] and asphaltene content), or color. These techniques, however, suffer from several shortcomings, such as tool drifting, dependence on end-member filtrate, and formation fluid properties. In all techniques, the measured properties are assumed to asymptotically approach the clean fluid properties.
This paper proposes a method to estimate the drilling fluid contamination levels and characterize reservoir fluid in real time using the formation tester tool measurements of the fluid. Although the equation-of-state (EOS) method has been previously proposed, in this approach, a combination of a multi-point EOS, distribution function of formation fluid, inherent geochemistry principles, and empirical correlations are used. This EOS method has been validated with laboratory and field data.
The algorithm developed uses an inverse method to compute reservoir fluid contamination. It takes as inputs the downhole fluid composition, including C1, C2, C3, C4-C5, and C6+ saturate, aromatic, resin, and asphaltene fractions, and fluid properties, such as GOR, density, bubble-point pressure, compressibility, and mud filtrate composition. Using an iterative process, an optimum combination of formation and filtrate fluids whose properties best match the fluid properties supplied to the algorithm is determined.