The accurate determination of fluid properties and contamination while sampling with a wireline pump-out formation tester is essential to achieve the primary objective of obtaining representative reservoir fluid samples with minimum rig time. Despite advancement in fluid identification sensors, sampling in mixed phases, especially immiscible fluids, still poses great challenges. In many cases, apparent erratic sensor responses are attributed to sensor noise and considered to be uninterpretable. However, careful study reveals that the sensors are actually showing the true nature of the multi-phase fluid flow. If this multi-phase behavior is not taken into consideration, it is difficult to determine fluid type and the level of contamination. This work addresses the development of new numerical and analytical models that make it possible to not only understand the cleaning behavior of formation fluids, but also quantitatively determine fluid-sample contamination in real time. The techniques shown in this paper highlight the variables that play important roles in guiding the cleanup process. The methods predict the contamination level in both the time and fluid volume pumped. Furthermore, the required pumping time, or volume needed to achieve the desired level of contamination, is also calculated with this method. The statistical uncertainties of these estimates are also considered based on how well the sensor data matches the analytical contamination model. Examples are provided to illustrate the efficiency of this technique in oil-based-mud (OBM) and water-based-mud (WBM) contaminated examples for both hydrocarbon fluid and formation water samples. The new analysis technique is applied to a highresolution fluid-density sensor that monitors the change in the resonance frequency of a vibrating tube-carrying fluid sample. The same interpretation method is also applied to a capacitance sensor and a resistivity sensor to further confirm the results derived from the density sensor. Once samples chambers are recovered at the surface, non-invasive measurements are performed on the samples to verify the fluid type and sample quality. This is achieved through precise measurements of sample density and compressibility without opening the chamber. When these measurements are compared with dowhole measurements, including mud-filtrate density, the contamination can be predicted. This technology enables verification of downhole measurements and ensures that sample integrity is maintained when they are retrieved at the surface. This, in turn, enables wellsite decisions to be made regarding which samples are the most representative before shipment to PVT lab.
展开▼
