Viscosity is driven by asphaltene content and is a key parameter in the development of heavy oil fields. Understanding fluid composition and temperature and pressure-induced changes in fluid viscosity is vital for an optimized production strategy and surface facility design. A recent field and laboratory study exemplifies the steps necessary to obtain the fit-for-purpose data from heavy oil samples. This paper presents the case study of a new downhole optical composition analysis sensor used during real-time downhole fluid analysis and sampling for the first time in a Kuwait heavy oil formation.
The primary objectives of a sampling program are to confirm fluid indications on the openhole logs and collect crucial pressure/volume/temperature (PVT) samples. The downhole optical composition analysis sensor provides the information necessary to estimate a sample contamination level. It also indicates when the sample is sufficiently clean for PVT analysis. The samples should be acquired from the reservoir and maintained as single phase throughout transport to the laboratory. The pressure should be maintained higher than the asphaltene precipitation onset pressure and much higher than the bubblepoint. If the sample is not maintained higher than the asphaltene onset pressure, asphaltenes precipitate in the sample chamber and cannot be reconstituted as single phase in the laboratory. The new optical composition analyzer can also identify fluid stream components and their relative concentration in real time with laboratory-quality accuracy downhole.
Near-infrared (NIR) sensors are most commonly used to identify fluid in the wireline formation tester (WFT). The sensors work well in light hydrocarbons. However, in heavy oil, the sensor performance degrades and fails to identify the contamination level accurately. The new multivariate optical computing (MOC) technique for downhole optical composition analysis overcomes this by performing a photometric detection with the entire relevant spectral range compared to spectroscopic analysis, which is only performed over a narrow band or sparse set of channels while traditional sensors are configured. The MOC sensor also recognizes in real time the chemical nature (optical fingerprint) of analytes (e.g., methane, ethane, propane, carbon dioxide, hydrogen sulfide, water, asphaltene, aromatics, and saturates) using all of the useful information in the optical spectrum. The real-time analyte chemical composition provided by the sensor is comparable to laboratory tests conducted on the collected PVT sample.
Laboratory measurements on representative fluid samples from the correct locations early in the field development stage help develop an optimal field-development strategy. At the same time, sample integrity is maintained from the reservoir to the laboratory, which is vital. This paper discusses how the new optical compositional analysis sensor in combination with a high-resolution fluid identification sensor provides comprehensive and accurate downhole fluid composition in real time. This compares well with the laboratory-measured PVT analysis of heavy oil samples. The compositional analysis sensor optimizes pumpout time, thus helping obtain practically ideal contamination levels to begin the single-phase sampling process, which saves valuable rig time.