Search

Robust downhole fluid analysis highly relies on quality measurements and data mapping from various optical sensors implemented into modern formation testers. The direct modeling used to determine a multivariate correlation between optical sensor responses and diverse fluid compositions or properties is often cost-prohibitive with sensor-based calibration. This paper presents a novel method based on concatenated optical computing neural networks (COCN), which link sensor-specific signal transformation to generic fluid characterization through validated data mapping. The COCN models are built separately in the in-house laboratory using optical data transformation networks and multi-analyte fluid characterization networks. To evaluate the uncertainty of sensor signal mapping, additional sensor data prediction networks, which exchange inputs and outputs of sensor data transformation networks to provide reverse data mapping, can be built to produce the simulated sensor responses. The actual sensor measurements are then compared to the simulated sensor responses to validate the optical signal transformation. The degree of matching can be used to identify issues associated with the quality of the sensor measurements and the range of calibration and testing data, and to select the transformation networks in ruggedizing predictions of fluid compositions and properties. This paper presents the method used and case studies to exemplify the application of the proposed approach for reliable fluid sample characterization using laboratory and field measurement data. Calibrating sensor-independent fluid predictive networks on a large database usually pertains to low uncertainty because the analyzed measurements can be collected from diverse reservoir fluid samples. The construction of data transformation networks, however, is only realistic by using a small number of reference fluids because of their high calibration cost, which conversely challenges the quality of data mapping with new field sensor measurements. In this study, sensor data transformation networks and sensor data prediction networks form two groups of mutually complementing models that can validate one another, making the quality of data mapping observable before applying fluid characterization models. The results of the case studies demonstrate that the consistency between the actual and simulated sensor responses is a reasonable performance index in the prediction of fluid compositions and properties. The laboratory and field examples also justified the importance of using simulation data appropriately to overcome the limitation of sensor measurement data in the calibration of transformation networks.

Phase behavior characterization (PVT) and geochemical compositional analysis of petroleum samples play a crucial role in the reservoir evaluation process to help determine producible reserves and the best production strategy. Openhole samples are the most valuable types of samples for PVT and geochemical analysis. Unfortunately, traditional openhole sampling methods are costly and limited to ten to twenty samples, thereby restricting the scope of characterization in a well section. This study summarizes a new microsampling technique for logging while drilling (LWD) and a corresponding wellsite technique to provide compositional interpretation, contamination assessment, reservoir fluid compositional grading, and reservoir compartmentalization assessment. This microscale approach allows fast analysis with a field or near-field deployment of the analytical tool, providing fast turnaround time for analysis. The results inform planning for wireline sample retrieval, if necessary. The microsampler used in the downhole tool is capable of collecting reservoir fluid in small quantities, suitable for compositional analysis. Because of its small size, the microsampler can gather multiple fluids at various reservoir depths, while PVT sampling requires larger volumes and has more constraints. However, when used in combination with conventional PVT-grade samples, the microsamples can provide significant chemical profiling. The quantity of 40 microliters (μl) provides the opportunity to collect many more samples than the conventional PVT sample size of 200 to 1,000 milliliters (ml). Additionally, 40 microliters provides more than enough of a sample for a complete chemical analysis using a liquid chromatograph or gas chromatograph coupled to either a mass spectrometer for biomarker analysis or a flame ionization detector for a complete assay. Isotope analysis is also possible. Recovery to surface of fluid samples collected at reservoir temperature and pressure allows for analysis with an automated gas chromatograph (GC) deployed in the field, providing reduced labor and rapid analysis. The unique injection chamber of the GC is designed with the injection port and valve configured to withstand pressure up to 5,000 psi, which is approximately five times higher than standard GC injection valves. This allows for injection of the microsample with a solvent carrier as a single-phase fluid so that analysis can provide composition and fluid properties, such as gas to oil ratio, without a flash. The GC has two detectors including a flame ionization detector (FID) for hydrocarbon components and thermal conductivity detector (TCD) for inorganic gas components, such as carbon dioxide, nitrogen, and hydrogen sulfide. The system can quantify hydrocarbon components from C1-C36 and perform contamination studies of oil samples with drilling fluids. This study provides a new technique for reservoir engineers to characterize a reservoir completely, without limit to the number of acquired samples. In combination with conventional PVT samples, it is possible to extrapolate the PVT properties to all pump-out stations, and conduct a complete geochemical profile of the reservoir.