Search

Abstract: Predicting the time for which a petroleum mixture has been exposed to weathering effects could have many applications. Historically, research on the evaporation rates of hydrocarbon mixtures has focused on forensic oil spill identification. Relatively little attention has focused on estimating the exposure time for a weathered petroleum sample based on the observed composition at a given time, and assuming a prior composition. A hierarchical application of multivariate techniques was used to estimate the time for which a hydrocarbon mixture was exposed to evaporative weathering. Partial least squares discriminant analysis (PLS‐DA) could predict whether a sample was relatively fresh (< 12 hours exposure time) or highly weathered (> 20 hours exposure time). Subsequent regression models for these classes were evaluated for accuracy using the root mean square error of prediction (RMSEP). One nonlinear regression model in particular was found to satisfactorily estimate the age of weathered petroleum samples.

Micromachined Joule-Thomson (J-T) coolers have applications ranging from cryosurgery to cooling infrared detectors. With the absence of cold moving parts, the J-T coolers can be implemented with simple structures that are suitable for silicon/glass microfabrication. The investigation proposed in this thesis focuses on the development of micromachined Si/glass heat exchangers used in the J-T coolers that operate at 200-225K when the gas pressure is 1-2MPa. The heat exchangers must maintain good stream-to-stream heat conductance between the high- and low-pressure streams while restricting stream-wise conduction to achieve a high effectiveness. Two heat exchangers were designed, fabricated and tested. The first, a planar design, uses rows of high-conductivity silicon fins bonded onto a 100µm thick low-conductivity glass base plate. It was fabricated using a five-mask process including Si/glass/Si anodic bonding, two-step DRIE, and HF glass etching, etc. The second, a perforated-plate design, uses numerous silicon plates alternated with glass spacers. It was fabricated using a four-mask process including KOH on (110) silicon wafers, HF glass etching and anodic bonding. Platinum resistance temperature detectors were integrated into the heat exchanger for in-situ temperature sensing. Whereas the performance of the planar heat exchanger was limited by its ability to accommodate a pressure differential across the base plate, the perforated-plate heat exchangers demonstrated a high effectiveness (0.912) in at 237-252K in effectiveness tests and good robustness at high pressures (1MPa) in J-T self-cooling tests. The temperature distribution along the heat exchanger was measured by integrated resistance temperature detectors with sensitivities of 0.26-0.30%/K at 205-296K. A J-T system using the perforated-plate heat exchanger achieved 218.7K at steady state and 200.3K in a transient state. The system provided 200mW cooling power at 228K and 1W at 239K with an estimated parasitic heat load of 300-500mW. Finally, a flow-controlled J-T system using a perforated-plate heat exchanger and a piezoelectric microvalve was demonstrated. By modulating the flow, the microvalve could vary the cooling temperature by 5-8K around the operating points, which were 254.5K at 430kPa pressure difference in steady state, and 234K at 710kPa in transient state, without an added heat load.