Wax deposition tests under single-phase and two-phase dispersed water-in-oil at 16 and 35% water cut (WC) were conducted in a mini pilot-scale flow loop by using South Pelto crude oil. The initial inner wall temperature (T̅ w,ini ) and the bulk fluid temperature (T b ) were controlled to be 85−87 °F and 105−107 °F, respectively, throughout the experimental program. The initial Reynolds number (Re ini ) of tests conducted in rectangular (TS1) and circular (TS2) test sections was in the range of 400−4300 and 1000−5300, respectively. A high temperature gas chromatograph with flame ionization detector (HTGC-FID) and a differential scanning calorimeter (DSC) were used to analyze wax deposits. The results revealed that the 16 and 35% water cut tests yielded deposits containing only a trace amount of water (less than 2% w/w). The crossover behavior of the thickness versus time trends for different flow conditions was observed. The initial thickness growth rate increases with the flow rate. After the crossover period ended, the deposit formed under a lower flow rate condition had a higher thickness. In this experimental program, velocity, wall shear stress, Reynolds number, and heat transfer coefficient are coupled together. The initial heat transfer coefficient was found to correlate to the 48 h deposit thickness best compared to the initial wall shear stress, the initial Reynolds number, and the initial velocity. The effect of shear on the deposit composition was observed.

The process of crystallization is of great interest in a wide range of industries. In the oil industry, a major interest is the deposition of wax onto subsea oil pipelines, a costly phenomenon that hinders the production of crude oil. It is known that these deposits are a volume spanning network of orthorhombic, lamellar wax crystals consisting primarily of n-alkanes that entrap some of the crude oil to form a gel. The presence of other materials in a crystallizing system can have an impact on both thermodynamic and kinetic parameters. To analyze the effects of how n-alkanes impact the crystallization of one another, three different types of apparatus (differential scanning calorimetry, densitometer, and a coldfinger apparatus) were used to explore a wide range of crystallization and deposition properties. The results of these experiments showed that longer chained n-alkanes greatly influence the crystallization properties of shorter n-alkanes, whereas shorter n-alkanes only slightly impact the crystallization properties of longer chained n-alkanes. This impact is directly related to the amount of cocrystallization that exists between the n-alkanes, which is dictated by the carbon number difference, solubility differences, and cooling rate. Cocrystallization shifts the temperature at which crystallization occurs and reduces the heat that is released by the system. Polydispersity and cocrystallization also reduce the mass and wax fraction of a deposit formed using a coldfinger apparatus.

Paraffin or wax deposition under two-phase gas-oil slug flow in horizontal pipes was studied experimentally. The experiments were conducted using a 2.067-in large-scale multiphase flow loop under an operating pressure of 350 psig. Testing fluids were Garden Banks condensate and natural gas. Hydrodynamic experiments were performed prior to the wax deposition experiments to verify the flow patterns and examine the flow behavior. The hydrodynamic and heat-transfer variables were estimated using the unified mechanistic model (Zhang et al., 2003). Wax deposition experiments were conducted under single-phase and slug flows with parametric variation of oil and gas superficial velocities and testing durations of 4, 12, and 24 hrs. The bulk fluid and initial pipe wall temperatures at a removable spool piece were kept below WAT and relatively constant to control the initial concentration driving force. In this study, unprecedented detailed measurement and analysis of the circumferential local samples were conducted. A new pigging tool was designed and constructed for selective wax sampling at top, side, and bottom quarters of the circumference of the removable spool piece. Local deposit thicknesses were calculated from the direct measurements of deposit mass and density. Wax samples were analyzed by using DSC and HTGC for wax content and composition. The impact of the physics of the slug flow on wax deposition was investigated. The results indicated that deposit thickness, wax content, and wax mass were affected by the change in superficial velocities or the flow rates of the phases. It was found that the deposit thickness increased with time. The thickness decreased with increasing superficial liquid velocity; whereas, it increased with increasing superficial gas velocity. The trends of the thickness were found to crossover each other at a certain time for different superficial gas velocities. In slug flow, circumferential variations of the deposit characteristics were analyzed. The deposit at the top wall was thicker, softer, and contained more oil than the bottom. Wax fraction increased with time reflecting the aging. Moreover, wax fraction increased with higher superficial liquid and gas velocities at a given time. The crossover of the wax fraction trends with different superficial gas velocities occurred. In slug flow, deposit at the top wall always had lower wax fraction than the bottom. Average carbon number and WAT of the deposit relatively increased with wax fraction. Wax deposits under slug flow had longer chain paraffins compared to the single-phase flow with the same wax fraction. Wax mass at the top wall was higher than the bottom. This new set of experimental data can be used as a verification tool for further development of wax deposition mechanistic model under such conditions.