Large gravity separation tanks play an essential role in crude oil production in many fields worldwide. These tanks are used to separate water from an oil-rich stream before safely returning it to the environment. The oil/water dispersion enters the tanks through a feed spreader consisting of an array of pipes with small effluent nozzles. A major challenge is being able to predict oil/water dispersion distribution along the spreader as well as, the maximum water droplet size exiting through the effluent nozzles, under a given set of conditions. The capacity of the studied tank is 80,000 barrels (12,719 m~3). Current feed stream is about 60,000 bpd (9,540 m~3/day) of wet crude containing about 20% water by volume. A significant increase in flow rates and water volume fraction is anticipated [7], as more wells are added and existing ones mature. This work is aimed at investigating the separation performance of these tanks under current and future flow conditions; focusing primarily on the flow phenomena and droplet size distribution inside the spreader. The main objective is then to identify the impact of the spreader's geometry and piping configuration on flow behavior and tank's separation efficiency. The final product provides key information needed for mechanistic modeling the tank separation performance and optimizing tank components' design.The feed spreader is simulated using Computational Fluid Dynamics (CFD) to assess oil/water flow distribution inside the network. Droplet size distribution along branch-pipes effluent nozzles in, including droplet breakup and coalescence has beenstudied using the Gomez mechanistic model [2] with input from CFD results. An experimental investigation of the spreader using a scaled prototype was also conducted to better understand flow phenomena and verify the CFD models.Results confirm the occurrence of significant maldistribution of the water and oil phases along the spreader that could impair separation efficiency.
展开▼
