Archive | 2019

The study of down-hole hydro-cyclone efficiency in oil wells using computational fluid dynamics

 

Abstract


Water is by far the largest waste stream associated with oil and gas production and it is unimaginable to find an oil reservoir absolutely free from connate water. Water management has become an important issue of hydrocarbon production, since the produced water increases as the field grows older and the cost of water handling, such as separation, treatment and repair is dramatically increasing. Down-Hole Separation (DOWS) and re-injection results in the production of oil to the surface, while portions of water are injected to the underground formation without ever being lifted to the surface DOWS depends on geological characteristics of the producing and injection formation, down-hole conditions, well configuration and equipment, installation procedures, hydrocyclone arrangement and water disposal. With their compact size, Liquid-Liquid Hydro-Cyclones (LLHC) can perform as an integral part of a down-hole oil water separation system. Despite performance and functionality of cyclones not being fully understood, they have created new possibilities for separating fluid down-hole for the producing formation and injecting separated water far away from the production interval. Before placing cyclones into practice, a study of fluid properties, well geometry, and characteristics of the formation can yield valuable information about the applicability and efficiency of DOWS. In this study, Computational Fluid Dynamic (CFD) was used to determine the effect of changes in API oil gravity, flow rates and cyclone geometry in order to understand the behaviour of LLHC in down-hole conditions. CFD is a tool that can predict the quantity of oil flow into surface, the quality of injected water into formation, split ratio, separation efficiency, mass transfer and related phenomena by solving numerically set of governing equations defining the fluid behavior in a DOWS.

Volume None
Pages None
DOI 10.33915/etd.3272
Language English
Journal None

Full Text