Effects of uncertainty in the design parameters during the design of horizontal gravity oil, gas and water separators

作     者：Chilala, Abubakar Juma Stanko, Milan Mkinga, Oras J. 

作者机构：Univ Dar Es Salaam Dept Petr Sci & Engn Dar Es Salaam Tanzania Norwegian Univ Sci & Technol NTNU Dept Geosci & Petr Trondheim Norway 

出 版 物：《JOURNAL OF PETROLEUM EXPLORATION AND PRODUCTION TECHNOLOGY》 (J. Pet. Explor. Prod. Technol.)

年 卷 期：2025年第15卷第2期

页      面：1-16页

核心收录：

学科分类：0820[工学-石油与天然气工程] 08[工学] 0708[理学-地球物理学] 

基　　金：Norwegian Program for Capacity Development in Higher Education and Research for Development in Energy and Petroleum (EnPe) at the University of Dar es Salaam 

主　　题：Optimization Bulk separators Stochastic programming Decision support method Uncertainties 

摘      要：The design of bulk separators for oil, gas, and water vessel-type gravity separators significantly affects the cost of topside and subsea oil- and gas-processing installations. There are often many uncertainties during the design process, such as the separator performance, fluid properties, and variation and evolution of the inlet conditions. The industry employs mechanistic models, empirical correlations, numerical studies using Computational Fluid Dynamics, and extensive and expensive experimental campaigns with separator prototypes. However, owing to time, personnel, and information integration constraints in the design phase, design evaluation, and validation are often limited to performing comprehensive sensitivity studies. This often leads to suboptimal designs and early bottlenecks. This study proposes using stochastic mathematical programming and optimization to include the effects of uncertainties in the design parameters during the separator design process. The proposed solution involves the development of a three-phase horizontal separator design model. The model was implemented using Microsoft Excel Visual Basic for Applications. The study then considered the effects of the uncertainties in the oil, gas, and water flow rates and slug volume on the optimal design using stochastic optimization. Conservative, optimistic, and ultimate design parameters (P10, P50, and P90) were obtained for each scenario. The results of this study provide a method and tool to aid topside design teams while expanding on previous research in the literature regarding separator design.