Sensitivity analysis of effects of design parameters and decision variables on optimization of natural gas liquefaction process. (1st September 2020)

• Record Type: Journal Article
• Authors: Tak, Kyungjae; Choi, Jiwon; Ryu, Jun-Hyung; Moon, Il
• Is Part Of: Energy. Volume 206(2020)
• Publication Date: 2020-09-01
• Subjects: LNG -- SMR process -- Deterministic optimization -- Sensitivity analysis -- Exergy; Power resources -- Periodicals; Power (Mechanics) -- Periodicals; Energy consumption -- Periodicals; 333.7905
• DOI: 10.1016/j.energy.2020.118132
• Languages: English
• ISSNs: 0360-5442

Abstract

Abstract: Liquefied natural gas (LNG) plants require large amounts of energy to liquefy natural gas (NG). Therefore, many optimization studies have been conducted to minimize this energy consumption. Such studies have usually focused on optimization approaches to overcome the high nonlinearity of NG liquefaction process models. By contrast, decision variables and design bases have barely been investigated. In this study, an NG liquefaction process is modeled to perform sensitivity analysis of the design parameters and decision variables to determine their effects on the optimal operating conditions and process efficiency. A base case optimization is performed to investigate the convergence rate. Among 120 optimization runs, 57.5% are converged and 15.4% of the converged results show less than 0.1% difference in specific work compared to the best result. The effects of 11 decision variables and four design parameters are studied to obtain sensitivities. Among the decision variables, methane fraction and outlet temperature of a hot stream in an LNG heat exchanger strongly influence process efficiency. When changing the values of the design parameters within the ranges mentioned in the literature, specific work can vary from 724 kJ/kg LNG to 1509 kJ/kg LNG. Irreversibility in the coolers are the major reason to this variation. Highlights: A single mixed refrigerant process is modeled to identify optimal operating conditions. Convergence rate of the model is investigated when using a deterministic algorithm. Sensitivity analyses of design parameters and decision variables are performed. Exergy analysis indicates that exergy losses in coolers are highly sensitive. Specific work varied from 724 to 1509 kJ/kg LNG according to design basis.