Evolution and optimization of the dual mixed refrigerant process of natural gas liquefaction

•Evolution and optimization of DMR process are presented.•Recent trends in DMR are high process safety/efficiency, utilization of cold energy.•Single objective study minimizing specific compression energy and UA is performed.•Multi-objective optimization is performed employing controlled elitist gen...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:Applied thermal engineering 2016-03, Vol.96, p.320-329
Hauptverfasser: Khan, Mohd Shariq, Karimi, I.A., Lee, Moonyong
Format: Artikel
Sprache:eng
Schlagworte:
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Beschreibung
Zusammenfassung:•Evolution and optimization of DMR process are presented.•Recent trends in DMR are high process safety/efficiency, utilization of cold energy.•Single objective study minimizing specific compression energy and UA is performed.•Multi-objective optimization is performed employing controlled elitist genetic algorithm. This study unfolds the important developments in the evolution of dual mixed refrigerant (DMR) process of natural gas (NG) liquefaction followed by its optimization. The initial designs of DMR process involve direct intermixing of non-equilibrium streams that causes thermodynamic irreversibility and reduces efficiency. Major developments that improved DMR process efficiency were the use of coil-wind type heat exchangers followed by three stage throttling in NG pre-cooling and direct utilization of cold energy available to the boil-off gas. After enumerating major developments a generic design of DMR process is selected and optimized for specific compression energy (SCE) and overall heat transfer coefficient (UA) using Box methodology and controlled elitist genetic algorithm. Single objective optimization of SCE and UA with Box methodology results in savings of 36% and 15% respectively. There exists a partial trade-off between SCE and UA, thus the savings in SCE are offset by the increase of UA. Consequently a multi-objective optimization is performed that results in a simultaneous decrease of 24% and 3% in the SCE and UA values, compared to the base case.
ISSN:1359-4311
DOI:10.1016/j.applthermaleng.2015.11.092