Thermodynamic comparison of three small-scale gas liquefaction systems

Abstract

Natural gas liquefaction systems are based on refrigeration cycles, which can be subdivided into the cascade, mixed refrigerant and expander-based processes. They differ in their configurations, components and working fluids, and have therefore various operating conditions and equipment inventory. The present work investigates three configurations suitable for small-scale applications because of their simplicity and compactness: the single-mixed refrigerant, single and dual reverse Brayton cycles. The impact of different feed compositions and refrigerant properties is analysed. A detailed assessment of the energy and exergy flows is conducted, and the most promising cycle layouts are identified by performing multi-objective optimisation procedures. The findings illustrate the resulting trade-offs between the system performance and size in different operating conditions. Mixed-refrigerant processes prove to be more efficient (1000–2000kJ/kgLNG) than expander-based ones (2500–5000kJ/kgLNG) over larger ranges of operating conditions, at the expense of a greater system complexity and higher thermal conductance (250–500kW/K against 80–160kW/K). The results show that the use of different thermodynamic models leads to relative deviations of up to 1% for the power consumption and 20% for the network conductance. Particular caution should thus be exercised when extrapolating the results of process models to the design of actual gas liquefaction systems.