A novel three-stage direct expansion cycle with optimal internal heat recovery and splitting-mixing processes for utilising LNG’s cold energy

Published in Energy Conversion and Management: X , 2026

The growing global demand for liquefied natural gas (LNG) emphasises the need to effectively utilise its cold energy during regasification. This study proposes a novel three-stage direct expansion cycle (DEC) to maximise LNG’s cold energy recovery by addressing inherent exergy losses of conventional systems. The proposed cycle integrates multi-stage expansion, recuperative heat recovery, and stream splitting–mixing strategies, ensuring optimal thermodynamic interaction among system components. A differential evolution algorithm maximises the net work output while adhering to regasified LNG (RLNG) delivery requirements. Under optimal conditions, the system achieves a net work output of 193.72 kJ/kg-LNG and a second-law efficiency of 37.40% at a distribution pressure of 3.50 MPa. Compared to the state-of-the-art three-stage DEC system, the proposed configuration delivers 27.28% higher net work output and lower levelised cost of energy (LCOE) under identical conditions. Thermodynamic analyses for optimal conditions show that the recuperation strategies considerably reduce irreversibility, while sensitivity analyses emphasise the crucial role of the optimal intermediate pressure levels. This cycle provides consistent performance across a range of RLNG distribution pressures (2.50 to 4.50 MPa), with a work recovery of 214.97 kJ/kg-LNG at a delivery pressure of 2.50 MPa. The results indicate that the proposed cycle enables substantially higher work recovery, offering a technologically superior and scalable pathway for sustainable LNG cold energy utilisation.

Recommended citation: M. S. R. Ayon, Niloy Deb, M. M. Rahman, M. Z. Haq. A novel three-stage direct expansion cycle with optimal internal heat recovery and splitting-mixing processes for utilizing LNGs cold energy. Energy Conversion and Management: X, Article 101738, 2026.
Download Paper