Optimization of modified single mixed refrigerant process of natural gas liquefaction using multivariate Coggin’s algorithm combined with process knowledge

The optimization of a mixed refrigerant liquefaction process is a challenge because of its non-linear characteristics with stringent multiple process constraints. This study proposes a novel hybrid approach for the optimization of a newly developed, modified single mixed refrigerant process of natural gas liquefaction targeted for offshore applications. This contribution focuses on interpreting the geometric pattern of a plot of the temperature difference between the hot and cold composite curves in a cryogenic heat exchanger to understand the profound effects of the flow rates of the individual refrigerant components and the operating pressure on the liquefaction efficiency. From this, an effective method to generate a proper initial approach temperature profile was developed to ensure robust convergence of the main optimization step. An enhanced coordinate descent methodology was implemented in the main optimization procedure to accelerate the optimization of the modified single mixed refrigerant liquefaction process. The proposed knowledge-inspired hybrid optimization approach showed a robust convergence on determining the optimal design condition. The total energy requirement for natural gas liquefaction cycle was reduced by 21.9% compared to the base case. The proposed methodology can be extended directly to solve optimization problems for other mixed refrigerant based natural gas liquefaction processes