Thermo-economic investigation and AI-driven multi-objective optimization of advanced three-stage cascade refrigeration system with flash tank, ejector, and suction-line heat exchanger

• A novel triple-cascade refrigeration configuration implementing ejector is proposed. • Thermo-economic analysis is conducted under various operating conditions. • TOPSIS analysis is used to find highest performing system configuration. • Artificial Neural Network based system behavior prediction model is developed. • Genetic Algorithm based multi-objective optimization is conducted on the system. The critical applications of ultra-low temperature (ULT) cooling systems make them a point of deep interest in research, as these systems account for significant power consumption from the power generated worldwide. This study addresses the issue by introducing Ejector-enhanced Advanced Triple Cascade Refrigeration System (EATCRS), incorporating a suction line heat exchanger (SLHX) and a flash tank (FLT) as well with 1-Butene, R-12 and Toluene used as the refrigerants in low, medium and high temperature circuits respectively. With the improvements already provided by FLT and SLHX in reducing the overall compressor load and enhanced cooling capacity, the inclusion of ejector can significantly improve the overall system performance by mitigating the expansion loss issue. The maximum thermo-economic performance potential of the system can be extracted by using FES configuration verified by detailed TOPSIS analysis. The results from detailed python modeled thermo-economic analysis reveal superiority of proposed EATCRS over conventional triple cascade refrigeration system (TCRS). While operating at an identical condition, coefficient of performance (COP) and exergy efficiency are enhanced by about 33% and 27% respectively at an insignificant annual cost increase of about 5%. Parametric analysis reveals that condenser temperature of low-temperature and high temperature circuit ( T lc and T mc ) to be the most critical parameters for system performance, whereas economic performance was heavily influenced by condenser temperature (T cond ). Furthermore, the ejector suction nozzle pressure drop ( ΔP) is proven to be a critical operational parameter for both thermodynamic and economic performance of the system. In order to identify the best tradeoff between cost and performance, a machine learning enhanced multi objective optimization is conducted on the proposed model. Overall, the proposed EATCRS demonstrates superior thermo-economic performance offering a potentially high-performance and sustainable alternative to conventional ULT systems.