Objective The aim is to optimize and adjust the key process parameters of an L-BOG deep-cooling helium extraction plant.

Method Aspen HYSYS was used to establish a process model to analyze the effects of distillation column temperature, reflux ratio and other parameters on energy consumption and recovery effectiveness, and then the Box-Behnken response surface analysis was used to design and establish a combination of interaction conditions and simulate the optimization and adjustment of the combination of parameters.

Result The process simulation reflected that the distillation column pressure had minimal effect on the process, lowering temperature, increasing reflux ratio and at higher tower plate feeding all improved helium recovery, and the response surface analysis determined the significance order of the effect of each variable on helium volume fraction, condenser power and reboiler power. After optimizing model, the distillation tower feed temperature is set to −160 ℃, the tower top reflux ratio is 0.655, and the feed position is 5^# tower plate, the total energy consumption of the plant is reduced by 20.31%, and the helium recovery rate is increased by 0.12 percentage points.

Conclusion A multifactor predictive regression model based on HYSYS simulation coupled with Box-Behnken response surface methodology can provide a reasonable combination of parameter operating values for the L-BOG deep-cooling helium extraction plant, and the method can better realize the reduction of energy consumption and the improvement of helium recovery rate.