Abstract: In order to optimize the solution mining of salt-cavern gas storages, a mathematical model was established for optimizing the parameters of multi-well solution mining technology on the basis of non-linear programming theories and solution mining practices after comprehensive analysis was made on factors related with solution mining. Based on the calculation, two multi-well solution mining parameter optimization programs were prepared by considering the maximum cavern volume program and the minimum flowing back rate to the tank program. It is shown that, for purpose of energy conservation, the volume of brine flowing back to the tank shall be reduced, but also the concentration in the brine tank shall be kept at low level. When fresh water injection rate is given, the more the wells for reverse circulation are adopted, the larger the caverns are. And the volume of the caverns can be increased remarkably if the injection rate is high, but the energy consumption is higher because the concentration in the brine buffer tank is kept at high level. Generally, the program of normal circulation with low injection rate is used. Based on this program, the backflow is reduced on condition that fresh water consumption and brine delivery are satisfied. In this way, the concentration in the brine buffer tank is kept at lower level, so energy consumption drops. With this model, a practically optimal multi-well mining parameter arrangement program is developed by deploying the number of wells for solution mining rationally and arranging the circulation mode, injection working medium and injection rate for each well scientifically. The program plays an instructive role for field operation.