Abstract

The transition from conventional evaporation ponds to direct lithium extraction (DLE) processes in lithium production faces challenges in brine management and freshwater consumption. While the evaporative method loses approximately 125 m3 of water per ton of lithium carbonate equivalent, DLE requires high-quality fresh water with variable consumption rates. Additionally, several DLE processes require a pre-concentration stage to remove impurities and increase lithium concentration. This work explores simulating and optimizing a membrane distillation and crystallization (MDCr) process design as a solution for pre-concentrating multicomponent lithium brines and recovering freshwater. An optimization methodology based on a phenomenological model is proposed, incorporating NSGA-II and multi-criteria decision-making to design large-scale air gap membrane distillation modules. The performance is evaluated in a continuous plant with energy recovery and heat pump assistance, considering salt crystallization. Results showed that a module with a packing factor of 0.31, 3,595 porous fibers, 7,480 dense fibers, a length of 2.90 m, and an initial velocity of 0.85 m/s, combined with an average logarithmic temperature difference of 19.3 °C, achieved an optimal balance between water production and economic costs. This design resulted in a Capex of 78 US$/(m3/y) and an Opex of 4.3 US$/m3, with a specific electrical energy consumption of 71 kWh/m3, doubling lithium concentration and recovering 50 % of the water. These findings highlight the viability of MDCr technology as a competitive and sustainable option for brine pre-treatment, proving how mathematical modeling and optimization tools, grounded in validated phenomenology, can support the scaling-up of MDCr to large industrial applications.