Thermal extraction of water ice from the lunar surface II - vapor yields for an improved regolith model

This work focuses on thermal water extraction on the lunar surface. We previously developed a three-dimensional finite element model (FEM) implementing heat and gas diffusion in the porous granular medium that is icy lunar regolith. Here, we present an improved version of this work in which we implemented a more realistic regolith model. In particular, we addressed previous model simplifications on regolith emissivity and porosity, water sublimation rate, as well as regolith and water ice thermal conductivity and permeability. Incorporating recent modeling and experimental work from the literature, we investigated the effect of these soil properties on the outcome of our simulations, with a particular interest in the yield of the thermal extraction process. Aiming at understanding what thermal water extraction would produce if heating the lunar surface directly, we also studied the effect of open borders on extraction yields.

We find that the crude icy regolith approximation we implemented in Paper I provided a lower estimation of water vapor yields upon heating. Overall and using the same heating methods (surface heating as well as inserted drills), our more accurate regolith model implementation extracted more water from the simulation volume. With this new model, we observed that extraction yields depended mostly on the ice content of the regolith, and to a lesser extent on the heating configuration (number of drills) and power. In two specific configurations, 16 and 25 drills at $10^4$ W in 1%vol icy regolith, heating allowed the extraction of nearby ice, efficiently desiccating the entire simulation volume. Apart from these two cases, the highest extraction yields were obtained for $10^4$ W surface heating of a volume with closed borders with values over 80%. In open border volumes, highest yields were around 70% achieved for the highest number of drills (16 and 25), at the highest power ($10^4$ W) in the regolith with the largest icy fraction. Extraction masses started being noticeable around a few minutes, but reaching most of the maximum possible yields took up to several days in some cases.

Defining an extraction efficiency by combining the yield and extraction times, we found that the best compromise between hardware complexity, time, and yield would be working in open border environments, using dense drill configurations in ice-rich regolith, and loose drill configurations in ice-poor regolith. In both cases, extraction efficiencies were similar at $10^2$ W and $10^3$ W per drill, indicating that low power solutions would yield similar results than higher power ones. Overall, our results support the viability of thermal water extraction in future ISRU architectures.