(479f) Polymeric Membrane-Based Water Harvesting Device for Semi-Arid Regions: Technical Design and Economic Assessment

Water scarcity constitutes a major issue for the human society. A number of strategies have been proposed to recover water from a variety of sources and to make it suitable to human needs. Atmospheric water harvesting is an attractive strategy to supply drinking water in water-scarce regions, as it allows the production of water in a decentralised fashion. Thanks to the natural hydrological cycle, water is always available and extractable from atmosphere, and the produced water is clean, similarly to natural precipitation. However, only a few atmospheric water recovery systems are commercially available, because either they need high humidity levels to work or they are highly energy intensive processes.

We propose to use a water harvesting device based on high-performance membranes, which presents a single membrane stage to increase the purity of water vapour in air and a condenser to produce liquid water. The presence of the membrane stage allows to reduce significantly energy consumption and costs and makes the system suitable even to semi-arid regions with low humidity levels. For the first time, we performed an optimisation of the operating conditions of the process (i.e., water recovery, temperature and pressure in the condenser) to minimise energy demand and costs. For this scope, we referred to high-performance polymeric membranes for atmospheric water harvesting with performance parameters in line with those reported in the literature (H₂O permeance of 10⁴ GPU and H₂O/air selectivity of 10⁴).

We found attractive minimum water costs of 3.2 cent $/L when water fraction in air is 3% and 5.0 cent $/L when water fraction in air is 2%. These values were found at optimal recovery values in the membrane stage between 30 and 35%, at a temperature in the condenser of 20 ºC (selected to reduce the cooling energy consumption) and at a pressure in the condenser of 0.03 bar (optimal as it can reduce energy consumption for compression after the membrane stage).

We analysed the impact of permeance and selectivity of the membranes, by including also the impact of non-ideal effects, i.e., pressure drops and concentration polarisation. We found that an increase of permeance up to 5×10⁴ GPU is highly beneficial to reduce costs, even if the effect of concentration polarisation grows. In fact, the increase in water costs due to the larger membrane area required can be limited via a proper design of the membrane module and results below 10%. On the contrary, increasing the selectivity beyond 5×10³ is not advantageous because the water cost plateaus.

Finally, we simulated the yearly operation of a water harvesting device in a hot semi-arid region in Chad and we found that the system can stand variations in the inlet water concentration while producing significant volumes of water throughout the year with promising energy consumption and costs (lower than 300 kWh/m³ and 5 cent $/L, respectively, for most of the year). Furthermore, prospective costs of electricity by photovoltaic modules and of membranes make this system particularly appealing as the water cost goes down to a minimum of 2.6 cent $/L.

Overall, our results showed how membrane-based water harvesting device is an attractive option for water production on a decentralised basis especially for arid and semi-arid regions where provision of water is a real challenge for the population.