Inland disposal of reject brine, be it produced water from gas wells, reject brine from a reverse osmosis (RO) desalination plant or a process plant blowdown, poses insurmountable difficulty as deep well injections are being gradually phased out. Brine concentration followed by crystallization is a viable option but the process of concentrating the brine to around 300,000 mg/L total dissolved solids (TDS) prior to crystallization is a very energy intensive prerequisite which, in turn, is responsible for atmospheric CO₂ emission. Membrane distillation and multi-stage steam distillation are two such concentration processes and both are essentially vapor pressure driven processes i.e., they operate in close proximity to the boiling temperature of the brine. Hence, both are associated with operational problems of scaling and fouling.

Here we present for the first time a brine concentration process that is driven solely by the relative humidity of the air and can operate at ambient temperature. The fundamental scientific tenet of the Evaporative Ion Exchange (EIX) process rests on the premise that water uptake onto or water release from an ion exchange resin is greatly influenced by the relative humidity of air it is in contact with. In a two-step process at 35⁰C, a waste brine of 50,000 mg/L TDS from a gas well in California was concentrated by EIX to over 300,000 mg/L without fouling. Although thermodynamically consistent, the need for waste heat or thermal energy was avoided altogether in the EIX process. More experimental validations will also be presented.

Lead (Pb) is one of the most toxic and hazardous metals widely employed in various industrial processes and water pipelines. To protect human health, it is essential to thoroughly remove minute amounts of Pb²⁺ from drinking water. On the other hand, healthy cations in drinking water such as calcium (Ca) and magnesium (Mg) do not need to be fully eliminated as they are beneficial in reasonable amounts. Capacitive deionization (CDI) is a low-energy and low-cost water purification and treatment option. The electrodes used in the CDI cell play a crucial role in removing metal ions and must possess appropriate adsorption/desorption properties. Activated carbon and graphene, known for their high mechanical durability, conductivity, and specific surface area, are commonly used as CDI electrodes. In this study, we aimed to improve the selectivity and removal efficiency of Pb²⁺ via surface modification of the activated carbon and graphene oxide electrodes. Using both computational modeling and experimental validation, we examined the effect of different molecular groups (-SH, -COOH, and -NH₂) functionalized on the electrode surface on the selectivity and removal efficiency of Pb²⁺. Results showed that -COOH and -SH groups displayed strong affinity to Pb²⁺, resulting in high selectivity for Pb removal compared to Ca²⁺ and Mg²⁺. This research provides a better understanding of the interactions between functional groups and Pb²⁺, enabling us to design a sustainable, adaptable, and reliable CDI prototype for selective and fast separation of target cations and simultaneous output of clean water.

Inland desalination with reverse osmosis (RO) offers a promising alternative to increase potable water resources. However, large volumes of concentrated brine are a byproduct of the RO process. Inland concentrate disposal displaces valuable water resources and can make up to 33% of the total cost of desalination. Conventional disposal systems are constrained by location hydrogeology, climate, and policies, thus limiting the implementation of desalination facilities in water-stressed regions. Additionally, few conventional concentrate management approaches place value on resource recovery. Membrane distillation (MD) is an alternative concentrate management system that can simultaneously minimize disposal volume and maximize water recovery. MD uses thermal energy gradients to desalinate concentrate streams, achieving near zero-liquid discharge. Furthermore, MD can utilize low-grade heat and includes internal heat recovery, making MD an energy efficient alternative. A techno-economic and life-cycle assessment (TEA-LCA) of air-gap MD (AGMD) for RO concentrate management was performed and compared to conventional concentrate management technologies, including evaporation ponds, deep-well injection (DWI), and concentration-crystallization. TEA results indicate that compared to DWI at $1.09/m3 and concentrator-crystallizers at $4.65/m3, AGMD is a competitive concentrate management technology producing water at $1.05/m3 when low grade heat is available and at $1.74/m3 when external thermal energy supply is needed. However, MD has higher environmental impacts than evaporation ponds and DWI when low grade heat is not available. Results of this study highlight the application of MD for brine management, contributing towards a circular economy that prolongs the useful life of potable water resources while avoiding excessive use of energy resources.

Global water stress has prompted interest in desalination systems for treatment of saline and other impaired water sources for potable use. Seawater reverse osmosis (SWRO) facilities are often co-located with wastewater treatment facilities that also discharge a stream to the ocean. Recently, the prospect of pre-RO blending of seawater with wastewater effluent, to dilute the salinity of seawater and contaminant load of wastewater, has been proposed. However, the chemistry of disinfecting a blended seawater-wastewater stream have not been explored. Toxic byproducts, including N-nitrosodimethylamine (NDMA), are expected to form during chlorine disinfection of a blended stream, and the extent of byproduct formation will likely be a function of which stream is chlorinated first and whether disinfection happens before or after blending. In this work, three scenarios were modeled and experimentally evaluated using synthetic and authentic waters: (i) disinfecting wastewater before blending with seawater, (ii) disinfecting after blending, and (iii) disinfecting seawater before blending with wastewater. In both synthetic and authentic water, chlorinating wastewater prior to blending, which modeling results indicated would form the most dichloramine, produced the most NDMA. Bromochloramine, previously implicated in NDMA formation during chlorination of bromide-rich waters, was predominantly formed when the stream was disinfected after blending, however, this scenario produced NDMA concentrations that were an order of magnitude lower than wastewater disinfection in synthetic waters and 2.5 times less in authentic waters. This suggests that efforts to minimize NDMA formation should focus on pretreatment that minimizes dichloramine formation, likely by disinfecting seawater first, or after blending.

Optimization and design of full-scale membrane distillation (MD) systems usually require Sherwood and Nusselt correlations that are developed from lab-scale systems. However, entrance effects in lab-scale systems can significantly impact heat, mass and momentum transfer in the reactor, therefore affect the accuracy of the developed experimental Sherwood and Nusselt} correlations. Here, Computational Fluid Dynamics (CFD) simulations using OpenFOAM are performed to understand the effects of right-angled bends and inlet design on flow dynamics, temperature and concentration polarization in MD systems. Simulation results show that the presence of right-angled bends and inlets with sudden expansions leads to the formation of Dean vortices. Dean vortices enhance perpendicular mixing in MD systems and reduce both temperature and concentration polarization. Temperature and concentration polarization coefficients in MD systems with right-angled bends and inlets with sudden expansions vary significantly for the same volumetric flow rate. Our studies show that lab-scale systems with the same volumetric flow rate but different designs lead to significantly different Nusselt and Sherwood correlations. This study demonstrates the importance of CFD-informed design of lab-scale systems to minimize entrance effects and suppress Dean vortices for consistent model development and calibration across multiple scales.