On the cost of electrodialysis for the desalination of high salinity feeds

Abstract

We propose the use of electrodialysis to desalinate produced waters from shale formations in order to facilitate water reuse in subsequent hydraulic fracturing processes. We focus on establishing the energy and equipment size required for the desalination of feed waters containing total dissolved solids of up to 192,000 ppm, and we do this by experimentally replicating the performance of a 10-stage electrodialysis system. We find that energy requirements are similar to current vapour compression desalination processes for feedwaters ranging between roughly 40,000-90,000 TDS, but we project water costs to potentially be lower. We also find that the cost per unit salt removed is significantly lower when removed from a high salinity stream as opposed to a low salinity stream, pointing towards the potential of ED to operate as a partial desalination process for high salinity waters. We then develop a numerical model for the system, validate it against experimental results and use this model to minimise salt removal costs by optimising the stack voltage. We find that the higher the salinity of the water from which salt is removed the smaller should be the ratio of the electrical current to its limiting value. We conclude, on the basis of energy and equipment costs, that electrodialysis processes are potentially feasible for the desalination of high salinity waters but require further investigation of robustness to fouling under field conditions.

Optimal design and operation of electrodialysis for brackish-water desalination and for high-salinity brine concentration

“…Figure 9 shows that the decreasing area and the increasing velocity result in an increase in the optimal length with the cost ratio. We note that the membrane areas and the lengths reported are the effective area and length, and that real systems usually only use 65% of the total area effectively for mass transfer [26][27][28], so we expect the physical membrane area of these systems to be approximately 50% larger than the effective values reported above. We also note that the reported cost modeling is not dependent on the fraction of the area used effectively because it uses the effective cell-pair area.…”

“…This minimum is limited by practical difficulties in designing sealing frames and spacers [15]. For the following results, the channel height is set to 0.5 mm, which is a thickness already used in actual systems [26]. In addition, for each system size, the velocity is set such that the total power consumption is minimized.…”

“…where the subscript 's' denotes the salt, the subscript 'w' denotes the water, J is the molar flux, i is the current density, c s,C,m is the salt concentration at the membrane interface in the concentrate channel, π D,m is the osmotic pressure at the membrane interface in the diluate channel, T s is the salt transport number, [26]. Each stage is a batch system which takes the diluate from a given salinity to another set salinity.…”

“…In modeling high-salinity brine concentration, the model developed by Fidaleo and Moresi [25] is used with coefficients measured by McGovern et al [26] at high salinity. The major points of this model are summarized in Appendix B.2.…”

“…were measured at high salinity by McGovern et al [26]. Finally, the potential drop across the membranes is calculated as the sum of the membrane ohmic potential drops and the Donnan potential:…”