SELECTIVE CRYSTALLISATION OF MAGNESIUM AND CALCIUM HYDROXIDES FROM INDUSTRIAL WASTE BRINES: A PILOT STUDY
- Authors: F. Vassallo, D. La Corte, N. Cancilla , M. Bevacqua, A. Tamburini, A. Cipollina, G. Micale
- Publication year: 2021
- Type: Abstract in atti di convegno pubblicato in volume
- OA Link: http://hdl.handle.net/10447/528622
Abstract
In the last years, a rising interest has been focused on the valorization of waste brines from industrial processes, such as desalination plants. In most current scenarios, waste brines are disposed directly, or after being treated, into a receiving water bodies, often causing environmental concerns and, more importantly, renouncing to the possibility of using such a waste as source of valuable raw materials. In fact, these brines are typically rich in useful salts and minerals, whose recovery can increase the environmental and economical sustainability of the industrial process [1-2]. The ZERO BRINE EU-H2020 project aims at the development of technologies for the recovery of salts, minerals, and clean water from industrial waste brines. This is framed within the concept of circular economy by integrating several different technologies for the circular reuse and recovery of raw materials from the spent brine. This work presents the development, construction and testing of a pilot system for the selective reactive crystallization of Mg and Ca hydroxides within a demo plant aiming at the regeneration of waste brines produced during the regeneration phase of ions exchange resins for water softening. The integrated demo plant, whose process layout is depicted in Figure 1, consists of the following units: (i) Nano-Filtration (NF), (ii) crystallization, and (iii) multi-effect distillation (MED). Figure 1. Schematics of the treatment chain proposed within the ZERO BRINE EU-H2020 project for the valorization of brines from Ion Exchange resins water softening plants. According to the process scheme, the waste brine reaches the NF unit, which produces a concentrate enriched in bivalent ions, i.e. magnesium and calcium ions, while the permeate passing through the membrane contains a high concentration of sodium chloride. Then, the concentrate is processes within the selective crystallization unit to recover magnesium hydroxide, in a first step, and calcium hydroxide in a second one. Both precipitations occurred adding an alkaline solution (NaOH solution, 0.5M and 2M for the 1st and 2nd precipitation step, respectively). The produced slurries from the crystallizer are then filtered in a vacuum rotating drum filter, thus generating a solids cake and a filtered solution, containing NaCl and a pH between more than 13, is eventually neutralized by addition of HCl. Finally, the permeate from NF and the neutralized brine are mixed and sent to the evaporation unit to increase the sodium chloride concentration until reaching the target value suitable for ion exchange resins regeneration. The crystallizer is a tubular reactor, made up of two coaxial plexiglass tubes. The alkaline solution flowed inside the annular section, while the brine in the internal one. The brine is injected in the annular section via sixteen holes equally distributed along the tube length, that allow for a controlled and distributed mixing of brine in the alkaline reactant, thus being named as Multiple Feed Plug Flow Reactor (MF-PFR). The process flow diagram (PFD) of the magnesium and calcium recovery unit is reported in Figure 2, while a picture of the MF-PFR reactor is shown in Figure 3. Figure 2, Process Flow Diagram (PFD) of magnesium and calcium recovery unit. Figure 3, Picture of the MF-PFR reactor. An wide experimental campaign has been carried out through long-run tests aiming at investigating the stability of the MF-PFR unit in terms of operating conditions and reaction pH, which strongly affect purity and recovery of both precipitation steps. The experimental campaign has been carried out analysing three different inlet brine composition. In table 1, average compositions of tested brines are reported. Figure 4 shows brine and alkaline inlet flow rates, and reaction pH trends for the case B. A very stable behaviour of the system was encountered for the pH and both flow rates over the whole experiment run (8 hours) for the two precipita