(524b) Measuring Crystallization Kinetics of Sparingly Soluble Salts from Supersaturated Waters Using CSTRs-in-Series: Studies with CaCO3

Measuring crystallization kinetics of sparingly soluble salts
from supersaturated waters using CSTRs-in-series: studies with CaCO₃

Sankaranarayanan A. Ravichandran^a, Jordan Krist^b,
Dakota Edwards^a, Saied Delagah^c, John Pellegrino^a,*

¹Department of Mechanical Engineering, University
of Colorado, Boulder, CO USA

²Department of Chemical and Biological
Engineering, University of Colorado, Boulder, CO USA

³Bureau of Reclamation, Denver, CO USA

   
^*Corresponding author: john.pellegrino@colorado.edu

   Abstract

We present an approach for studying crystallization kinetics
of sparingly soluble salts, particularly supersaturated reject streams from
reverse osmosis (RO) membrane processes. Our focus was to study crystallization
kinetics to develop heuristics to improve the design of crystallizers used in
water treatment applications. We used a system of six continuously stirred tank
reactors in series (CSTRs-in-series), the fluid between CSTRs is transported
using peristaltic pumps. We used turbidity and pH changes to track
crystallization. Our steady-state approach provides a better signal to noise
ratio than extant batch mode methods to study the kinetics of crystallization processes.
Moreover, steady-state bench scale approaches are more realistic in describing
large scale continuous processes. We used a realistic calcium carbonate
supersaturated stream modeled on an RO reject stream found at a brackish water treatment
facility at Brighton, Colorado.  The residence times in each CSTR ranged
between ~3 to 11 min. The maximum hard water inlet feed was ~0.6 L/min for the
shortest residence time. For six CSTRs-in-series the total system residence
time of ~18 to ~ 66 min and showed supersaturation relief in the range of ~ 25
to over 50 percent respectively without added chemicals. We used a metric of 5
NTU turbidity to describe a point of discernible crystal crystallization where
crystal nucleation is significant. We studied the effects of impeller mixing
and the mixing in the peristaltic tubing using the Kolmogorov mixing theory.
Interestingly, for all residence times studied, a peristaltic recirculation
loop in the 1st CSTR where mixing takes place in a smaller surface to volume
ratio was required to induce discernible crystallization with turbidity values
>> 5 NTU. Moreover, increased impeller mixing corresponding to mixing
energy inputs higher than that of peristaltic recirculation loops could not
induce discernible crystallization in the 1st CSTR. We developed a parametric
semi-empirical model to co-relate mixing and, turbidity in the 1st CSTR using
Kolmogorov mixing theory and, laws of mass action. Our studies show that
optimizing the surface to volume ratio along with mixing energy input is key to
controlling crystallization kinetics. We were able to extend the scope of this
study for a more complex supersaturated system involving co-precipitation of
calcium carbonate and calcium sulfate.  

Figure 1. CSTR-in-series setup, observe the evolution of
turbidity in successive CSTRs