Computational chemistry methods for unveiling high recovery reverse osmosis scaling

INTRODUCTION
Water treatment with conventional reverse osmosis (RO) remains ahead of high recovery RO (HRRO). With over 600 installations of HRRO by 2021 [1], the adoption of HRRO is limited due to many reasons. One of the reasons is that there are few players in the market with tracked verifiable records. Another reason is that promised high recoveries of at least 90% are not practically achieved, and many times the high recoveries are limited to around 85%. Thus, the competitive advantage of HRRO can be deselected by users finding conventional RO multistage alternatives. The best-in-class multistage RO may guarantee recoveries of around 85%, which may also be the case for HRRO at similar water feed conditions.
HRRO needs better scaling predictions to achieve simulated recoveries. By determining appropriate scaling limits based on saturation index (SI) and calcium carbonate precipitation potential (CCPP) the systems can be better designed. In this study 4 types of water were selected aiming to reach recoveries of 90%. Moreover, the study investigated the automation options of computational chemistry methods into real-time water treatment systems.
METHODOLOGY
Water samples were collected from 4 different sources (river, seawater, wastewater effluent and tap water) and analysed for chemical parameters by a certified laboratory (Eurofins) and in the laboratory of Via University College in Horsens.
The model for estimation of scaling limits (SI and CCPP) was adapted from proven PHREEQC simulations applied to field RO data and validated by sensor measurements [2]. For the automation part, one alternative was evaluated, by using a Python version adaptation of PHREEQC [3] developed by Vitens engineers, VIPHREEQC [4].
RESULTS
A sensitivity analyses of SI and CCPP revealed that pH and temperature have a significant effect on both scaling indexes, with pH having a larger effect than temperature. At 90% recovery, the river water presented the highest calcite SI, followed by wastewater effluent, groundwater and seawater.
In a semi-batch operation of a HRRO pilot system, calculations of pH and SI in the retentate offered an early detection possibility for determining maximum recoveries before the onset of scaling.
CONCLUSIONS
By utilizing computation methods for chemistry analysis, it could be demonstrated that scaling indexes can be used to select water types that may not limit the application of HRRO for water desalination and water reuse.
REFERENCES
[1] Bluetech Horizon Scan. High Recovery Reverse Osmosis, 2021
[2] S. Hager, M. Meinardus, T. Hoffmann and K. Glas. CaCO3 deposits in reverse osmosis: Part II – Simulation model for hydrochemical predictions of reverse osmosis retentates and scaling propensity. Brewing Science 75 (2022): 54 – 68
[3] United States Geological Survey (USGS) Science for a changing World. PHEREEQC Version 3. https://www.usgs.gov/software/phreeqc-version-3/
[4] Vitens Copyright License, Licensed under the Apache License, Version 2.0