Novel Sensors for Real-time Detection of Membrane Fouling in RO Desalination Systems

Abstract:

The dominant technology for desalination is reverse osmosis (RO). However, like all membrane-based separation processes, RO is subject to membrane fouling, which results in degraded performance, increased energy consumption and higher operating costs. Inorganic fouling (scaling) is a common form of fouling observed in membrane-based RO desalination due to the variety of salts present in typical brackish and seawater feeds. Real-time detection of early-stage membrane fouling could provide important benefits for commercial desalination by making it possible to address fouling when remediation methods are most effective. In contrast to current detection methods, the novel methodology described here provides accurate and sensitive real-time chemical identification of the foulants.

Our methodology utilizes Raman spectroscopy, an optical sensing technique based on Raman scattering that can provide a precise chemical fingerprint of fouling species and the underlying membrane with sub-micron spatial resolution. This methodology is independent of desalination operating conditions, and also shows promise for real-time monitoring of fouling removal (cleaning) with high sensitivity.

To demonstrate the capabilities of this technique, we designed a custom bench-scale flat-sheet RO system. The flow cell of the RO system is outfitted with optical access to the membrane and interfaces with a Raman microscope. The Raman-based sensor acquires real-time Raman spectra as the RO system runs under realistic operating conditions. Using CaSO₄ and CaCO₃ as model scalants, we have demonstrated local scaling detection with higher sensitivity than global scaling metrics such as permeate flux changes. Additionally, with a modified sampling strategy, we show that the methodology can distinguish between CaSO₄ and CaCO₃ scaling. We also describe the ability of this Raman-based technique to monitor cleaning of CaSO₄-scaled membranes, and show that permeate flux recovery temporally lags Raman-based metrics. Initial results obtained using this Raman-based sensor methodology show promise for performance characterization in RO desalination plants and smart water management for water security.