Ultrasound, gravimetric, and SEM studies of inorganic fouling in spiral-wound membrane modules

Ultrasonic time-domain reflectometry (UTDR) is combined with permeate-flux, gravimetric, and scanning electron microscopy (SEM) measurements for studying calcium-sulfate dihydrate fouling in a commercial spiralwound reverse osmosis membrane module. The UTDR amplitude response is shown to be more sensitive than permeate-flux decline for detecting the early stages of fouling because it is a local rather than integral measurement. Hence, appropriately located acoustic transducers can be used to detect the onset of fouling well before it impairs module performance so that appropriate remediation can be undertaken. The UTDR arrival-time response in principle can be used to determine the fouling layer thickness. However, this study indicates that arrival-time measurements in a spiral-wound module can be compromised by movement of the membrane envelope layers as is the case for the Koch 2521 spiral-wound module used in this study. The estimated module expansion of 210 µm inferred from the UTDR measurements exceeds the thickness of the fouling layer deposits estimated from the gravimetric measurements. The gravimetric studies reveal a 6 to 10-fold increase in the thickness of the fouling layer deposits in the feed-flow direction owing to concentration polarization. A particularly interesting result of the gravimetric studies is a 16 to 27-fold increase in the thickness of the fouling layer deposits in the permeate-flow direction presumably owing to expansion of the outer relative to the inner feed channels. The SEM studies indicate that fouling is initiated along the feed-spacer mesh presumably due to dead-flow regions behind mesh elements transverse to the flow. Overall, the present study demonstrates that when UTDR is combined with other independent measurement techniques, it can provide significant insights regarding fundamental fouling and membrane module behavior.

This research firmly establishes that UTDR can provide valuable insight concerning inorganic fouling in spiral-wound membrane modules. The use of a focused UTDR transducer permitted studying the change in arrival time and amplitude of the ultrasonic waveform reflections from specific membrane envelope interfaces such as the 2nd and 3rd of the 12 membrane envelope layers in a commercial spiral-wound module. The results of this study complement those of Zhang et al. [13] who used the “acoustic signature” to detect membrane fouling from several membrane envelope layers at a particular axial position along the module. The acoustic signature was a composite of the UTDR response owing to multiple reflections from several membrane layers as well as the feed-spacer and permeate channels. However, the present study provides information on the state of fouling on specific interfaces. This study confirms that UTDR can indeed be adapted to the more complex spiral-wound geometry and its multiple interfaces. The methodology also can provide a more sensitive indicator of membrane fouling since UTDR provides a local measurement in contrast to permeate flux decline, which provides an integral measure of fouling throughout the module. As such, the UTDR technique can be used to assess the initiation of fouling by optimally positioning the UTDR transducer so that remediation measures can be taken before the module performance is seriously impaired. The UTDR methodology can also be used to optimally configure a membrane module to minimize fouling vis-à-vis feed-spacer design and hydrodynamics. The gravimetric and SEM analyses indicated that the fouling increased in the feed-flow direction owing to the increased concentration polarization. However, a surprising observation from these gravimetric and SEM analyses was that the fouling increased dramatically towards the center of the module. This is believed to be caused by a larger feed-channel spacing and associated shear in the outer relative to the inner membrane envelope layers owing to the high pressure. This implies that the feed flow in the Koch 2521 spiral-wound membrane module used in this study is non-uniform rather than purely axial as has been widely assumed. An important conclusion is that UTDR analysis can provide a useful tool for optimally designing the feed-spacer for spiral-wound modules.