Need for membranes with a better selectivity, better module design and higher log removal credits

In the last ten years, the membrane community experienced a tremendous surge in research of synthesizing novel membranes. New reverse osmosis (RO) membranes are developed in which the active layer is 3D-printed on novel support layers, alternative materials such as graphene and graphene oxide, alternative polymer chemistry and the use of mixed matrix materials such as zeolites and metal organic frameworks are increasingly used to synthesize membranes to be used in water treatment. The aim of these novel membranes is to increase the membrane performance by boosting the water production at a lower membrane area, ultimately to decrease the required energy for water treatment and desalination.

In a plenary speech at the 9th International Water Association (IWA) Membrane Technology Conference & Exhibition for Water and Wastewater Treatment and Reuse, prof. E.R. Cornelissen highlights some of the limitations of this membrane production enhancing strategy.

Plenary speakers.

Need for membranes with improved solute selectivity

As a result of increasing efforts in R&D, promising RO membranes are developed on laboratory scale with impressive results related to water flux (water production per membrane area). A very important aspect of membrane filtration is the ability to retain undesired components from the feed water, particularly membrane selectivity towards contaminants of emerging concern. Membrane selectivity of small polar organic compounds such as pyrazole, N-Nitrosodimethylamine (NDMA), Methyl-tert-butylether (MTBE) and benzotriazole is known to be very limited and membrane passages of <50% are reported for state of the art membranes. It is important to understand and improve the membrane selectivity of these specific compounds by developing new membranes targeting an improvement of membrane selectivity. New membranes should focus on improved selectivity and not (only) on improved water flux.

Develop membranes for brackish water sources

The energy consumption of RO is relatively high, particularly for salty feed water sources such as seawater. To produce clean water the osmotic pressure of the feed water needs to be overcome by hydrostatic pressure pumps. Developing new membranes with a higher productivity does not necessarily lead to a substantial decrease in energy requirement since this is limited by the osmotic pressure of the feed water. When the feed water contains less salts, such as for surface water, ground water and even wastewater, improving the membrane productivity will lead to much higher reductions in energy consumption. New membranes targeting an energy consumption reduction should focus more on lower salt feed water and not (only) on membranes for seawater desalination.

Presentation by Emile Cornelissen.

Need for better membrane module design

Commercial RO membranes are predominantly constructed as spiral wound membrane elements, in which membranes are separated by a feed spacer creating mixing in the feed channel. This design is already more than 40 years old which remained more or less the same since then with only minor incremental improvements. This old fashioned design is not ideal for the new membranes with a higher production, because mixing is limited by the use of feed spacers. This design can therefore not exploit the benefits of the new RO membranes currently being developed. Apart from this, the feed spacer is a source for fouling problems connected with limited mixing in the elements. New – possibly spacer-free – membrane elements should be developed to benefit from the advantages of new membranes.

Panel members.

Need for an on-line membrane integrity monitoring tool

Despite the extremely robust nature of RO membranes, the credited log-removal value (LRV) for viruses of RO processes in water treatment schemes is limited to only two. The reason is the lack of available methods for monitoring real virus removal of full-scale RO installations without the undesired addition of markers/surrogates. A new method based on DNA biomonitoring was recently developed at KWR using natural viruses present in sources for drinking water, which can be used a membrane integrity tool. With this novel method a LRV of seven was obtained using surface water as a feed for RO. This method is currently still off-line and ex-situ, and there is a need for an on-line membrane integrity monitoring tool.

New RO membranes are increasingly developed on the laboratory, and some are finding their way on the market. There is a strong need to develop RO membranes targeted on an increased selectivity for brackish water sources and a more appropriate module design. Also an on-line membrane integrity monitoring tool is desired to improve the credited LRV in full-scale plants.