Nanofiltration Concentrate Treatment For Increasing The
Water Recovery In Drinking Water Production
S. Van Geluwe, C. Vinckier, L. Braeken, B. Van der Bruggen
* Laboratory of Physical Chemistry and Environmental Technology
W. de Croylaan 46, PO Box 02423
B-3001 Leuven
** Laboratory of Molecular Design and Synthesis
Celestijnenlaan 200F, PO Box 02404
B-3001 Leuven
Keywords: membrane filtration; ozone; electrodialysis
Problem sketch
Nanofiltration is an effective and reliable method for the combined removal of a broad
range of pollutants in surface water. However, fouling of the membrane limits the
water recovery for this application to about 80%. As problems with water scarcity are
expected to grow worse in the coming decades, even in regions currently considered
water-rich, it cannot be tolerated that 20% of the feed water is wasted.
Therefore, it is necessary to develop technologies that make the discharge of
concentrate streams superfluous. The general concept of this study is to remove
specific pollutants in the concentrate stream so that this stream can be returned to
the feed side of the membrane without increased membrane fouling. The natural
organic matter (NOM) in the concentrate is decomposed by ozone oxidation, if
necessary combined with hydrogen peroxide. The salts in the concentrate are
removed by electrodialysis. In this way, a closed cycle with a recovery of almost
100% may be obtained, as shown in figure 1.1. The aim of this research is to find the
optimal process parameters that make the recycle of the concentrate stream
possible.
Figure 1.1 Diagram of the proposed process for the treatment of nanofiltration concentrates in order to
increase the water recovery of nanofiltration in the drinking water industry.
Ozone oxidation
Due to their hydrophobic chemical structure, humic acids are often considered as the
most severe membrane foulants in nanofiltration of surface water. Concentrated
solutions of natural humic acids were oxidized by ozone for ten minutes. A
substantial decrease of the chemical oxygen demand (COD) by 40% could be
achieved, even at low ozone concentrations in the gas phase (< 20 g O 3/Nm3)(~ 2 g
O3/ g COD). The COD was further reduced at higher ozone concentrations or
reaction times, but only to a small extent, which could not be justified by the higher
treatment costs. A COD reduction of 70-80% is required in order to avoid the
accumulation of organic matter in the closed cycle. It was not possible to achieve this
requirement with pure ozone oxidation. Ozone reacts selectively with the
hydrophobic fraction of the organic matter, which is responsible for severe membrane
fouling, and could reduce the hydrophobic COD by 70% at low ozone doses.
These observations are explained by the fact that ozone preferentially oxidizes
electrophilic aromatic groups to oxygenated functional groups, such as aldehydic,
ketonic and especially carboxylic groups. These saturated compounds react typically
very inefficiently with ozone, so they are not further mineralized into carbon dioxide
and water. The oxidation of these reaction products rather than the primary organic
matter dictates the required oxidants doses and treatment cost. The mineralization of
the carboxylic reaction products was enhanced by the addition of hydrogen peroxide
to the ozonated solution. Mineralization could be further improved but this
improvement was limited, as shown by the data in table 1.1. Futher optimisation of
the hydrogen peroxide dose, pH and reaction time is necessary to make this process
economically attractive.
Table 1.1 Procentual UV absoption (UVA)(280 nm) removal during ozone oxidation (ozone gas phase
3
concentration: 18 g O3/Nm , pH: 7-8, initial UVA: 0.932, alkalinity: 420 mg/L NaHCO3). After ten
minutes, H2O2 was continuously added (molar ratio of H2O2 to O3 dose was 0.72)
Ozonation Perozonation
Reaction time (min) 0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30
% UVA removal 0% 20% 33% 43% 47% 50% 56% 57% 60% 61% 62% 62% 64% 64% 65% 67%
Electrodialysis
Electrodialysis has to be able to reduce the concentration of the main ions (Na +,
Ca2+, Mg2+, Cl-, SO42-, NO3-) in the nanofiltration concentrate sufficiently in order to
prevent salt accumulation in the closed cycle. Due to their high rejection by
nanofiltration membranes, divalent ions have a high concentration in the membrane
concentrate and consequently require a high removal degree in the electrodialysis
step. Batch-scale experiments were conducted with the Fumasep-membranes FTCM
and FTAM of Fumatech. Experiments with single salt solutions proved that these
membranes are aselective. The desalination of salt mixtures with a typical
composition of nanofiltration concentrates, showed unexpected but very valuable
results. The permselectivity of SO42- was higher compared to Cl- or NO3-, although
steric hindrance typically obstructs the transport of SO42- ions through ion selective
membranes. Thus, the characteristics of the Fumasep membranes, especially the
high crosslinking density, were effective for changing the permselectivity between
monovalent and divalent ions, which makes these membranes attractive for this
application.
Significance and impact
Combinations of chemical oxidation and membrane filtration are very promising to
tackle problems encountered in both technologies. For instance, ozone oxidation can
mitigate membrane fouling by natural organic matter and membranes are important
to resolve catalyst recovery problems in heterogeneous photocatalysis. Research
about hybrid oxidation/membrane systems is necessary to reduce the cost of both
technologies and make them more attractive in drinking water and wastewater
industry. For electrodialysis, the desalination of other salts than NaCl is hardly
investigated. Although the presence of other ions in the feed solution has a large
impact on the separation efficiency, the effect of a mixture of monovalent and
multivalent ions on the performance of electrodialysis is unknown.