Technical Field

The present invention relates to a system and process for membrane fouling control and a calcium and/or magnesium additive for membrane fouling control.

Prior art and background of the present invention

Membranes are used for wafer treatment in a number of industries, including municipal drinking wafer supply and wastewater treatment, food and beverage processes, pharmaceutical, oil&gas, and many more.

The main operational problem of membrane filtration is fouling. Foulanfs carried in the liquid †o be filtered are deposited on the membrane and lead †o a degradation of the membrane performance. There are different types of fouling. Colloidal/parficulafe fouling is formed by an accumulation of foulanfs on the membrane surface and inside the pores reducing membrane permeability. Organic fouling is caused by natural or synthetic organic matter from the source liquid. Biofouling is stemming from aquatic organisms such as bacteria or algae which can form colonies on the membrane surface and cause biofouling. Inorganic fouling or scaling is formation of hard mineral deposits on the membrane surface as supersafurafion limits of sparingly soluble salts is exceeded. Therefore, most membrane applications require fouling control measures †o minimize the negative impacts of fouling on membrane operation / lifetime.

If was found that fouling in each application depends largely on the feed wafer qualify, the membrane and the operating conditions. One strategy †o minimize fouling is thus the correct choice of the membrane and of the operating conditions depending on the application, i.e. depending on the type of foulanfs in the liquid †o be filtered. However, this requires a large variety of membrane filters for handling the large variety of different fouling situations. This increases the difficulty in projecting a filter application as each application is slightly different. Another solution against fouling is periodic cleaning of the membrane. This can be achieved by physical cleaning, biological cleaning or chemical cleaning. Physical cleaning can be gas scour, sponges, wafer jets or backwashing using permeate. Physical cleaning could also include some abrasive cleaning agents added in the liquid †o be filtered. Chemical cleaning involves the use of chemicals, e.g. acids, bases, oxidants, enzymatic components, surfactants, complexing agents and formulated detergents †o remove foulanfs and impurities. Chemical cleaning is normally more effective †o remove fouling which cannot be removed by physical cleaning. However, chemical cleaning creates additional waste and has thus a negative environmental impact. Physical cleaning might be better for the environment but needs †o be performed more often or sometimes might no† be sufficient †o remove the foulants from the membrane. The mentioned membrane cleanings are also called membrane remediations, i.e. measures †o re-establish the membrane functionalities after fouling has occurred.

Physical cleaning and chemical cleaning can also be combined. An example can be found in document EP1920821 using amongst other, water insoluble calcium carbonate.

A further solution is the pre†rea†men† of the water †o be filtered. The pre†rea†men† reduces the foulants in the water †o be filtered before they arrive a† the membrane or changes the characteristics of the foulants in such a way as †o reduce or limit their interaction with the membrane material and structure. Examples of a pre†rea†men† are coagulation, adsorption, oxidation, magnetic ion exchange (MIEX), biological treatment or some integrated pre†rea†men†s. Normally this is achieved by adding a fouling control additive into the water †o be filtered †o condition the water †o be filtered, i.e. †o modify or reduce the amount of foulants in the water †o be filtered before reaching the membrane. However, the efficiency of the pre†rea†men† of the water depends on many factors like the type of agents (coagulant, adsorbent, flocculan†, oxidizer, ...), the dosage, the dosing modes (continuous or intermittent), the dosing point, the mixing efficiency, temperature, properties of the foulants (hydrophobicity, charge density, molecular weigh†, and molecular size) and characteristics of the membrane (membrane charge, hydrophobicity, and surface morphology). Therefore, prefreafmenf methods are highly sensitive to the application conditions and need to be individually planned, optimized and monitored.

Conditioning by coagulation/flocculation is commonly used in conventional filtration technology (e.g., media filtration). The conditioning aims to prepare the liquid to be filtered to enable the correct functioning of the media filtration equipment by eliminating in particular the problems of media bed clogging and surface cake formation or filter blinding.

Although membrane technologies were originally designed as a higher performing and chemical-free alternative to conventional filtration technology, practical experience shows that various pretreatment processes (including physico-chemical processes) may be required to ensure stable membrane operation. Consequently, coagulation was adopted also as fouling control with a pretreatment of the water to be filtered through the membrane. A coagulation additive is added to the water to be filtered to promote the aggregation of foulants into larger floes to reduce or eliminate blocking of the membrane pores. However, pin-sized floes were deemed sufficient as the membrane pores are 2-3 orders of magnitude smaller than the pores in conventional filtration technology. Coagulants can be organic or mineral. Mineral coagulants are for example aluminum or iron salts. Coagulation is now the most widely used fouling control chemical in membrane-based water treatment systems. The coagulant acts here as the fouling control additive.

One problem of the coagulant is that unreacted coagulants can reach the membrane and react on the surface of the membrane or even in the membrane porous structure. Membranes may suffer from fiber plugging due to high coagulant use, or pore blocking by smaller monomers, dimers, and trimers of the hydrolyzing metal coagulant. So, the coagulant could become a source of fouling instead of solving the fouling. Fouling by Fe ²⁺ and Mn ²⁺ may also occur with coagulants of low grade. Fouling by iron requires a specifically tailored and intensive cleaning protocol to recover permeability. In addition, water generated from periodic membrane cleaning that contains the spent coagulant can have an environmental impact if disposed back directly to the environment without adequate treatment. While the above-described pretreatment measures condition the water to be filtered †o control the fouling, if is also known †o condition the membrane. For example, W02001027036A1 and W02002044091 A2 disclose a filtration aid which is added in each filter cycle info the wafer †o be filtered for conditioning the membrane. The filtration aid added in the wafer †o be filtered contains fouling control particles which form a deposit layer on the membrane due †o the permeate flux. The deposit layer protects the membrane from fouling by retaining the foulanfs in the deposit layer. A† the end of each cleaning cycle, the deposit layer with the retained foulanfs is cleaned from the membrane, e.g. by backwashing. Ion-exchange resin particles and suspensions of iron hydroxides, of aluminum oxides and hydroxides, of sintered iron oxide particles, of pulverized activated carbon, of clay particles and of other mineral particles are proposed as membrane fouling control additive. If is suggested †o use fouling control particles with a size between 0.5 and 20 micrometer (miti). However, all materials suggested for the fouling control particles had severe side effects like being reactive, being abrasive, being toxic or environmentally no† desirable which caused damages in the membrane, pumps and/or caused problems with the handling of the wastewater created by the cleaning. Consequently, this technology has no† found application a† industrial scale.

A similar system with the described membrane fouling control additive is also described in WO2017009792A1 . In addition, it is suggested †o have a stimuli responsive layer on the membrane and reactive fouling control particles such †ha† the dynamic protective layer formed by the fouling control particles remains attached †o the stimuli responsive layer even during the cleaning step until a stimulus is created with the stimuli responsive layer †o remove the fouling control particles. I† is suggested †o use fouling control particles with a particle size smaller than 1 m, preferably between 100 and 200 nanometer. This technology might lead †o some reduced amount of fouling control particles needed. I† did however no† solve the problems of the fouling control particles proposed being reactive and/or abrasive and/or being environmentally no† desirable. The stimuli responsive layer creates an additional process step associated with increased cos† and complexity of the proposed solution, including waste handling and disposal. Brief summary of the invention

I† is therefore the object of this invention †o provide a system and a process for membrane fouling control as well as a calcium and/or magnesium additive as a fouling control additive which avoids the disadvantages of the state of the art and/or which is provided for an easy and efficient membrane fouling control.

According †o the invention, this object is solved by a process for membrane fouling control comprising the steps of: Conducfing/filfering a liquid †o be filtered through a membrane; and adding fouling control particles upstream of the membrane and/or info the liquid †o be filtered †o form a dynamic protective layer on the membrane. The process is characterized by one or more of the subsequent embodiments.

Further, According †o the invention, this object is also solved by a system for membrane fouling control comprising: a firs† conduct portion; a second conduct portion; a membrane arranged between the firs† conduct portion and the second conduct portion and/or configured †o filter a liquid by conducting it from the firs† conduct portion †o the second conduct portion; and a fouling control means filled with fouling control particles and/or configured for adding the fouling control particles in the firs† conduct portion, wherein the fouling control particles and/or the fouling control means is configured such †ha† a dynamic protective layer is formed on the membrane by the fouling control particles added by the fouling control means for protecting the membrane from foulants. The system is characterized by one or more of the subsequent embodiments.

According †o the invention, this object is further solved by a calcium and/or magnesium additive (for membrane fouling control), preferably comprising particles for forming a dynamic protective layer on the membrane for fouling control of the membrane, when added in the liquid flowing through the membrane.

The calcium and/or magnesium additive (for membrane fouling control) is characterized in †ha† the particles comprise synthetic mineral precipitate particles based on calcium and/or magnesium chosen amongst ultrafine synthetic mineral precipitate particles based on calcium and/or magnesium having a particle size distribution dso lower than or equal to 1 m, d9o lower than or equal to 10 pm and a dio comprised between 50 nm and 500 nm and microfine synthetic mineral precipitate particles based on calcium and/or magnesium having a particle size distribution dso lower than or equal to 5 pm, d9o lower than or equal to 15 pm and a dio comprised between 200 nm and 3 pm.

Accordingly, the additive according to the present invention is a synthetic mineral precipitate based on calcium and/or magnesium having strictly controlled particle size distribution showing adequacy with the membrane fouling control application in that the membrane permeability was nearly completely restored after each cleaning of the dynamic protective layer while no† being detrimental †o the membrane. The calcium and/or magnesium additive according †o the present invention is able †o permit high initial liquid flow, provide pore space †o trap and contain the foulants and do / does no† adversely affect the membrane permeability, especially due †o the fact †ha† the particle size distribution of synthetic mineral precipitate based on calcium and/or magnesium is tailored †o create coating layers †ha† provide highest flow rates through the membranes bu† also having properties a† the layer level.

According †o the present invention, the particle size distribution (PSD) is measured according †o the standard IS013320:2020(E), item 5 by laser light scattering (commonly also called laser diffraction) of the calcium and/or magnesium additive dispersed in water. Preferably, the PSD is analyzed/measured for the range smaller than 0,4 m by laser light scattering using light a† 450 nm, 600 nm, 900 nm with two polarizations for each wavelength and for the range equal †o and larger than 0,4 pm by a laser of 780 nm wavelength. The PSD is generated by a Mie Diffraction model, according †o the standard IS013320:2020(E), item 5. Water is used as dispersion liquid. Sonication is used †o reach complete dispersion of the calcium and/or magnesium additive in the water before the measurement with the laser diffraction. In the used measurement method for obtaining the PSD below 0.4 pm, a white light emitting diode (LED) is used with three different filters †o obtain the light measurement a† the three wavelengths 450 nm, 600 nm, 900 nm and with two polarizers †o obtain for each of the three wavelengths two (orthogonal) polarizations (resulting in a total of six light measurements). In the used measurement method for obtaining the PSD at and above 0.4 miti, laser light at 780 nm (red laser) is used. For the red laser, no polarizers are used. The PSD for the range below 0.4 miti is generated by the Mie Diffraction model based on the six measurements resulting from the three wavelength measurements a† 450 nm, 600 nm, 900 nm each measured with two orthogonal polarizations. The PSD for the range a† and above 0.4 miti is generated by the Mie Diffraction model based on the measurement with the red laser.

The values d _y of a particle size distribution, as explained below, refers †o the size a† which y% of the distribution is below the size d _y. The values d _y in this application refer †o a number-based particle size distribution meaning that y% of the number of particles are below d _y and/or (100%-y%) of the number of particles are above d _y (neglecting the particles exactly a† d _y ). The values d _y of a particle size distribution are determined from the PSD determined by laser light scattering as explained above.

Synthetic mineral precipitate based on calcium and/or magnesium are mineral precipitates for which the manufacturing is controlled †o reach specific feature in the synthetic precipitate such as the particle size distribution, the BJH porosity and BET specific surface area. Such synthetic mineral precipitate distinguishes typically over naturally occurring mineral precipitate by the constraints applied during the manufacturing process yielding typically †o a narrower particle size distribution for the synthetic mineral precipitate since the process is controlled while naturally occurring mineral precipitate have typically a very broad particle size distribution since crystallization is no† performed under constraints and most often quite slow.

Synthetic mineral precipitate particles are for example disclosed in US20150044469 or in US20180170765.

The process, system and/or additive is characterized by one or more of the following embodiments.

In one embodiment, the fouling control particles are precipitates, preferably mineral or metal precipitates. Precipitate particles, in particular the family of mineral precipitate particles and of metal precipitate particles allows †o tune the properties of the fouling control particles like their size or size distribution during the manufacturing. In one embodiment, the calcium and/or magnesium particles are/comprise synthetic mineral precipitates based on calcium and/or magnesium carbonate. Preferred examples of this family are precipitated calcium carbonate and precipitated hydromagnesife. This is a versatile mineral- based material that can be tuned in terms of morphology, particle size and polydispersify †o provide a range of functionalities for different applications. For the purpose of membrane fouling control, we found a surprisingly good effect with this material, whereby membrane permeability was nearly completely restored after each cleaning of the dynamic protective layer. In addition, this material is no† detrimental †o the membrane material or other system components (no† abrasive, no† corrosive) and does no† pose an environmental hazard (can be disposed of with minimal handling). Ultrafine and/or microfine precipitated calcium carbonate or/and precipitated hydromagnesite is/are chemically inert and is/are able †o form high porosity filter cakes †ha† permit high initial liquid flow, provide pore space †o trap and contain the foulants and do / does no† adversely affect the membrane permeability. Particle size distribution of precipitates of this material can be tailored †o create coating layers †ha† provide highest flow rates through the membranes.

In one embodiment, the fouling control particles are composite particles (made of a composition) of a firs† material and a second material. The composite particles comprise preferably a core of the firs† material and a shell of the second material. The firs† material is preferably a mineral and/or a mineral precipitate, preferably a calcium-based mineral (precipitate), preferably a calcium carbonate, preferably mainly calcite with a minor portion of aragonite. The second material is preferably a mineral and/or a mineral precipitate, preferably a magnesium-based mineral (precipitate), preferably a hydromagnesite, preferably hydromagnesite with (some traces of) nesquehonite. These composite particles can be easily tuned in size based on the material of the core and/or of the firs† material and/or can be easily functionalized based on the material of the shell and/or of the second material. This allows †o functionalize the dynamic protective layer, e.g. for improving the adhesion of the foulants †o the dynamic protective layer, and thus further improve the filter function of the dynamic protective layer. The shell material has preferably different properties than the core material and would absorb different kind of foulanfs and †o different surfaces. In addition, the shell material can be porous, allowing absorption info the pores, which keeps the spaces (pores) between the particles open. On the other side, this group of composite particles are environmentally much more compatible than the reactive fouling control particles used in the state of the art.

In one embodiment, the fouling control particles are inert. Within the meaning of the present invention, the term “inert” means that the fouling control particles do not change their properties in the dynamic protective layer on the membrane during the operation mode. In the state of the art, only reactive fouling control particles were suggested to react with the foulants. This has a negative effect on the environment and needs special waste treatment. Often the reactive fouling control particles are also aggressive to the membrane and other system equipment like pumps, e.g. abrasive or corrosive.

In one embodiment, the fouling control particles are reactive or functionalized. Within the meaning of the present invention, the terms “reactive or functionalized particles” means preferably that the surface of the fouling control particles have specific charges or groups which absorb foulants and/or which attach to the membrane surface.

In one embodiment, the synthetic mineral precipitate particles based on calcium and/or magnesium have a polydispersity value greater than or equal to 0.1 and/or smaller than or equal to 1 .5. Preferably, the polydispersity value is larger than or equal to 0.2, preferably larger than or equal to 0.3, preferably larger than or equal to 0.4, preferably larger than or equal to 0.5. Preferably, the polydispersity value is smaller than or equal to 1.4, preferably smaller than or equal to 1.2, preferably smaller than or equal to 1.1 , preferably smaller than or equal to 1.0. Preferably, the ultrafine synthetic mineral particles have a polydispersity value comprised in the range 0.3 to 1.2, preferably between 0.5 to 1.0. Preferably, the microfine synthetic mineral particles have a polydispersity value comprised in the range 0.7 to 1 .5, preferably between 0.9 to 1.3. The polydispersity value of the synthetic mineral precipitate particles according to the present invention allows to characterize the distribution width (also sometimes referred to a certain polydispersity of the population of the particles in the mineral precipitate particles) considered between the particle size d9o and dio divided by the dso particle size preferably larger than or equal †o 0.1 , preferably larger than or equal †o 0.15, preferably larger than or equal †o 0.2. The polydispersify value of the calcium and/or magnesium particles is preferably smaller than or equal †o 1 .5, preferably smaller than or equal †o 1 .2, preferably smaller than or equal †o 1 .0, preferably smaller than or equal †o 0.8, preferably smaller than or equal †o 0.6, preferably smaller than or equal †o 0.4, preferably smaller than or equal †o 0.35, preferably smaller than or equal †o 0.3, preferably smaller than or equal to 0.25. By the terms “polydispersify value”, if is mean†, within the meaning of the present invention the distribution width between d9o and dio divided by the dso. The polydispersify value is calculated according †o the formula below: where d9o, dio, dso values are the d9o, dio, dso of the synthetic mineral precipitate particles based on calcium and/or magnesium. As defined above, the d9o, dio, dso are obtained from a number-based distribution. Such polydispersity value obtained from number-based distributions is used †o characterize the narrow particle size distribution of the synthetic mineral precipitate particles according †o the present invention and is disclosed in literature. Even if polydispersity value can also be obtained from volume-based distribution, the polydispersity value calculated according †o the present invention is obtained from number-based distribution. I† has indeed been realized †ha† the polydispersity value plays a significant role in the formation of the dynamic protective layer by fouling control particles, because of the risk of pore blockage by too fine fouling control particles on the one hand and poor and inhomogeneous membrane coverage by large fouling control particles on the other hand.

In one embodiment, the fouling control particles have an absolute value of †he ze†a potential smaller than 50 mV, preferably than 45 mV, preferably than 40 mV. This has the advantage †ha† the fouling control particles can form a sacrificial layer, i.e. a layer †ha† can be easily detached from the membrane. In one embodiment particles have a positive zeta potential smaller than or equal to 50 mV, preferably smaller than or equal †o 45 mV, preferably smaller than or equal †o 40 mV. In another embodiment, the particles have a negative zefa potential greater than -50 mV, preferably greater than or equal †o -45 mV, preferably greater than or equal †o -40 mV.

According †o the present invention, the zefa potential of the particles was measured according †o the following experimental procedure. The suspensions of synthetic mineral precipitate particles have been subjected †o a magnetic agitation for 30 minutes, before each zefa potential measurement. The suspensions have been then subjected †o a dispersion step during 30 seconds with an ultrasound sonicafion probe Hielscher-UP4005 (with an amplitude of 50% and a cycle of 0.5) followed by a further magnetic agitation during 30 minutes before a new series of measurements. The zefa potential measurements have been performed with electroacoustic spectrometer DT-1200 (Dispersion Technology). The electroacoustic probe for measuring the zefa potential was calibrated with a colloidal silica suspension showing a zefa potential of -38 (±1 ) mV. For each sample, 5 zefa potential measurements are realized.

In one embodiment, the fouling control particles have a specific surface area measured by nitrogen adsorption manometry and calculated according †o the BET method larger than 10 m ² /g, preferably than 15 m ² /g, preferably larger than or equal †o 20 m ² /g. This results in a better deposition on the membrane surface as well as larger affinity for foulanfs capture from the liquid †o be filtered.

In one embodiment, the Scherrer diameter of the fouling control particles are larger than 20 nm, preferably than 30 nm, preferably than 40 nm, preferably than 50 nm, preferably than 60 nm. In one embodiment, the Scherrer diameter of the fouling control particles are smaller than 150 nm, preferably than 140 nm, preferably than 130 nm, preferably than 120 nm, preferably than 1 10 nm, preferably than 100 nm. Preferably, the Scherrer diameter of the fouling control particles are between 50 nm and 100 nm. The Scherrer diameter indicates the mean size of the ordered (crystalline) domains and is calculated according †o the Scherrer equation, preferably based on measurements with X-ray diffraction. X-ray powder diffraction (XRD) is an analytical technique used for phase identification and crystals size (Schemer Diameter) of a crystalline material. Diffraction is based on generation of X-rays in an X-ray tube. These X-rays are filtered and collimated in order to produce a concentrated monochromatic radiation, which is directed towards the sample. This analysis is performed using a Bruker - D8 advance instrument. Copper K-a is an x-ray energy frequently used on lab-scale x-ray instruments. The energy is 8.04 keV, which corresponds to an x- ray wavelength of 1 .5406 A.

In crystalline materials, the scattered X-rays undergo constructive and destructive interferences. This process of diffraction is described by Bragg’s Law (nA=2d sin Q). The diffracted X-rays are detected, processed, and counted.

As the directions of possible diffractions depend on the size and shape of the crystals composing the sample and the intensities of the diffracted waves depend on the kind and arrangement of atoms in the crystal structure, this method enables to determine the chemical composition of a sample. By applying Schemer’s equation to peak broadening for a crystal structure, it is also possible to determine the size of crystals in the sample.

Typically, we would use the Lorentz fit (for the diffraction peak) to calculate the Schemer diameter and we apply this to the peaks at ca. 29.45°, 39.46° and 43.21 ° of the diffraction pattern.

In one embodiment, the fouling control particles and/or the additive is under the form of a slurry with a solids content and/or a fouling control particles content preferably lower than or equal to 20 w†%, preferably lower than or equal to 15 w†%, preferably lower than or equal to 12 w†%, preferably lower than or equal to 1 1 w†%, preferably lower than or equal to 10 w†%. In one embodiment, the fouling control particles and/or the additive is in a form of a slurry with a solids content and/or a fouling control particles content preferably greater than or equal to 1 w†%, preferably greater than or equal to 1.5 w†%, preferably greater than or equal to 3 w†%, preferably greater than or equal to 5 w†%, preferably greater than or equal to 6 w†%, preferably greater than or equal to 7 w†%. The higher the solid content of the fouling control particles, the less volume and weight the additive and/or the fouling control particles require for storage and transport. In one embodiment, the ultrafine synthetic mineral precipitate particles based on calcium and/or magnesium having a particle size distribution dso lower than or equal to 1 m, d9o lower than or equal to 10 pm and a dio comprised between 50 nm and 500 nm. The ultrafine mineral precipitate particles according †o the present invention are particularly suitable for ulfrafilfrafion.

In one embodiment, the microfine synthetic mineral precipitate particles based on calcium and/or magnesium having a particle size distribution dso lower than or equal †o 5 pm, d9o lower than or equal †o 15 pm and a dio comprised between 200 nm and 3 pm. The microfine mineral precipitate particles according †o the present invention are particularly suitable for microfilfrafion.

Typically, according †o the present invention, the dio of the particle size distribution of the calcium and/or magnesium particles is larger than a defined pore size of the membrane. Preferably, the dio of the size distribution of the fouling control particles is 3 times, preferably five times, preferably seven times preferably nine times, preferably fen times larger than the defined pore size. If is shown that the selection of the lower size limit of the calcium and/or magnesium particles dependent on the defined pore size of the membrane increases significantly the performance of the fouling control method. The calcium and/or magnesium particles are selected larger than the defined pore size of the membrane so that the calcium and/or magnesium particles cannot enter the membrane and the narrowing of the pores of the membrane is avoided. This explains likely the positive effect of this measure. In the state of the art, only the average particle size was controlled and the average particle size was no† selected depending on the membrane defined pore size. This explains the limited performances of the membrane fouling control by the fouling control particles forming a dynamic protective layer on the membrane of the state of the ar†. In addition, the fact †ha† in the fouling control particles of the state of the ar† only the average particle size was controlled neglected the often very large size distributions of the fouling control particles in those small dimensions which lead for average particle sizes above the defined pore size still †o a large amount of fouling control particles below the defined pore size leading therefore still †o a narrowing of the pores. This measure is particularly advantageous with the subsequent embodiment. In one embodiment, depending on the microfiltration membrane, the d9o of the size distribution of the calcium and/or magnesium particles is lower than or equal †o 15 miti, preferably lower than or equal †o 10 miti, preferably lower than or equal †o 5 miti, preferably lower than or equal †o 1 miti, preferably lower than or equal †o 0.7 miti, preferably lower than or equal to 0.5 miti.

In one embodiment, depending on the ulfrafilfrafion membrane, the d9o of the size distribution of the calcium and/or magnesium particles is lower than or equal †o 10 miti, preferably lower than or equal †o 5 miti, preferably lower than or equal †o 1 miti, preferably lower than or equal †o 0.7 miti, preferably lower than or equal to 0.5 miti.

If is shown that the main influence of the upper size limit of the calcium and/or magnesium particles for the performance of the fouling control method is the homogeneity of the deposition of the calcium and/or magnesium particles. The homogeneity depends mainly on the absolute particle size and less on the dependence on the defined pore size of the membrane. The above- mentioned upper size limits have been proved †o significantly reduce the heterogeneity of the deposited calcium and/or magnesium particles leading thus †o a high performance dynamic protective layer on the membrane. As already mentioned above, if was further found out that if is no† only important †o select the average particle size, bu† more important †o select the d9o of the size distribution of the calcium and/or magnesium particles. There might be also some effects of the upper limit of the particle size which depend on the defined pore size like full blocking of the pores. However, i† was shown †ha† these defined pore size dependent effects have a reduced influence compared †o the mentioned homogeneity which depends on the absolute size of the particles.

In a preferred embodiment, the dio of the particle size distribution of the calcium and/or magnesium particles is chosen depending on the defined pore size, while the d9o of the size distribution of the calcium and/or magnesium particles is chosen as an absolute value. This combination significantly improved the performance of the membrane fouling control compared †o the ones known in the state of the ar†.

In one embodiment, the dio of the particle size distribution of the calcium and/or magnesium particles for ul†rafil†ra†ion is greater than or equal †o 50 nm, preferably greater than or equal †o 70 nm, preferably greater than or equal †o 100 nm, preferably greater than or equal †o 200 nm.

In one embodiment, the dio of the particle size distribution of the calcium and/or magnesium particles for ultrafiltration is lower than or equal to 500 nm, preferably lower than or equal to 400 nm, preferably lower than or equal to 350 nm, preferably greater than or equal to 320 nm.

In one embodiment, the dio of the particle size distribution of the calcium and/or magnesium particles for microfiltration is greater than or equal to 200 nm, preferably greater than or equal to 250 nm, preferably greater than or equal to 300 nm, preferably greater than or equal to 500 nm.

In one embodiment, the dio of the particle size distribution of the calcium and/or magnesium particles for microfiltration is lower than or equal to 1000 nm, preferably lower than or equal to 900 nm, preferably lower than or equal to 800 nm, preferably lower than or equal to 750 nm.

In one embodiment, the process comprises the further step of forming the calcium and/or magnesium particles in situ and/or the fouling control means comprises a fouling control particle forming means for forming the fouling control particles in situ or in the fouling control means. The fouling control particles are preferably formed before adding the fouling control particles into the liquid to be filtered.

The calcium and/or magnesium particles are preferably formed by a physical action and/or a chemical reaction. The chemical reaction uses preferably at least one raw material or substance, preferably at least two raw material or substances to obtain by the chemical reaction the calcium and/or magnesium particles. The chemical reaction is preferably a precipitation. The chemical reaction is preferably a carbonation, preferably a carbonation of calcium hydroxide and/or magnesium hydroxide. Preferably, precipitated calcium carbonate is obtained by the chemical reaction. By producing the calcium and/or magnesium particles in situ, some stability problems of the size distribution of some calcium and/or magnesium particles can be reduced, the raw materials used for the chemical reaction might be easier to handle and process. The physical action can be a size reduction action and/or a classification.

In one embodiment, in the process according †o the present invention, the dynamic protective layer is formed directly against the membrane.

In one embodiment, in the process according †o the present invention, the dynamic protective layer formed on the membrane is a porous layer which allows the liquid †o pass through †o the membrane and which retains the foulanfs.

In one embodiment, in the process according †o the present invention, the dynamic protective layer is formed and/or maintained on the membrane by a fransmembrane pressure generated across the membrane as a result of conducting the liquid †o be filtered through the membrane and/or by the permeate flux, i.e. by the permeation drag or the hydraulic force that transports the particles †o the membrane surface.

In one embodiment, the process according †o the present invention comprises the further step of and/or the system according †o the present invention is configured for cleaning the dynamic protective layer with the filfered/refained foulanfs from the membrane.

In one embodiment, the process according †o the present invention performs a plurality of filter cycles and/or the system according †o the present invention is configured †o perform a plurality of filter cycles. One or more or all filter cycles of the plurality of filter cycles comprise:

• Adding the fouling control particles info the liquid †o be filtered for forming the dynamic protective layer on the membrane;

• filtering the liquid †o be filtered conducted through the membrane while the foulanfs are retained by the dynamic protective layer; and cleaning the dynamic protective layer with the retained foulanfs from the membrane.

In the process according †o the present invention, the membrane fouling control additive is a calcium and/or magnesium additive as described above. In one embodiment, the membrane of the process according †o the present invention and/or of the system according †o the present invention is a microfilfrafion membrane or ulfrafilfrafion membrane.

In one embodiment, in the process according †o the present invention, said liquid conducted through the membrane is an aqueous liquid, preferably municipal wastewater, seawater, industrial process wafer and industrial wastewater, source wafer for municipal drinking wafer, brackish wafer, fresh ground wafer, fresh surface wafer.

In one embodiment, in the process according †o the present invention, the calcium and/or magnesium particles added info the liquid †o be filtered are forming a dynamic protective layer on the membrane such that, when the liquid †o be filtered contains foulanfs, the foulanfs are retained in the dynamic protective layer before the liquid †o be filtered is conducted through the membrane.

Other embodiments according †o the present invention are mentioned in the appended claims, the subsequent description of

Brief description of the Drawings

Figure 1 shows a system for fouling control.

Figure 2 shows a step of a method for fouling control of introducing a membrane fouling control additive in a liquid †o be filtered.

Figure 3 shows a step of the method for fouling control of forming a dynamic protective layer on the membrane.

Figure 4 shows a step of the method for fouling control of retaining foulanfs in the dynamic protective layer.

Figure 5 shows a step of the method for fouling control of cleaning the dynamic protective layer from the membrane.

Figure 6 shows an embodiment of a size distribution of the calcium and/or magnesium particles.

Figure 7 shows an idealized membrane with the fouling control particles of a firs† size.

Figure 8 shows the idealized membrane with the fouling control particles of a second size. Figure 9 shows the pressure over different filtration cycles for a membrane fouling control using coagulants.

Figure 10 shows the pressure over different filtration cycles for a membrane fouling control using precipitated calcium carbonate as fouling control particles.

Figure 1 1 show a translation table for the defined pore size and the

MWCO.

Figure 12 shows the pressure over different filtration cycles for a membrane fouling control once using precipitated calcium carbonate as fouling control particles and once using no fouling control particles.

In the drawings, the same reference numbers have been allocated †o the same or analogue element.

Detailed description of an embodiment of the invention

Other characteristics and advantages of the present invention will be derived from the non-limi†a†ive following description, and by making reference †o the drawings and the examples.

Fig. 1 shows a system for membrane fouling control comprising a membrane 1 , a firs† conduct portion 2, a second conduct portion 3 and a fouling control means 4. Preferably, the system comprises further a cleaning outlet 9 and/or cleaning outlet control means.

The system for membrane fouling control is configured †o filter a liquid †o be filtered through a membrane 1 †o obtain a filtered liquid while controlling the fouling of the membrane 1. The application of the system for membrane fouling control can be in municipal drinking water applications, municipal wastewater applications, applications for desalination or reuse of water, applications for industrial wastewater like oily waters, sludge, waters with chemical substances, etc. Plowever also in industrial process water like in the beverage industry the present system can be applied.

The liquid †o be filtered is the liquid in the firs† conduct portion 2 or the liquid which is in the operation mode upstream of the membrane 1. The filtered liquid is the liquid in the second conduct portion 3 or the liquid which is downstream of the membrane 1 in the operation mode. The filtered liquid obtained by the liquid †o be filtered conducted through the membrane 1 is also called the permeate. The liquid to be filtered and the filtered liquid is normally a wafer-based liquid (aqueous liquid). However, if is also possible that the liquid †o be filtered is no† a water-based liquid, bu† another liquid.

The liquid †o be filtered is normally a liquid including contaminants (contaminated liquid). The contaminants are intended †o be filtered ou† by the membrane 1. The filtered liquid is the liquid with a reduced amount of contaminants, preferably substantially depleted from the contaminants. Of course, it is also possible †ha† the liquid †o be filtered is a clean liquid, i.e. a liquid without contaminants. This can be for certain operation modes or because simply the liquid †o be filtered is no† always contaminated. The liquid †o be filtered is normally a water-based liquid (aqueous liquid), i.e. a contaminated water or water with contaminants and the filtered liquid is normally a clean water or a water with a reduced amount of contaminants, preferably free of contaminants. The contaminated water could be a water source for municipal drinking water and/or the filtered liquid could be the municipal drinking water or a pre- processed municipal drinking water. The contaminated water could be a wastewater, like municipal wastewater with biological or other contaminants or like industrial wastewater with chemical, mineral or other contaminants, or like sludge and/or the filtered water could be filtered wastewater or clean water. The contaminated water could be sal† or seawater with the contaminants being salts, minerals and/or biological contaminants. The contaminated water could be industrial process water.

The contaminants are normally undesired particles in the liquid †o be filtered and the filtered liquid (also called permeate) is the desired product of the membrane process. However, in some applications, the contaminants retained by the membrane 1 or the re†en†a†e, i.e. the liquid with the retained contaminants, could be the desired product of the process. The contaminants are preferably particles which can and/or should be retained by the membrane 1.

The liquid †o be filtered contains further foulants 10. Foulants are contaminants which create membrane fouling. The foulants can be in a solid form (fouling particles) in the liquid †o be filtered and/or can be in a dissolved form in the liquid †o be filtered. Different types of foulants 10 and different types of membrane fouling were already described in the introduction and are no† repeated here. All or some foulants 10 can be contaminants. All or some contaminants can be foulants 10. This depends on the application, the liquid to be filtered, the membrane 1 and the process conditions. The subsequent description focuses mainly on the foulants 10 without specifying, if the foulants are contaminants or not.

The membrane 1 is configured to filter contaminants out of the liquid to be filtered to obtain the filtered liquid, when the liquid to be filtered traverses the membrane 1. The filtering of the contaminants means that the amount of contaminants in the filtered liquid is smaller than the amount of contaminants in the liquid to be filtered, preferably reduced by more than 50%, preferably by more than 60%, preferably more than 80%, preferably more than 90%. Most preferably all contaminants are retained by the membrane 1 , when the liquid to be filtered traverses the membrane 1. The membrane 1 has a first side 1.1 and a second side 1 .2. The first side 1 .1 is connected with the second side 1 .2 through pores 5 of the membrane 1 . The first side 1 .1 provides the membrane surface of the membrane 1 . The first side 1.1 or the membrane surface is provided depending on the design of the membrane 1 by tubes, hollow fibers or other geometries.

The membrane 1 has preferably a defined pore size defining the retaining capability of a membrane 1. The defined pore size is a value resulting from the pore size distribution of the pores of the membrane 1 and/or resulting from the particle size for which particles with this particle size are retained at a certain percentage X %. An example of a pore size resulting from the pore size distribution is the pore size most common in the distribution or the average pore size or the d9o of the pore size distribution. d _z of a size distribution refers to the size at which z% of the distribution is below the size d _z . In some membranes 1 , the defined pore size of a membrane 1 refers to a particle size for which the particles are retained by the membrane 1 (with a certain likelihood). The defined pore size of a membrane 1 refers preferably to a particle size at which X % of the particles of this particle size are retained. The percentage X is preferably larger than or equal to 50%, preferably larger than 60%, preferably than 70%, preferably than 80%. A preferred definition of the defined pore size is X being 90%. Some membranes 1 define the defined pore size as a molecular weigh† cut-off (MWCO) with units in Dalton (Da). If is defined as the minimum molecular weigh† of a globular molecule that is retained †o 90% by the membrane 1. The MWCO can be converted for the purpose of the invention †o the defined pore size according †o the fable shown in Fig. 1 1 . The second column shows the MWCO in kilo Dalton (kDA) and the firs† column show the corresponding translated defined pore sizes in nanometer (nm). The third column shows the corresponding type of the membrane: Reverse osmosis, nanofilfrafion, ulfrafilfrafion and microfilfrafion. Any MWCO value of a membrane 1 can be translated info its corresponding defined pore size of the membrane 1 according †o the invention by a linear interpolation of the next bigger MWCO value and the next smaller MWCO value. This translation method might not always be very precise, but is sufficient for the current invention to define clearly a defined pore size of a membrane 1 based on its MWCO value, if the membrane 1 is defined by its MWCO value instead of its pore size.

The invention is particularly advantageous for membranes 1 with pore sizes larger than 5 nm, preferably larger than 10 nm, preferably larger than 50 nm, in some embodiments larger than 100 nm. The invention is particularly advantageous for membranes 1 with pore sizes smaller than 50 miti, preferably smaller than 20 miti. The membrane 1 is preferably a microfiltration membrane or an ultrafiltration membrane. A microfiltration membrane is considered a membrane with a pore size between 150 nm and 20 miti. An ultrafiltration membrane is considered a membrane with a pore size between 5 nm and 150 nm. However, the invention is not limited to those pore sizes and could also be applied to other membrane types, maybe larger nanofiltration membrane pore sizes. In a preferred embodiment, the membrane 1 is not suitable for removing monovalent and/or divalent ions.

The membrane 1 can be any membrane type like monoliths, tubular, flat sheet (plate, frame or spiral wound), hollow fibre or any other membrane type. The membrane 1 can be of any material like ceramics, polymers or others. The membrane 1 shown in Fig. 1 is arranged in a dead-end configuration. The membrane 1 can however also be arranged in a cross- filtration configuration. The membrane works preferably with a low-pressure membrane technology.

The first conduct portion 2 is configured to store or conduct the liquid to be filtered such that it can be conducted through the membrane 1 . The first conduct portion 2 has therefore a contact surface with the first side 1 .1 of the membrane 1 . The first conduct portion 2 can be a vessel, a tube, a tank or any other liquid storing or conduct means. The first conduct portion 2 can comprise a pump to create a transmembrane pressure between the first side 1.1 of the membrane 1 and the second side 1 .2 of the membrane 2 and/or to transport the liquid to be filtered to the membrane 1 and/or to create a hydraulic force that transports the liquid to be filtered to the membrane surface. The created liquid flow is preferably perpendicular to the membrane surface. The transmembrane pressure is the resultant pressure formed in the system from the flow through the membrane according to Darcy’s law. The first side 1.1 and second side 1 .2 of the membrane 1 are not necessarily in geometrical terms, but rather in functional terms of the membrane 1 .

The second conduct portion 3 is configured to store or conduct the filtered liquid. The liquid to be filtered flows thus through the membrane 1 to become the filtered liquid and enters the second conduct portion 3. The second conduct portion 3 has therefore a contact surface with the second side 1.2 of the membrane 1 . The second conduct portion 3 can be a vessel, a tube, a tank or any other liquid storing or conducting means. The second conduct portion 3 can comprise a pump to transport the filtered liquid away from the membrane 1 and/or to create a transmembrane pressure between the first side 1.1 of the membrane 1 and the second side 1 .2 of the membrane 1 .

The system or the first conduct portion 2, the membrane 1 and/or the second conduct portion 3 is/are preferably configured (in the operation mode) to create a transmembrane pressure or liquid flow which conducts the liquid to be filtered from the first conduct portion 2 through the membrane 1 to the second conduct portion 3 to become the filtered liquid. The transmembrane pressure can be achieved by a pump in the first conduct portion 2 or in the second conduct portion 3. The transmembrane pressure or the liquid flow can also be created by gravity or by any other force on the liquid. In some embodiment, the system or the first conduct portion 2, the membrane 1 and/or the second conduct portion 3 is/are preferably configured (e.g. in the cleaning mode) †o create a reverse fransmembrane pressure ora reverse liquid flow which conducts the filtered liquid or the liquid in the second conduct portion 2 through the membrane 1 info the firs† conduct portion 2. This reverse transmembrane pressure or reverse liquid flow can be applied in a cleaning mode †o create a backwash of the membrane 1 †o clean the membrane 1 from a dynamic protective layer 7, calcium and/or magnesium particles 6 and/or foulants.

The fouling control means 4 is configured †o add calcium and/or magnesium particles 6 and/or a membrane fouling control additive into the liquid †o be filtered and/or into the firs† conduct portion 2. The fouling control means 4 comprises preferably a container or vessel for storing the calcium and/or magnesium particles 6 and/or the membrane fouling control additive. The fouling control means 4 comprises an opening or a connection towards the firs† conduct portion 2 for adding the calcium and/or magnesium particles 6 and/or the membrane fouling control additive into the liquid †o be filtered and/or into the firs† conduct portion 2. The opening or connection can preferably be opened and closed †o control the amount of opening or a connection towards the firs† conduct portion 2 for adding the calcium and/or magnesium particles 6 and/or the membrane fouling control additive added. Preferably, the fouling control means 4 contains the calcium and/or magnesium particles 6 and/or the membrane fouling control additive.

The membrane fouling control additive is an additive †o prevent or control membrane fouling by conditioning the membrane, in particular, by forming a dynamic protective layer 7 of calcium and/or magnesium particles 6 on the membrane 1. The membrane fouling control additive is an additive added during the operation mode of the system/method (no† during the cleaning mode). The membrane fouling control additive comprises the calcium and/or magnesium particles 6. The calcium and/or magnesium particles 6 are configured †o form the dynamic protective layer 7 on the membrane 1 , when the calcium and/or magnesium particles 6 are added in the liquid †o be filtered.

The membrane fouling control additive and/or the calcium and/or magnesium particles 6 are preferably in a form of a slurry. The slurry contains preferably a liquid, preferably wafer, and solid particles (solids). The solids comprise the calcium and/or magnesium particles 6, preferably consist of af leas† 50 w†% calcium and/or magnesium particles, preferably of af leas† 60 w†% calcium and/or magnesium particles, preferably of af leas† 70 w†% calcium and/or magnesium particles, preferably of af leas† 80 w†% calcium and/or magnesium particles, preferably of af leas† 90 w†% calcium and/or magnesium particles. Most preferably, the solids are the calcium and/or magnesium particles 6. The solids content and/or the calcium and/or magnesium particles 6 content of the slurry is preferably smaller than 20 weigh† percent (w†%), preferably smaller than 15 w†%, preferably smaller than 12 w†%. The solids content and/or the calcium and/or magnesium particles 6 content of the slurry is preferably larger than 1 w†%, preferably larger than 2 w†%, preferably larger than 4 w†%, preferably larger than 5 w†%, preferably larger than 6 w†%, preferably larger than 7 w†%.

In an alternative embodiment, the membrane fouling control additive and/or the calcium and/or magnesium particles 6 are carried to the fouling control means 4, filled into the fouling control means 4, stored in the fouling control means 4 and/or added into the first conduct portion 2 and/or the liquid to be filtered in a dry formulation, i.e. as a powder. It is also possible to mix this powder with a liquid to obtain a slurry between one of the mentioned steps, e.g. before or when filling the powder into the fouling control means 4 to add the membrane fouling control additive and/or the calcium and/or magnesium particles 6 as a slurry into the first conduct portion 2 and/or the liquid to be filtered.

In the above-described embodiment, the membrane fouling control additive and/or the calcium and/or magnesium particles 6 are filled in the fouling control means 4 in the same form as they are entered in the first conduct portion 2 and/or into the liquid to be filtered. In an alternative embodiment, one or more raw materials are added to the fouling control means 4 to obtain the membrane fouling control additive and/or the calcium and/or magnesium particles 6 as entered in the first conduct portion 2 and/or into the liquid to be filtered. This can be simply a mixing step for example to mix the membrane fouling control additive and/or the calcium and/or magnesium particles 6 as a powder with a liquid to obtain the above-mentioned membrane fouling control additive and/or the calcium and/or magnesium particles 6 as a slurry.

In one embodiment, the membrane fouling control additive contains the calcium and/or magnesium particles 6 and a dispersant. The dispersant keeps the calcium and/or magnesium particles 6 from agglomerating. The dispersant can be polyether-polycarboxylate dispersants, polyacrylate dispersants or other dispersants.

In one embodiment, fouling control means 4 comprises a calcium and/or magnesium particles manufacturing device for forming the calcium and/or magnesium particles 6 or the membrane fouling control additive in situ by a chemical reaction. Thus, one or more raw materials are inserted in the fouling control means 4, in particular in the calcium and/or magnesium particles manufacturing device to obtain in situ by the chemical reaction the membrane fouling control additive and/or the calcium and/or magnesium particles 6. The membrane fouling control additive and/or the calcium and/or magnesium particles 6 are further fed into the first conduct portion 2 and/or the liquid to be filtered. Preferably, the chemical reaction comprises a gas-liquid reaction. Preferably, the chemical reaction comprises a precipitation to obtain synthetic precipitate particles as calcium and/or magnesium particles.

Preferably, the chemical reaction and/or the precipitation is a carbonation. The result of the carbonation is preferably a synthetic mineral precipitate. The chemical reaction is preferably realized by introducing gaseous carbon dioxide to a milk of calcium/magnesium hydrates under controlled conditions to obtain the final calcium and/or magnesium particles. Like this a synthetic mineral precipitate can be manufactured. The controlled conditions comprise the flow rates of the carbon dioxide, the temperature, the pressure and/or other conditions. The calcium and/or magnesium particles 6 and/or the additive can be manufactured before adding them into the fluid to be filtered.

One embodiment to manufacture for example precipitated calcium carbonate in situ will be described. In the calcium and/or magnesium particles manufacturing device, a milk of lime (slaked lime or calcium dihydrate) having a solid content comprised between 1 and 15 w†% is set at a temperature of 0 to 20°C. A gas mixture containing gaseous carbon dioxide with a volume fraction of 5 †o 40%vol is injected under pressure of 0.1 †o 0.5 MPa af a gas flow rate comprised between 1 and 5 (normal) Itr CC /min/ltr suspension. The feed of gaseous carbon dioxide is stopped when a pH of 8,3 or lower is obtained. The resulting suspension comprises af leas† a precipitated calcium carbonate content of 5 to 20%wt with respect to the total weight of the slurry and a precipitated calcium carbonate content with respect to the total solid content of 95%wt. The resulting suspension is a slurry of calcium and/or magnesium particles which is then further fed in the first conduit 2.

The system is preferably configured to operate in different modes.

In the operation mode, a transmembrane pressure is created over the membrane 1 and/or a liquid flow of the liquid to be filtered is created from the first conduct portion 2 through the membrane 1 into the second conduct portion 3 to obtain the filtered liquid.

In a cleaning mode, the membrane 1 is cleaned from the dynamic protective layer 7, calcium and/or magnesium particles 6 and/or foulants 10. Preferably, the cleaning is realized by a backwash (also called backflush), i.e. the transmembrane pressure and/or the flow direction over the membrane 1 is inverted. Preferably, the filtered liquid is conducted back through the membrane 1 to clean the membrane 1 . However, it is also possible to use different cleaning mechanisms for cleaning the membrane 1 from the dynamic protective layer 7, calcium and/or magnesium particles 6 and/or foulants 10. This could be other cleaning mechanisms as the ones mentioned in the introduction. Preferably, the cleaning outlet 9 is opened during the cleaning mode to conduct the cleaning liquid comprising the dynamic protective layer 7, calcium and/or magnesium particles 6 and/or foulants 10 through the cleaning outlet 9 out of the first conduct portion 2. The cleaning outlet control means preferably closes the membrane 1 from the source of the liquid to be filtered so that the liquid to be filtered is not polluted even more. The cleaning outlet control means is preferably configured to close the cleaning outlet 9 during the operation mode. However, in another embodiment it is also possible to conduct the cleaning liquid in the cleaning mode back into the first conduct portion 2. In this case, the cleaning outlet 9 and the cleaning outlet control means are not necessary. The system is preferably configured to run different filter cycles. Each filter cycle comprises an operation mode and a cleaning mode.

Subsequently, the process for membrane fouling control is explained at the example steps shown in Fig. 2 to 5 which represent one exemplary filter cycle. Subsequently, the calcium and/or magnesium particles 6 and the membrane fouling control additive are for the sake of brevity no† mentioned each time as alternatives explicitly, bu† it shall be clarified †ha† they are always interchangeable in the subsequent †ex†.

Fig. 2 shows the step of adding the calcium and/or magnesium particles 6 into the firs† conduct portion 2 and/or into the liquid †o be filtered. During this step, the system is preferably in the operation mode. During this step, the liquid †o be filtered is preferably conducted through the membrane 1 . During this step, the liquid in the firs† conduct portion 2 is conducted through the membrane 1 into the second conduct portion 3. Preferably, the calcium and/or magnesium particles 6 are added such †ha† they are well distributed, preferably equally distributed in the liquid †o be filtered, when it arrives a† the membrane 1 . This can be achieved by adding the calcium and/or magnesium particles 6, where there is a non-laminar flow or where there are turbulences, e.g. in or close †o a pump. As described before, preferably the calcium and/or magnesium particles 6 are added in the operation mode, i.e. when the liquid †o be filtered is conducted through the membrane 1. Plowever, it may also be possible †o add the calcium and/or magnesium particles 6 partly or completely, before starting the flow through the membrane 1. Preferably, the calcium and/or magnesium particles 6 are added a† the beginning of the filter cycle, preferably a† the beginning of the operation mode †o create immediately the dynamic protective layer 7 on the membrane 1 before too much foulants can come into contact with the membrane 1 . The calcium and/or magnesium particles 6 are preferably added within a period of the firs† 120 seconds (sec), preferably of the firs† 90 sec, preferably of the firs† 60 sec of the filtration cycle. Preferably, the adding of the calcium and/or magnesium particles 6 and/or of the membrane fouling control additive is interrupted after a sufficient amount of calcium and/or magnesium particles 6 have been added †o create the dynamic protective layer 7. E.g. after 30 to 120 sec depending on the concentration of the calcium and/or magnesium particles 6 added into the liquid to be filtered. For some applications, if showed advantageous †o star† again †o add calcium and/or magnesium particles 6 in the operation mode after the adding has been interrupted for a while (no† shown in the figures), e.g. a† the end of the operation mode. This helped in some applications for the removal of the dynamic protective layer 7 from the membrane 1 in the cleaning mode. In some applications, it showed beneficial †o continuously add the calcium and/or magnesium particles 6 during the whole operation mode.

Fig. 3 shows a step of forming the dynamic protective layer 7 on the membrane 7. The dynamic protective layer 7 is formed by the calcium and/or magnesium particles 6 added into the liquid †o be filtered. The flow of the liquid †o be filtered transports the calcium and/or magnesium particles 6 to the membrane 1 which retains the calcium and/or magnesium particles 6 such †ha† they form the dynamic protective layer 7 similar †o a filtration cake (see also Fig. 8). The dynamic protective layer 7 is formed and/or held in place by the transmembrane pressure created by the flow of the liquid †o be filtered through the membrane 1 . The calcium and/or magnesium particles 6 are configured †o form on the membrane 1 a porous dynamic protective layer 7 so †ha† the liquid †o be filtered can continue †o flow through the dynamic protective layer 7 to the membrane 1 and then through the membrane 1.

The step of adding the calcium and/or magnesium particles 6 and the step of forming the dynamic protective layer 7 happen preferably a† leas† partly overlapping. While the calcium and/or magnesium particles 6 are added in the flow of the liquid †o be filtered, the dynamic protective layer 7 is formed on the membrane 1 by the liquid flow and/or the transmembrane pressure. The dynamic protective layer 7 is formed on the firs† side 1 .1 of the membrane 1 . The period of adding the calcium and/or magnesium particles 6 and of forming the dynamic protective layer 7 can maybe divided in three periods. A firs† time period in which the calcium and/or magnesium particles 6 are added into the liquid †o be filtered and the calcium and/or magnesium particles 6 added are transported †o the membrane 1. A second time period in which the calcium and/or magnesium particles 6 are continued †o be added into the liquid †o be filtered and the dynamic protective layer 7 starts †o be formed on the membrane 1 by the calcium and/or magnesium particles 6 arriving on the membrane 1 . In this second time period, the dynamic protective layer 7 continues to grow by the newly added calcium and/or magnesium particles 6. A third time period, in which the addition of the calcium and/or magnesium particles 6 is stopped and the forming of the dynamic protective layer 7 is finished while the calcium and/or magnesium particles 6 remained in the liquid †o be filtered are transported †o the membrane 1 †o be deposited on the dynamic protective layer 7. The third time period is obviously omitted, if the calcium and/or magnesium particles 6 are continuously added over the complete time of the operation mode. If is also understood that the calcium and/or magnesium particles 6 can also deposit no† only on the membrane 1 , bu† also on the walls of the firs† conduct portion 2. Later it will however be described how this can be avoided or reduced best.

The liquid †o be filtered and/or the liquid in the firs† conduct portion

2 into which the calcium and/or magnesium particles 6 are added is preferably the same liquid †o be filtered containing the foulants and/or contaminants filtered in the step shown in Fig. 4. This has the advantage †ha† no clean liquid needs †o be wasted for depositing/forming the dynamic protective layer 7. However, for certain cases like for very aggressive contaminants and/or foulants which shall a† all costs no† touch the membrane 1 , it would also be possible †o use in/during the s†ep(s) of adding the calcium and/or magnesium particles 6 and/or of forming the dynamic protective layer 7 on the membrane 1 for the liquid †o be filtered a different liquid than in the step of filtering the liquid †o be filtered shown in Fig. 4. This has the advantage †ha† a very clean dynamic protective layer 7 can be formed without any foulants and/or contaminants in between. However, this makes the system for membrane fouling control more complex and increases the waste of clean liquid, and thus in many cases reduces the amount of clean liquid produced by the process.

Fig. 4 shows the step of filtering the liquid †o be filtered through the membrane 1 †o obtain the filtered liquid and/or the step of conducting the liquid †o be filtered from the firs† conduct portion 2 through the membrane 2 †o the second conduct portion 3. The liquid †o be filtered needs †o traverse firs† the dynamic protective layer 7 before it traverses the membrane 1. The foulants 10 in the liquid †o be filtered are retained by the dynamic protective layer 7 before reaching the membrane 1 and thus preventing membrane fouling of the membrane 1 . If is obviously possible that a certain percentage of foulanfs 10 could pass the dynamic protective layer 7, but in any case, the foulanfs reaching the membrane 1 is significantly reduced so that membrane fouling is reduced or avoided. The dynamic protective layer 7 can also retain the contaminants of the liquid. The dynamic protective layer 7 can thus be considered like a filtration aid which helps the membrane 1 †o filter the liquid †o be filtered. The longer the operation mode lasts the more foulanfs 10 and/or contaminants are retained in the dynamic protective layer 7. Thus, the filtration cake created by the dynamic protective layer 17, the foulanfs 10 and/or the contaminants grow over time. This reduces the flow rate of the liquid †o be filtered through the membrane 1 and/or increases the amount of energy necessary †o maintain the same flow rate.

Fig. 5 shows a cleaning step †o clean the dynamic protective layer 7 with the retained foulanfs 10 and/or contaminants. The cleaning step removes/cleans the filtration cake created by the dynamic protective layer 17, the foulanfs 10 and/or the contaminants. Preferably, the cleaning step creates a cleaning flow of the liquid in the firs† conduct portion 2 with the filtration cake ou† of the firs† conduct portion 2 into a cleaning outlet 9. This cleaning flow is preferably realized by a backwash or backwash operation as shown in Fig. 5 which conducts the filtered liquid from the second conduct portion 3 back through the membrane 1 into the firs† conduct portion 2. One of the selection criteria for the calcium and/or magnesium particles 6 is the easy cleanability from the membrane 1. Preferably, the cleaning operation is done by a physical cleaning like the backwash. Plowever, it is also possible †o use a chemical cleaning step. Most preferably, the cleaning step of the filter cycles are physical cleaning steps and once in a while a chemical cleaning step is used †o remove some irreversible fouling created notwithstanding the dynamic protective layer 7. Before starting the cleaning s†ep/mode, the operation mode is preferably ended and the liquid †o be filtered is no† conducted any more from the firs† conduct portion 2 †o the second conduct portion 3. After the cleaning s†ep/mode has ended, the cleaning outlet 9 is closed. In another less preferred embodiment, it is also possible †o conduct the filtration cake back into the firs† conduct portion 2 instead of ou† of the cleaning outlet 9. This would avoid the cleaning outlet 9 but would worsen the quality of the liquid †o be filtered in the firs† conduct portion 2. After the cleaning s†ep/mode a new filter cycle can star†.

Preferably, the system or process comprises a plurality of subsequent filter cycles as described above.

In one embodiment, the process could comprise the optional step of forming the calcium and/or magnesium particles in situ, before adding them †o the liquid †o be filtered. This could be achieved by a chemical reaction as explained in more detail above.

In one embodiment, the ultrafine synthetic mineral precipitate particles based on calcium and/or magnesium having a particle size distribution dso lower than or equal †o 1 m, d9o lower than or equal †o 10 pm and a dio comprised between 50 nm and 500 nm. The ultrafine mineral precipitate particles according †o the present invention are particularly suitable for ul†rafil†ra†ion.

In one embodiment, the microfine synthetic mineral precipitate particles based on calcium and/or magnesium having a particle size distribution dso lower than or equal †o 5 pm, d9o lower than or equal †o 15 pm and a dio comprised between 200 nm and 3 pm. The microfine mineral precipitate particles according †o the present invention are particularly suitable for microfil†ra†ion.

In one embodiment, the dio of the particle size distribution of the calcium and/or magnesium particles for ul†rafil†ra†ion is greater than or equal †o 50 nm, preferably greater than or equal †o 70 nm, preferably greater than or equal †o 100 nm, preferably greater than or equal †o 200 nm.

In one embodiment, the dio of the particle size distribution of the calcium and/or magnesium particles for ul†rafil†ra†ion is lower than or equal †o 500 nm, preferably lower than or equal †o 400 nm, preferably lower than or equal †o 350 nm, preferably greater than or equal †o 320 nm.

In one embodiment, the dio of the particle size distribution of the calcium and/or magnesium particles for microfil†ra†ion is greater than or equal †o 200 nm, preferably greater than or equal †o 250 nm, preferably greater than or equal †o 300 nm, preferably greater than or equal †o 500 nm.

In one embodiment, the dio of the particle size distribution of the calcium and/or magnesium particles for microfil†ra†ion is lower than or equal †o 1000 nm, preferably lower than or equal †o 900 nm, preferably lower than or equal †o 800 nm, preferably lower than or equal †o 750 nm.

The smaller the particles are, the more homogenous is the deposition of the calcium and/or magnesium particles 6 †o form the dynamic protective layer 7. Preferably, the average particle size of the calcium and/or magnesium particles 6 is larger than 0.05 miti, preferably than 0.1 miti, preferably than 0.15 miti. However, if showed that the average particle size is actually less important than the boundaries of the size distribution like the dio and the d9o. This is because particles at this average size dimensions tend to have a very broad size distribution. Without controlling the size distribution, the average particle size might not help to reach the function of the aimed average size. Fig. 6 shows an exemplary size distribution with dio and the d9o for the calcium and/or magnesium particles 6.

The calcium and/or magnesium particles 6 have preferably a size distribution with a dio larger than the defined pore size of the membrane 1 , preferably than twice the defined pore size of the membrane 1 , preferably than three times the defined pore size of the membrane 1 , preferably than five times the defined pore size of the membrane 1 , preferably than seven times the defined pore size of the membrane 1 , preferably than nine times the defined pore size of the membrane 1 , preferably than ten times the defined pore size of the membrane 1. Fig. 7 and 8 show for illustration purposes a schematic membrane 1 with one single defined pore size and one single fouling control particle 6 size, i.e. with a Dirac distribution of the pore size and of the fouling control particle size. In Fig. 7, it is shown what happens, if the fouling control particle 6 size is smaller than the defined pore size. The calcium and/or magnesium particles 6 enter the pores 5 of the membrane 1 and deposit in the pores 5. This narrows the pores 5 of the membrane 1 and negatively influences the flow of the liquid to be filtered through the membrane 1. This is also called standard blocking in membrane technology. Fig. 8 shows schematically a dynamic protective layer 7 on the membrane 1 which does not block the pores 5, because the calcium and/or magnesium particles 6 do not lead to standard blocking in the pores 5. The calcium and/or magnesium particles 6 have preferably a size distribution with a d9o smaller than 10 miti, preferably than 5 miti, preferably than 1 miti, preferably than 0.5 miti.

More preferably, in one embodiment, depending on the microfilfrafion membrane, the d9o of the size distribution of the calcium and/or magnesium particles is lower than or equal †o 15 miti, preferably lower than or equal †o 10 miti, preferably lower than or equal †o 5 miti, preferably lower than or equal to 1 miti, preferably lower than or equal †o 0.7 miti, preferably lower than or equal to 0.5 miti.

In one embodiment, depending on the ulfrafilfrafion membrane, the d9o of the size distribution of the calcium and/or magnesium particles is lower than or equal †o 10 miti, preferably lower than or equal †o 5 miti, preferably lower than or equal †o 1 miti, preferably lower than or equal †o 0.7 miti, preferably lower than or equal to 0.5 miti.

This helps †o guarantee that the calcium and/or magnesium particles 6 form the dynamic protective layer 7 homogenously on the membrane surface. If the d9o of the calcium and/or magnesium particles 6 becomes too large, the deposition of the calcium and/or magnesium particles 6 becomes too heterogenous. This disturbs the functioning of the dynamic protective layer 7 and could reduce the operative membrane surface. For example, hollow fiber membranes 1 or tubular membranes 1 provide the membrane surface 1.1 within long hollow fibers or tubes with the pores 5 in the fiber or tube walls. If the calcium and/or magnesium particles 6 have a d9o larger than mentioned above, if leads †o a heterogeneous deposition of the calcium and/or magnesium particles 6 with the larger particles being mainly concentrated of the end of the tubes or fibers. This leads †o an insufficient creation of the dynamic protective layer 7 for effective fouling control of the beginning of the tubes or fibers and a dynamic protective layer 7 of the end of the tubes or fibers which is too thick and blocks the flow through the tubes or fibers and/or through the dynamic protective layer 7 to the pores 5. A small average particle size of the calcium and/or magnesium particles 6 alone does no† guarantee a homogenous deposition. If the calcium and/or magnesium particles have a wide size distribution or a large d9o they tend †o a heterogenous deposition of the calcium and/or magnesium particles 6. A dynamic protective layer 7 made out of heterogeneously deposited calcium and/or magnesium particles 6 blocks the pores 5 of the membrane 1 and reduce the efficiency of the fouling control. Therefore, the calcium and/or magnesium particles 6 having a d9o below the above-mentioned values, will significantly improve the qualify and performance of the dynamic protective layer 7 for the membrane fouling control of the membrane 1 .

The calcium and/or magnesium particles 6 are preferably synthetic precipitate particles. Synthetic precipitate particles have the advantage that their characteristics like size, thermal stability, mechanical stability, polydispersify index, zefa potential, BET surface are and/or maybe others can be well controlled when they are manufactured.

Preferably, the calcium and/or magnesium particles 6 and/or the synthetic precipitate particles comprise/are metal precipitate particles and/or mineral precipitate particles. Preferably, the calcium and/or magnesium particles 6 and/or the synthetic mineral precipitate particles are based on calcium and/or magnesium. This family of mineral precipitates offers tunable and environmentally compatible particles.

In one embodiment, the calcium and/or magnesium particles 6 and/or the synthetic mineral precipitate particles comprise/are precipitated calcium carbonate (PCC).

In one embodiment, the calcium and/or magnesium particles 6 and/or the synthetic mineral precipitate particles comprise/are precipitated hydromagnesife.

In one embodiment, the calcium and/or magnesium particles 6 and/or the synthetic mineral precipitate particles comprise/are mineral composite particles containing a mixed solid phase of two minerals, preferably of calcium and of magnesium, preferably of synthetic calcium carbonate and of synthetic magnesium carbonate, preferably of (synthetic precipitated) calcium carbonate and of (synthetic) hydromagnesife.

Synthetic calcium/magnesium carbonate originates from natural calcium/magnesium carbonates which have been calcined †o produce calcium/magnesium oxide by removing CO2 during calcination. Natural calcium carbonate is generally called limestone. Natural magnesium carbonate is generally called magnesite, but can exist under several hydrated forms. Natural calcium-magnesium carbonate is generally called dolomite. When the calcium/magnesium oxide is produced by calcination, if can be further hydrated †o form a hydrated form or a slaked form, which can be further carbonated †o form precipitates of calcium/magnesium carbonates (under pressure or no†). Typically, the molar ratio of calcium †o magnesium in natural dolomite is comprised between 0.8 and 1 .2. This ratio can be modified by slaking the oxide in the presence of magnesium hydroxide or calcium hydroxide †o about 0.1 to 10.

The precipitated calcium/magnesium carbonate can also contain impurities coming from the naturally occurring form. The impurities notably comprise all those which are encountered in natural limestones and dolomites, such as clays of the silico-aluminate type, silica, impurities based on iron or manganese a† an amount of maximum 10 w†%, preferably maximum 5 w†%, preferably maximum 1 w†%.

Generally, the CaC03, MgCOs, Ca(OH)2 and Mg(OH)2 contents in calcium-magnesium compounds may easily be determined with conventional methods.

The composite particle comprises preferably a core of the firs† mineral material and/ora shell of the second material. Preferably the firs† mineral material is a mineral precipitate, preferably a calcium-based mineral, preferably a PCC.

The PCC core comprises preferably mainly calcite with some trace of aragonite. Calcite and aragonite are two crystalline structures of calcium carbonate CaC03.

In another embodiment of the invention, the PCC core may further contain calcium hydroxide Ca(OH) 2 , e.g. in por†landi†e form. I† is †o be noted †ha† the PCC core always has a majority weigh† proportion of calcite, aragonite or a mixture of both.

PCC can be obtained by controlled carbonation of quicklime as explained before and the literature reported some carbonation of a milk of lime in presence of additives, such as for example for paper manufacturing (see US20189170765). Preferably, the second mineral material is a magnesium-based material, preferably comprises hydromagnesite, preferably hydromagnesite with only some traces of nesquehonite. Hydromagnesite, also called basic magnesium carbonate has the formula Mgs(C03)4(0H)24H20 or 4MgCC> ₃ -Mg(0H)2-4H20. In the International Centre for Diffraction Data database (ICDD), hydromagnesite corresponds to the datasheets carrying references 00-025-0513 (monoclinic) or 01-070-1 177 (orthorhombic). In one embodiment, the monoclinic hydromagnesite is used. In another embodiment, the orthorhombic hydromagnesite is used. Hydromagnesite must not be confused with magnesite, a magnesium carbonate of formula MgCCb, or with nesquehonite, a hydrated magnesium carbonate of formula MgCC> ₃ .3H20, which are to be avoided according to the invention.

In one embodiment of the invention, the second mineral material may further contain periclase MgO and/or brucite Mg(OH)2 . The proportions of these different components in the second mineral material of the invention, in addition to the synthetic calcium carbonates and hydromagnesite, can be related to the operating conditions and properties of the hydrated dolomite used for the carbonation method allowing the obtaining of the mixed solid phase in the second mineral material according to the invention described below. Fully hydrated dolomite (hydrated under pressure to prevent the residual presence of MgO in the hydrated dolomite) will lead to higher contents of Mg(OH) 2 in the mixed solid phase of the second mineral material according to the invention than if a partially hydrated dolomite is used, the latter possibly leading to the presence of MgO in the mixed solid phase of the second mineral material according to the invention.

An example about a way to produce a mineral composition particles containing a mixed solid phase of calcium carbonate and of magnesium carbonate is described in detail in WO2013/139957 and WO2015/039994 which are incorporated by reference in here. According to the state-of-the-art documents, the synthetic mineral precipitate does not present the particle size distribution making the synthetic mineral precipitate suitable for fouling control of membrane and providing the aforementioned advantages. However, according to the present invention, the manufacturing process is adapted by controlling the growth of the mixed solid phase of calcium carbonate in such a way if shows a particle size distribution characterized by dso < 5 miti, d9o < 15 miti and a dio comprised between 200 nm and 3 miti. Optionally, grinding or sieving step can be performed †o remove too large particles or concentrate in very fine particles.

Preferably, the calcium and/or magnesium particles containing a mixed solid phase (of calcium carbonate and of magnesium carbonate) comprise PCC as one of the two minerals or the firs† of the two minerals.

Preferably, the calcium and/or magnesium particles containing a mixed solid phase (of calcium carbonate and of magnesium carbonate) comprise hydromagnesite as one of the two minerals or the second of the two minerals.

Such a mixed solid phase has a further unexpected advantage in †ha† its specific surface area is larger than †ha† of usual synthetic calcium carbonates which have a specific surface area in the order of 4 †o 15 m ² /g. In one advantageous embodiment of the present invention, the mixed solid phase has a specific surface area of 15 m ² /g or larger, more particularly larger than 20 m ² /g and preferably 25 m ² /g or larger, possibly reaching 35 m ² /g. By “specific surface area” used in the present invention is mean† the specific surface area measured by manometric nitrogen adsorption and calculated using the Brunauer, Emme†† and Teller model (BET method) after degassing a† 190° C.

The mixed solid phase present preferably a bulk density equal †o or lower than 250 kg/m ³ and equal †o or higher than 80 kg/m ³ measured in accordance with standard EN 459.2

The calcium and/or magnesium particles are in one embodiment inert, like for PCC or precipitated hydromagnesite or for other inert mineral precipitates. In another embodiment, the calcium and/or magnesium particles are reactive or functionalized. The composite particles described before are preferably functionalized.

The proposed solutions targe† †o enhance the hydraulic performance and rejection capacity of membrane 1 by applying a dynamic protective layer on the surface of the membrane 1 a† each filtration cycle, preferably a† its star†. The dynamic protective layer will ac† as a filtration aid †o help control flow through the membranes. The dynamic protective layer consists of preferably of particles of ulfrafine precipitated calcium carbonate (i.e., nano- PCC). The dynamic protective layer forms a porous layer on the membrane surface and becomes the filtering medium that traps the solids and prevents them from fouling the membrane surface and pores. Dynamic protective layer filtration is principally mechanical, no† chemical in nature; particles interlace and overlay leaving a large network of interstitial void space †ha† can allow the flow of water.

The proposed invention of membrane pre†rea†men† was tested against water pre†rea†men† based on a coagulation solution (i.e., using FeCb for pre†rea†men† of the liquid †o be filtered). For all experiments, canal water collected from Arquennes was used as liquid †o be filtered. A lab-scale membrane module based on an ul†rafil†ra†ion membrane 1 was used for the testing. The molecular weigh† cut-off (MWCO) of the membrane 1 used was 150 kDa, i.e. a defined pore size of 0.02 m. The surface area for filtration was approximately 0.08 m ² . Filtration was performed in a dead-end mode for 10 consecutive cycles of filtration and backwash. Filtration flux was se† a† a constant rate of 190 dm ³ /(m ² .h) (dm ³ per square meter per hour) for a duration of 15 minutes per cycle. After each filtration cycle, a backwash step was performed a† a flux of 250 dm ³ /(m ² .h) for 1 minute. Fig. 9 shows the result of the coagulation solution, while Fig. 10 shows the result of the solution proposed by the present invention. The transmembrane pressure (TMP) was recorded throughout the consecutive cycles as an operational indicator of membrane performance and is plotted in Fig. 9 and 10. The second Y-axis on the right shows the TMP in bar (lower curve). A† the end of 10 consecutive cycles of filtration and backwash, a chemically enhanced backwash step was performed using caustic soda, oxidant and acid, followed by rinsing with demineralized water. Each chemical was introduced into the membrane system a† 250 dm ³ /(m ² .h) for a duration of 1 minute, followed by soaking for 10 minutes and rinsing with demineralized water for 1 minute. The firs† Y-axis on the let† shows the feed volume per hour in dm ³ /h (or litres per hour - l/h upper curve).

The coagulant solution was established by treating the liquid †o be filtered with an equivalent concentration of 1 mgFe ³⁺ /dm ³ . The coagulant was added †o the feed tank and mixing was applied a† 350 rpm for 1 minute to ensure rapid dispersion of coagulant and destabilization of colloids / particulate matter in the liquid †o be filtered. Thereafter a slow mixing intensify of 50 rpm was applied for the remaining duration of the testing. Results are illustrated in Fig. 10. The TMP begins a† 0.35 bar a† the star† of the experiment and climbs up †o a value of 0.85 bar after 10 consecutive cycles of filtration and backwash. Hydraulic backwashing is no† sufficient †o recover the TMP a† the end of each cycle.

The experiment based on the solution according †o the invention was performed using as calcium and/or magnesium particles 6 ultrafine PCC of a dio of the size distribution of the calcium and/or magnesium particles of 0.08 m and a d9o of the size distribution of the calcium and/or magnesium particles 6 of 0.15 pm. The PCC stock solution was diluted †o a solids concentration of 0.5% and stirred continuously a† 50 rpm †o avoid sedimentation. The calcium and/or magnesium particles 6 were introduced into liquid †o be filtered a† the star† of each filtration cycle via the feed line. Thereafter, canal water was filtered through the membrane for 15 minutes and the cycle was ended by a backwash †o remove the PCC dynamic protective layer 7 and the foulants 10. Chemically enhanced backwash was performed after 10 filtration cycles. Results are given in Fig. 10. The TMP is approximately constant throughout the duration of the experiment a† 0.45 bar. Pressure increase during a filtration cycle is minimal and the permeability can almost be completely recovered a† the end of the cycle. This allows †o make significantly longer filtration cycles and †o significantly increase the number of filtration cycles with a physical cleaning before a chemical cleaning is needed. This reduces power needed for establishing the TMP, increases operation time of the membrane 1 and reduces the amount of chemicals needed for cleaning.

Another experiment was performed with the following experimental protocol. Surface water membrane filtration experiments were performed on an automated pilot, where feed, permeate, re†en†a†e and backwash valves and pumps are automatically switched. Continuous readings of flow rates and pressures were performed. An ul†rafil†ra†ion membrane module with a total surface of 0.8 m ² was used for these experiments with a defined pore size of 30 nm. The quality of the surface water used has been characterized by a pH of 8, a conductivity of 670 m$/ati, a DOC of 9 mg/L, a TOC of 12 mg/L, a Ca ²⁺ of 97 mg/L, a Mg ²⁺ of 10 mg/L and an Alkalinity of 5 mmol/L. A total of 7 cycles was performed per PCC product (315 min). With the pressure readings, trans membrane pressure (TMP) was calculated, as well as permeability, K (Flux = 82.5 L/m ² .h for these experiments). Permeability variation across the †es† (dK/dt) was measured for a duration of 6 cycles (270 min), taking permeability a† the beginning of cycles 1 and 7. Each filter cycle comprises around 45 min of operation mode and subsequently a cleaning mode. Considering a critical permeability value ( Kcnticai) of 150 L/m ² h (value below which a chemical backwash is needed †o continue operation), the chemically enhanced backwash frequency (CEBF) was calculated. The higher this value, the better the filtration performance.

Trans-membrane pressure: TMP [bar] = P _eed - P, permeate

L Flux

Permeability: K [ Q m ² .h.bar] ⁱ = TMP AxTMP

K initial ~ K critical

CEBF [h] = dK_ dt

These parameters have been used for the following four measurements.

In a first measurement, a PCC suspension af 0.6 w†% solid content was prepared, put under agitation, and injected onto the membrane surface with a dosing of 25 mg solids/L of filtered wafer af the beginning of the operation mode of the filter cycle, resulting in a PCC dynamic protective layer 7 thickness of 0.58 miti (firs† portion of the operation mode). This was followed by a surface water filtration cycle of 45 min (second portion of the operation mode), a† the end of which the membrane was backwashed with permeate water (cleaning mode). A† †ha† moment, the PCC layer (sacrificial layer) was washed ou† together with any contaminants. The PCC particles used had a dlO of 0.05 miti, a d50 of 0.07 miti and a d90 of 0.12 miti.

In a second measurement, the same PCC suspension was mixed a† the beginning in the feed water corresponding †o the same amount of membrane fouling control additive added less concentrated and more continuously over the complete operation mode as it is done in the state of the ar† for the purpose of coagulation OR feed water conditioning (no† shown). In a third measurement, the same experiment has been performed without any fouling control particles added.

In a fourth measurement, a sfafe-of-fhe-arf PCC has been used injected on the membrane 1 as in the firs† measurement (no† shown). Fig. 12 shows the TMP 16 over time for the firs† measurement and the TMP 15 over time for the third measurement. I† can be clearly seen †ha† without the PCC dynamic protective layer 7, the TMP of the third measurement 15 increases much more each cycle than the TMP of the firs† measurement 16 with the dynamic protective layer 7. The CEBF frequency was calculated for the measurement 16 with 29 hours while it was only 8-9 hours for the third measurement 15. Also, the CEBF for the second measurement was only 15h. That means †ha† the continuous injection of the fouling control particles has a reduced effect compared †o a concentrated injection a† the beginning of the operation mode. The state-oMhe-ar† PCC showed only a CEBF of 1 1 h and performed thus much worse than the PCC in the inventive value ranges of the particle distribution.

I† should be understood †ha† the present invention is no† limited †o the described embodiments and †ha† variations can be applied without going outside of the scope of the claims.