WO2014038935A1

WO2014038935A1 - Purification of aqueous liquids

Info

Publication number: WO2014038935A1
Application number: PCT/NL2013/050636
Authority: WO
Inventors: Hans-Jürgen RAPP; Vinodnarain Veedprakash Bhikhi
Original assignee: Fujifilm Manufacturing Europe B.V.; Deukum Gmbh
Priority date: 2012-09-04
Filing date: 2013-09-04
Publication date: 2014-03-13

Abstract

A process for purifying an aqueous feed liquid comprising chloride ions, comprising the steps of: a) performing a first electrodialysis step wherein the feed liquid is passed through diluate and concentrate compartments of leading ED module(s) to give a first diluate having a reduced chloride content and a first concentrate having a chloride content greater than that of the feed liquid; b) performing a second electrodialysis step wherein the said first diluate is passed through diluate compartments of following ED module(s) to give a second diluate having a further reduced chloride content and a second concentrate having a chloride content greater than that of the first diluate; c) contacting the second concentrate with an electrode to generate chlorine and/or hypochlorous acid; and d) disinfecting at least a part of the feed liquid using the chlorine and/or hypochlorous acid generated in step c); wherein the leading and following ED modules are connected in series.

Description

Purification of Aqueous Liquids

This invention relates to a process for purifying aqueous liquids comprising chloride ions, e.g. desalination of brackish water or sea water.

Most of the world's water contains such high levels of salts that it is not suitable for human or animal consumption, or for irrigation, without purification. In many parts of the world there is a great need for fresh water.

Fresh water may be obtained from salty water by a process known as desalination. The choice of the desalination process used in a particular place depends on many factors, including the amount of salts present in the water, the amount of purified water needed and the form and availability of energy for performing the desalination process. Known desalination processes include reverse osmosis (RO) and electrodialysis (ED).

Currently RO is the favourite process for large scale desalination; for small scale desalination RO is not energetically favourable.

ED is generally used for purifying brackish water and waste water, but it is often considered too expensive for the purification or seawater due to the electrical energy needed to transport ions through the membranes.

We have now devised a process for purifying aqueous liquids comprising chloride ions (e.g. brackish, sea and/or waste water) which can be used in places where good quality water is in short supply.

According to the present invention there is provided a process for purifying an aqueous feed liquid comprising chloride ions, comprising the steps of:

a) performing a first electrodialysis step wherein the feed liquid is passed through diluate and concentrate compartments of leading ED module(s) to give a first diluate having a reduced chloride content and a first concentrate having a chloride content greater than that of the feed liquid;

b) performing a second electrodialysis step wherein at least part of the said first diluate is passed through diluate compartments of following ED module(s) to give a second diluate having a further reduced chloride content and a second concentrate having a chloride content greater than that of the first diluate;

c) contacting the second concentrate with an electrode to generate chlorine and/or hypochlorous acid; and

d) disinfecting at least a part of the feed liquid using the chlorine and/or hypochlorous acid generated in step c);

wherein the leading and following ED modules are connected in series. A preferred embodiment comprises passing at least part of the first diluate through concentrate compartments of following ED module(s), although one may use an alternative chloride-containing liquid for this purpose as desired.

The aqueous feed liquid is preferably salty water, e.g. water comprising sodium and/or potassium cations and chloride anions, for example brackish water, waste water or sea water.

The leading and following ED modules are connected in series with each other to enable the output of diluate obtained from the leading ED module(s) to be passed through the following ED module(s) to further reduce its chloride content.

In step a) the concentration chloride ions in the feed liquid may be reduced by recirculating the feed liquid through the first ED device until the desired chloride concentration is reached. To maintain electrical neutrality, cations are simultaneously removed, but through the membrane opposite to the membrane through which the chloride ions pass.

In step a) the first diluate has a reduced concentration of chloride ions relative to the feed liquid.

In step b) the second diluate has a further reduced concentration of chloride ions relative to the feed liquid. Drawings

Fig. 1 is a schematic flow diagram showing an ED module.

Fig. 2 is a schematic flow diagram showing two leading ED modules connected in parallel, in series with one following ED module.

Fig. 3 is a schematic flow diagram of a process according to the present invention.

Description

In Fig 1 , the ED module comprises an anode and a cathode at opposite ends, with alternating anion exchange membranes (having a positive charge) and cation exchange membranes (having a negative charge) in between. Each membrane is water-impermeable, permeable to ions of opposite charge and impermeable to ions of identical charge.

Each membrane and the space between it and the next membrane define a compartment. Compartments which cause the ion concentration of the feed liquid to increase are referred to as concentrate compartments and the compartments which cause the ion concentration of the feed liquid to decrease are referred to as diluate compartments. Two adjacent compartments (i.e. one concentrate compartment and the adjacent diluate compartment) are referred to as a "cell".

In Fig.1 , the feed liquid (typically a stream of salty water) enters the ED module at the top and flows downwards through the five compartments. As the feed liquid flows through the compartments, positively charged ions (e.g. Na⁺) are attracted out of the second and fourth compartments, through the adjacent negatively charged membrane and towards the cathode (-). Similarly, negatively charged ions (e.g. CI^") are attracted out of the second and fourth compartments, through the adjacent positively charged membrane towards the anode. In this way, the feed liquid passing through the second and fourth compartments is depleted in ions to give a "diluate stream". In contrast, the compartments either side (i.e. the first, third and fifth compartments) become enriched in ions (obtained from the diluate stream) to create a

"concentrate stream". In ED modules having the arrangement shown in Fig. 1 , the anions and cations which pass out of the diluate streams and into the concentrate streams are unable to reach the anode or cathode because they are rejected by the identically charged, outer membrane which stands between them and the electrode to which they are attracted.

Other ED module arrangements may also be used, for example one may use

ED modules having more (or less) membranes than the number shown in the schematic arrangement of Fig. 1. to create additional (or fewer) compartments and cells. Each membrane has an opposite charge to the previous membrane, thereby creating an ED module having alternate concentrate and diluate compartments.

Preferred ED modules typically comprise a large number of cells e.g. 50 to

1000 cells, or in some cases more than 1000 cells (each cell comprising a concentrate compartment and a diluate compartment). The number of cells is selected as desired, for example taking account of the desired throughput of the ED module, practical handleability and so forth.

ln_Fig. 2, each D represents a diluate compartment, each C represents a concentrate compartment and each R represents a rinsing compartment housing an electrode (not shown). The diluate streams from the two (parallel) leading ED modules feed into the diluate and the concentrate compartments of the following ED module.

Fig. 3 is described in more detail in the Example.

Although one may use only one leading ED module if desired, typically step a) is performed using two, or more than two, leading ED modules connected in parallel (i.e. with each other), for example as shown schematically in Fig. 2. In this way the productivity of the process is enhanced. Similarly, one may perform step b) using two, or more than two, following ED modules connected in parallel (i.e. with each other).

Preferably the process further comprises the additional step of filtering the feed liquid before it enters the leading ED module. In this way, one may remove unwanted particulate matter, for example algae, bacteria, plant matter, dirt, sand, and other debris, e.g. by sand filtration. If desired one may use a cascade of two or more filters having decreasing molecular weight cut off in order to efficiently remove unwanted matter from the aqueous feed liquid, e.g. one may microfilter (e.g. with a filter having an average pore size >0.1 μηι), ultrafilter (e.g. with a filter having an average pore size of >0.01 μηι to 0.1 μηι) and even nanofilter (e.g. with a filter having an average pore size of 0.01 μηι or less) the aqueous feed liquid before it enters the leading ED device if desired.

For ultrafiltration, polymeric membranes and ceramic membranes are particularly suitable. Suitable materials for UF membranes include polyethersulfone, polyacrylonitrile, polyvinylidenefluoride, aluminia, zirconia, titania and silicon carbide. The ultrafiltration system may for example be a cross flow ultrafiltration system, a dead-end ultrafiltration system or a submerged ultrafiltration system. Examples of ultrafiltration modules include hollow fibre modules, spiral wound modules and plate and frame modules.

The ED modules typically comprise at least one diluate compartment (where the diluate is formed) and at least two concentrate compartments (where the concentrate is formed, either side of the diluate compartment). These compartments typically comprise a wall comprising a cation or anion exchange membrane and one or more spacers which keep the membrane a desired distance from the next membrane while allowing the relevant stream to pass between the two membranes. The spacers therefore provide a gap between the cationic and anionic membranes through which the diluate or concentrate may pass in order to be depleted or enriched with ions. At the start of the process, the feed liquid passes through the diluate and concentrate compartments and the volume ratio of the feed to the diluate and concentrate compartments may be controlled in order to optimise the process.

Preferably each spacer creates a gap between a cation exchange membrane and an anion exchange membrane of 80 to 800 μηι, more preferably 100 to 400 μηι. The gap created by a spacer is preferably large enough to ensure the pressure drop through the module is not too great, while at the same time being small enough to ensure electrical resistance is not too high. In a preferred embodiment, the leading ED module comprises diluate compartment(s) and concentrate compartment(s) and the feed liquid is fed into the diluate and concentrate compartments in a volume ratio of 0.6: 1 to 4:1 respectively, preferably 1 : 1 to 2.5: 1 respectively.

Furthermore, the following ED module comprises diluate compartment(s) and concentrate compartment(s) and preferably the first diluate is fed into the diluate and concentrate compartment(s) in a volume ratio of 2:1 to 20: 1 respectively, more preferably 4: 1 to 10:1 respectively.

Seawater typically has a conductivity of about 54 mS/cm, corresponding to about 35 g/l of total salts, of which sodium chloride is the main component. In step a), the chloride content of the first diluate is preferably reduced to a conductivity level of 20 to 3 mS/cm. In other words, in step a) the total salts content of the first diluate is preferably reduced to about 14 to 2 g/l.

The process is preferably configured to maximise the yield of purified water and to minimise the energy consumption. The optimum conductivity value for the resultant purified water depends, among other things, on the composition of the feed liquid, the membranes used, the configuration of the modules and the applied current density (or voltage per cell). More preferably the conductivity of the first diluate is reduced to a level of 10 to 4 mS/cm or a total salts content of about 6.7 to 2.7 g/l.

In step b) the chloride content of the second diluate is preferably reduced to a value suitable for human consumption, e.g. below 1 mS/cm, preferably below 0.5 mS/cm. Thus in step b) the total salts content of the second diluate is preferably reduced to below 0.6 g/l, more preferably below 0.3 g/l.

Electrical resistance and permselectivity are important membrane parameters to be optimized for optimal functioning of the process. The permselectivity is preferably higher than 90%, more preferably higher than 95%.

In step a), the leading ED module(s) preferably comprise membranes having low electrical resistance. Because the conductivity of the feed liquid is high (corresponding to a low electrical resistance), the electrical resistance of the membrane in the leading ED module(s) is an important contributor to the total resistance. Preferably the leading ED module(s) comprise membranes having an electrical resistance below 5 ohm. cm², more preferably below 3 ohm. cm². Relatively high current densities can be applied to achieve a fast desalination without the risk of water splitting. The process therefore preferably comprises applying a voltage per cell of the leading ED module(s) of 0.1 to 1.8 V/cell, more preferably 0.5 to 1.3 V/cell. Alternatively a fixed current density of 30 to 200 A/m² is applied, more preferably 50 to 120 A/m².

In step b), the conductivity of the first diluate passing through the following ED module(s) is much lower than that of the feed liquid which passed through the leading ED module(s), resulting in a relatively high resistance. To prevent water splitting into hydrogen and oxygen, it is preferred to apply a lower current density in the following ED module(s) than that used in the leading ED module(s). The process therefore preferably comprises applying a voltage per cell of the following ED module(s) of 0.05 to 1.0 V/cell, more preferably 0.1 to 0.8 V/cell. Alternatively a fixed current density of 3 to 30 A/m² is applied, more preferably 5 to 20 A/m². Preferably the current density or voltage per cell of the following ED module(s) is lower than that of the leading ED module(s).

The cation exchange membranes carry a negatively charged group, for example a sulphonate, phosphonate and/or carboxylate group. These membranes allow cations to pass through them but not anions. Suitable membranes and their preparation are described in WO 2007/018421 , WO 2008/016302, W011/027138, W01 1/073639 and W01 1/073637. Such membranes may also be obtained from FUJIFILM.

The anion exchange membranes carry a positively charged group, for example an ammonium and/or quaternary ammonium group. These membranes allow anions to pass through them but not cations. Suitable membranes and their preparation are described in W01 1/073638, W01 1/027138 and W01 1/073641. Such membranes may also be obtained from FUJIFILM.

Optionally the process further comprises the step of recirculating the feed liquid through the leading ED module(s) (e.g. via one or more stock tanks) two or more times (e.g. 2 to 20, preferably 2 to 10 times) before diverting the resultant first diluate into the following ED module(s). Similarly, the process optionally further comprises the step of recirculating the first diluate through the following ED module (e.g. via one or more stock tanks) two or more times (e.g. 2 to 20, preferably 2 to 10 times) before collecting the second diluate.

Recirculation is especially preferred when the process is performed on a relatively small scale, for example when producing < 500 m³ of purified water per day. For larger scale production, the process is preferably performed as a continuous process, e.g. the water to be purified passes through the modules only once (single pass). To enable single pass and achieve the desired degree of desalination one may place a plurality of leading modules in series and a plurality of following modules in series.

Preferably the process is performed such that the feed liquid and the first diluate pass through the relevant modules and along the surface of the membranes at a velocity of 0.1 to 5 cm/s, more preferably 0.5 to 1.5 cm/s.

The rate at which the process purifies the feed liquid depends to some extent on the number of ED modules run in parallel, the number of cells in each module, the surface area of the ion exchange membranes in each module and the properties of the ion exchange membranes.

The purification rate (i.e. m³ of liquid purified per hour) can be increased by operating more than one leading ED module in parallel (e.g. from 2 to 10 leading ED modules in parallel) and by operating more than one following ED module in a parallel (e.g. from 2 to 10 following ED modules in parallel).

One may monitor the chloride content of the diluates and concentrates by standard conductivity techniques, e.g. using an in-line conductivity meter. Typically the conductivity meter is first calibrated using stock solutions of known chloride content.

In step a), when the desired chloride content is reached the first diluate is fed into the following ED module(s) (optionally via one or more stock tanks) and step b) is performed. In one embodiment, the first diluate is fed into both the diluate compartment(s) and the concentrate compartments of the following ED module(s). In this way, the part of the first diluate which passes through the diluate compartment(s) of the following ED module(s) is further purified, whereas the part of the first diluate which passes through the concentrate compartments of the following ED module is 'sacrificial' and becomes less pure. In an alternative embodiment, the first diluate is fed into the diluate compartment(s) of the following ED module(s) and the feed liquid (i.e. that which fed into the leading ED module) is fed into the concentrate modules of the following ED module(s). In this way, the part of the first diluate which passes through the diluate compartment(s) of the following ED module(s) is further purified, whereas the original feed liquid which passes through the concentrate compartments of the following ED module(s) becomes less pure. This latter embodiment is preferred, especially when the feed liquid is brackish water, because it results in a higher yield of purified water. However this latter embodiment is not always possible, e.g. when the feed liquid is sea water its chloride concentration is sometimes too high for it to be of practical use in the concentrate compartments of the following ED module(s). In order to achieve the lower levels of chloride in the following ED module(s), one may use different membranes, spacers and/or use a different current density in the following ED unit compared to the leading ED unit (the latter being preferred).

As the first diluate already has quite low levels of chloride (as a result of step a)), the second concentrate may have a sufficiently low chloride content for a part of it to be recirculated back to the first ED module(s) and used as a part of the first concentrate stream.

At least a part of the second concentrate is used in step c) to generate chlorine and/or hypochlorous acid. Also the complete volume of the second concentrate may be used to generate chlorine and/or hypochlorous acid which may then be used for disinfection.

Preferably the first concentrate is used to produce energy, e.g. by reverse electrodialysis as described below.

Alternatively one may dispose of the first and/or any of the second concentrate not used in step c) (if any), or allow it to dry to produce salt (e.g. NaCI for industrial use or, if sufficiently pure, for domestic use).

The electricity required to perform the electrodialysis steps can be provided by any source, including mains electricity and electricity made by a generator (e.g. a petrol or gas fuelled generator). The presently claimed process is of particular value for providing desalinated water in coastal areas, e.g. in developing countries. Such areas often lack infra-structure, with sporadic or no mains electricity supply. Petrol or gas may be too expensive for the local population to afford, or at times they may be unavailable in remote areas. Therefore the process preferably further comprises the step of generating at least a part of the electricity used in one or both of the electrodialysis steps using a renewable energy source, e.g. using a wind turbine, solar cells, fuel cells and/or reverse electrodialysis.

Hydrogen gas produced in the ED modules may be used to generate electricity, for example in a fuel cell.

An advantage of the current process is that it can be operated with DC power; e.g. there is no need to convert DC power to AC power when solar energy is used.

To provide a continuous operation, the process preferably comprises the storage of electricity in one or more batteries. Thus in periods of low wind or low sunlight, the purification may continue by using electricity generated previously and stored in batteries. Suitable batteries include lithium-based batteries, e.g. lithium ion batteries and lithium-sulphur batteries; lead-acid batteries, especially deep cycle lead- acid batteries of flooded (wet), gelled, and AGM (Absorbed Glass Mat) type; redox flow batteries, e.g. vanadium flow batteries and polysulfide bromide batteries; sodium sulphur batteries and zinc based batteries, e.g. zinc-bromine and zinc-cerium batteries.

As the concentrates have elevated chloride levels, some of the energy used to generate the diluate (and therefore also the concentrate) may be recovered by performing reverse electrodialysis ("RED") using one or more of the concentrates. As described in WO 2010/106356, currently known RED processes are typically used in areas where there are large supplies of salty water (e.g. sea water or brackish water) and less salty water (e.g. brackish water or fresh water) in close proximity. For example, sea water and fresh or brackish water may be passed through an RED module comprising two types of membrane, namely one that is selectively permeable for positive ions and one that is selectively permeable for negative ions. The sea or brackish water loses anions and cations as they pass through a membrane of opposite charge into a flow of less salty water (e.g. fresh water), thereby generating a potential difference that can be utilized directly as electrical energy.

Thus the present process preferably comprises the use of RED to generate at least a part of the electricity used in the first and/or second electrodialysis step, e.g. by a process comprising passing salty water and less salty water through an RED module. In one embodiment the salty water is sea water and the less salty water is brackish water or fresh water. In another embodiment the salty water is brackish water and the less salty water is fresh water. However, the use of fresh water in these embodiments result in wastage of fresh water, which may be acceptable during rainy seasons, but will often run counter to the objective of the present invention.

In a preferred embodiment, the salty water used in the abovementioned RED process is or comprises the first and/or second concentrate. For example, one may pass the first and/or second concentrate through an RED module alongside less salty water such as the feed liquid or rain water. In this way, one recovers some of the energy used to create the concentrate stream, leading to a more efficient process for purifying water.

Thus a preferred process according to the invention further comprises the step of generating electricity by a reverse electrodialysis process comprising the generation of electricity by the transfer of ions from the first concentrate and/or the second concentrate through a membrane to a liquid having a lower concentration of said ions. In this process, the transfer of ions is preferably from:

a) the first concentrate to a liquid having substantially the same composition as the feed liquid, and/or b) the second concentrate to a liquid having a lower concentration of ions than the second concentrate.

If fresh water is available, e.g. rain water, river water or waste water, then such water may be used (sacrificial ly) in a RED module as a dilute stream, receiving ions from the first concentrate and/or the second concentrate. In another embodiment the feed liquid is pre-treated in a RED module to reduce the total salts content of the feed liquid before it is transferred to the leading ED module(s) and thereby electricity is generated by the RED module. In order to make desalinated water potable, it is often necessary (or at least desirable) to disinfect the water. Typically chemical disinfectants have been used in the past. However chemical disinfectants can be expensive and place an additional financial burden on the user, which may be unaffordable in certain regions of the world. Furthermore, chemical disinfectants may at times be unavailable in remote regions, even if they could be afforded.

An advantage of the present process is that steps c) and d) reduce or completely remove the need for chemical disinfectants. Instead one generates chlorine and/or hypochlorous acid from the second concentrate and uses that to disinfect at least a part of the feed liquid, which in turn leads to purified water which is also disinfected.

A particularly useful aspect of the present invention is the production of chlorine and/or hypochlorous acid specifically from the second concentrate which may then be used for disinfection. The chloride content of the first concentrate is generally too high to obtain a suitable level of chlorine, especially when sea water is used as feed liquid; chlorine formation during the first electrodialysis step may be prevented by rinsing the electrode compartments using a liquid low in chloride content (e.g. using sodium sulphate solution). Therefore in a preferred embodiment, the process is performed such that the first concentrate is not contacted with an electrode. Chlorine formation may be achieved during the second electrodialysis step by rinsing the electrode compartments with the second concentrate, thereby continuously generating chlorine. So preferably the process is performed such that most of the chlorine and/or hypochlorous acid generated in the process (e.g. at least 90 vol%, more preferably at least 95 vol%, especially all, of the total amount of chlorine and/or hypochlorous acid generated in the process) is generated in step c). By tuning the parameters that influence the chlorine formation the desired chlorine level for disinfection can be obtained. The amount of chlorine and/or hypochlorous acid present in the purified liquid obtained from the process (e.g. potable water) can be adjusted to the desired level by varying any of a number of parameters. For example, one may adjust the amount of second concentrate that is fed back to the feed liquid or one may adjust the chlorine and/or hypochlorous acid concentration in the second concentrate, e.g. by varying the starting chloride concentration at the second electrodialysis step, the volume ratio of second diluate : second concentrate streams, electrode potential, electrode material etc.

We have found that there is a direct relationship between the chloride concentration at the start of the second electrodialysis step and the amount of chlorine formed during the second electrodialysis step. One may use this relationship in order to control the amount of chlorine generated during the process. The chloride concentration at the start of the second electrodialysis step can be reduced, for instance, by lowering the chloride concentration at the end of the first electrodialysis step (e.g. by continuing the first electrodialysis step until a lower conductivity value is reached) and/or by keeping a certain volume of purified liquid in the tank at the end of the second electrodialysis step.

Typically step d) is performed by re-circulating a part of or all part of the product arising from step c) (which contains chlorine and/or hypochlorous acid) into the feed liquid. The second concentrate (or a part of it) may be used to disinfect the feed liquid before it is filtered or after it is filtered. Disinfecting the feed liquid has the advantage of ensuring the apparatus is disinfected and kept clean during the process.

The chlorine and/or hypochlorous acid may be generated by circulating the second concentrate through electrode compartments of the second ED device, e.g. in parallel or series with the flow of the second concentrate through the concentrate compartments of the following ED module.

Step d) is preferably performed such that one introduces a free concentration of chlorine and hypochlorous acid (recognising that either or both may be present) into the feed liquid of 0.1 to 1.5 mg/l, more preferably 0.3 to 0.8 mg/l, especially about 0.5 mg/l.

One may determine the concentration of chlorine and/or hypochlorous acid introduced into the feed liquid by a (free and/or total) chlorine analyser, e.g. ExStik® CL200 chlorine meter from Extech® Instruments, USA, or by using test strips, e.g. Merckoquant® Chlorine Test from Merck, Germany.

One may also influence the amounts of chlorine (and/or hypochlorous acid) produced by selecting the material used to make the electrode in the following ED module. For example, electrodes made of titanium and mixed oxides produce much more chlorine than electrodes made of, for example, titanium/platinum or diamond. The electrode may be in any suitable form, e.g. plate form, rod form, wire or tubular form, mesh form or cylindrical mesh form.

The type of electrode used in the leading ED module(s) is not particularly critical, especially when the process comprises the step of rinsing the electrode(s) of the leading ED module(s) with a halide-free salt solution (e.g. a sodium sulphate solution). Mixed metal oxide coated titanium electrodes may comprise oxides of Iridium, ruthenium, zirconium, niobium, tantalum, platinum, titanium, or combination of two or more thereof. Suitable electrodes for the leading ED module(s) include ruthenium-titanium coated electrodes, e.g. (TiO₂+40% Ru0₂)-activated perforated Ti electrodes, ruthenium-iridium-titanium coated electrodes, iridium-tantalum-titanium coated electrodes, iridium-tin-titanium coated electrodes, ruthenium-manganese- titanium coated electrodes and iridium-titanium electrodes.

Preferably the following ED module comprises one or more electrodes comprising platinum, diamond, and/or a conductive polymer (e.g. polypyrrole, polyaniline, poly(p-phenylene), poly(p-phenylenesulfide), polynaphtalene, polypyrene, polyfluorene, poly-o-aminophenol, polythiophene, poly(3,4-ethylenedioxythiophene), poly(3-alkylthiophene), poly-acetylene, poly-(phenylenevinylene), polyfuran, polyindole, polyazulene, polyanisidene, polyazepine, polycarbazole and/or other poly(heteroaromatic vinylenes)). Conductive polymeric electrodes are preferably doped, e.g. with iodine, bromine, arsenic pentafluoride, mineral acid or base. The abovementioned electrodes are preferred because they generally provide a stable and suitably low level of chlorine and/or hypochlorous acid for disinfection without compromising the potability of the resultant purified water. Suitable electrodes for the following ED module(s) include Pt-coated perforated Ti electrodes. Also suitable are electrodes coated with noble metals, e.g. gold, palladium, rhodium, ruthenium, iridium and the like, and mixtures comprising two or more thereof, or with doped diamond. The electrode preferably has a titanium core, although niobium, tantalum or molybdenum may also be used.

In one embodiment the following ED module comprises the same (or similar) type of electrode to the electrode used in the leading ED module. This is especially applicable when brackish water is used as feed liquid or when the chloride concentration at the start of the second electrodialysis step is low. Preferably inexpensive electrodes are used in the leading and following ED units.

The following module preferably comprises cation exchange membranes comprising polytetrafluoroethylene (PTFE) or PTFE having anionic groups, especially the membranes forming the electrode compartments. This is because such membranes are particularly resistant to damage from chlorine and hypochlorous acid. Commercially available PTFE-based membranes include Nafion™ membranes from DuPont's and Fumapem™ membranes from Fumatech.

Preferably the modules are constructed from corrosion-resistant materials, for example polyethylene, polypropylene, polymethylmethacrylate, polyethyleneterephthalate, polyethylene (PE), polypropylene (PP), polymethylmethacrylate (PMMA), polyether ether ketone (PEEK), polyethylene terephthalate (PET), composite materials and combinations thereof. The modules preferably comprise a housing made from an electrically non-conductive material as this reduces the risk of users receiving an electric shock.

In a preferred process according to the present invention each ED module comprises the components:

a) an anode,

b) a first cation exchange membrane,

c) a first anion exchange membrane,

d) a second cation exchange membrane,

e) a second anion exchange membrane,

f) optionally a third cation exchange membrane, and

g) a cathode;

wherein said components are arranged in parallel, in the order a) to g), or in the order g) to a); and

wherein each membrane is separated from the next by one or more separators to define a compartment.

To prevent or reduce fouling, scaling and the formation of chlorine and/or hypochlorous acid during step a), the process optionally comprises the step of rinsing the anode and cathode of the leading ED module(s), e.g. using a liquid low in chloride content (e.g. using sodium sulphate solution). A low amount of chloride ions are allowed, e.g. for disinfecting the rinsing liquid.

The liquid used to rinse the electrode can be chosen freely as long as the conductivity of the liquid is high enough to obtain an acceptable electrical conductance, e.g. at least 10 mS/cm. The rinse liquid may be stored in a rinse liquid stock tank and is preferably circulated continuously through the electrode compartments of the leading ED module. The rinse liquid may be fed through the anode compartment first and then through the cathode compartment, then back to a rinse liquid stock tank. In the anode compartment, protons are formed and this can make the rinse liquid acidic; in the cathode compartment, hydroxide ions are formed, neutralizing the protons formed in the anode compartment. The order of rinsing the anode and then the cathode is preferred because this reduces the chance of inorganic salts being precipitated.

The anode and cathode in the following ED module may be rinsed during step c) as described above. .

In one embodiment, the process comprises a switching step comprising reversing the polarity of the electrodes in the leading and/or following ED module and also swapping the diluate and concentrate compartments within an ED module (i.e. so that the diluate compartment is used as a concentrate compartment and vice versa). This has the advantage of reducing membrane scaling and fouling. Preferably the switching step is performed regularly during the process, for example at intervals of from 10 to 120 minutes. When the liquid is recirculated in a batch process the switching step is preferably done in between cycles.

In a preferred embodiment, the feed liquid is passed through the diluate and concentrate compartments of the leading ED module at substantially the same rate (e.g. the same volume per minute +/- 10%). This is because only one motor is then needed to pump the feed liquid into the leading ED module. By analogy, an analogous preference applies to the following ED module too. For devices aimed at purifying on a small scale (e.g. < 20 m³ of purified liquid per day) the cost benefit of only needing one motor for the leading ED device and one motor for the following ED device (which may be the same or different motors) is highly advantageous.

Preferably the duration of the step b) is shorter than the duration of step a), although this is not essential because at the end of step a) one may store the first diluate in a first diluate stock tank if desired until the following ED module is ready to receive the first diluate.

In one embodiment the process is free from the step of bringing the diluates (or any of the diluates) into contact with ion exchange beads.

The process may optionally be further improved by including the step of selectively removing multivalent ions, e.g. calcium, magnesium and/or sulphate ions. By "selectively" we mean without removing significant amounts of monovalent ions. Removal of multivalent ions may reduce the chance of scale formation which could otherwise block the ED modules. One may selectively removing multivalent ions by nanofiltration, ion exchange or by any other method which removes multivalent ions more efficiently than monovalent ions. The ED modules preferably comprise n anion exchange membranes and n or (n+1) cation exchange membranes, wherein n is 50 to 1000, more preferably 100 to 500. This corresponds to 50 to 1000 (or 100 to 500) cells (and twice that number of compartments in the ED module). The membranes and spacers in the ED modules may be placed in a vertical position or in a horizontal position or in a tilted position, i.e. between vertical and horizontal.

The leading and following ED modules optionally comprise an identical number of cells to each other or different number of cells to each other.

Steps a) and b) may be conducted at the same temperature or at different temperatures. Ambient of slightly above ambient temperatures are preferred, e.g. below 70°C, especially 15 to 50°C. Higher temperatures are not preferred because of the cost involved in achieving them and the potentially detrimental effects high temperatures can have on membranes and spacers.

The process preferably also comprises the step of monitoring the chloride and/or hypochlorous acid concentration of the product of step c).

Optionally the quality of the feed liquid at the start of the process is monitored before and/or after the filtering step e.g. to check for chlorophyll, heavy metals and/or biological oxygen demand. Examples - Purification of Sea Water

Leading and a following electrodialysis modules were connected in series as shown in Fig. 3, wherein the various components are as described in Table 1 below:

Table 1

system)

Pre-treatment stock tank S1 was filled with sea water. Unwanted particulate matter was removed from the sea water using the ultrafiltration unit PT and the filtered sea water was stored in the feed liquid stock tank S2.

Flow rates were 2.4 m³/hr for all flows through the ED modules. A voltage of 55.8 V was applied to all modules, corresponding to 0.186 V/cell for step a) and with 0.28 V/cell for step b) of the process. The 'total chlorine' levels (including hypochlorous acid) were measured using an ExStik® CL200 chlorine meter from Extech® Instruments, USA.

Example 1

Step a)

The first diluate stock tank DST1 and the first concentrate stock tank CST1 were each filled with 240 L of feed liquid (UF-treated sea water having a conductivity of 47.5 mS/cm) from feed liquid stock tank S2. The liquid in DST1 was recirculated through the diluate compartments of the leading ED device ED1 (comprising two ED modules) and the liquid in CST1 was recirculated through the concentrate compartments of the leading ED device ED1 to give a first diluate having a conductivity (in DST1) of 5 mS/cm and to give a first concentrate. At this point, recirculation of the liquids through the leading ED device was stopped and 100 L of the first diluate from DST1 was transferred to the second diluate stock tank DST2 and 50 L of the first diluate was transferred to the second concentrate stock tank CST2. The first concentrate was disposed of.

During step a), the electrode compartments of the leading ED device were continually rinsed with a sodium sulphate solution at a flow rate of 2.4 m³/hr.

Steps b) and c)

The liquid present in DST2 was recirculated through the diluate compartment of the following ED device ED2 and the liquid present in CST2 was recirculated through the concentrate compartment of the following ED device ED2 to give a second diluate having a conductivity of 0.5mS/cm. At this point, recirculation of the liquids through the following ED device was stopped and the resultant second diluate was transferred to the product stock tank PST. During step b), the electrode compartments were rinsed with the second concentrate from CST2 whereby the liquid first passed the anode compartment and then the cathode compartment, after which the liquid was fed back to CST2. Chlorine (and/or hypochlorous acid) was formed in the cathode compartment of the following ED module during this step and dissolved in the second concentrate. The resultant second concentrate in CST2 contained 1.2 ppm of chorine.

Step d)

50 L of the second concentrate arising from step c) was fed back to feed liquid stock tank S2 containing feed liquid (700L of UF-treated sea water), thereby disinfecting a subsequent batch of feed liquid. The chlorine content of the feed liquid in S2 became 0.09 ppm.

Example 2

Example 1 was repeated except that the recirculation in step a) was stopped when the conductivity of the first diluate (in DST1) reached 10 mS/cm.

After completion of steps b) and c) the second concentrate contained 1.6 ppm of chorine. 50 L of the second concentrate (from CST2) was fed back to stock tank S2 containing the feed liquid (700 L of UF-treated sea water). The chlorine content of the feed liquid in S2 became 0.12 ppm.

Example 3

Example 1 was repeated except that the recirculation in step a) was stopped when the conductivity of the first diluate (in DST1) reached 15 mS/cm.

After completion of steps b) and c) the second concentrate contained 2.1 ppm of chorine. 50 L of the second concentrate (from CST2) was fed back to stock tank S2 containing the feed liquid (700 L of UF-treated sea water). The chlorine content of the feed liquid in S2 became 0.15 ppm.

The process as described in Examples 1 to 3 was performed in a batch-wise manner, successfully disinfecting the feed liquid in each cycle with a portion of the second concentrate containing chlorine and/or hypochlorous acid obtained from the previous cycle.

In each cycle during step a) the liquid level in feed liquid stock tank S2 was restored to a level of 700 L by adding feed liquid obtained from pre-treatment device PT. After step b) and c) 50 L of second concentrate from CST2 was added to S2 (final volume became 750 L). For step a) DST1 and CST1 were each filled up with feed liquid from S2 up to a level of 240 L. Step a) was performed as described above and was stopped when the conductivity of the first diluate reached 8 mS/cm. For step b) 100 L from DST1 was transferred to DST2 and 50 L from DST1 to CST2. Step b) was performed as described above until the conductivity of the second diluate reached 0.5 mS/cm. At this point, recirculation of the liquids through the following ED device was stopped and 100 L of the resultant second diluate (from DST2) was transferred to the product stock tank PST. 50 L of the resultant second concentrate (from CST2) was transferred to feed liquid stock tank S2 where it contributed to the disinfection of the feed liquid.

The polarity of all of the electrodes was reversed and the flows were changed after every fifth batch, such that the diluate compartments used above were now used as concentrate compartments and vice versa. The above process steps were then repeated (with the second concentrate from each cycle being used to disinfect the feed liquid in the next cycle) until the desired volume of purified water had been obtained.

Claims

1. A process for purifying an aqueous feed liquid comprising chloride ions, comprising the steps of:

b) performing a second electrodialysis step wherein the said first diluate is passed through diluate compartments of following ED module(s) to give a second diluate having a further reduced chloride content and a second concentrate having a chloride content greater than that of the first diluate;

wherein the leading and following ED module(s) are connected in series.

2. A process according to claim 1 wherein the leading ED module(s) comprise two, or more than two, ED modules connected to each other in parallel.

3. A process according to any one of the preceding claims wherein the following ED module(s) comprise two or more ED modules connected to each other in parallel.

4. A process according to any one of the preceding claims which further comprises the step of generating electricity by a reverse electrodialysis process comprising the generation of electricity by the transfer of ions from the first concentrate and/or the second concentrate through a membrane to a liquid having a lower concentration of said ions.

5. A process according to claim 4 wherein the transfer of ions is from:

a) the first concentrate to a liquid having substantially the same composition as the feed liquid, and/or

b) the second concentrate to a liquid having a lower concentration of ions than the second concentrate.

6. A process according to claim 4 or 5 wherein at least a part of the electricity generated by the reverse electrodialysis is used in the first and/or second electrodialysis step.

7. A process according to any one of the previous claims wherein the leading ED module comprises diluate compartment(s) and concentrate compartment(s) and the feed liquid is fed into the diluate and concentrate compartments in a volume ratio of 0.6: 1 to 4:1 respectively.

8. A process according to any one of the previous claims wherein the following ED module comprises diluate compartment(s) and concentrate compartment(s) and the first diluate is fed into the diluate and concentrate compartment(s) in a volume ratio of 2:1 to 20:1 respectively.

9. A process according to any one of the previous claims wherein the following ED module comprises a platinum, diamond and/or a conductive polymer electrode.

10. A process according to any one of the previous claims wherein the feed liquid is brackish water, waste water or sea water.

1 1. A process according to any one of the previous claims which is free from the step of bringing the diluates or any of the diluates into contact with ion exchange beads.

12. A process according to any one of the previous claims which comprises the further step of filtering the feed liquid to remove unwanted particulate matter.

13. A process according to any one of the previous claims wherein the ED modules each comprise n anion exchange membranes and n or (n+1) cation exchange membranes, wherein n is 50 to 1000.

14. A process according to any one of the previous claims wherein each ED module comprises the components:

a) an anode,

b) a first cation exchange membrane,

c) a first anion exchange membrane, d) a second cation exchange membrane,

e) a second anion exchange membrane,

f) optionally a third cation exchange membrane, and

g) a cathode;

15. A process according to any one of the preceding claims which is performed batchwise, with each batch comprising steps a) to d), optionally with a switching step comprising reversing the polarity of the electrodes in the leading and/or following ED module(s) and swapping the streams flowing into the diluate and concentrate compartments within an ED module being performed after step d) of one batch and before step a) of the next batch.