AND/OR FILTRATION DEVICE

FIELD OF THE INVENTION

The present invention relates to stability testing of a fluid purification and/or filtration device, including water purifiers with chemical and/or physical filtration steps as well as filtration devices used for medical applications, for example used for cell separation and dialysis.

BACKGROUND OF THE INVENTION Membrane filtration is used in water purification, for example as published in international patent application WO2009/019592 by Vestergaard SA, as well as in medical cell separation, for example as disclosed in international patent application W095/13829 by Braun Melsungen. Despite different purposes, the filtration principles, for example when using hollow fibre assemblies, are similar. The proper functioning of the filtration devices is essential in both fields, as a defective cell separation device or dialysis filter can be fatal for the patient, as well as a defective water purification device can be fatal for the user in areas, where the water contains pathogens. Quality criteria have to be high in both cases, and it should be assured that the filtration devices can fulfil the promised performance standards not only at the factory site but also after de- livery to the user. For this reason, there are performed functional tests for the corresponding filtration devices, and also similarly for other kind of purification devices, for example, water purification devices including chemical cleaning steps as disclosed in international patent application WO2008/067816 by Vestergaard SA. However, a global standard does not yet exist.

In connection with portable water purification devices, special attention is given to purification devices that are used in rural areas or emergency situations where clean water is scarce. The purification devices have to function reliably over months or even years, as these devices, often, are the only means available for guaranteed purified wa- ter, and no frequent supply of such devices is available. Typically, the purification de- vices are transported long distances for delivery in such areas, and the devices are subsequently exposed to rough handling during long term usage.

People in emergency areas or other rural areas, as described above, should be guaran- teed reliable water purification means living up to high standards, as malfunctioning is fatal. For this reason, global testing methods are required. Although there are different testing methods of materials in general, for example for measuring Young's modulus as disclosed in US2009/000378 by Dill et al, there is a lock of proper global testing methods for water purification devices with respect to their overall stability.

DESCRIPTION / SUMMARY OF THE INVENTION

It is therefore the object of the invention to provide steps for a reliable test procedure for liquid purification devices, preferably, a global standard test procedure. The test procedure should also be useful for testing liquid filtration devices in other areas, for example, devices for medical usage in medical separation processes.

This purpose is achieved with a number of handling simulations of a liquid purification and/or filtration device as explained in the following.

The term filtration has to be understood as a mechanical separation process, where a membrane or a fibrous, adsorptive filtration material retains particulate matter. This retaining of particulate matter may be used for purification of a liquid or for increasing the concentration in a liquid, the latter especially used for biological or medical pur- poses, for example in treatment of blood. The term purification may cover these filtration processes as well if the purpose is purification, however, it also covers non- filtrative liquid purification as it is known from chemical treatment of water with, optionally, subsequent exposure to adsorptive granular activated carbon (GAC), in order to kill and remove microbes.

In a preferred embodiment, the liquid filtration and/or purification device is a portable water filtration device with a housing having a liquid inlet upstream of a microporous filter and a clean water outlet downstream of the microporous filter. For example, the filter is a bundle of hollow microporous membranes. An example of such filtration devices has a tubular housing, for example made of polycarbonate or polypropylene, containing the microporous membrane. For example, the housing has a length of less than 50 cm, rather in the interval of 20 to 40 cm, and a cross section of the tubular housing of less than 6 cm, rather in the interval of 2 to 4 cm. An example of such a filtration device is commercially available under the name LifeStraw®. Examples of such water purification and/or filtration devices are disclosed in WO2008/025358, WO2008/067817, WO2008/067816, WO2008/110166, and WO2008/110172. In the following, a number of simulations are explained. These are simulations testing the stability of the device when exposed to

A- Truck drive into rural areas

B- Drop of a bulk box containing the device

C- Weathering and aging of the device

D- Drop of a single device

E- Impacts

F-Static loads

G- Long term use of moving parts of the device

H- pressure stability

These procedures may be combined into sequences. Advantageous sequences are sequences with the steps A and B, A and B and C, A and B and D, A and B and E, A and B and C and E, A and B and D and E, A and B and C and D and E, where the sequence of steps can be subject of permutation, for example, the sequence "A and B and C" includes the sequences A+B+C, A+C+B, B+A+C, B+C+A, C+A+B, and C+B+A.

If a sequence of simulations is used, advantageously, although not strictly necessary, the device is inspected after each of the simulations. An alternative is inspection of the device after a selection of simulations in the sequence or just a single inspection after the entire sequence of simulations.

One of the possible test methods implies creating overpressure of gas in the device and testing, whether the gas leaves the device. Gas is a good detection method, because gas finds its way through even very small cracks and can be easily verified by observation of bubbles in a test where the device is kept under water or other liquid. Such tests are described below as test J. Another possible test method is a measurement of the log reduction of microbes. Such an evaluative test is described below as test K.

For example, the purification and/or filtration device may be of the kind having a liquid inlet and a liquid outlet and a liquid filter or a non-filtrative purification means or both included in the flow path of the liquid between the liquid inlet and the liquid outlet. In the following, the filtration and/or purification device will be denoted "device" for sake of simplicity.

If the device has a microporous filter, the device may, optionally, be equipped with a forward flush and/or backflush facility. For portable devices, which are predominantly envisaged here, the backflush facility advantageously includes a manually compressible balloon connected to the downstream clean water side of the filter. Exerting pressure on such balloon creates overpressure on the downstream side of the filter and forces clean water backwards through the membrane of the filter in order to remove scale and clogging material from the influent side of the membrane.

Each of the following simulations or tests A though J is in its character independent from the other simulations or tests among A through J described herein. However, the simulations or tests among A though J may advantageously be combined.

Simulation A - Truck transport simulation

The simulation has the purpose of simulating transport of boxes containing one or more of the devices. The simulation contains vibrations that are believed to be trans- ferred to the box when transported by a truck. For this reason, one or more transport boxes, for example two boxes, are filled with assembled and packed devices for the simulation. The box is then exposed to vibration. Optionally, the device in the box is heated to a temperature T _A i in the interval of 60°C to 100°C or cooled to a temperature T _A2 in the interval of -30°C to 10°C for the vibration simulation. In order to reach a temperature in this range, the box is heated to a temperature above ambient temperature. For the lower temperature range, it is envis- aged that the device is cooled to below ambient temperature. For the heating and/or the cooling, a heating apparatus or a cooling apparatus can be used.

For example, the vibrations having an amplitude in the interval of 10 mm to 20 mm or 10 mm to 15mm, for example in the interval of 12 to 14 mm, possibly 13 mm.

An optional frequency range for the simulation is 1 to 100 Hz or even 1 to 200 Hz. For example, the box is exposed to the entire range by continuously changing the frequency over that range. Alternatively, continuous sub-ranges of frequencies may be selected for the simulations. As a further alternative, the entire range or one or more selected subranges may be selected and the box exposed to discrete frequency points in the range or selected sub-ranges. For example, the box may be exposed to a vibrational sequence of lHz, 2Hz, 3Hz, etc. up to 100 or 200 Hz.

In a further embodiment, the vibration is randomly changed within the vibration range of 1 to 200 Hz; or in one or more subranges within this interval of 1 to 200 Hz; or with discrete frequencies selected in this interval.

In one embodiment, the method for testing the device being a liquid filtration device or a non- filtering liquid purification device comprises

- heating the device to a temperature T _A i in the interval of 60°C to 100°C or cooling the device to a temperature T _A 2 in the interval of -30°C to 10°C,

- performing a vibration test, wherein the vibration frequency is within the range of 1 to 200 Hz. Optionally, this test is performed with the device being packed in a transport box, as this is typically the case when transporting such devices. An optional time length t _A i for the exposure to vibrations is at least 30 minutes or at least 1 hour, for example 30 minutes to 12 hours, or 1 to 5 hours or 1 to 4 hours or 2 to 4 hours or 3 hours. An example of vibrational intensity is 0.2 to 0.8 G (rms, root mean square value), where G is the gravitation of 9.82m/s ² , for example 0.4 to 0.7 G (rms) or 0.5 to 0.6 G (rms) or 0.52 G (rms).

In a special embodiment, a random vibration test denoted "Simulation A3", is per- formed according to the transport norm ASTM D-4169 for a truck drive simulation with assurance level II for a time span of t _A 2 of in the interval of 30 minutes to 6 hours, for example 1 to 5 hours or 2 to 4 hours or 3 hours. In a Simulation Al, t _A 2 is 3 hours and, optionally, the vibration has an intensity of 0.52 G (rms). Alternatively or in subsequent addition, the simulation may imply a second simulation, Simulation A4, of a loose vibration simulation following the transport norm ASTM D- 999 for a time span of t _A4 of in the interval of 30 minutes to 6 hours for example 1 to 5 hours or 1 to 2 hours or 1 hour, at Frequency A in the interval of 3 to 6 Hz and an Amplitude Al in the interval of 10 to 40 mm. In Simulation A2, Frequency A is 4.5 Hz, Amplitude Al is 25 mm, and t _A4 is one hour for a simulation according to the transport norm ASTM D-999.

In a further embodiment, Simulation A5, the simulation Al and simulation A2 are serially combined.

These vibration simulations can be performed at room temperature. A harder test for the material, however, is obtained by performing the simulations under heated or cooled conditions, for example by first performing the simulation at heated conditions and then cooled conditions. Examples of heated conditions imply temperatures in the range of 60°C to 100°C, for example 70°C or 80°C. Examples of cooled conditions imply temperatures in the range -30°C to 10°C, for example at -20°C or 10°C. In case of combination of heating and cooling, an example of a suitable temperature combina- tion is -20°C and 80°C for a rough simulation and 10°C and 70°C for a standard simulation.

Especially, a test may imply the above simulation Al combined with A2 at a tempera- ture of -20 degrees and repeated by the same test at 80°C, or performed at 80°C first and at -20°C subsequently. Alternatively, the simulations Al is performed at these two temperatures followed by A2 performed at these two temperatures.

An example of a simulation sequence for a box containing a device or rather a plurality of devices is performed in the following way:

- conditioning at a first temperature T _A i of 60°C to 100°C the box for a time span of 1- 5 hours

- performing Simulation A3 for a time span 1-5 hours

- re-conditioned at T _A i of 60°C to 100°C for a time span of 1-5 hours

- perform Simulation A4 for a time span of 0.5 - 2 hours

- conditioned at a second temperature T _A2 of -30°C to 10°C for a time span of 1-5 hours

- perform Simulation A3 for a time span of 1-5 hours

- re- conditioning at T _A 2 -30°C to 10°C for a time span of 1-5 hours

- perform Simulation A4 for a time span of 0.5 - 2 hours

The term "conditioning" implies that the temperature actually is attained by the device, despite any optional insulating package inside the box. An example of a Sequence A for a box containing a device or rather a plurality of devices is performed in the following way:

- conditioning the box for 3 hours at a first temperature T _A i

- performing Simulation Al

- re-conditioned at T _A i for 3 hours,

- perform Simulation A2

- conditioned at a second temperature T _A2 for 3 hours

- perform Simulation Al - re- conditioned at T _A2 for 3 hours

- perform Simulation A2

For example, the temperatures are (T _A i , T _A 2)=(80 °C, -20°C) or (T _A i , (70°C, 10°C).

Simulation B - Drop test bulk box

This simulation B has the aim of simulating a drop of a box of packaged devices that can occur during transport.

In this Simulation B, a transport box with at least one device is dropped onto a hard floor from a height of H _B , for example a height H _B in the interval of 1 to 4 meter, for example 2 to 3 meter or 2.5 meter.

Optionally, this simulation is performed at elevated or lowered temperature. Examples of heated conditions imply temperatures in the rage of 60°C to 100°C, for example 50°C or 80°C. Examples of cooled conditions imply temperatures in the range -30°C to 10°C, for example at 0°C or 10°C. In case of serial combination of heating and cool- ing, an example of temperature combinations is 0°C and 80°C for a rough simulation and 10°C and 50°C for a standard simulation.

The preferred simulation Sequence B is a first drop after conditioning of the box at T _B i and once again after conditioning at T _B 2, where the temperatures are (T _B i,T _B2 )=(80°C, 0°C) or (T _B1 ,T _B2 )= (50°C, 10°C).

Alternatively, the temperatures are (T _B i,T _B2 )=(0°C, 80°C) or (T _B i,T _B2 )= (10°C, 50°C)

Simulation C - Weathering and aging

This simulation is used for provoking aging of the plastic under elevated temperature, moisture, or UV light exposure, or a combination of these. In this simulation, the device without transport box is exposed to a temperature of 40°C to 80°C for a period of 100 to 1000 hours, optionally combined with exposure to humidity, for example a humidity RH of 30% to 80%, for example 30% to 50% and optionally with exposure to UV radiation.

In Simulation CI, the device is exposed to elevated temperature for a time t _c and, optionally, with exposure to humidity. The latter can, optionally, be achieved by humidifying the device repeatedly during the simulation time t _c . For example t _c is 100 to 1000 hours or 200 to 800 hours or 400 to 600 hours, or 500 hours.

In Simulation C2, the device is exposed to UV light for a certain time t _c , optionally, with exposure to humidity. The latter can be achieved by humidifying the device by water exposure repeatedly during the simulation time t _c . Optionally, the repetition of the humidification has a frequency of 1 to 4 hours. An exampled for a UV light source is a Xenon arc lamp certified for ISO 4892.

The Simulations CI and C2 may, advantageously be combined, such that the device is exposed to UV light under elevated temperature conditions at a temperature T _c and exposed to humidification For example, T _c is in the interval of 40°C to 60°C; a good value is 50°C.

An example of a simulation is a 500 hours exposure to UV light of the above stated type at a temperature of 50°C with regular or steady humidification.

For example, the simulation may be performed by exposure of the device to a temperature of 40°C to 60°C for 100 to 1000 or 200 to 800, optionally including exposure to UV light, for example according of the above described type. The sequence comprises 100 to 300 cycles. Each cycle contains exposure to 40°C to 60°C and a humidity RH of 10% to 40% or 40% to 60% followed by humidifying, for example by water from a shower. The RH levels are switched between 10%> to 40%> and 40%> to 60%> every 6 to 48 hours. A further simulation Sequence C3 is exposure to 50°C for 500 hours, optionally including exposure to UV light, for example according of the above described type. The sequence comprises 150 cycles of 120 minutes. Each cycle contains 102 minutes at 50°C and a humidity RH of 30% or 50% followed by + 18 minutes humidifying, foer example by water from a shower. The RH levels are switched between 30% and 50% every 24 hours.

Simulation D - Drop test single product This check simulates a drop that can occur in a normal using environment.

The device without any package is dropped from a height of 1 to 3 meter, optionally 1.8 m onto a hard floor. For example, the temperature is 20-30°C. For a rougher check, the temperature of the device is lowered to a temperature of -30°C to 10°C or raised to an elevated temperature of 40°C to 80°C for the drop onto a hard floor. For a more soft check, the device is dropped on a cardboard covered floor.

In one embodiment, the simulation includes 10 drops at an elevated temperature of 40°C to 80°C, for example 40°C to 80°C, followed by 10 drops at a lowered tempera- ture of -30°C to 10°C, all onto a hard floor from 1.8 meter.

A preferred simulation Sequence Dl includes 10 drops at 50°C followed by 10 drops at 10°C, all onto a hard floor from 1.8 meter. In another simulation, Sequence D2, the device is dropped at room temperature 5 times onto a cardboard covered floor from 1.8 meter.

Simulation E - Impact check

The impact check simulates a drop from a height H _E , for example H _E being 1 to 2.5 meter, for example 1.8 m, with the most vulnerable parts of the device hitting the floor first. The simulation uses the setup, where a steel ball with a weight of 500 to 1000 grams, for example 600 to 800 grams or 700 grams is attached to a cord of length H _E of 1 to 3 meter, for example 1.5 and 2.5 meter or H _E being 1.8 meter. The ball with the cord is fastened to a ceiling and held horizontal as a start position, after which it is released in order to accelerate on its way downward, where it hits the device, especially at vulnerable parts, for example flush valves or backflush balloon connectors.

The check may be repeated several times. For example, it may be executed 2 to 8 times, or 3 or 5 times, optionally at different parts for each check or series of checks. For example, the ball may hit the device consecutively 3 times as described, each time at a different vulnerable part or the same part under different angles.

In order to increase the effect of this check, the device may be cooled to a temperature T _EI below room temperature or heated to an elevated temperature T _E2 . Alternatively, the device may be cooled to T _E i in a first check or series of checks and then be exposed to a second check or series of checks at T _E2 . Non limiting examples of temperatures T _EI are -10°C to 15°C or 0°C to 15°C, for example 8°C to 12°C or 10°C. Non limiting examples of T _E2 are 40°C to 80°C, or 50°C. In a simulation Sequence E, a steel ball with 700 grams on a cord of 1.8 m is hitting the device subsequently at three vulnerable parts while the device has been conditioned to 50°C followed by three corresponding hits at 10°C.

Simulation F - Static load

The static load check simulates a static loads applied to the product during normal use.

In one embodiment, a simulation is performed with static load of 10 - 40 kg, for example 15 to 30 kg, at a temperature of 20°C to 80°C, for example 40°C to 60°C.

A first specific type of simulation, Simulation Fl, is performed with static load of 15 kg at a temperature of 40°C. A second, harder type of simulation, Simulation F2, uses a static load of 30 kg. An even more challenging simulation, Simulation F3, uses a static load of 30 kg after heating of the device to 50°C.

If the device is tubular, the static load is provided normal to the longitudinal axis of the tube. In a practical embodiment, the device is placed on a smooth plane surface. Optionally, the static load is provided as a steel disc, for example having a diameter of 30- 100 mm or 40-60 mm or 50 mm. The first contact is made before the force is applied, and the force is increased within 1 to 10 seconds until full load. A time for the static test is between 1 minute and 2 hours, for example between 5 and 10 minutes.

Simulation G - Endurance check of moving parts This simulation is a check of the endurance of the moving parts of the device, for example turn valves, which are typically used on water purification devices.

In a first simulation, a valve is opened and closed a number of times as estimated to be approximately equal to the number of times that the valve is opened during the lifetime of the device, for example 10,000 to 100,000 times at a temperature of 0°C to 80°C. For example the valve is opened and closed 25,000 to 75,000 times, for example 50,000 times.

A suitable operation temperature for this simulation is 0°C to 80°C, for example 10°C to 50°C or 20°C to 30°C. If the device is a water filtration device with a clean water outlet valve, the simulation would be appropriate for such a valve. Therefore, for a water filtration or purification device with a clean water valve, in a simulation Sequence Gl, the valve is opened and closed 50,000 times at a temperature of 20°C to 30°C.

In a second simulation, a valve is opened and closed a number of times as estimated to be equal to the number of times that the valve is opened during the lifetime of the device, for example 1,000 to 2,000 times or 1.100 or 1,300 times. A suitable operation temperature for this simulation is 0°C to 80°C, for example 10°C to 50°C or 20°C to 30°C. If the device is a water filtration device with a microporous membrane filter, for example with a hollow fiber membrane bundle, and configured for forward flushing and/or backward flushing though a flush water outlet valve, the check for such a flush valve should be different from the clean water valve, as this flush valve, during normal operation, is not used as frequently as a clean water valve. For this reason, this simulation is appropriate for a flush valve. For a more realistic simulation, the device may be filled with turbid water for the flush water outlet valve simulation test. Accordingly, with such turbid water in the device at 20°C to 30°C, the flush valve is opened and closed 1300 times in a simulation Sequence G2 and 1100 times in a simulation Sequence G3.

If the device is a water filtration device with backflush capabilities, and a compressible balloon is provided for the backflush facility, an option for a check of the balloon is the following. In this simulation, the balloon is squeezed 1000 to 5000 times when containing liquid. In simulation Sequence G4, the balloon is squeezed with tap water at 20°C to 30°C for 4,000 times, and in simulation Sequence G5, it is squeezed 3300 times according. All Sequences G1-G5 include a visible inspection of the moving parts for damages. Simulation H - Pressure stability.

This simulation is used for checking the reliability of the product when put under pres- sure. The device is filled with gas, for example air, and subjected to internal pressure of P _H for a time t _H . Suitable values for P _H are 0.5 to 2 bar above atmospheric pressure, for example 0.8 to 1.5 bar or 1 bar. Suitable values for t _H are 5 to 60 seconds, for example 10 seconds. A liquid pump may be used for creating the liquid pressure inside the device, while all taps are closed. If the device is a liquid filtration device with an internal filter, the pressure should be applied to the upstream side of the filter and also to the downstream side of the filter. For example, the pressure may be provided to the upstream side first and then to the downstream side.

If the device is a tubular water filtration device, the pressure may be applied to a water entrance connector in order to check the upstream side. It may then be connected to the water exit or, if present, to the connector of a back fiush balloon. In a Simulation Sequence H, the device is connected with its water inlet to a water pump to create pressure of 1 bar inside the device for 10 seconds, after which the pressure is applied to the connector for the backflush balloon to create a pressure of 1 bar inside the device for 10 seconds.

Simulation J - Microbiological challenge

If the device is a water filtration device, or otherwise water purification device, it is advantageously checked for microbiological performance. For this reason, its capability for reduction of microbe concentration is checked by exposing the device to a liquid containing microbes which should be removed to a sufficient degree. The degree of microbe removal is typically measured in the reduction in terms of the logarithm of 10. Thus, a log 3 reduction of bacteria means that, after the filtering, the concentration of living bacteria is reduced to 0.1 % .

In a Check Jl, one criterion for passing the test is a log reduction of parasites of at least 3.

In a Check J2, one criterion for passing the test is a log reduction of bacteria of at least 6.

In a Check J3, one criterion for passing the test is a log reduction of virus of at least 4. In a Check J4, the criteria Jl, J2, and J3 are combined. : Leaking check

For a water filtration device with a water inlet upstream of a microporous filter, a clean water outlet with clean water valve, a backflush balloon, and a backflush water outlet, the following leaking checks may advantageously be used for checking the device after one or more of the above simulations, for example after each simulation. For example, the temperature for the leaking check is in the interval of 15°C to 25°C.

Check Kl is used for checking possible leaks in welds and at the backflush water tap. The taps of the device are closed. The balloon has to be connected, and a compressor is connected to the water inlet exposing the device to pressurized gas. The device is kept under water, and it is checked for leaking of bubbles.

Check K2 is used in a search for possible leaks in the clean water outlet. A compressor is connected to the backflush balloon connector, and the clean water outlet valve is closed. The device is kept under water, and it is checked for leaking of bubbles. This check may also be used to check the tightness of a microporous membrane inside the device. Check K3 is used for checking the balloon. A compressor is connected to a clean water outlet with open outlet valve. The device is kept under water, and it is checked for leaking of bubbles.

A pressure value for the pressurised gas is in the interval of 0.3 to 4 bar above atmos- pheric pressure, for example in the interval of 0.3 to 1 bar above atmospheric pressure or 0.5 bar.

In Sequence K4, the three checks, Kl, K2, and K3 are combined as a serial sequence. Combination Check

In the following, a series of simulations are tabularised for a standard test and a worst- case-testing of a number of boxes, optionally 2 boxes, with sequential steps A-J. Optionally, each step apart from step J is followed by a check according to Sequence K4.

Worst-case simulation Standard simulation

A Sequence A with Sequence A with (T _AI , T _A2 )=(80 °C, -20°C) (TAI, T _A2 )= (70°C, 10°C)

B Sequence B with H _B =2.5 m Sequence B with H _B =2.5 m

(T _BI ,T _B2 )=(80 ^o C,0 ^o C) (T _B i,T _B2 )= (50°C,10°C)

C Sequence C3 with environment temSequence C3 with environment temperature: 50°C perature: 50°C

including UV exposure ISO 4892

D Sequence Dl with height 1.8 m Sequence D2 with height 1.8 m

E Sequence E no simulation E

F Simulation F3 Simulation Fl

G Sequence d, G2, G4 Sequence Gl, G3, G5

H Simulation H Simulation H

J Check J4 Check J4

It should be emphasized that the term "in the interval of for parameter intervals includes the endpoints of the interval. SHORT DESCRIPTION OF THE DRAWINGS

The invention will generally be explained further with reference to the drawing, where FIG. 1 illustrates a water filtration device;

FIG. 2 illustrates a water filtration device with greater detail of the valves;

FIG. 3 illustrates an impact simulation E for testing the mechanical stability of the device; in FIG. 3a, the ball hits the clean water valve laterally to the longitudinal axis of the tubular device, whereas in FIG. 3b, the ball hits the flush water valve parallel to the longitudinal axis of the tubular device;

FIG. 4 illustrates a static load on the filtration device.

DETAILED DESCRIPTION / PREFERRED EMBODIMENT

FIG. 1 illustrates a portable water filtration device 1 for the proposed test. The filtration device is similar to the marketed water filtration device LifeStraw®. The filtration device 1 comprises a housing 2 inside which a number of microporous capillaries 3 are contained. Water enters 4 through an inlet 5. The water flows through the capillaries 3 into an outlet chamber 6 in the lower end, from which it can be released through a flush water valve 7 at a flush outlet 8 in the case of forward flush. If the flush water valve 7 at the flush outlet 8 is closed, the pressure on the water drives the water through the walls 10 of the capillaries 3 and into the interspaces 11 between the capillaries 3. From the interspaces 11 , the water can be released for consumption through a clean water outlet 12 having a clean water outlet valve 13. In addition, the filtration device 1 has a compressible balloon 14 in which clean water is accumulated. As the balloon 14 is located lower than the clean water outlet 3, when the device is in proper orientation, it is filled with clean water before water is released through the clean water outlet 3. When the clean water outlet 12 is closed by the clean water valve 13, and pressure is exerted on the balloon 14, pressure drives the water from the balloon through the capillary walls 10 and back into the inner volume 15 of the capillaries 3. This back flush presses microbes and other particles out of the capillary pores and away from the inner surface 16 of the capillaries 3. A subsequent or simultaneous forward flush through flush outlet 7 removes the microbes and particles from the filtration device 1.

A different type of water filtration device in connection with the invention comprises a flat membrane instead of capillaries, for example as illustrated in International patent application WO2008/110172 by Vestergaard SA.

FIG. 2 illustrates a water filtration device in greater detail with respect to the clean water valve 13 at the upper end 17 and the flush valve 7 at the lower end 18. The connector 19 is used for attachment of a balloon under normal operation condition.

Although the simulations are described in connection with a water filtration device, the test methods are of a general nature may also be applied for medical filtration devices, for example those that are used for concentration or filtration of body fluids. FIG. 3 illustrates the above mentioned Simulation E. A steel ball 20 with a weight of 500 to 1000 grams, for example 600 to 800 grams or 700 grams is attached to a cord 21 of length H _E , for example H _E being 1.8 m. The ball 20 with the cord 21 is fastened to a ceiling 22 and held horizontal as a start position, after which it is released in order to accelerates on its way downward, where it hits the device 1 , especially at vulnerable parts, for example flush valves 7 or backflush balloon connectors 19.

FIG. 4 illustrates the above mentioned static load Simulation F. The water filtration device 1 has a tubular housing 2, and the static load 23 is provided normal to the longitudinal axis 24 of the tubular housing 2. In a practical embodiment, the device 1 is placed on a smooth plane surface 25 and the static load 23 is provided by a steel disc 26, for example having a diameter of 50 mm.