DK166435B1 - Method for removal of microorganisms during microfiltration of a material of a primary membrane filter without significant formation of a secondary membrane, as well as apparatus for use during performance of the method - Google Patents

Description

DK 166435 Bl i

The present invention relates to a method for removing microorganisms by microfiltrating a material on a primary membrane filter without substantial formation of a secondary membrane, applying on the primary membrane at frequent intervals 5, opposite opposite short-duration pressure pulses and maintaining a stream of the starting material extending along the surface of the primary membrane and an apparatus for use in the practice of a conventional microfiltration system including a microfiltration membrane separating a first chamber where there is starting material access and retentate discharge; second chamber with permeate discharge.

In membrane filtration, the filtration is usually carried out in the manner of using a membrane, designated a primary membrane having a larger pore size than that which is theoretically needed, ie. so that the microorganisms and substances desired to be retained on the retentate side can penetrate the primary membrane at the onset of filtration. After an initial filtration, where the permeate may optionally be returned, a retentate deposit is formed on the membrane. This deposition acts as a secondary membrane which substantially retains all the undesirable substances in the permeate. 1 2 3 4 5 6 7 8 9 10 11

However, this membrane filtration technique is associated with 2 the problem that the filtration process itself occurs on a portion of 3 cell layers which is formed partly due to concentration polarization 4, ie. forming a larger and larger concentration of 5 particles the closer you get to the membrane on the retentate side of the membrane 6, and partly because of fouling, ie. dirt 7 with solid material in the pores of the membrane itself. This implies that the filtering capabilities of this technique are highly dependent 9 on the nature of the material, and the "active membrane" has a 10 less well defined and larger pore size distribution compared to if the filtration process occurred on the primary membrane alone.

2 DK 166435 B1

Particularly for microfiltration processes where the purpose is to remove microorganisms, this filtration technique causes the disadvantage that during the process of building the secondary membrane unwanted microorganisms will penetrate the membrane.

Such penetration of undesirable microorganisms will cause a microbiological contamination of the permeate system, which is precisely desired to be free of microorganisms until the secondary membrane is constructed to act in the desired manner for retention of microorganisms. This initial microbial contamination of the permeate side of the apparatus is extremely unfortunate in that it may cause the undesirable microbial flora in the permeate to propagate.

Furthermore, this filtration technique results in a particular change during the build-up phase of the secondary membrane, but also during the effective filtration period, the filtration properties of the secondary membrane are constantly changed, which can result in a uniform composition of permeate material.

Namely, when the desired secondary membrane is formed, a further deposition occurs which gradually reduces the flux of the membrane. This implies that a defined effective filtration period is obtained which lasts until it becomes necessary to interrupt the filtration and remove the material deposited on the membrane.

Membrane filtration is divided into three types: Reverse osmosis filtration (RO), ultrafiltration (UF), and mi body in 1 tration (MF). .30 All three types of filtration are characterized by pressure. In ultrafiltration and microfiltration, substantially the same pressure is used, while reverse osmosis filtration takes place at a somewhat higher pressure. MF membranes have larger pores than UF membranes, which in turn have larger pores than RO membranes.

Microphobia 1 is characterized by being pressurized through a microphobia 1 membrane with a pore size in the range of 0.1 to 10 µυι. The pressure drop across the membrane should usually be as small as possible, preferably on the order of 20 to 400 kPa. Although increasing the pressure, a rise in flux occurs, this increase is very short-lived as the amount of deposited material will also increase.

Many of the materials desired to be treated by microfiltration have a high tendency for so-called fouling, which means that microorganisms, particles and / or molecules in the material are deposited on or in the membrane or have some other effect with membrane. This fouling provides a substantial clogging of the membrane and low flux values appear after a short time of actual filtration. Examples of such particularly fouling materials are liquids 15 known from the dairy industry, such as e.g. whole milk, skimmed milk, whey and cheese salad. Examples of particularly fouling materials, known in other industries, are beer, wine and juice. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16

From U.S. Patent No. 4,105,547 (Alfa Laval), a file 2 filtration method is known, especially for ultrafiltration, where a liquid is to be filtered by means of a pump along one side of the filter 4 so that a significant pressure drop occurs in the flow direction. Part of the liquid is returned to the pump and 6 is passed back along the filter. At the same time, a liquid 7 flow is maintained in the same direction, as well as a recirculation of an 8 part of the liquid, on the opposite side of the filter to form a corresponding pressure drop in the flow direction, thereby keeping the pressure difference between the two membrane sides largely constant. constantly over the entire filter area. Thus, by this method 12 a reduced and constant filtration pressure is obtained, and at the same time 13 is avoided too fast clogging of the filter due to liquid flow 14 while "washing" effect. This method is particularly suitable for file 15 trimming of skimmed milk, but in studies conducted 16 at Statens Mejer in experiments, Hillerød, Denmark, in 1989, not very good results were obtained by filtering e.g. cheese brine.

4 DK 166435 B1 AU-B-34,400 / 84 (Memtec Ltd.) describes a method for cleaning a filter by backwashing with a pressurized gas stream.

GB-A-1,535,832 (Brown et al.) Also discloses a method of purifying a filter using one backbone 11 using a pressurized gas stream, interrupting the liquid flow to the filtration system during backwashing. .

WO 88/00494 (Memtec Ltd.) describes a method of cleaning a filter using periodic "backwashing cycles" with a pressurized gas stream. In Example 3, lines 8-10, a filtration time between the backwashes of 7 minutes and a duration for each backwash totaling approx. 1 minute and 30 seconds.

From U.S. Patent No. 3,630,360 (Davis et al.), A fine suspension filtration system is known in which the filter is purified by back pressure pulses. The filtration is based on the so-called "dead end" principle, where the liquid flow is applied perpendicular to the filter surface and forced through the filter by applying a pressure difference across the filter, thus forming a secondary membrane of deposited material. has a size of 200 x 1150 mesh per inch and is flexible so that it assumes a concave shape during the filtration and a convex shape during the purification by the back pressure pulses.This change of shape will cause the removal of deposited material on the filter. 30, a pressure of 2.5-6 inches of mercury, corresponding to 9-20 kPa, is applied, and the pressure pulses are applied every 15-30 seconds, ie 2-4 times per minute (see claim 1). filtration of a suspension of algae in water and filtration of a suspension of calcium oxide.

In the article "Ecrémage et épuration bactérienne du lait antier cru par microfiltration sur membrane en flux tangential" in 35 DK 166435 B1 5

Technique Laitiere & Marketing - No. 1016 by Piot et al., It is recognized that cross-flow membrane filtration with permeate discharge shut-off for short periods can be used for microfiltration of milk. Work is done here with time periods of 1 minute duration of 1 minute and intervals of 6-9 minutes.

Among those skilled in the art, prejudice exists against the backwash method, and the method has found no major practical application, as the loosened dry matter is rapidly re-deposited on the filter, resulting in a loss of capacity. Thus, for example, in the article "MICROFILTRATION - STATE OF THE ART" by Rune Glimenius (Alfa-Laval), cited in Desalination, 53, pages 363-372, 1985, published by Elsevier Science Publishers BV, Amsterdam, The Netherlands, it is stated that "Rinse has been sought for in 15 cases. In some cases it is the only way to go, but in principle it must be said that flushing only means treating the symptoms of the disease and certainly not curing the disease itself. Rinse means that we achieve our permeate with great difficulty and then spend both time and 20 energy to force it back and lose it. Rinse back means we lose both operating time and capacity ".

Thus, although various principles for purifying and increasing capacity of filters are known, there is still a desire to improve the capacity of membrane filtration for removal of microorganisms, especially for microfiltration of aqueous solutions with a relatively small content of proteins such as cheese brine and whey, where more efficient membrane filtration systems are desired. Such increased efficiency 30 will provide significant economic benefits, as the acquisition and operation of such membrane systems are associated with high costs.

It has now been found that by combining the so-called "cross flow" principle or cross flow principle, the product flow is characterized by proceeding "washing" along the diaphragm, with frequent and short-term pressurization 6 DK 166435 B1 pulses are capable of conducting a microfiltration on a membrane with such a small pore size that an effective retention of undesirable microorganisms is achieved throughout the filtration period, while at the same time achieving a large increase in filtration capacity.

The cross-flow principle is understood to mean a filtration where the retentate under pressure, e.g. by means of a pump, flows along the diaphragm with a pressure drop in the flow direction 10 as a result. During this cross flow, some deposition of deposited material on the membrane takes place. However, one can not overcome, but only reduce the problems of fouling and clogging of the membrane, since in practice some particles have been found to fit nicely into the pores of the membrane and clogging them that other 15 form large aggregates, that some are stuck on the membrane, and yet others interact with the membrane and are therefore not removed by the cross-flow filtration.

Using back pressure pulses, periodically 20 permeate are pressed against the filtering direction, thereby loosening the filter cake from the filter and maintaining the filter capacity.

The invention thus relates to a method of the kind mentioned in the introduction, which is characterized in that the opposite pressure pulses have a duration of 1 to 5 seconds, a frequency of 1-10 pulses per second. per minute and a pressure difference of 100 - 1000 kPa across the membrane.

The process of the invention provides that, by selecting the pore size of the membrane, it will be possible to allow the pore size of the primary membrane to be so small that undesirable microorganisms are effectively retained throughout the filtration period, while obtaining high flux values. , the microfiltration being performed on a well-defined, uniformly effective filtration membrane.

Furthermore, the ability to use this well-defined, uniformly effective filtration membrane ensures a uniform permeate and simplifies the process adaptation, which, knowing the exact pore size and pore size distribution in the filtration membrane, allows different kinds of molecules to be fractionated. Thus, this allows for the production of products with new functional properties and means that it is increasingly possible to use microfiltration in separation processes, where centrifugation has previously been used with more or less effective result.

10

The use of back pressure pulses enables higher pressure to be applied across the diaphragm without the previously mentioned problems of too rapid clogging of the diaphragm, which results in a significant increase in capacity. Interruption of the filtration process to clean or replace clogged membranes is also avoided or greatly reduced.

In a preferred embodiment of the invention, the oppositely directed pressure pulses have a duration of 1-2 seconds, a frequency of 3-6 pulses per second. and a pressure difference of 100-1000 kPa across the membrane.

When using pressure pulses with such frequency and strength, particularly good results are obtained, keeping the membrane continuously clean of deposited material.

The reason for the good results obtained by the method according to the invention is in particular the use of frequent and short-duration retraction pulses.

30

It is important that the total time consumption per impulse is as small as possible, as the permeate flow is stopped when back pressure pulses are performed, i.e. capacity is diminished. The use of time can be divided into two parts. One is the amount of time spent on the shock itself. This advantageously has a duration of approx. 1 sec The second time is the time taken to return the amount of permeate returned during the shock to the permeate again. This time depends on the volume / pressure under pressure in the concentrate and on the flux.

In order to achieve good results, it has been found that it is important to apply as many pulses as possible, taking into account the time consumed by each pulse. A frequency of 1-10 pulses per second. per minute, preferably 3-6 pulses per minute. minute, gives the best results. By increasing the number of pulses, any increase in permeate current will be offset by the additional time required by several pulses. Furthermore, the choice of the number of impulses will depend on the material to be filtered on the membrane. The greater the tendency of the material to coat the membrane, the greater the number of pulses per minute. time unit be.

15

The force of the impulse must be as large as possible in order to maximize the impact of the membrane coating. A high strength means that the shock time is shorter, thereby stopping the permeate flow for a shorter time. In the process according to the invention 20 a differential pressure of 100 - 1000 kPa is used.

The amount of permeate sent from the permeate side to the retentate side during the impulses must be considered a negative capacity. Therefore, this volume should be as small as possible, 25 but so large that permeate will pass through to loosen the coatings on the membrane.

The filtration membrane used in the process according to the invention is advantageously made of a ceramic material. Membranes made of ceramic materials have a very high strength and are able to withstand frequent back pressure pulses. In addition, it is possible to make membranes with a small pore size distribution, and it is relatively easy to make asymmetric membranes, ie. membranes consisting of several layers of different pore diameters, with one side of the membranes acting as the actual separating element, and where a support or support structure below has a high permeability. Such asymmetric membranes are today the most commonly used for industrial use. Membranes made of polymeric materials and metallic membranes could also be used.

5

In a preferred embodiment, the pore size of the filtration membrane is 0.1-10 mm, preferably 0.5-1.0 mm. Filtration membranes with a pore size of approx. 0.8 mm are particularly suitable for use in the dairy industry for the removal of unwanted microorganisms from a material by microfiltration according to the method of the invention. In the said case, this pore size simultaneously provides the desired high filtration capacity.

Advantageously, the process of the invention can be carried out using an apparatus comprising a conventional microfiltration system which includes a microfiltration membrane separating a first chamber having starting material access and retentate discharge from a second permeate chamber 20, which apparatus is characterized in that the permeate discharge is provided with a shut-off means and at least one means for providing pressure pulses, and that the starting material supply and the retentate discharge are designed in such a way that during the filtration a flow of material from the starting material supply to the retentate outlet is maintained, as in the first chamber. extends along the surface of the membrane.

Microfiltration by the process of the invention opens the possibility of producing products with improved shelf life, removing microorganisms that would impair shelf life. In addition, it is possible to produce products under very low production costs, as the process according to the invention provides an increased capacity compared to conventional membrane filtration processes. The process according to the invention also allows for the production of products with new functional properties, whereby a more well-defined filtration can be obtained, and it is possible to produce products with a high content of ingredients in their naturally occurring state. , as one can often omit the heat treatment previously needed to remove microorganisms.

5

The process according to the invention is particularly suitable for microfiltration of aqueous solutions with a relatively small content of proteins or with relatively small molecules, e.g. brine and whey.

10

The process according to the invention has a wide range of applications in dairy industry. Examples include: 15 Removal of microoganisms from whey in order to produce whey products such as e.g. whey protein concentrates, containing no living or dead microorganisms. In such products, the properties of the proteins are retained as they are not heat-denatured.

20

Removal of microorganisms from cheese brine without otherwise altering the composition of the brine, especially with regard to mineral content and protein content.

25 Removal of microorganisms from skimmed milk to produce long-lasting milk, where the constituents, in particular protein, enzymes and salts, are kept in their naturally occurring state and to improve the quality of whole milk, cheese and dry milk products.

30

In addition, the method can be used generally in all industrial branches to concentrate bacterial cultures.

In particular, in the dairy industry, the process can also be used

Separation and isolation of casein micelles and whey proteins from milk in their naturally occurring state, DK 166435 B1 11 clarification and removal of residual fat in the whey to optimize a subsequent ultrafiltration process to produce low fat whey protein concentrate products in order to produce high quality phospholipid products.

In addition, the method can be used for clearing and germinating water, juice, cooks, beer and wine.

The scope of the applicability of the invention will become apparent from the following detailed description.

Example 1

Experiments to compare microfibre 1 with and without back pressure pulses.

The present experiment was carried out with microfiltration of a cheese salt layer which had not previously been microfiltered, and this brine therefore contained more fouling material than would normally be the case with industrial application of this technique, since a daily microfibre 1 is treating the entire amount of brine.

The material used was a brine which was pre-filtered through a Jesma screen with an average pore size of 6 pm (supplier Jesma-Matador AS, Niels Finsensvej 4, 7100 30 Vejle, Denmark).

The experiment was carried out under the following experimental parameters:

feed pressure: 50 kPa 35 temperature: 20 ° C

flow: 6.0 m3 / hour.

12 DK 166435 B1

Retentate valve1: almost closed corresponding to a retentate amount of 3% of the starting material flow or a concentration of 1:33.

5 The body used in 1 tethering membrane was an asymmetric ceramic membrane with an average pore size of 0.8 µm and a membrane area of 0.2 m2 with the trade name MEMBRALOX® supplied by SCT Department Membranes Ceramiques, B.P. 113 - 65001 Tarbes Cédex, France.

10

The following test variables were used: back pressure pulses 0 or 6 times per minute (compressive strength 400 kPa), 15 back pressure on permutation outlet 0 or 50 kPa.

The experiment carried out was divided into four sections as follows: 20 Experiments: Backpressure pulses (BP) Back pressure on permissible discharge 1. - + 2. - cleaning 3. + + 2 5 4. +

The results obtained in the experiment are shown in Table 1. 1 35 13 UK b l <υ Ο) c • r- X X D)

<- 3 Τ-Ί >> CM >> CX

Λί r- S- t- -r- 3 £ _ H- D> + JO) -L> i -> - (¾ Ό 4-> Q "UQ" DQ µε c s. (/) OQ- (/) ο 0-Wed

0) (0 (0 L- Σ CO £ - E CO CC

CQ> +> Ο O <1) <0 to LL + I LL I I a.> «X ECO Ο OOOOHH C- O Mi> 3 · 33« Ν. CD CO 00 CO ¢ 7) 00 «t r-ι O« 0 «tt— i — CDli) t N Η Η Η H CM HH O) U- HW« "· L> (0 Q) - E (0 t 30.010 oooooo o <D 0L _X vj- ΙΛ U) IO (OU) U) ΙΟ O OO «t Q. - I 2 a) o I— I— Tt 33 M-

O OOOOOO O OO

f— £ _ (/) - + -> * ------ '' '

Ο) -r- 0) \ O (O (C l £) l £> ΙΟ Φ (O (O CD (O O

yes ϋ C n

(0 <U · <- E

I- 03 £ - - (0

30-00 OOOOOO O O

Cl yL σ> t- co o co oo oo co co ιι o moo ^ oooooo ο o • rQ-OCO o cm cm cm cm cm CM II o µ £ L H CM CO CO CO CO CO CO CO «- * ffl -

-H

c <D.

+> CL'- 'CEO (3) 00 iflCOOH I Ο Ο II £ 0 ® 0) 0 rlH CM TI CM CM CM CM * - · 03 I-- ro ΟΙ ε <UQ - -DC- Ο O <=> Q + J <4_ ο ΙΛ oooooo oo 0 - (/) 0-1-1 ΙΟ (Ο ΙΟ ΙΟ Λ Λ (Λ II i-1

in Ο Ιβ O li) O U) CD Ο LO

73 ΙΟ ΙΟ li) Ο Ο Ν Μ «it« - * CO

• Γ ....... * * 'J— c * - C- · C— CO ¢ 0 CO CO CO 0) 0¾ DK 166435 B1 14 £ _ 0)

O) JsC .IC

C CO>> • i- £ _ I_ C O) D> + »Q TJ S Ό t- to o o. Tn o Q- 8 £. ε cq t- ε co ε o o tv u_ + i u. i +

CQ

+> s

«U) σι 'ί" Ϊ CO N CM LO CM

X B mcoo'inn cm cm cm cm

3 \ li) Ifl LO ^ »t« Ϊ LO LO LO

a, - + j ro <u —- ε ro s_ 3 o. oooooo 0J CL Y to LO LO LO LO LO ooo CL - +> ro is (0 tt) o -H i— r— £ - 3 M- O OC I <- OOOOOO ooo n- £ _tn +> ~ ^ »- - - • i-OJ'v to to to to (O (O (O to O to rH OC« tu ·· - ε 1 - CE £ _ tt)) n ro - i- ro o

3 CL O I I I I I O I I C

cl .X co t- ro “σι £ _ + -tu ro ro oo tn +> • f- a. cm iiiii cm iii— ro +> Q. ^ CO CO 3 tt) α ε +> ε ι- c · ι- α) <ϋ · Λί tn α.

+> α.— .χ

C ε Ο Ο) Ο Η Ο Η τ—! L Η Η rt L 7 ^ ° TO

(1) 0) 0 rH CM CM CM CM CM CM CM CM + J £ _ £ _ O.

QC 1— t— tn +> +> ___0) cn tn β) a n c o) D)> · - ro ro c ro s- ro + j οι ro o +>

Q. 73 o)> - -D

I Ε .V I— c 3 -r o tt) G (-— OOOOOO OOO -1--1-3 + (3 * UC- LO LO LO LO LO LO LOLOIO +>

O + - »* + - II II II II

U- tn CL t_ ___0) 4J> LO OLOOLOOLO (D C- o o c_ * t LO LO t — 1 CO LO CM CM ^ CM C +> Ό · ......... 13 73

• 1- Ο Ο O t — i rH t — t CM CM CM O r 3 CL O

I— rH rH rH t-Η rH rH rH HHH IL CL Q. ffl Σ DK 166435 Bl 15

When filtering cheese brine, it is clearly seen that back pressure pulses (BP) have a significant effect. This is seen when comparing the results for permeate flux from Experiments 1 and 2 with the results in Experiments 3 and 4, respectively. With no backpressure filtration, no significant flux drop occurs after the backpressure pulse unit is activated (Experiment 4).

Example 2

Attempts to determine retention of microorganisms.

The test medium was as in Example 1 cheese salt layer which was pre-filtered through a Jesma screen with a pore size of 2 μιη. The test parameters used were similar to Example 1, and the microfiltration membrane used was also the same as in Example 1. Prior to and during the test, samples were taken for bacteriological analysis of the germ content. The seed count was determined as the number of total germs / ml at 30 ° C at standard plate count. The results are shown in Tables 2 and 3.

25 1 35 DK 166435 B1 16 i- c Φ -r-

U) E

C N

X C

.x to ro in. - Q- ffi -X 4-- E CL i.

ro fun m o + V) A-> s «X E OO'i ^ lCNONrl

3 \ lOfflNlDNlflNOH

r- i— OlttMDIfllO'J'i'i'i IL -

P

ro ro - E ro

L 3 CL

<u cl .x

CL - O O I 1 O O I I I

2 ro o I- I— CM 3> + -

X I - «OOOOOOOOO

r— L (/) 4-

{U 'i-TOV. IoiD {Dffl <D {S (OIOIO

• Ω O C n ro ro · - e I— OC S- - ro

3 CL O O I I O O I I I

CL X C- D- «3 tD

ro o o o o

• r · CL CM CM I I CM CM I I I

Ή 0. -X CO CO CO CO CO

ro - +> c ro · u a— «CEO co co i i o o i i o

(1) Q) O t-h i-t CM CM CM

o: i- -— ro

CL

I Ξ -X ro s v-

“D L O O I I O O I 1 O

Q + j u- co to to ro to

LL ro CL

tOOOOOOtOIOlOlO

t-tCMCMCO ^ HMCJlOCO

Ό ..........

• t- CCt'CMDO®®H

H

Table 3.

17 DK 166435 B1

Bacteriological Assays 5 Sample No._Total Kim / ml_ _1 _._ Starting Material (1000 1) _248 x 10a_ _2 _._ permeate 1_41 x 101_ _3 _._ permeate 2_106 x 101_ _4 _._ permeate 3_93 x 101_ 10 5._permeat 4_144 x 101_

It can be seen from Table 3 that a significant germ reduction is obtained by microfiltration with back pressure pulses. The germ reduction corresponds to a bacterial retention of 99.4 - 99.8%. It should be noted that the permeate side was not sterilized prior to the start of the experiment.

Example 3 Microfiltration of whey with and without back pressure pulses. The following 2 attempts were made:

Experiment A: Continuous operation without back pressure pulses.

Experiment B: Continuous operation with back pressure pulses, stroke volume 50 ml, (i.e. the amount of permeate pushed back through the membrane at each back pressure pulse), frequency 2 x / minute.

30 Both tests were carried out with a feed pressure of 50 kPa and a recirculated flow rate of 6 m3 / h.

The microfiltration membrane used is the same as in Examples 1 and 2. The results of Experiment A and Experiment B are shown in Tables 4-8 below.

Table 4 18 DK 166435 B1

Supplies 1 Backpressure Flux Avg Retention% Material Supplies Impulses Continuous Hold See impact. freq. equal operating amount Total Valle- _mj_x / min. l / ma / t_Fat protein protein A - - 124 1:22 18 6 7 10 B 50 * _2_912_1: 31 10_4_3 * The 10 lbag pressure pulse pressure is approx. 300 kPa at a stroke volume of 50 ml.

When comparing Experiment A and Experiment B, it can be seen that the reverse pressure pulse valve has a distinct effect. Experiment A, which is without the use of back pressure pulses, clearly has poorer flux in continuous operation than in Experiment B performed with back pressure pulses.

20

Furthermore, the ratio of retentate to permeate in experiment A is only 1:22 compared to a ratio of 1:31 in experiment B. 1 2 3 4 5 6 7 8 9 10 11

From the table it can be seen that in microfibra- tion 1 whey with up to 2 baking pressure pulses a significantly smaller percentage for 3 retention of fat and protein is obtained. This shows that the use of retraction pulses significantly impedes the formation of a secondary membrane.

6 7

Trials with back pressure pulses at a frequency of 2 x / min.

8 and 0.5 x / min. shows that the back pressure pulse frequency of 2 x / 9 min. gives the best flux. Furthermore, a stroke volume of 10 gives 50 ml corresponding to a back pressure pulse pressure of approx. 300 kPa 11 a better flux than a stroke volume of 30 ml corresponding to a ten 1-baking pulse pressure of approx. 100 kPa: 19 DK 166435 B1

Stroke Volume Frequency Flux (l / m2 / h) 30 ml 0.5 x / min. 350 50 ml 0.5 x / min. 408 30 ml 2 x / min. 480 5 50 ml 2 x / min. 912 10 15 20 25 30 35 20 DK 166435 B1

Results of Experiment A Table 5 _ Feed flow__Retentate__Permeat

Reci rkule- 5 Time Temp. Pf Temp. Pi Pu ring flow Pu Flux __ ° C kPa_ ° C kPa kPa m3 / t_ kPa l / mg / t

On start-up phase

Start 10.05 - 50 47 260 60 6.0 10 1440 10.15 - 50 50 250 60 6.0 - 1098 10.25 - 50 '53 260 60 6.0 0 912 10.40 - 50 48 260 60 6.0 5 732 10.50 52 50 48 260 60 6.0 0 522 11.05 52 50 48 260 60 6.0 - 432 11.15 51 50 48 250 60 6.0 - 372 11.30 51 50 48 250 60 6.0 - 282 11.45 50 50 48 260 60 6.0 - 270 15 12.00 49 50 48 260 60 6.0 - 240 12.30 48 50 49 250 60 6.0 - 189 13.00 48 50 49 250 60 6.0 - 159

Continuous operation

Start 2Q 13.16 48 50 49 250 60 6.0 - 159 13.30 48 50 48 250 60 6.0 - 144 13.50 48 50 50 250 60 6.0 - 135 14.00 47 50 50 250 50 6.0 - 129 14.15 47 50 50 250 60 6.0 - 129 14.30 46 50 50 250 60 6.0 - 117 15.00 46 50 50 250 60 6.0 - 111 15.15 45 50 50 250 60 6.0 - 96 25 15.45 45 50 51 250 60 6.0 - 89 1 35 DK 166435 Pg 21

Results for Experiment B Table 6 _First Cream__Retentate__Permeat

RECIRC.

5 Time Temp. Pf Temp. Pi Pu dP flow Pu Flux __ »C kPa_ ° C_kPa kPa kPa m3 / t I kPa l / m2 / t

Water 100 2331 f 1 hour Start-up phase Start 10 9.27 48 50 48 280 60 220 6.0 50 1920 9.45 52 50 49 270 50 220 6.0 40 1537 10.00 52 50 49 280 50 220 6.0 40 1335 10.20 51 50 49 270 50 220 6.0 40 1200 10.40 50 50 49 270 50 220 6.0 30 1058 10.55 50 50 49 270 50 220 6.1 30 998 15 Continuous Operation Start 11.15 50 50 49 270 50 220 6.0 30 953 11.35 50 50 49 270 50 220 6.0 30 930 11.55 50 50 50 270 50 220 6.0 30 945 12.15 47 50 50 270 50 220 6.0 30 893 12.35 45 50 49 270 50 220 6.0 30 945 20 12.55 45 50 45 270 50 220 6.0 20 780 13.15 44 50 47 270 50 220 6.0 30 773 13.35 44 50 49 270 50 220 6.0 20 773 13.55 44 50 49 270 50 220 6.0 20 25 1 35 DK 166435 B1 22 H-µ (0 (0 O) O)

C C

* r— -r- C c µ µ 3 3 I— <- CO CO lO 00 CO o tn in * t i-ι o io t- <o * t- Cf> V1_ O) * - * * * * praise co th co o (os * ooo io co σι in c- i in co o in Ό i- - »I« C “- T3 £ _

SO r-l O CO U) SO

> 4-> µ I c C s <u £

£ µ E

+> E (0 (0 O) ro ro σ> cocot ~ «cf µ in c co co o Tt c- µ in c c- co to i σι c -> - co to n ιο σι C'r · - - - - IQ) Γ h ^^ i.

S · C T-lOOlO µ tø µ OOOtOlf) µ in µ s c ffi s c s ir o in ir o in <CQ___ σ ___ σ s s tn in o) £ - i- -c 0 o in · ι ~

4-4 ° C

µ -µ co co <£> ιο σ> gr- s- 4- OUDlOtO I »4- µ to (OO Ί 4 th E> 01 t ~ -r- µ to CM OI CO I SCO ·> - - - - - - v QJ ^ • µ (- - ^ · * µ c- co ooo (O m µ in (0r- Ό-'ίΟΟΟίΛ (0 f— Ό in -o • µ ro µs S3

<“-O U) I— Ώ Ul JO

310 -I- 3 (0 -I- O) co ro tnt -> - c- in h - µ to lo µ t_ o) i- µ co s £ _l_ - * s 4- t_ Si- i «· i £ _ SS cg O ιο Ό S 3 S CM LO S 3 μ 1— Ό in c μ in c 4- oc> · “4-> -r- S JC ro» - μ SW ID 4 I— μ Ο'ΪΙΟΙΟ Ό > ro c µ tn co o rr o ro c µ c- co o co c CO - --- 1-1-1 c O - - - - it -l- 1_ <i ^ i-iooOLO v <; «c Tt o oo mv µω ---- Ό Ό S r-

Tt µ O

µ Tt o µ o Tt in * t jz

t'T-i t- co o in OS

co LO v,) {1 - - - - t- (so

Ol S i-ι OOOWV> c tn m c .γιο cm ro o μ μ co τϊ μ μ o c-> σ> in to -i in c- in .co μ μ cm - iii lo μ μ i_ ii rot . o roi-so in -µ Q) roi_ s ro µs c E µs o in e µ µ o «t ~ o s t-ιημμιο Tf i_ in s µ c- in o in s α.µ> - i so. - - μ

Beer. o s τ-t o i i t in 0.0 ^ 00 m s s_

—-- LJ__E O.

µ s c Ό σι cm co co in S ο ^ ε-τϊΐοοφμ

S IO CO O Tf Tt O Ό (D CO O Tt LO O OS

IL - - - - - f- Q - - - - - C0 S) ooo to in cm o. Ooo in lo m i_ to s μ de de> 13 • i- 3 'i * de i de * de σ oo cr Er- oo: e> - cs ι-ι-ωμ · γ · ε lc co 4- · - e ro> ute o.de w ojis adeiT-w oyxs I— S μι— I— r— SS μι— r Cl .

(0 cr— c μ tn ro ro ci— μ μ inioro cool μ · ι - «- · γ-ό t_ μ μ μ-ri— οό ι_ μ μ ιη osrossxsoc osrot-sxsoc * Η-μ> μα.ο .κ-ι- «ι- μ> ou cu- ι- ιο * *

Claims (6)

A method for removing microorganisms by microfiltrating a material onto a primary membrane filter without substantially forming a secondary membrane, applying on the primary membrane at frequent intervals opposite directed short-term pressure pulses and maintaining a stream of the starting material extending along the surface of the primary diaphragm 10, characterized in that the oppositely directed pressure pulses have a duration of 1 to 5 seconds, a frequency of 1-10 pulses per second. per minute and a pressure difference of 100 - 1000 kPa across the membrane. Method according to claim 1, characterized in that the opposite directed pulses have a duration of 1-2 seconds, a frequency of 3-6 pulses per second. per minute and a pressure difference of 100 - 1000 kPa across the membrane. Process according to claim 1, characterized in that the primary filtration membrane is made of a ceramic material. Method according to claim 1, characterized in that the pore size of the primary filtration membrane is 0.1-10 μιη. Process according to claim 4, characterized in that the pore size of the primary filtration membrane is 0.5-1.0 µm. £ Apparatus for use in practicing the method of claim 1 comprising a conventional microfiltration plant including a microfiltration membrane separating a first chamber of starting material access and retentate discharge from a second chamber of permeate discharge, characterized by that the permeate outlet is provided with a shut-off means and at least one means for providing pressure pulses, and that the starting material supply and the retentate outlet are designed in such a way that during the filtration a flow of material from the starting material supply to the retentate outlet is maintained, as in the first chamber extends along the surface of the membrane. 5 10 15 20 25 30 35