Patent Application
20050013912
The invention relates to a method for reducing the total bacteria count with the features of the preamble of claim 1.
In a technical article published in the year 2000 entitled “Developmental trends in cheese production (H.-P. Bachmann, H. Schaer, R. Sieber, H. Winkler, F. Rentsch) of the Federal Institute for Dairy Farming, Switzerland, “New technologies in cheese production” are described under Section 3. Here it is emphasized that in the new technology for cheese production the treatment of the milk is of prime importance. On the one hand it is a matter of making the milk as free of bacteria as possible and on the other hand the chemical composition of the milk is to be changed in a purposeful manner.
As processes for sterilization of milk in addition to the various known heat treatment processes (thermization, pasteurization) microfiltration using separation membranes is presented, as a result of the pore diameter of the membrane of 1.4 μm all additional skim milk components pass through the membrane and only the bacteria and spores are retained. Since the bacteria and the fat globules have approximately the same size, in membrane filtration the fat must be centrifuged off before microfiltration and only the skim milk is microfiltered. The resulting retained material (bacteria concentrate) and the cream are flash pasteurized with an ultraflash pasteurizing temperature system at 130° C. for 4 seconds and then returned again to the process milk. In this way milk which is largely free of spores can be processed. It has been shown using a pilot system with a capacity of 500 l milk per hour that the concentration factor is 20:1 (5% retained material) and that the butyric acid bacteria and spores during microfiltration by means of separation membranes are separated with very high efficiency, the number of other bacteria also being reduced.
In addition to the aforementioned membrane separation processes the aforementioned technical article also discloses sterilization separation, so-called bactofugation being the best known process for reducing bacteria. In the pertinent process which has been known for more than 20 years, use is made of the effect that the spores are specifically heavier than the milk, in order to improve efficiency the sterilization separation often being done twice in succession. Other sterilization processes are so-called high pressure processes in which the treatment of the milk is done with hydrostatic pressure from 1,000 to 10,000 bar for two to 60 minutes at 20 to 30° C. In these processes the vegetative microorganisms in particular are effectively killed by the pressure on their membrane. Spores and enzymes are however largely resistant. Moreover the casein micelles and whey proteins are adversely changed.
Furthermore ultrafiltration is known as another membrane separation process in which using a pressure difference high-molecular substances are separated from low-molecular substances and concentrated. In ultrafiltration of skim milk the proteins (caseins, whey proteins) of the milk are retained by the membrane in the retained material, while salts, lactose, and low-molecular nitrogen compounds of the membrane pass into the permeated substance. With an increasing degree of concentration the dry mass and protein content in the UF retained material rise as a consequence, the ratio of casein to whey protein however remains unchanged. The pertinent ultrafiltration membrane separation process in a pretreatment stage of the milk however allows an increase of the incorporation of native whey proteins; this benefits the quality of the milk.
The above described sterilization processes are expensive to use in terms of process engineering, in particular if they use separation membranes, to which the high price of the membrane to be used also contributes and the milk fat (lipids) generally is to be removed before the actual separation process, in order to prevent clogging of the membrane; this is also called blocking. Commercial use with large amounts of aqueous dispersions, such as milk for sterile filtration, therefore cannot be considered.
In order to improve the process steps on a commercial scale and in particular to ensure fully automated operation, under the trademark TETRA ALCROSS™ MB microfiltration systems with ceramic membranes of the Tetra Pak company are used; they likewise resort to microfiltration by means of separation membranes which however are made in this case as ceramic membranes of aluminum oxide and/or other metal oxides, with a pore size between 0.1 to 1.4 μm. As a result of using ceramic membranes they can be sterilized with hot water, can be highly loaded, are chemically resistant in a wide pH range, and have high bursting pressures. The criteria for selection of ceramic membranes are formed by the retention rate for protein and the desired effects of protein fractionation, for example, separation of casein and whey proteins and the separation of microorganisms. The ceramic membranes used produce a low, constant and uniform transmembrane pressure along the entire membrane surface; this allows a high and constant throughput. With the pertinent microfiltration systems a generic process may be carried out, fully automatic operation being attainable with high throughputs. As a result of using ceramic membranes as the separation membranes however the known process is also more expensive and complex in implementation Furthermore the ceramic membranes likewise tend to blocking.
PCT-WO 96/32021 discloses a generic process which however allows only bacteria reduction of milk serum, i.e., only of the continuous phase of a dispersion. Therefore the disclosed process according to the PCT publication relates to filtration of a homogenous aqueous solution, charged (polar) deep-bed filters with their filter media being used as the deep-bed filter means. With the pertinent known process using charged deep-bed filter media, the sterilization of two phase systems of the type of an aqueous liquid-liquid dispersion as a two-phase mixture such as milk would not be possible and in this connection, as suggested by the PCT publication, should filter aids be used, their immediate blocking and thus their becoming unusable would have to be expected.
A similar process is also the subject matter of EP-A-0 798 003 which allows titer reduction of viruses in an aqueous solution, likewise using electrically charged deep-bed filter media. With the respective approach case the bacteria reduction of a nonhomogeneous two-phase or multiphase mixture, such as milk or skim milk, likewise is not possible. If in the disclosed approach polar (charged) filter media were to be used in two-phase mixtures, in static filtration, immediate blocking of the filter medium would occur.
DE 40 26 365 A discloses a sterile-filtered, casein-containing colostral milk and a process for its production. The colostral milk is set to a pH of 2.5-3.5 and then clarified and sterile-filtered. Preferably the colostral milk is defatted before this treatment and after acid treatment it is set to a pH of 5.5-8 prior to the filtration measures. In the known approach it was surprisingly found that colostral milk can be clarified and sterile-filtered without removing the casein fraction beforehand, if the optionally defatted colostral milk is acidified to a pH of less than 3.5. When the pH is reduced the casein precipitates at a pH 4 to 5, but goes back into solution by a further drop of the pH. This acid solution is changed in its properties compared to the colostral milk to the extent that it can now be sterile-filtered.
The known approach does not indicate whether the deep-bed filter used for example for clarification is made polar or apolar. But the technical article of Hou K. et al.: “Capture of latex beads, bacteria, endotoxin and viruses by charge-modified filters”, Applied and Environmental Microbiology, Washington, D.C., US, Vol. 40, No. 5, November 1980 (1980-11), pages 892-896, XP001053164, ISSN: 0099-2240 discloses that the charge of the filter necessarily changes at low pH so that the deep-bed filter which is optionally used apolarly according to the above-referenced citation (DE 40 26 365 A) becomes polar in any case during use.
In the indicated disclosure after Hou K. et al. it is essentially a matter of separation of viruses from aqueous solutions, latex beads being used for simulation of viruses, in order to have scientifically precise particle diameters. The latex results can thus be applied to viruses. In the known approach an experiment with an uncharged filter is described in which only the so-called “sieve effect” of the filter is working. Hou K et al. describes the general relationship of the charge of the filter and the charge of the components which are to be filtered. Depending on what is present in the aqueous dispersion, a charge which is different in each instance or no charge of the filter is appropriate in order to prevent rapid blocking of the filter. But the charge only plays a role when it is a matter of separation of particles with a diameter which is smaller than the “sieve effect” of the filter. While therefore Hou K. et al. are concerned with the reduction of the total bacteria count in aqueous dispersions which are subject to the sieve effect of the filter, as a result of the complexity of the systems encompassing milk and milk products, the results obtained cannot be applied in this regard.
On the basis of this state of the art, the object of the invention is to further improve the known processes such that efficient reduction of the total bacteria count in aqueous dispersions (two-phase and multiphase systems) by sterile filtration is possible economically and with high throughputs on a commercial scale, avoiding blockages on the filter means and prior separations of lipids and/or proteins before actual sterile filtration. The pertinent object is achieved by a process with the features of claim 1 in its entirety.
In that, as specified in the characterizing part of claim 1, the aqueous dispersions are nonhomogeneous two-phase or multiphase mixtures, specifically in the form of animal milk and products obtained from it, such as skim milk with a definable content of dispersants, in particular lipids and proteins, and in that the deep-bed filters (18, 22) of the respective deep-bed filter means are electrically uncharged (apolar), in that the total bacteria count is reduced with an effectiveness which corresponds to a logarithmic bacteria retention value (LRV value) from 2 to 5, because the filter media are provided with a nominal separation rate of <1 μm, in particular <0.5 μm, preferably between 0.2 and 0.3 μm, and because the temperature of the respective dispersion which is to be filtered is chosen to be <60° C., preferably 40° C. to 50° C., but >10° C., sterile filtration of dispersions in the form of animal milk is achieved, on a commercial scale with high throughput amounts.
Since the deep-bed filter means allow continuing filtration and there is no division into retained material and permeated material and in particular taking into account the fact that the deep-bed filter media are more resistant than the known membrane layers, very high differential pressures (between 0 and 5 bar) can be maintained; this is decisive for increasing the amount of throughput to be treated.
Since moreover the filter media in deep-bed filter means are apolar (electrically uncharged), blockages, as in the separation filter membranes and deep-bed filter means used which are otherwise known, do not occur, nor is it necessary to separate the protein and/or lipid parts of the aqueous dispersion beforehand in order to arrive at good sterilization results. The deep-bed filter means can be produced in a plurality of embodiments on the scale of large series and can thus be obtained relatively economically; this in turn clearly reduces the costs of the overall process for reducing the total bacteria count in the aqueous dispersions. If the filters of the respective deep-bed filter means are clogged, they can be easily and economically backflushed and thus they can be regenerated again and again so that the device for carrying out the process as claimed in the invention largely acquires the nature of a production technology and not that of a test setup using membrane filter layers. In particular, compared to ceramic separation filter media, the costs for the use of deep-bad filter means are clearly reduced.
It is surprising to one skilled in the art in the area of sterile filtration that he can obtain much improved steriliation by using deep-bed filter means with deep-bed filters which are clearly more economical than the known separation filter membranes at high throughputs of the unfiltered material and that at the same time the lipid and/or protein content of the dispersion remains essentially unchanged by deep-bed filtration, in particular is not adversely affected. In practical tests it has been shown that the total bacteria count of the filtrate with the filtrate with the process as claimed in the invention is <200 colony-forming units per millimeter (CFU/ml).
Other advantageous embodiments are the subject matter of the other dependent claims.
The process as claimed in the invention is detailed below using a device for its implementation according to the drawings The figures are schematic and not to scale.
The process as claimed in the invention for reducing the total bacteria count (TBC) in aqueous dispersions takes place by sterile filtration of the indicated dispersions. The pertinent dispersion can for example be present in the form of a suspension of animal milk and products obtained from it, such as skim milk with a definable content of dispersants, in particular in the form of lipids and proteins. The task of the process as claimed in the invention is to reduce the bioburden by a deep-bed filtration process without adversely changing the lipid and/or protein content for the product.
Before starting filtration, first the two filters 18 and 22 are sterilized with steam. Furthermore the two filters 18, 22 together with the complete sterile filtration system are started up with hot water and only then switched to raw milk. The required filtration pressure was produced exclusively by the separator 12. In automatic “desludging” the pressure drops to 0 bar differential pressure and builds up again immediately after cleaning. When starting up and rinsing the two filters 18, 22, there is no initial pressure loss.
Sterile filtration of the raw milk is undertaken by means of deep-bed filter means and the filters 18, 22 are accordingly made as deep-bed filters, in particular of the type of deep-bed filter cartridges. In deep-bed filtration the unfiltered material flows from the inside to the outside, in this case however preferably from the outside to the inside, through the respective deep-bed filter cartridge 18, 22 and the filtrate is discharged from the interior of the respective filter 18 and 22 for further use. In deep-bed filtration therefore permanent filtration takes place and the deep-bed filter cartridges used allow high flow rates of the filtering medium (unfiltered material).
With the deep-bed filter means shown in
The filter media used in the deep-bed filters 18, 22 are those with a nominal separation rate of 0.2 to 0.3 μm. Furthermore temperatures for the dispersion to be filtered are chosen to be between 40° C. to 50° C.; this leads to good bacteria reduction numbers. Furthermore in the indicated temperature range the result is that the casein and/or fat in the milk do not block the deep-bed filters 18, 22 and as a result make it at least partially unusable. Deep-bed filters which have proven particularly effective are those carried in the product program of the applicant under BECO-PROTECT KTB 373 with 0.3 μm nominal separation rate, as well as BECO-PROTECT KTB 273 with a nominal separation rate of 0.2 μm Here logarithmic bacteria retention values from 0.8 to 2.3 and from 1.6 to 2 have easily resulted. The pertinent values arise at incident flow rates of 300 liters/hr for a 30 inch deep-bed filter cartridge so that the incident flow velocity against the surface of the cartridge itself is approximately V=1200 l per m2 and hour. Very good values have also resulted when skim milk instead of raw milk is used for the two-stage filtration process.
The deep-bed filter cartridges used in milk filtration with a nominal separation rate between 0.2 and 0.3 μm consist preferably of polypropylene meltblown layers 32 as are shown in
The indicated meltblown nonwoven layers 32 are wound around an inner support pipe 32 which, provided with perforations, is used for the filtrate to drain out of the deep-bed filter cartridge 18, 22. To protect the meltblown nonwoven layers 32 to the outside the respective deep-bed filter cartridge 18, 22 on the outer peripheral side has a jacket 36 which is provided likewise with perforations, for example in the form of lengthwise slots, and in this way allows entry of the unfiltered material into the respective filter cartridge 18, 22. The throughflow direction of the unfiltered substance is shown in
Like sterilization of raw milk, by way of sterile filtration the pertinent filtration can also be easily carried out with skim milk as well, and then a bacteria burden of <100, preferably 10 colony-forming units per millimeter filtrate, can be expected. Thus with 30 inch polypropylene deep-bed filter cartridges at a volumetric flow of approximately 300 l/hr skim milk, the bacteria in the skim milk can be reduced by at least approximately 99%. The important components of the milk, such as proteins and lipids, are preserved in spite of filtration. Sterile filtration by a deep-bed filter allows high flow rates and can be accomplished very economically. The process is therefore also particularly suitable for sterile filtration of whey.
The process as claimed in the invention is detailed below using two sample applications:
According to experience, skim milk has a residual fat content of approximately 0.1% and is thus a highly diluted oil/water emulsion (two-phase mixture). In preliminary tests with polar (charged) filter media (both deep-bed filter media of cellulose and also membrane filters (0.45 μm)) with static filtration immediate blocking of the filter medium is observed.
In many tests, relative to the process as claimed in the invention a two-stage filtration system was integrated into an existing milk processing line. A partial flow of skim milk originating from the separator was statically filtered. The temperature of the skim milk is 42° C.. Filtration is carried out over a time interval of 200 minutes with continuous recordation of the throughput and of the pressure increase during filtration. During the test microbiological samples were taken upstream and downstream from the filter unit at an interval of 60 minutes.
Microbiological Evaluation:
The determination of the total bacteria count in the samples was carried out using the Koch bacteria count method (Methodenbuch [Process Manual] Volume IV. M 6.3.1, 1985 with 4th supplement 1996). To determine the thermodurable (heat-resistant) bacteria the samples were heated in a water bath for 30 minutes to 65° C. and then cooled, further handling such as the Koch bacteria count method. Psychotrophic bacteria (subsisting at cold temperatures) were in turn prepared after the Koch bacteria count process, but incubated for 7 days at 10° C.
Results:
During the test interval only a small pressure rise of 0.3 bar on the respective deep-bed filter unit was observed. The bacteria reduction rates are summarized in the following table:
No depletion of the fat content of the milk was observed during the test (fat content of the milk approximately 0.1%).
The completed process showed that in all examined bacteria classes according to the outlines above a reduction of greater than 99% was achieved. Since in particular thermophilic and psychotrophic bacteria are responsible for the spoilage of fresh milk, they were successfully separated from the skim milk with the described process.
In the running process of a milk processing line, downstream from a separator a whole milk stream of 1000 l/hr was branched off and statically filtered. This partial flow-homogenized whole milk has a fat proportion of 3.5% which is present as an oily phase in an oil/water emulsion (two-phase mixture). Based on the size of the fat globules, each polar deep-bed filter and each conventional filtration membrane (for example 0.45 μm) will immediately become blocked during static filtration according to the known process.
The filtration temperature was 55° C.. The homogenization pressure was 170 bar with single-stage homogenization. During the test interval the flow rate was kept constant and the pressure rise on the filter was recorded. At a 30 minute interval microbiological samples were drawn and studied according to the aforementioned example.
Results:
The pressure difference during filtration at the end of the filtration time of 140 minutes was only 0.2 bar.
Results of microbiological studies:
These results show that the partial flow-homogenized whole milk was sterilized with bacteria reduction values of an average 90%. By means of the downstream pasteurization step whole milk with bacteria numbers less than 10 BE/ml can be obtained. The storage quality of this whole milk is prolonged without the disadvantages of the otherwise conventional pasteurization arising.
With the process as claimed in the invention dead-end filtration can be carried out, which compared to the known filtration processes, also compared to the known crossflow filtration, also offers the advantage that the filter or filter elements can be backflushed for their regeneration and are otherwise electrically apolar; this helps reduce the cost of supplies for the filter elements.
| Number | Date | Country | Kind |
|---|---|---|---|
| 101 58 009.6 | Nov 2001 | DE | national |
| Filing Document | Filing Date | Country | Kind |
|---|---|---|---|
| PCT/EP02/12794 | 11/15/2002 | WO |
