GB2117267A - Ultrafiltration of fermentation effluent - Google Patents

Abstract

A method for the treatment of an effluent by-product from a fermentation process, such as spent wash or pot ale, comprises passing the effluent through an ultrafiltration membrane. The membrane may be of the tubular type, a plate and frame design or a wide gap spiral type and preferably has a molecular weight cut- off of 18,000. The membranes are preferably operated at a temperature of from 50-55 DEG C. The permeate may be passed to a reverse osmosis unit for further treatment. The concentrate obtained from ultrafiltration and/or reverse osmosis may be passed to an evaporator and/or drier to produce an enriched feed supplement in the form of distilled dark grains. <IMAGE>

Description

SPECIFICATION Improved process for the treatment of waste by-products of whisky manufacture This invention relates to the treatment of waste by-products of the whisky industry.

The major by-products of malt whisky manufacture are spent grains or draff and pot ale, the high boiling point residue from the distillation process. Wet draff and dried draff (light grains) have long been used for animal feed. Pot ale has in some instances been used as a fertilizer but is generally processed further by evaporation down to a syrup or completely dried to produce distillers dried solubles. In recent years the syrup has been added to wet draff and the mixture dried to produce distillers dark grains. The products because of their inherent high protein content have been acceptable to the animal feed compounding industry and have produced financial benefits for the distilleries.

The process for the production of such products is however, highly energy intensive and the rising costs of operating such plant is of major concern to the distilleries. The cost of running a multi-stage evaporator to concentrate the solids being of the order of two thousand pounds per week for a medium size distillery.

Supplies of pot ale are so large, in the region of 100,000 gallons per distillery per week, that if unutiiized a serious disposal problem arises. The problem is compounded by the fact that the pot ale as an effluent exhibits a very high bio-chemical oxygen demand (B.O.D.) and may contain relatively high levels of metallic contamination. "Effluents from Malting 8 Brewing" by P.C. Isaac and G. Anderson.

Journal Inst. of Brewing Vol. 79. 1973.

The high protein content of the solids in pot ale make them an attractive proposition for incorporation in animal feedstuffs, provided they can be recovered in an economic manner. Increasing capital and operating costs of processing and drying plant are restrictive and the feasibility of continuing to use this equipment economically appears doubtful. Evaporators at present in use in the distilleries are estimated to cost 1 800 per week to run on fuel alone to produce pot ale syrup at 40% dry matter. The present invention proposes an alternative low energy process for the concentration and recovery of the valuable solids from the pot ale by-product of the distillery process. Assuming all distillers use such a process for concentration, the potential energy savings of the process of this invention are considerable.

Pot ale is the liquid residue that remains in the first (wash) still. The distillate enters a second (spirit) still from which the final product is condensed, leaving spent lees as the liquid residue. At some distilleries the pot ale and spent lees are combined together and called pot ale and this may also be further diluted by water used for washing equipment. Pot ale will therefore differ in character and composition between distilleries.

Pot ale consists largely of suspension of dead yeast cells in a colloidal solution. Amino acids, polypeptides, minerals, vitamins, fuel oils and unfermented carbohydrates are also present. The significant amounts of yeast and lactic acid bacteria contributes to riboflavin, thiamine and pathothenic acid. The amino acid content is comparable to dried yeast and large amounts of arginine, histidine and lysine are a result of yeast synthesis. The nutritive quality of the solid would be high and balanced if the product was supplemented with a methionine rich food such as marine products.

Characteristics of a typical pot ale pH 3.5 Dry matter 2 < 5% Crude protein 30% B.O.D. 20,000 mg/litre Copper content 5 mgrams/kg Characteristics of a resulting pot ale syrup after evaporation pH 3.5 Dry matter 40% Crude protein 30% Cu 65 1 120 mgrams/kg Price 45 per Tonne The strongly acidic nature of the pot ale and syrup has obvious disadvantages for incorporation in feedstuffs and for disposal as an effluent.

In addition pot ale solution exhibits the characteristics of a lyophobic sol which presents difficulties in recovery of solids since they are suspended in a colloidal state.

A discussion of the problems associated with flocculation of colloidal yeasts may be found in "Effect of some Monovalent and Diva lent lons on the Flocculation of Brewers Yeast Strains" by G. G. Steward and T. E. Goring. Journal Inst. of Brewing Vol. 82. 1976.

Scotch malt whisky is produced in a number of relatively small distilleries, about one hundred.

The brands which are not straight malt whisky are marketed as a blend of malt whiskies with another type of whisky produced in Scotland called Scotch grain whisky, or because it is distilled continuously in a Coffey-type patent still, patent-still whisky. Most Scotch whiskies available on the international market consist of blends with 3040% malt whisky. Within the blend, there may be as many as 20-30 individual malt whiskies and grain whiskies.

The cereals used in the manufacture of Scotch grain whisky are malted barley along with a high proportion (up to 90% of maize (called corn in North America) and occasionally some rye. Scotch grain whisky is manufactured in a continuous distillation process.

Some distillers filter their worts from the mash before fermentation, but others, especially in North America, ferment the whole mash, i.e. ('in grains' fermentation).

In such a continuous process, the raw material maize, is ground and cooked under pressure. A limited quantity of malted barley (approx T wt. of maize feed) is added to provide the necessary enzymes for hydrolysis of the starch to simple carbohydrates. The resultant mash is passed to the mash .tun wherein the temperature and pressure are lower and the bulk of the conversion of the starch occurs here.

The contents are optionally filtered to separate out the wet draff or the whole contents are cooled and fed to the fermenter. Yeast is added and the fermenting wash is fed to the distillation columns which comprise a wash column and a rectifying column in series. The raw whisky is recovered from the top of the rectifying column and spent wash is collected as a bottoms product from the wash column.

The spent wash is comparable with the pot ale produced in batch processes. it is a somewhat more dilute, continuous production giving approximately 9.5 gallons spent wash per gallon proof compared with approximately 5 gallons pot ale per gallon proof in a batch process.

The spent wash could therefore be regarded a dilute pot ale. incidentally, whereas 'spent wash' is terminology used in the Scottish art, it may be equated with the term 'thin stillage" used by others in the art to describe the liquid effluent remaining after the separation of bulk solids from 'whole stillage' i.e. that portion containing 'distillers solubles'.

In general, spent wash has a solids content of approximately 3.6% before screening and 2.5% after, of which 2/3 are considered to be dissolved whilst the balance are suspended.

The pH of this spent wash is from 3.3 to 3.8 and the biochemical oxygen demand (B.O.D.) after screening amounts to about 10,000 p.p.m.

In the past this has been subjected to concentration by evaporation, neutralisation and spray drying to 3% moisture to give distillers dried solubles (D.D.S.).

An object of the present invention is to obviate-or mitigate the aforesaid disadvantages and thus provide a more economic process for effective utilization of pot ale.

According to the present invention there is provided a method for the treatment of an effluent byproduct from a fermentation process which comprises passing the effluent through an ultrafiltration membrane.

Optionally the effluent may be gravity-settled or centrifuged and decanted before passing through the ultrafiltration membrane to obtain enhanced concentration.

The particular source of effluent will have a bearing on the treatment required, e.g. an in-grains fermentation spent wash will require a different treatment from a pot ale from a malt distillery.

Preferably the ultrafiltration membrane is of the tubular type of a plate and frame design.

A suitable alternative in most cases, except for an in-grains spent wash, is a wide gap spiral ultrafiltration membrane.

Although it has been found that adequate concentration (de-watering) and substantially total removal of solids and high molecular weight solubles can be achieved with ultrafiltration, further concentration may be achieved by passing the concentrate from the ultrafiltration plant to conventional evaporators and driers.

In some cases it may be advantageous to follow ultrafiltration with reverse osmosis in order to further reduce the water content and B.O.D. of the effluent.

Use of ultrafiltration gives many advantages, for instance, pot ale may be reduced to tankertransportable volumes for treatment at a processing plant remote from the distillery.

The ultrafiltration plant can be used to process liquor from a draff press in conjunction with pot ale to produce a permeate which would be fed to a conventional digester reduce the B.O.D., and the concentrate could be mixed with the draff to produce an enriched feed supplement in the form of distillers' dark grains. Energy from the digester can be used to supplement that necessary for the evaporator or drier.

The invention may be further understood by reference to the following preliminary test results, pilot plant data and drawings in which: Fig. 1 shows a block diagram indicating schematically the general layout of the pilot plant Fig. 2 is a graph relating to the pilot trials on fresh pot ale; Fig. 3 is a further graph relating to the pilot trials on stored pot ale; Fig. 4 is a graph relating to pilot trials on pot ale using a decanter/centrifuge prior to ultrafiltration; Fig. 5 is a graph relating to pilot trails on pot ale from a different source; Fig. 6 is a graph relating to pilot trials on spent wash (uncentrifuged); and Figs. 7-10 show process flow charts which illustrate some of the systems possible incorporating ultrafiltration with ancillary equipment.

In the following preliminary tests three ultrafiltration (UF) members were tested, each being a five foot long tube containing 0.1 m2 of filter area.

The first had a nominal molecular weight (MW) cut-off circa 5,000; the second a molecular weight cut-off circa 18,000; and the third which had limited caustic and chlorine resistance, also had a molecular weight cut-off circa 1 8,000.

Of these the second gave the most promising results.

Test 1 -- Clarified Pot Ale A sample of the clarified pot ale passed through the first membrane gave a dark lager coloured permeate, but had a low permeate flow rate of 30 1 m-2h-' (standardized) The second membrane gave a much darker permeate and a flow rate of 44 1 m-2h-1 (standardized) The third membrane gave a similar permeate at a flow rate of 47 1 m-2h-' (standardized) Due to equipment limitations a concentration ratio of only 2.5:1 was obtained.

Test 2 - Raw Pot Ale Pot ale which had cooled to 250C was processed using the second and third UF membranes described above at a pH of 3.6, with initial fluxes of 95 1 m-2h-' and 110 1 m-2h-l respectively (standardized) After 21 hours of processing a concentration factor of 4 was obtained, the final corrected fluxes being 70 1 m-2h-1 and 75 1 m-2h-4 respectively.

Table I overleaf gives the results of sample analysis.

It was projected that in a full scale process operation, a concentration ratio of 10:1 with a flux ratio of 75 1 m-2h-' using a tubular membrane with a cut-off of 18,000 may be obtained.

TABLE I

Dry Crude Matter Protein Calcium Sodium Copper Description PH g/kg g/kg g/kg g/kg g/kg Clarified pot ale Solids 8.2 166.3 79.0 3.02 3.06 75.8 Supernatant 8.2 38.4 11.5 0.03 2.65 1.1 UF Conc.

Supernatant 8.5 48.3 14.1 0.02 4.21 36.3 Raw pot ale UF Concentrate 3.6 57.5 28* - - 20.6 UF Effluent 3.6 29.2 8.9 0.19 0.10 1.6 R. O. Concen. 3.6 31.2 23.0 - - 1.6 R. O. Effluent 3.6 Negl. Negl. 0.03 0.03 3.6 * Estimated Value Pilot Plant Trials Pilot plant trials were carried out on line at a distillery to test the use of ultrafiltration on a larger scale. Twelve tubular membrane units were connected in series to provide a total membrane area of 2.4 square metres.

Many of the tests were carried out on pot ale that had been allowed to settle in the washback, thereby providing artificially high values of solids in the feed material. It was considered that such a procedure would provide a "worst case" condition for the equipment and accordingly due note of this fact should be taken when assessing results.

The general pilot plant layout is indicated schematically in Fig. 1. The heat exchanger was introduced to the system because whereas pot ale straight from the still was at about 800C the temperature limitation of the UF equipment was 550C, and a delay was required for cooling.

Batch 1: 1200 gallons of pot ale were fed straight from the still to the washback at 760C, after initial tests were run the contents were allowed to cool at 550C.

Initial Concentration Initially a sample of 200-300 gallons of Batch 1 were processed to evaluate the membrane characteristics. The final concentrate was sampled and analysis gave a dry matter value of 17.9%. The initial concentration tests gave an initial flowrate of 150 1 m-2h-' and fell to 5 o 10 1 m-2h-' at cut-off, Fig. 2 Run 1 (Aged) giving an average flux of 80 1 m-2h-'. They also indicated that a concentration ratio of 6 or 7 times could be expected, Fig. 3 Run 1. It was estimated that to obtain 10 times concentration of the full batch would take 20 hours with the membrane area available, 2.4 m2.

Full Batch Concentration The concentrate obtained from the initial tests was returned to the washback via the recirculation loop. Full concentration run was attempted. The permeate was transferred to the milk tank for processing by the reverse osmosis equipment.

The data from the full batch concentration is shown in Table II and flux rate is plotted in Fig. 2 Run 2.

TABLE II

Temp. Flowrate Time Pressures (OC) (seconds) Flux (1 m-2h-') 0202 3.96/1.66 34 1.51/35 64 0215 4.0/1.86 43 1.5/35 64 0224 4.0/1.26 46 1.5/35 64 0300 4.0/1.0 50 1.5/35 62.5 0400 4.2/1.5 48 1.5/43 52.0 0500 4.0/1.0 48 1.5/50 45.0 0600 4.0/1.0 47 1.5/55 40.9 0700 4.0/1.0 46 1.5/60 37.5 The drop off in flowrate can be seen, after attention by the technician the flowrate was restored, but by 1 500 hours the flux had fallen to 50 1 m-2h-'. Ity was considered that the reduction in flux from the expected 80 1 m-2h-', was probably due to bacterial growth as the pot ale aged. This could cause membrane clogging if the bacteria lie down in strings on the membrane. The run was aborted at 1 700 hours, because of the reduction in flow.

Batch 2: A second test batch of pot ale was held in the still for a day and was transferred to the washback the next day after passing through the newly installed heat exchanger loop. The temperature of the pot ale in the washback was 54.50C7A sample (2S8E) of the new pot ale feed was taken and a concentration run over a 7 hour period undertaken. At this time it was appreciated that it was impractical to attempt a full batch concentration with the limited equipment available.

Samples of permeate (2S6) and concentrate (2S2) were taken for analysis. The flux rates and concentration ratio for this test are shown in Fig. 2 Run 3 and Fig. 3 Run 3 respectively.

Batch 3 A further batch was cooled and transferred to washback and another concentration run undertaken, samples were not taken for analysis, the flux rate and concentration rate are, however, shown in Fig. 2, Run 4 and Fig. 3 Run 4 respectively.

TABLE Ill Summary of Data from Runs 1-4 re Flux Rate and Concentration Ratio:

Initial Final Average Concentration Dry g/kg Batch Run Flux 1 m-2h-l Flux 1 m-2h-' Flux 1 mlh-l Ratio Matter 1 1 145 10 75 7 179.0 1 2 65 48 52 N/A N/A 2 3 82 23 64 8 189.0 3 4 118 10 60 5-6 N/A The data shows the initial flux variation between runs and that this initial flux variation is quickly smoothed out. The flux depends also closely on the volume concentration ratio.The concentration factor is the most useful way of plotting the available data. The flux is probably limited either by the soluble protein content of the pot ale or the viscosity of the feed material due to fine suspended constituents.

The ageing of the pot ale appears to effect the flowrates, this may be due to bacterial build up on the membrane, as bacteria may be lying down in strings and clogging the membrane pores. Membrane cleaning however, proved to be easier than expected, and should not be an obstacle to the use of the membranes.

For fresh pot ale it appears that one could expect average flux rates of 60-80 1 m~2h~1 and a concentration factor of the order of 7 times.

Analyses of batches 1 and 2 are shown in Table IV overleaf.

Observations 1. High level of suspended solids in feed, due to settling of solids in the tank, for saturation of UF membrane.

2. Copper values as received, not as in dry matter.

3. Crude protein value in UF concentrate high at 50% and 51%.

4. Suspended solids in total dry matter only 68%, balance 32% will be dissolved solids with molecular weight greater than 1 8000 i.e. proteins etc. (Compare decanter).

5. Overall retention by membrane a) 87% for total solids b) 91 % for total protein c) 100% for suspended solids and calloids by definition.

Further Ultrafiltration Tests were performed using the same ultrafiltration equipment.

Ultrafiltration of Centrate from Decanter -- Centrifuge On the 1 9 November a sample of the centrate from a decanter was transferred to a tank and processed through the UF equipment to determine flux rates and concentration ratios. Samples of centrate, UF concentrate and permeate were taken for analysis. The suspended solids were low and a far higher concentration would have been possible than with raw pot ale, which would seem to suggest that viscosity factors dominate.

600 litres of centrate were concentrated at a temperature of 550C to approx 27 litres, Table V gives the data.

TABLE IV Results obtained from pilot tests on pot ale Analysis of Batches 1 and 2

DRY SUSPENDED CRUDE CARBO MATTER SOLIDS PROTEIN HYDRATE CALCIUM SODIUM COPPER Copper DESCRIPTION PH g/kg g/kg g/kg g/kg mg/kg mg/kg mg/kg in D/M UF Liquor Concentrate 3.7 179 - 89.9 14.8 - - 34.9 195.02 UF Liquor Effluent 3.7 24 - 7.6 3.9 - - 34.6 RO Concentration 3.6 58 - 18.6 1.2 160 130 2.1 Effluent R.O. Effluent 4.3 0.77 - 0.2 Nil 15 15 Trace Raw Pot Ale 3.6 58 30 27.7 4.3 13.3 34.9 7.2 125.09 UF Liquor Concentrate 3.7 189 129 97.4 12.2 - - 121 641.14 UF Effluent 3.6 24 8.9 3.1 - - 7.5 TABLE V Flux Rates Decanter Centrate

Time t secs 1.5 1/Sec Pressures Temp. Flux 1m-2h-1 1315 0 18.5 4.0/1.1 500C 121.6 1335 20 19.19 4.0/1.1 500C 117.3 1355 40 20.58 4.0/1.05 500C 109.3 1415 60 21.97 4.0/1.05 500C 102.4 1435 80 23.95 4.05/1.1 490C 94.0 1555 100 25.95 4.0/1.1 470C 86.7 1500 105 Stopped Topping Up 1510 115 28.75 4.0/1,0 460C 78.15 1520 125 29.33 4.0/1.0 460C 76.7 1530 135 30.25 4.0/1.0 450C 74.4 1535 140 33.17 4.0/1.0 450C 67.83 The plot of this data is shown on Fig. 4, as run 1 The test was repeated on the centrate from a different batch of pot ale, which contained some solids carry over. The system concentrated 374 litres to 30 litres. The batch of centrate was stored overnight and therefore, was not fresh, however as can be seen from Fig. 4 run 2 a concentration factor of 9 times was achieved. The upper limit of concentration was not obtained, but will probably be of the order of 20 to 25 times (V/V). No samples for analysis were taken from either batch.

Ultrafiltration of Pot Ale from a Different Source The UF equipment was used to concentrate 1 50 gallons of pot ale from a different distillery. The data obtained is shown in Table VI and flux and concentration ratio in Fig. 5.

TABLE VI

Time t (secs) Temp ( C) Permeate/1 .51 Pressures 1230 0 50 22 secs 4.0 1.0 1300 30 49 27 " 4.0 1.05 1330 60 49.5 32 " 3.9 1.0 1400 90 50 31 " 4.0 1.05 1430 120 49.5 30 " 4.0 1.05 1500 150 50 31 " 3.5 1.0 1530 180 50 32 " 3.8 1.0 1600 210 50.5 36 " 3.9 1.0 The results indicate that this pot ale can be concentrated 14:1 (V/V) and that the suspended solids in the feed were lower, however this could be a function of sampling method. No analysis was performed on these samples.

Ultrafiltration of Biological Processing Plant Effluent The objective of these tests was to check on the likely flux if the UF equipment were to be used as a filter on the outflow from an Anarobic Digestion Plant. The UF plant being used to reduce the B.O.D. of the effluent from the plant and improve the amount of methane generated. In addition any surges of biodegradable material passing from the distillery through the digester would be restrained by the UF plant from going into the environment. The tests were run on cold effluent and indicated that a flux of the order of 100--125 lm-2h-' could be expected. No samples were taken for analysis, except for two B.O.D. samples. It is obvious from the results that the UF equipment could be used for this purpose.

Conclusions UF Tests on Pot Ale 1. On the basis of laboratory tests and small scale pilot trials it would appear that the ultrafiltration membrane could be used successfully to concentrate raw pot ale to a dry matter of the order of 20%.

The performance of the equipment is such that one could expect flux rates of the order of 60-80 1 m~2h-1 and a concentration factor of the order of 6 to 8 times.

2. Limited sample analysis indicates that the membrane having a MW cut-off at 1 8000 retains 87% of the total solids and 91% of the total protein. The physical operation of the membrane is such that suspended solids and colloids should be retained totally.

3. The crude protein content in the concentrate is high (50%) compared to equivalent products at approximately 30%. The significant factor, however is that only 68% of the total dry matter is suspended solids. The balance 32% will be dissolved solids with a molecular weight greater than .18000 and may contain a large proportion of protein.

4. The permeate from the UF equipment can he processed with reverse osmosis equipment.

5. No difficulty was experienced in cleaning the membranes with raw pot ale as the feed material.

6. At present it is not understood why the initial flux rates vary from run to run, or whether the flux is limited by the soluble protein content of the pot ale or the viscosity of the feed material.

7. The UF system can be used successfully to concentrate the centrate from the decanter, results indicated that a concentration ratio of 20 to 25 times (v/v) could be expected. The voluble soluble protein fraction in the centrate should be recovered by the UF treatment.

8. Tests on pot ale from three separate distilleries would appear to indicate that the use of ultrafiltration techniques may have general application 9 most malt distilleries. Flux rates and concentration factors, however, will probably vary with distillery characteristics.

9. The use of an ultrafiltration system for treatment of the effluent from an anerobic digestion plant is feasible, flux rates of the order of 100--125 1 m-2h-l could be expected. The system used as a filter could reduce effluent BOD values, possibly increase methane generation and smooth out surges in plant operation.

Ultrafiltration of Spent Wash (Grains-in) Uncentrifuged spent wash from a grain distillery was tested on the ultrafiltration (UF) equipment, the permeate was further processed using the reverse osmosis (RO) unit. The fluxes obtained on this initial test was limited by temperature, the spent wash being at room temperature 2O-250C and rising to 350C at the end of the run. Optimum temperature of operation of membranes is 50-550C.

The flux and concentration ratios achieved are shown in Fig. 6. The permeate from the UF unit was a clear light yellow liquid, which was processed easily with the RO unit. As with the pot ale permeate, the liquid will be warm and sterile. The results of the analysis of samples taken is given in the Table Vli below.

Typical Spent Wash: (in grains fermentation) Total solids = 89% Dissolved solids = 3% Suspended solids = 56% TABLE VII Results:

Total Susp. Crude Solids Solids Prot. Ca Na Cu Description PH g/kg g/kg g/kg mg/kg mg/kg mg/kg Uncentrifuged Raw Spent Wash 4.1 95.3 - 26.5 3.8 3.3 0.06 UFConcentrate 4.1 116.4 - 33.2 - 0.27 UF Effluent 4.1 17.6 4.2 - - 0.01 R O Concentrate ex UF 4.0 64.0 -13.0 --0.03 R O Permeate ex UF 4.5 0.2 - Nil 1.9 7.9 Trace B.O.D.'s RO Permeate = 14.8 ppm Ultrafiltration of Spent Wash (grains-out) 45 gallons of spent wash at a temperature of 350C was processed using 14 tunes (5') giving 1.4 m2 total membrane area over a period of approximately 2 hours. The system was very viscous and towards the end of the run pump vibration was evident. Table VIII overleaf shows data relating to the process.

Air drying at 70-100 to determine dry matter in samples taken periodically gave the following results shown in Table IX overleaf. Freeze drying would give higher results.

TABLE VIII Results:

Time Permeate Flow AP (hours) for 1 litre (secs) bar Temp C 1m-2h-1 11.55 33.6 4.6/1.0 350C 76.5 12.00 35.5 4.6/1.0 34 72.4 12.12 35.0 5.0/1.5 35 73.5 12.21 35.0 5.0/1.0 - 73.5 12.34 37.0 - 40 69.5 12.41 39.0 5.0/1.0 - 65.9 12.52 40.0 5.0/1.0 36 64.3 13.02 42.0 4.5/1.0 38 61.2 13.20 40.0 4.5/1.0 42 64.3 13.31 44.0 4.2/1.0 40 58.4 13.33 44.0 4.5/1.0 40 58.4 13.38 48.0 4.0/0.5 40 53.6 13.43 75.0 3.25/0 40 34.3 13.48 110.0 2.5/0 40 23.4 13.50 118.0 2.5/0 40 21.8 TABLE IX

Time (mins) D.M. (%) 00 2.86 26 4.1 52 4.1 91 8.6 112 10.4 118 11.0 Conclusions UF Tests on Spent Wash 1. The ultrafiltration on uncentrifugal spent wash from an in-grains fermentation process was also successful, producing a clear yellow sterile permeate. Further tests are required to establish flux rates and concentration factors, although it was estimated a concentration factor of the order of 3 could be expected and flux rates were higher than for pot ale.

2. The implications of the use of UF in grain distilleries are most important, because of the high volume throughput of spent wash and the fact that separated permeates are fed back to the mashing tanks. The UF permeates, should be sterile thereby reducing contamination build-up in mashes.

Energy savings in the production of by-products could also be very significant.

3. Ultra-filtration on the spent wash from a grain distillery, with a grains-out fermentation process, has proved to be very effective due to the fact that such spent wash has a lower suspended solids content.

The main advantages of adopting this alternative technology for the treatment of pot ale or spent wash lie in the energy saving by replacement or reduced use of evaporators. A real energy saving of up to 62% on present running costs can be achieved.

Use of ultrafiltration alone or in tandem with a reverse osmosis system leads to the production of novel liquid products suitable for use as animal feed, this providing higher % energy savings and investment recovery.

The following analysis indicates the potential savings of this system in comparison with conventional evaporators.

Present Method by Evaporation (Consider Input of 100 kg of Pot Ale to System) Typical Pot Ale Composition - 1% S.S. (suspended solids) 3% D.S. (dissolved solids) This is at present reduced at distillery site to P.A.S.

(Pot Ale Syrup) at approx. 40% D.M. (Dry Matter) rate to either (a) sell off site (b) as feed-stream for production of D.D.G. (Distillers Dark Grains).

Water

as foul viaE.O.T Plant . . . * condensate Via to O. D. Plant .: E-f toenviroment 2000 > to ... Evaporation X l 3 1 2.8 (9.5kg.P.A.S &commat; 40goD. M.) Raw Pot Ale Thus 96-5.7 = 90.3 kg./water removed &commat; 0.4p./kg = 36.12 p.

and 9.5 kg./P.A.S. produced &commat; cost of 36.12 p. = 3.80p./kg.

38.0/tonne Proposed Alternative Method:- (For 100 kg Pot Ale Input) ULTRA-FILTRATION REVERSE OSMOSIS

R 89. 33 keg. / R /. wat /Ipermeate 80. 81 kg. /water 2.33 D. S. 0.2D.M.

F permeate via B. O. D.

13 plant to enviroment Raw Pot Ale Cone ntrate C centrate LC - 'I iliilE 'iliffiTH s w .66 2.1 3 U/F conc. line at 20% D.M. R/O conc. line at 20% D.M.

(up to 50% protein) (sugars and carbohydrates) Option 1 (a) Combine U/F concentrate and R/O concentrate, use existing evaporator to produce 40% D.M.

(i.e. P.A.S.) i.e.

h + Wi = R-nrfLrator r % \ \ 0 t 7&num; 11 ll 1 .66 2.13 2. 79 12. 79 (19.97% D.M.) (40% D.M. P.A.S.) Thus Ultra-filtration process 89.33/kg. water ( .02 p/kg. = 1.79 p.

Reverse-osmosis process 80.81/kg. water &commat; .01 p/kg. = 8.10 p.

Evaporator process 9.5 /kg. water &commat;.40 p/kg. = 3.80 p.

Total = 13.69 p.

So 9.48. of P.A.S. produced for 13.69 p = 1.44 p/kg.

=#14.4/tonne.

Option 2 Combie @@/F concentrate and R/O concentrate and @@ Combine U/F concentrate and R/O concentrate and sell off site at nominal 20% D.M. rate: Product is 18.98 ka. (l 19.97% D.M. rate.

Cost 1.79 p. + 8.10 p. = 9.89 p.

Thus 18.98 kg. costs 9.89 p. to produce = 0.52 p./kg.

= 5.20/tonne.

The above schemes relate to energy costs only, excluded are capital equipment costs, maintenance, etc.

Claims (9)

1. A method for the treatment of an effluent by-product from a fermentation process which comprises passing the effluent through an ultrafiltration membrane.

2. A method according to claim 1 wherein the membrane is of the tubular type.

3. A method according to claim 1 wherein the membrane is of the plate and frame design.

4. A method according to claim 1 wherein the membrane is of the wide gap spiral type.

5. A method according to claim 1 or claim 2 or claim 3 or claim 4 wherein the membrane has a molecular weight cut-off of 18,000.

6. A method according to any one of the preceding claims wherein the membranes are operated at a temperature of from 50-550C.

7. A method according to any one of the preceding claims wherein the permeate is passed to a reverse osmosis unit.

8. A method according to any one of the preceding claims wherein the concentrate obtained from ultrafiltration and/or reverse osmosis is passed to an evaporator and/or drier.

9. A method according to any one of the preceding claims wherein the effluent is subjected to a clarification treatment prior to ultrafiltration.