SYMPOSIUM PAPER

The quality of milk in relation to cheese manufacture*

JEAN BANKS Hannah Research Institute, Ayr KA6 5HL

The gross composition and bacteriological quality of milk used in cheese manufacture can huve a significant influence on both the yield and the quality of the cheese produced. Within the pa.rt I0 years there have been changes in terms of milk production and utilization in the dairy industry i i the United Kingdom which could influence both the composition of milk and its bacteriological quulity. Methods to eliminate any undesirable effects of changes in milk quality are available. The advantages of standardizing the casein to fat ratio in milk and extending the storage life ofmiik by thermization and deep cooling are discussed in relation to cheese manufacture,

INTRODUCTION Within the past decade there have been significant changes in methods of milk production and utilization within the dairy industry in the United Kingdom. These changes have arisen mainly through the implementation of the regulations of the European Economic Community on the common organization of the market in milk and milk products, which came intoeffect in the UK in 1984.

While in terms of milk production the introduction of quotas resulted in a reduction in the overall volume of milk produced, to some extent it has also led to changes in the gross compositional quality of milk in the UK. Changes in both the composition of milk and its availability are of particular significance to the cheese manufacturer. The composition of milk used in cheese manufacture will affect both the yield and the quality of cheese. Additionally, the lack of availability of milk for manufacture and the requirement for maximum efficiency in plant operation may necessitate the storage of milk for extended periods prior to cheese manufacture. This again may influence both the yield and the quality of cheese.

This paper highlights some of the effects which the introduction of quotas may have had on the suitability of milk for cheese manufacture. The changes evident in the gross compositional quality of the milk supply are considered together with a review of the problems encountered in the use of cold stored milk in cheesemaking. Methods are proposed to maximize the efficiency of cheese production considering the inevitable changes in both the gross composition and the bacteriological quality of the milk.

CHEESE PRODUCTION IN THE UK, 1979-89 Within the past decade, the volume of milk produced in the U K has been reduced from a peak value of 16.5 million litres in 1983-84 to just over 14 million litres in 1988-89 (Table I ) . This decline in milk production has led to a gradual reduction in the volume of milk going to manufacture. However, within the last five year period, the percentage volume of milk utilized in cheese production has increased significantly. In 1984, 26.7% of the total volume of milk for manufacture was used to produce cheese. By 1989, this figure had risen to almost 40%. This emphasizes the need to maximize efficiency within the cheese sector in order to ensure the economic security of the industry as a whole.

‘Paper presented at symposium on ‘Milk Utilization’. 17 May 1989, Dumfries and Galloway.

Year

TABLE 1 Utilization of milk for cbeese manufacture

in the UK, 1979-89

Volume‘ % Volume Volume+ 96 Volume of of milk of milk of milk manufactured

produced to manufacture to cheese milk to cheese

198 1-82 1982-83 1983-84 1984-85 1985-86 1986-87 1987-88 1988-89

1979-80 15,161 51.9 2377 30.2 1980-8 1 15,212 53. I 2335 28.9

15.149 53.4 2448 30.2 16,360 57.5 2478 26.3 16.42 I 57.6 2526 26.7 15,227 54.5 2417 29.1 15.276 54.8 255 1 30.5 . ,~ . 15,352 55.5 2613 30.7 14,436 52.8 272 1 35.1 14.010 51.8 2884 39.8

*Million litres Calculated from data iri Anon (1979-1989)

CHANGES IN MILK COMPOSITION 1979-89 The principal constituents influencing both the yield and the quality of cheese are rhe fat and casein content of the milk used in cheese manufacture. There are distinct seasonal trends in cheese yield (Banks & Tamime, 1987) which reflect the variation in the level of milk constituents. To some extent the fat content of the milk can be considered the major factor controlling cheese yield (Hawley et al, 1960), since the casein content of milk is comparatively constant in relation to the variation in fat, and since the fat accounts for nearly 60% of the ‘fat + casein’ value.

Within the last 10 years, the mean level of fat in the milk supply has altered to a certain extent. Values for Scotland and England and Wales are shown in Fig. I . In Scotland the level of fat increased from a minimum value of 3.69% in 1980 to 3.83% in 1983-84. By 1985, t?le mean level offat in milk reached 3.89%, and has since declined. A similar but less extreme effect was noted in England and Wales, with the decrease in fat levels in milk after 1986 being lower than those evident in Scotland. Some of the changes seen in fat levels within the past decade may be related to the introduction of quota volumes in 1984.

The compositional payment schemes in effect at that time encouraged producers to offset losses in terms of volume by delivering milk of a high fat content. In 1986, when the European Communi:y realized that the restriction on the volume of milk which can be produced as a result of the quota system was encouraging producers to boost their income by delivering milk with ii high fat content, a butterfat quota was introduced. In addition to the volume of milk a producer delivers to his purchaser each year being recorded the average

Journal of the Society of Dairy Technology, Vol. 43, No. 2, May 1990 35

Englaqd and Wales ' r c Scotland [MtIk Marketing Board only)

L O C

LL

36 t 3 L ,,t.. L .d. I I --

80 8' 82 83 8 L 85 86 87 88 89 Year

Fig. I . The mean tat content of the milk supply in Scotland and in England and Wales 1980-89. Data from Anon, Dairy Facts and Figures. 1980-89.

fat content of thc milk is also recorded, and both the volume of milk and the butterfat level are used t o establish whether or not a producer has exceeded his quota. T h e introduction of this scheme may have been responsible for the decrease in the level of butterfat produced in 1986 and 1987 where the respective values for fat in milk were 3.77 and 3.78%.

The need for maintaining levels of production in terms of both volume and composition throughout the year may bring about changes in the composition of milk which influence both the yield and the quality of cheese. In any one year climatic conditions will determine the level and quality of grazing. In 1988, in the southern half of England and Wales, conditions for grazing were favourable and many producers exceeded their production targets over the summer months (Anon, 1989). In order to reduce outputs in winter months, these producers cut feeding levels. Such practices are likely t o interfere with the production of milk showing typical seasonal quality, and the cheese manufacturer must be aware of these compositional changes if he is to produce cheese most efficiently. Fluctuations in the fat content of milk throughout the year have a significant effect on cheese yield since this is affected by the sum of the casein and fat. Additionally, changes in the level of fat a re responsible for variations in the casein to fat ratio. This in turn affects the quality of cheese (Scott, 1977). a n d the efficiency of fat recovery in curd (Banks & Tamime, 1987).

In a n extensive study of the relationship between milk compositional factors and the yield and quality of cheddar cheese (Banks et al, 1984) i t was noted that for much of the year the casein t o fat ratio is outside the recommended optimum (0.68-0.72) for production of quality cheese. While a high correlation was observed between the sum of the casein + fat and the yield of cheese (r = 0 . 9 9 , it was apparent that the variation in the casein t o fat ratio in the milk t o a certain extent influenced the level of milk fat retained in curd during cheesemaking (Banks & Tamime, 1987).

In this study it was noted that where casein to fat ratios were comparatively high in summer milk samples, the level of fat retention in curd was also high, in the order of 92.5%. In winter samples, where casein to fat ratios were low, the level of fat retention in curd was approximately 91%.

A linear increase in the level of fat retention in curd was noted as the fat content of milk decreased in proportion t o casein until a ratio of 0.72 was reached. Thereafter a high level of casein in proportion t o fat resulted in a decrease of fat retention in curd. This suggests that when the ratio is less than 0.70, the curd is t o some extent saturated with fat a n d consequently high losses o f fat are inevitable when curds a re cut during cheese manufacture. A value of0.72 appears t o be optimal in promoting the retention of fat in curd.

Published milk composition da ta (Anon, 1985-1989) have been used t o estimate changes in the casein t o fat ratio which

occurred in the milk supply in England and Wales throughout each month over the years 1985 to 1989. The published values !or protein were used to calculate the casein content of the milk by assuming that the proportion of milk protein which was casein was 76.3%. This was the mean value measured by Harding (1974) in a comprehensive study of the compositional quality of creamery milk in England and Wales, and this ratio has been shown to remain relatively constant throughout the year (Muir rial , 1978). The results arc shown in Fig. 2. From this it is apparent that with the exception of one month in the summers of 1985 and 1986, the casein to fat ratio of the milk supply is outside the recommended optimum for production of quality cheese, a n d it is consistently too low to ensure the most efficient recovery of fat in curd during cheesemaking. This is particularly evident between October and March, and in these months standardization of the casein t o fat ratio in the milk to 0.70-0.72 would be particularly beneficial.

Standardization of high fat milk with liquid skim for cheesemaking decreases the yield potential of milk by reducing the sum o f the casein t fat, and consequently may result in a small decrease in average yield over a season (Banks er al, 1984). However, this loss in yield is compensated for by a gain in the efficiency of retention of milk constituents, principally the fat.

0 8 0 v 1985-86 '= 1986-87

1987-88

I L-.- I 1 I--. ,.I . I A

A M J J A S O N D J F M A Month

Fig. 2. Seasonal changes in the casein to fat ratio in milk in England and Wales 1985-89. Calculated from data in Anon, Dairy Facts and Figures, 1985-89.

Although quota restrictions a re now effectively reducing the levels of fat in milk, there is a gross imbalance in the ratio of casein t o fat in the milk supply throughout the year, and it is essential that milk for cheese manufacture is standardized. This would encourage the economic conversion of milk constituents to high quality cheese of a standardized composition.

The standardization of milk for cheesemaking can be used to eliminate seasonal variations in the fat in dry matter content of cheddar cheese. The fat in dry matter content will influence the moisture in non-fat solids content of the cheese (Pearce, 1978), which is the principal factor influencing cheese body and flavour development.

COLD STORAGE O F MILK FOR CHEESEMAKING The overall reduction in the level of the milk supply to thc manufacturing sector of the dairy industry, together with the increasing need for maximizing the efficiency of plant operation, have resulted in the use of extensive bulking and storage of milk prior to processing. The effects of cold storage on milk a re twofold. The storage of milk at low temperatures promotes the selection and growth of psychrotrophic organisms, and also brings about changes in the physicochem- ical characteristics of milk constituents. These factors may

36 Journul of rhe Socier? of Dairy Technology. Voi. 43. KO. 2. iWuy 1990

influence the yield and quality of cheese produced from cold stored milk. I t is therefore important that when milk for cheese manufacture has to be cold stored, attempts must he made to minimize these effects. Methods to control the potential increase in psychrotroph load during storage include the heat treatment of milk a t subpasteuriiation levels, ie, thermization, prior to storage, and the use of deep cooling to 2°C. ie, the storage of milk at temperatures lower than those found in commercial creamery silos at the present time.

Although psychrotrophic organisms are generally destroyed by pasteurization (Witter, 1961) they produce extracellular enzymes which are thermostable (Law, 1979; Cousins, 1982) and are capable of withstanding high-temperature short-time pasteurization. The most important of these enzymes with regard to cheese manufacture are the proteases and lipases.

Heat stable proteases a re thought to be responsible for decreasing cheese yields by breaking down milk proteins, thereby causing increased nitrogen losses in whey (Cousins and Marth, 1977; Olson. 1977). Lipases from psychrotrophic bacteria which survive pasteurization may give rise to rancid flavour defects in cheese (Ohren and 'luckey, 1969; Law er at, 1976).

With regard to physicochemical effects, the extended storage of milk at low temperatures can cause a n increase in the level of soluble casein compared with micellar casein (O'Connor and Fox, 1973; Creamer er at, 1977). and this may bring about a reduction in cheese yield (Ali er at, 1980a.b).

In recent years a t the institute we have examined the effects on cheese yield and quality of thermizing (Banks et at, 1986) and of deep cooling milk (Banks er (11, 1988). In the thermization trials, a sample of bulk milk was obtained from a nearby creamery. This was divided into nine batches, three of which were thermizedat 55OCfor60secandanotherthreeat 65"Cfor 15sec. The remaining samples were untreated. All samples were stored at 4 5 ° C for I , 3 or 7 days prior tocheese manufacture on a pilot scale (Banks and Muir, 1984).

The initial psychrotroph counts of the raw milk prior to thermization averaged 1.7 X lo5, with the range in counts extending from 2.6 X 10' to I X lob. In the raw milk samples the psychrotroph counts reached almost 10' after 3 days' storage, and some samples had levels in excess of 10' after 7 days' storage (Table 2).

TABLt 2 The effect of thermization of creamery silo milk on

psychrotroph counts of milk stored at 5°C' .

log,,, psvchrotroph count Storage period 'Geometric mean count ar rhermirarion rrrutmenrt

(0hP.S) Non-themiired 5S"C/OO .r 05"C/I5 s I 5.28 4.76 2.77

3 6.94 6.04 3.53

7 7.92 7.69 6.56

( 5 26-5.32) 17.04-5.45) (2.08-4.20)

(6 87-7.(X)) (4.56-6.59) (2.15-5.34)

(7.36-8.26 I 16.56-8.20) 15.5 1-7.30 )

'Mean of nine trials: range in parentheses

Thermization of the milk at 55°C for 60 sec produced only a tenfold reduction in the psychrotroph count of raw milk, whereas treatment at 65°C for 15 sec produced a 3 log cycle reduction in count.

A significant reduction in the mean yield of cheese from raw milk was observed as the storage period increased from 1 to 7 days (Fig. 3). Decreases in yield were lower in the thermized samples, with significant reductions only after 7 days' storage, when counts were frequently in excess of 10' colony forming units/ml. There was no significant difference in yields between the two thermization treatments.

Storage time (days1 0 Control 55OC/60 Eeconds a 65OC/15 Eeconds

Fig. 3 The effect of thermiiation o n the yield of cheddar cheese. Data from Ba iks ei u/ (1986)

The quality of the cheese produced was evaluated by the institute's taste pane after the cheese had matured for 6months. Panel members wen: asked to evaluate the cheese in terms of flavour, texture and overall acceptability (Fig. 4). All cheese produced from milk stored for 7 days, irrespective of treatment, developed significant off-flavours. Cheese manufactured from untreated milk or thermized milk stored for I day were identi- cal in terms of flavour. However, a difference was noted in the level of off-flavour development in the mature cheese made from milk stored for 3 days. The cheese manufactured from milk heated at 65°C fo r 15 sec was found to be most acceptable.

C F I T X T O A OF

0 Control 55OC/60 seconds

a 65"C/15 seconds

Fig. 4. The effect of storage o f milk for three days following thermimion on cheese quality at six months. CFQ = cheddar flavour qualit:;; CFI = cheddar flavour intensity; TXT = texture; OA = overall acceptability; OF = off-flavour. Data from Banks er a1 (1986).

Thermization therefore can be used t o extend the shelf-life of milk which is to be used in cheese manufacture, and while there is no difference in t e r m of cheese yield in usinga heat treatment of 55°C for 60 sec or 65°C for 15 sec, the quality of cheddar flavour and the overall acceptability of cheese produced from milk heated a t 65°C for 15 sec a re superior to that of cheese produced from milk heated at 5 5 O C for 60 sec.

An alternative procedure t o the use of thermization of milk prior to storage to re.iuce the level of psychrotroph growth is a reduction in the temperature of the milk o n storage. It has been shown, for example, that o n lowering thestorage temperature of milk from 5°C tc, O O C , the shelf-life is increased by

.Journal of thc Society of Dairy Terhnology. Vol. 43, h'o 2. Mar 1990 37

approximately the Same number of days as would be achieved on cooling milk from 30°C to 5°C (Greene & Jezeski, 1954). However, reduction of the temperature of milk beyond that normally used in creameries could result in a n increase in the intensity of undesirable physicochemical effects.

In a second study carried out a t the institute, the advantages of storing milk at 2°C rather than 6°C were examined to determine how this would influence psychrotrophic growth and physicochemical parameters in relation to cheese manufacture.

Raw silo milk was obtained from a nearby creamery on six occasions over a 12 week period. A further six samples of bulk milk from the institute were also evaluated during the same period. Milk samples were divided into 50 litre aliquots, three of which were placed in storage at 2°C. while the remaining samples were stored at 6°C. Cheese was manufactured from the samples after storage for 0, 2 and 4 days, using a 45 litre scaled version of the pilot cheesemaking system described in Banks & Muir (1984).

The initial psychrotroph counts in raw milk samples ranged from approximately I X lo2 cfu/ml to 5 X IO'cfu/ml (Table 3). Counts were generally tenfold higher in creamery samples than in bulk tank samples. The mean value for the creamery samples was 7 . 2 X lo', while that for bulk tank samples was 7.2 X lo2 cfulml.

casein and fat contents of the creamery milks were consistently lower. Although cheese yields declined slightly over the 4 day storage period (approximately I % ) , this effect was seen at both 2°C and 6°C. The yield difference evident over the storage period was not statistically significant, and it is possible that the overall decline in cheese yield may have been associated with a slight increase in the level of soluble casein in milks, which had been stored and pasteurized.

Although high levels of soluble casein were noted in milks on storage, and this effect was greater at 2°C than at 6"C, it was evident that subsequent heating of the milk for cheesemaking effectively brought about a return of the soluble casein to the micellar form (Banks et a / , 1988).

The quality of cheddar cheese was found to improve on storing milk at the lower temperature (Fig. 6). The diffcrencc in quality between the samples was related to the presence of off- flavours, which were particularly strong in samples stored for 4 days at 6°C. Cheese produced from milk stored at 2°C obtained higher acceptability scores than that produced from milk stored a t 6°C. and this effect is particularly evident after milk has been stored for 4 days.

5 r-

1'ABl.t . 3 The effect of storage of milk at 2°C and 6°C: on the psychrotroph count

Geometric mean psychrotroph couni (log,, , psychrotroph count)

Farm hulk milk Creamery milk Storage period IdU,'S) 2 9 c 6OC F C' 60 C

0 2.86 2.86 3.86 3.86

2 3.24 4.61 4.63 5.69

4 4.62 6.32 5.55 7.46

(1.99-3.81) (1.99-3.81) (2.97-5.69) (2.97-5.69)

(2.93-3.66) (3.XX-5.46) (3.26-5.41) (4.15-6.26)

(3.54-5.30) (6.08-6.51) (4.23-6.1 I ) (6.26-9.31)

*Mean of six trials; range in parentheses

Milk stored for 4 days at 6°C showed a n increase in psychrotroph count of approximately 4 log cycles, and within this storage period the psychrotroph count of four of the creamery samples exceeded I X lo7 cfu/ml. Milk stored at 2°C showed an increase in psychrotroph count of less than I x los cfu/ml.

The mean yield of cheese obtained o n processing milk samples stored at 2°C and 6°C is shown in Fig. 5 . The cheese yields obtained using creamery milk samples were considerably lower than those obtained using institute bulk tank milk as the

Storage time (days) L

0 2 o c 6 O C

Fig. 5. The effect o f storing milk at 2°C and 6°C on cheese yield. Data from Banks et al (1986).

CFO CLI

2 o c

0 6OC

Fig. 6 . I he cffcct o f storing milk at 2°C and 6°C' lor four days on cheese quality at six monlhb. CFQ = cheddar Ilavour quality: ( .Fl = cheddar flavour intensit!; 'TXT = texture: OA = overall acceptability. Data from Banks et a l ( 1988).

CO I\; C L lJSl ON S

The restrictions o n milk production which have arisen through the EEC regulations that have become effective in the UK in recent years have brought about some changes in the compositional quality of milk going to manufacture. Imbalances in the casein to fat ratio of the milk supply should be corrected in milk which is to be used in cheese manufacture by standardization. This would ensure economic production of quality cheese. Additionally, there is the potential to diminish the adverse effects of extended cold storage of milk for cheesemaking through the use of thermization or deep cooling.

REFERENCES Ali A A , Andrews A T & Cheeseman C C (1980a) lnfluencc of storage

of milk o n casein distribution between the micellar and soluble phase5 and its relationship to cheesemaking parameters. Journal of Ik7ir.t Research 4 1 371-382.

Ali A A , Andrews A T & Cheeseman G C (19XOb) Factors inllucncing casein distribution in cold stored milk and their effects on checsc- making parameters. Journal of Dairy Research 47 383-391.

Anon (1979-89) Dairy Facts and Figures, IJK. Federation of llnitcd Kingdom Milk Marketing Boards.

Banks J M & Muir D D (1984) A laboratory technique for controlled production o f cheddar cheese. Journalof'Food Technology 19 593-604.

38 Journal of the Societ, of Darrr Technology, Vol. 43, 3'0. 2. Ma\ I990

Banks J M & Tamime A Y (1987) Seasonal trends in the efficiency of recovery of milk fat and casein in cheese manufacture. Journalof the Society of Dairy Technology 40 64-66,

Banks J M, Muir D D & Tamime A Y (19x4) A comparison of the quality of cheese produced from seasonal and standardiLed milk. Journal of rhe Sociery of Dairy Technology 37 88-92.

Banks J M, Griffiths M W, Phillips J D & Muir D I> (1986) The yield and quality of cheese produced from thermised milk. Dairy Indusrrie-3 Inrernational51 3 1-35.

Banks J M, Griffiths M W, Phillips J D & Muir D D (1988) A com- parison of the effects of storage of raw milk at 2OC and 6°C o n the yield and quality of cheddar cheese. Food Microbiology 5 9-16.

Cousins M A (1982) Presence and activity of psychrotrophic micro- organisms in milk and dairy products, a review. Journal of Food Prorecrion 45 172-207.

Cousins M A & Marth E H (1977) Cheddar cheese made from milk that was precultured from psychrotrophic bacteria. Journal of Dairy Science 60 1048-1056.

Creamer L K , Berry G & Mills 0 E (1977) A study of the dissociation of beta-casem from the bovine casein micelle at low temperature. New Zealand Journal of Dairy Science and Technology 12 58-66.

Greene V W & Jezeski J J (1954) Influence of temperature on the development of several psychrotrophic bacteria of dairy origin. Applied Microbiology 2 1 10- I I 7.

Harding F (1974) Variations in the nitrogen-containing fractions of

bulked milk in England and Wales during the period 1947-70. Dairy Industries Inrerwrionul 39 372-378.

Hawley H B, Dwyer T 8. Davis J Ci ( 1960) The yield of cheese from milk in Somerset, 1936- 1959. Dairy Industries Inrernational25 440-444.

Law B A (1979) Enzymes of psychrotrophic bacteria and their effect on milk and milk prtducts. Journal of Dairy Research 46 573-588.

Law H A, Sharpe M E & Chapman H R (1976) The effect of lipolytic gram negative psychiotrophs in stored milk on the development of rancidity in cheddar 2heese. Journal of Dairy Research 43 459-468.

Muir D D, Kelly M E, !'hillips J D & Wilson A G (1978) Journal of the Society of Dairy Technology 31 137-144.

O'Connor P & Fox P F (1973) Temperature dissociation of casein micelles from the milk of various species. Nerherlands Milk and Dair). Journal27 199-217.

Ohren J A & Tuckey !j I. (1969) Relation of flavour development in cheddar cheese to chemical changes in the fat of the cheese. Journalof Dairy Science 52 598-64l7.

Olson N F (1977) Faztors affecting cheese yields. Dairy Industries International 42 14. I 5 , 19.

Pearce K N (1978) The relationship between Tat and moisture in cheese. AeH: Zealand Journal of Dairy Science and Technology 13 59-60.

Scott R (1977) Factc,rs affecting cheese quality. Dairy Industries International 42 14-2 I .

Witter L D (1961) Psy:hrophilic bacteria-a review. Journal of Dairy Science 44 983- I0 15.

SECTION PAPER

Hygienic design of a dairy*

H G VINSON Express F o o d s Group Limited, Victoria Road, South Ruislip HA4 OHF

Various engineering principles of construction and layout may he employed t o minimize the risk of product contamination in a dairy. This paper summarizes the main requirements f o r a milk handling and powder production plant .

INTRODUCTION A number of guidelines and codes relating to hygienic manufacture have been published or are in the course of preparation. This paper identifies the engineering aspects of these guides.

While the paper deals particularly with milk product manufacture, the principles are applicable to a number of food manufacturing processes using liquid and powdered materials. In particular, the precautions and design criteria mentioned are necessary where the endproducts are to be consumed by sections of the population at high risk such as the sick, the very young and the elderly.

It is inevitable that there can be no separation between the requirements of good manufacturing practice and good engineering design. A number of operational issues are therefore covered.

The environment that we live in contains large numbers of pathogenic organisms such as Salmonella. They may be present in the faeces of birds, on roads, in airborne water droplets, in farm yard excrement and in standing pools of water.

'Paper proenlcd 10 Northern Ireland Sccilon in 1989

A number of the raw materials used for food production are naturally contaminated. It is the job of the food processor to manage the environment and treat materials in such a way as to produce hygienic edible products.

THE ?ROCESS ENVIRONMENT The site selected for food manufacturing premises should be situated in areas free from flooding, objectionable odours, smoke, dust and any other contaminants. Sites adjacent to waste disposal sites, sewage works and vermin-attracting areas should be avoided.

The site should be designed to segregate as far as possible vehicle movement into clean and dirty areas, thus establishing the overall principle of high and low risk areas.

In the milk industry, the vehicles collecting the milk from the farms will inevitab'y have travelled over farm tracks and into farmyards collecting mud and filth on their wheels. These vehicles should have a definitive route throughout the factory to the milk unloadin;% area and not travel on any road where finished goods are handled or despatched. This area, assumed to be contaminated, must not be a normal pedestrian thoroughfare and people workink: in thisarea must not haveaccess to high risk

Journal of the Sociery of Dairy Technology, Vol. 43, No. 2, May 1990 39