1. PRINCIPLES OF CANNING


1.1 Thermal Destruction of Bacteria
1.2 Thermal Processing Requirements for Canned Fishery Products
1.3 The Concept of Thermal Process Severity (Fo Value)

1.3.1 Determination of Fo values
1.3.2 The improved general method of Fo calculation
1.3.3 The trapezoidal integration and method

1.4 Specification of the Thermal Process Schedule
1.5 Application and Control of the Scheduled Process


The technology for preserving foods in cans was developed at the beginning of the nineteenth century when a Frenchman, Nicolas Appert, won a competition initiated by another great character in French history, Napoleon Bonaparte. Napoleon is better remembered for his feats as a conquering General, than he is for providing the stimulus for the development of a food preservation technique that was to mark the start of the canned food industry. Appert won his prize (12 000 francs) for demonstrating that foods which had been heated in air-tight (hermetic) metal cans, did not spoil, even when they were stored without refrigeration. Once the reliance on the refrigerated and/or frozen food chain had been broken, it was possible to open markets for shelf-stable canned products where no entrepreneur had ventured previously. In the time since Appert's success, the technology of canning has been modified and improved. however. The principles are as true today as they were when first enunciated. The success of the international fish canning industry rests on the sound application of these principles.

1.1 Thermal Destruction of Bacteria

When fish are landed they contain, in their gut and on their skin. Millions of bacteria which. if allowed to grow and multiply will cause a rapid loss of the "as fresh" quality and eventually result in spoilage. During post-harvest handling, in transit to the cannery, the fish inevitably become contaminated with other bacteria; these will further accelerate spoilage unless protective measures (such as icing) are employed. The purpose of canning is to use heat. alone or in combination with other means of preservation, to kill or inactivate all microbial contaminants, irrespective of their source, and to package the product in hermetically sealed containers so that it will be protected from recontamination. While prevention of spoilage underlies all cannery operations, the thermal process also cooks the fish and in some cases leads to bone softening; changes without which canned fishery products would not develop their characteristic sensory properties.

In order to make their products absolutely safe, canned fish manufacturers must be sure that the thermal processes given their products are sufficient to eliminate all pathogenic spoilage micro-organisms. Of these Clostridium botulinum is undoubtedly the most notorious, for if able to reproduce inside the sealed container, it can lead to the development of a potentially lethal toxin. Fortunately, outbreaks of botulism from canned fishery products are extremely rare. However, as those familiar with the 1978 and 1982 botulism outbreaks in canned salmon will testify, one mistake in a seasons production has the potential to undermine an entire industry. It is because the costs of failure are so prohibitive that canned fish manufacturers go to great lengths to assure the safety of their products. Safety for the end-user. and commercial success for the canner, can only be relied upon when all aspects of thermal processing are thoroughly understood and adequately controlled.

When bacteria are subjected to moist heat at lethal temperatures (as for instance in a can of fish during retorting), they undergo a logarithmic order of death. Shown in Figure 1 is a plot (known as the survivor curve) for bacterial spores being killed by heat at constant lethal temperature. It can be seen that the time interval required to bring about one decimal reduction (i.e., a 90% reduction) in the number of survivors is constant; this means that the time to reduce the spore population from 10 000 to 1 000 is the same as the time required to reduce the spore population from 1 000 to 100. This time interval is known as the decimal reduction time, or the "D value ". The D value for bacterial spores is independent of initial numbers, however, it is affected by the temperature of the heating medium. The higher the temperature the faster the rate of thermal destruction and the lower the D value - this is why thermal sterilization of canned fishery products relies on pressure cooking at elevated temperatures (>100C) rather than on cooking in steam or water which is open to the atmosphere. The unit of measurement for D is "minute" (the temperature is also specified, and in fish canning applications it can be assumed to be 121.1C).

Figure 1 Survivor curve for bacterial spores, characterized by a D value of 5 min, subjected to heat at constant lethal temperature

Another feature of the survivor curve is that it implies that no matter how man decimal reductions in spore numbers are brought about by a thermal process, there will always be some probability of spore survival. In practice, fish canners are satisfied if there is a sufficiently remote probability of pathogenic spore survival for there to be no significant associated public health risk; addition to this they accept, as a commercial risk, the greater probability of there being some non-pathogenic spoilage.

Shown in Table 1 are the reference D values for bacteria commonly found to be important in canning. Since it can be seen that not all bacterial spores have the same D values, a thermal process designed to, say, reduce the spore population of one species by a factor of 10 (i.e., 9 decimal reductions or a 9D process) will bring about a different order of destruction for spores of another species. The choice for the fish canner therefore becomes one of selecting the appropriate level of spore survival for each of the contaminating species. Thermophilic spores (those which germinate and outgrow in a temperature range of between 40 and 70C and have their optimum growth temperatures around 55C) are more heat resistant, and therefore have higher D values, than spores which have mesophilic optimum growth temperatures (i.e., at 15 to 40C). This means that raw materials in which there are high levels of thermophilic spores will require more severe thermal processes than will products containing only mesophilic spore formers, if the same degree of thermal destruction is to be achieved for each species.

1.2 Thermal Processing Requirements for Canned Fishery Products

From the point of view of preventing microbial deterioration in the finished product. there are two factors which must be considered when a fish canner selects thermal processing conditions. The first is consumer safety from botulism, and the second is the risk of non-pathogenic spoilage which is deemed commercially acceptable.

Table 1 Decimal reduction times (D values) for bacterial spores of importance in fish canning

Organism

Approximate optimum growth temp. (C)

D value
(min) a/

B. stearothermophilus

55

D121.1 4.0 - 5.0

C thermosaccharolyticum

55

D121.1 3.0 - 4.0

D. nigrificans

55

D121.1 2.0 - 3.0

C. botulinum (types A & B)

37

D121.1 0.1 - 0.23

C.sporogenes (PA 3679)

37

D121.1 0.1 - 1.5

B. coagulans

37

D121.1 0.01 - 0.07

C. botulinum type E

30 - 35 b/

D82.2 0.3 - 3.0

a/ D values quoted are those at the reference temperature of 121.1C, with the exception of that for C. botulinum type E, the spores of which are relatively heat sensitive, being killed at pasteurization temperatures (e.g., 82.2C)

b/ Although the temperature range for optimum growth of C. botulinum type E is 30-35 C, it has a minimum of 3.3C which means that it is able to grow at refrigeration temperatures

Safety from botulism caused by underprocessing means that the probability of C. botulinum spores surviving the thermal process must be sufficiently remote so as to present no significant health risk to consumers. Experience has shown that a process equivalent to twelve decimal reductions in the population of C. botulinum spores is sufficient for safety; this is referred to as a 12D process and assuming an initial spore load of 1 spore/g of product, it can be shown that, for such a process, the corresponding probability of C. botulinum spore survival is 10-12, or one in a million million. This implies that for every million million cans given a 12D process, and in which the initial load of C. botulinum spores was l/g, there will be only one can containing a surviving spore. Such a low probability of survival is commercially acceptable, as it does not represent a significant health risk. The excellent safety record of the canning industry, with respect to the incidence of botulism through underprocessing, confirms the validity of this judgement. In the United States over the period 1940-82, in which time it is estimated that 30 billion units of low-acid canned food were produced annually (and of these approximately one billion per year were canned seafoods), there have been two outbreaks (involving four cases and two deaths) of human botulism attributable to delivery of inadequate thermal processes in commercially canned food in metal containers. This corresponds to a rate of botulism outbreaks due to failure in the selection or delivery of the thermal process schedule of under 1 in l012 (0.6/l012 ).

Spoilage by non-pathogenic bacteria, although not presenting as serious a problem as botulism will, if repeated, eventually threaten the profitability and commercial viability of a canning operation. It is because of the commercial risks of product failure that canners ought to quantify the maximum tolerable spore survival levels for their canned products. As with the adoption of the 12D minimum process requirement for safety from botulism, experience is the best guide as to what constitutes an acceptable level of non-pathogenic spore survival. For mesophilic spores, other than those of C. botulinum, a 5D process is found adequate; while for thermophilic spores, process adequacy is generally assessed in terms of the probability of spore survival which is judged commercially acceptable. In other words what level of thermophilic spoilage can be tolerated bearing in mind the monetary costs of extending processes to eliminate spoilage, the quality costs arising from over-processing and finally the costs of failure in the market place, should surviving thermophilic spores cause spoilage. All things being considered, it is generally found acceptable if thermophilic spore levels are reduced to around 10-2 to 10-3/g. There are two reasons why higher risks of spoilage (arising through survival, germination and outgrowth of thermophilic spores) can he tolerated. First, given reasonable storage temperatures (i.e., <35 C) the survivors will not germinate; and secondly even if spoilage does arise it will not endanger public health.

If a thermal process is sufficient to fulfill the criteria of safety and prevention of non-pathogenic spoilage under normal conditions of transport and storage, the product is said to be "commercially sterile". In relation to canned foods, the FAO/WHO Codex Alimentarius Commission (1983) defines commercial sterility as "... the condition achieved by application of heat, sufficient, alone or in combination with other appropriate treatments, to render the food free from microorganisms capable of growing in the food at normal non-refrigerated conditions at which the food is likely to be held during distribution and storage". Although this definition specifically refers to "non-refrigerated" conditions and thereby excludes those semi-preserved and pasteurized foods in which refrigerated storage is recommended (and in many cases is obligatory in order to prevent growth of the pathogenic psychrophile C. botulinum type E -which can grow at temperatures as low as 3.3C ), publications by the Department of Health and Social Security in the United Kingdom and the Standards Association of Australia do not exclude refrigerated foods. According to these less restrictive interpretations, commercial sterility may then also encompass those foods which are intended to be stored at refrigeration temperatures; this implies that commercially sterile canned foods will be free from microorganisms capable of growing at ambient or refrigeration temperatures, whichever is considered normal. Whether the product is intended to be stable under refrigeration or at ambient temperatures, the attainment of commercial sterility is the common objective when manufacturing all canned fishery products. There are, however, circumstances in which a canner will select a process which is more severe than that required for commercial sterility, as for instance occurs when bone softening is required with salmon or mackerel.

1.3 The Concept of Thermal Process Severity (Fo Value)

A mathematical equation describing the thermal destruction of bacteria can be derived from the survivor curve shown in Figure 1. If the initial spore load is designated No and the surviving spore load after exposure to heat at constant. temperature is Ns , then the time (t) required to bring about a prescribed reduction in spore numbers can be calculated and is related to the D value of the species in question by the equation,

t = D(log No – log Ns )

From this equation it is apparent that the time required to bring about a reduction of spore levels can be calculated directly, once the spore level before, and the desired spore level after, the heat treatment are specified, and the D value of the spores under consideration is known. For instance, considering the generally recognized minimum process for prevention of botulism through under-processing of canned fishery products preserved by heat alone (which assumes that initial loads are of the order of 1 spore/g, and in line with good manufacturing practice guidelines, final loads shall be no more than 10-12 spore/g), the minimum time required to achieve commercial sterility (i.e., a 12D process) can be calculated from,

t = 0.23(log 1 – log 10-12)
= 0.23 x 12
~ 2.8 min

This means that the minimum thermal process required to provide safety from the survival of C. botulinum is equivalent, in sterilizing effect, to 2.8 min at 121.1C at the slowest heating point (the SHP) of the container. This process is commonly referred to as a "botulinum cook".

Having established the minimum process with respect to product safety, it remains to select a processing time and temperature regime which will reduce the numbers of spore forming contaminants (more heat resistant than those of C. botulinum) to an acceptable level. If, for instance, the canner is concerned at the possibility of C thermosaccharolyticum spore survival (because it is known that raw materials are contaminated with these spores and it is likely that the product will. be stored at thermophilic growth temperatures) and the No and Ns are 10 spore/g and 10-2 spore/g, respectively; the time required to achieve commercial sterility can be calculated as before,

t = 4.00 (log 10 - log 10-2)
= 4.00 (2 + 2)
= 16 min

Thus, in order to prevent commercial losses through thermophilic spoilage by C. thermosaccharolyticum the thermal process must be equivalent, in sterilizing effect, to 16 min at 121.1 C at the SHP of the container. This approach to calculating the thermal process requirements tends to be an oversimplification for two reasons:

  1. in practice it is not reasonable to assume that naturally occurring contaminants will be present only as pure cultures. However, because fish and other raw materials contain a mixed flora, canners assume "worst-case" conditions in order to develop a process which always provides adequate protection from all contaminants. It is customary, therefore, to assume that C. botulinum and other heat resistant spore forming bacteria are present: and then to select a thermal process, the severity of which is sufficient to reduce their probability of survival to commercially acceptable levels.
  1. The survivor curve (shown in Figure 1) assumes that the temperature of the heat treatment is constant (and in the cases considered, equal to 121.1 C), whereas during heating in a commercial retort, the SHP of the can experiences a lag in heating and in many cases may never reach retort temperature. Thus the equation that permits calculation of the time required at constant temperature to achieve a desired survivor level (i.e. , Ns) cannot be simply applied to the effects of heating at the SHP of a can. Consequently, the total sterilizing effect at the SHP of a can, which by convention is expressed as time at constant reference temperature, is not the same as the scheduled time for the thermal process (i.e., the time for which a batch retort might be held at operating temperature). To account for the influence on total sterilizing effect of heating lags it is necessary to integrate the lethal effects of all time/temperature combinations at the SHP during a thermal process and express their sum as being equivalent to time at reference temperature. In manufacture of shelf-stable canned fish it is standard practice to express the magnitude of the sterilizing effect of a thermal process in "minutes" at the reference temperature of 121.1 C. Following this convention, the symbol for the total sterilizing effect of a thermal process is designated as the Fo value; where Fo is defined as being equivalent, in sterilizing capacity, to the cumulative lethal effect of all time/temperature combinations experienced at the SHP of the container during the thermal process. Taking the examples considered above, this means that a botulinum cook must have an Fo value of at least 2.8 min, whereas freedom from thermophilic spoilage by C thermosaccharolyticum would necessitate an Fo value of at least 16 min.

1.3.1 Determination of Fo values

The Fo value of a thermal process can be determined by microbiological or physical means. The former method relies on quantifying the destructive effect of heating on bacterial numbers through their enumeration before and after thermal process; the latter method measures the change in temperature during thermal process at the SHP of the container and relates this to the rate of thermal destruction at a reference temperature. These techniques can be applied to measure the lethal effects of pasteurization processes (in which the target organisms are usually the relatively heat sensitive forms of bacteria, yeasts an moulds) or they may be used to assess the severity of sterilization processes (in which the target organisms are heat resistant spore-forming bacteria). In this text only the physical method of quantifying the lethal effect. of thermal processes will be described.

First, it is necessary to record heat penetration data with thermocouple probes which have been carefully placed to detect changes in product temperature at the thermal centres of the packs. There are many commercial brands of thermocouples available to suit most sizes of fish cans, glass jars and retortable pouches; they can also be constructed with copper/constantan thermocouple wire in which the hot junction is constructed by soldering together the ends of the two wires. The hot junction is coated with a thin laquer layer to insulate the exposed metal surfaces from the product (and thereby prevent surface corrosion which might otherwise interfere with the accuracy of the reading), and then it is carefully positioned at the SHP of the container. Once the thermocouples are in place and the process commenced, the temperature is recorded regularly throughout the heating and cooling phases of the thermal process. The heat penetration data so collected may be treated in a number of ways in order to calculate the Fo value of the process; however, only two of these methods are described in the following sections.

1.3.2 The improved general method of Fo calculation

A plot of temperature versus time is made on specially constructed lethal rate paper in which the temperature (on the vertical axis) is drawn on a semi-logarithmic scale and process time on the horizontal scale; also shown on the vertical axis (but usually, for convenience, on the right-hand side of the paper) is the corresponding lethal rate for the temperature which is on the adjacent left-hand vertical axis. By convention, the rate of thermal destruction (designated L) at product temperature (designated T) for bacteria, or their spores, important in canned fish sterilization is taken to be unity at 121.1 C; and further, the rate changes by a factor of ten for every 10 C that the temperature changes. Mathematically this relationship is expressed by the equation,

(1)

This means that. the rate of destruction for all temperatures can be related to the rate of destruction at the reference temperature (121.1 C). Thus the cumulative lethal effects, for all time-temperature combinations experienced at the SHP in a container, can be equated to time of exposure at 121.1 C.

Once the plot is drawn. the area under the graph is calculated (by counting squares or by using a planimeter) and divided by the area which is represented by 1 min at 121.1 C .i.e., an Fo value of 1 min. This yields the total sterilizing effect, or the Fo value, of the process. Shown in Figure 2 is an example of a temperature-time plot for a conduction heating pack processed at 121.1 C. In the worked example, the area under the graph is 70 "units", which when divided by the area corresponding to a Fo of 1 min, i.e. , 4 "units", yields 17.5 min, which is the Fo value for the process being evaluated.

Figure 2 Temperature-time plot for conduction heating pack processed at 121.1 C

It can be seen that the total sterilizing effect of the process is equivalent to 17.5 min at 121.1 C, even though the product. temperature never reached 121.1C, and neither did the retort operate at that temperature. Because it is possible to equate the rates of thermal destruction at any temperature, to the rates of destruction at the reference temperature of 121.1C, the effects of heating lags can be quantified.

1.3.3 The trapezoidal integration and method

This is a simplified mathematical method in which the time-temperature are used to record the changes in the lethal rates of spore destruction at the SHPs of containers during heating and cooling. If product temperature is recorded at regular time intervals, and assuming that this temperature constant for the period between measurements, the lethal rate applying for time interval can be computed (using equation 1). When the rates (applying over each time interval) are summed and multiplied by the time between measurements, the cumulative Fo value for the entire process can be found without the need graphical representation of the heating and cooling curves. The trapezoidal method also allows simple calculation of the contribution to total process lethality of the heating and cooling components of the process. In Table 2 is shown a worked example in which the product temperature was recorded at 5 min intervals during a process of 60 min at 121.1 C.

Table 2 Time, temperature, lethal rate, cumulative lethal rate and Fo value for a conduction heating product retorted at 121.1 C for 60 min.

Time
(min)

Temperature
(C)

Lethal rate

Cumulative lethal rate

Fo value
(min)

0

24.0

0

0

0

5

24.5

0

0

0

10

34.2

0

0

0

15

54.7

0

0

0

20

72.5

0

0

0

25

87.0

0

0

0

30

98.0

0.005

0.005

0.025

35

105.1

0.025

0.030

0.150

40

110.5

0.087

0.117

0.585

45

114.5

0.219

0.336

1.679

50

117.2

0.407

0.743

3.717

55

119.0

0.617

1.360

6.798

60
(steam off)

120.3

0.832

(1.776)

(8.880)

   

2.192

10.960

65

120.3

0.832

3.024

15.120

70

106.0

0.031

3.055

15.275

75

88.1

0.001

3.056

15.280

80

70.0

0

3.056

15.280

To calculate Fo for the total process: the sum of the L values gives 3.056 which when multiplied by five (the time interval between readings), gives an Fo value of 15.3 min. (Although the theoretical total Fo value for the process is 15.280 min, this can be rounded to 15.3 min as it is unrealistic to quote values beyond the first decimal place.)

To calculate Fo for the heating phase: the sum of the L values at times 25 min and 60 min (i.e., 0 and 0.832) is divided by two and this value (0.416) is added to the sum of the L values from 30 min to 55 min (1.360), so that the total accumulated lethal rate at the time the steam was cut (1.776) can be multiplied by five to give a total Fo value of 8.9 min at steam off. This feature of the trapezoidal method allows for simple calculation of the Fo value during thermal processing, as for instance may be required when the schedule calls for steam to be cut when the Fo reaches an assigned value.

1.4 Specification of the Thermal Process Schedule

Once target Fo values for canned fish products are specified. manufacturers must take steps to ensure that all cans receive the correct thermal process and that all factors affecting the rate of heat transfer to the SHP of every can are controlled. It is by these means that microbiological spoilage arising from under-processing can be prevented and the associated health and/or commercial risks avoided. The technique most frequently adopted to control delivery of the thermal process is to draw up a thermal process schedule which specifies those factors which. in any way. could affect delivery of the target Fo value to the SHP of the container. The Codex Alimentarius Commission (1983) destine scheduled process as "the thermal process chosen by the manufacturer for a given product and container size to achieve at least commercial sterility".

Government regulators in many countries adopt similar systems to monitor the scheduled processes of products sold under their jurisdiction. and of these perhaps one of the best known is that implemented by the United States Food and Drug Administration (FDA). In addition to requiring that those processors of acidified and low-acid canned foods sold in the United States register their establishments with the FDA. it is also necessary to file with FDA scheduled processes covering all canned foods which are destined for sale in the United States. Although these requirements will only be relevant for those canners supplying the United States market. the regulations identify several factors which form a useful checklist for canners who are formulating new canned fish scheduled processes. amending existing ones or wishing to review their control procedures. The information which should be specified in the scheduled process is summarized in Table 3.

Not all the items shown in Table 3 will be relevant for a single process. For instance, with some processes the number of retort baskets per retort load will remain constant, whereas with others. it may vary because of delays caused by fluctuations in the supply of fish to the canning line. Under "worst-case conditions" (i.e., with full loads) the steam requirements will be considerably greater than when the retort is only partially full; also. under these conditions steam circulation can be impaired so that the rate of heat transfer to the SHP of the containers is adversely affected. In a case such as this, that steam circulation is influenced by the load size, need be of no consequence. provided the effect is accounted for when calculating the scheduled temperature and duration of the thermal process.

Taking another example, specification of product fill weight may be important when filling solid style tuna or whole abalone into cans which are later to be topped-up with canning liquor; in both instances the convective currents in the brine favour rapid heat transfer to the boundaries of the solid product. there then follows conduction heating during which heat is transferred more slowly to the SHP of the container. However, should fill weight not be controlled. with the result that some cans contain more solid (and therefore less brine. given that the latter is added to a constant headspace). the rate of heat transfer to the SHP of containers will vary, being slower in those packs containing a higher ratio of solids to liquids. The effect of changing the solids to liquids ratio in a pack ought not be underestimated, and alterations should never be adopted without first confirming the adequacy of the process after the proposed change. This point has been demonstrated through trials in which fill weight for solid style tuna packed in 84 x 46.5 mm cans was increased by 10% over the maximum specified. the packs were then processed at 121.1 C, and in order to achieve a constant target Fo value of 10 min (for the standard and the overweight packs). it was found necessary to increase process time by 16% for the heavier pack. In this case, failure to compensate for overfilling would not significantly affect public health risks while the target Fo was of the order of 10 min (or more). although there would be an increased probability of survival for those spores more heat resistant than C botulinum and. associated with that, an increase in the commercial risk of non-pathogenic spoilage. However, public health risks arising from overfilling can increase for those manufacturers, who, being wary of the reduced yields and or losses in sensory quality caused by processing heat sensitive marine products (e.g. , oysters, mussels and scallops), select target Fo values closer to the minimum for low-acid canned foods (i.e., Fo = 2.8 min).

Table 3 Checklist of factors affecting delivery of the. scheduled processes for canned fishery products

Item

Reason for inclusion

Container dimensions Affects rate of heat transfer to SHP
Target Fo value Affects probability of under-processing spoilage
Process temperature Affects time required to achieve target Fo
Process time Affects temperature
Product initial temperature Affects time for product to reach temperatures lethal to spore-forming bacteria
Product fill weight, i.e. , extent of conduction or convection heating Affects mode of heat transfer to SHP
Product consistency (with homogenous packs) Affects rate of heat transfer to SHP
Liquids to solids ratio and particle size (with particulate packs) Affects rate of heat transfer to SHP
Packing style (e.g., horizontal or vertical alignment of pieces) Affects rate of heat transfer to SHP
Container stacking patterns in retort or retort baskets Affects rate of heat transfer to SHP
Number of retort baskets/retort Affects rate of heat transfer to SHP
Retort operation e.g., venting and/or condensate removal Affects temperature of heating medium
Cooling method Affects contribution to total process Fo of cooling phase

The rationale behind preparation of the thermal process schedule is to provide a standard format for identifying and specifying all those factors affecting the adequacy of the thermal process. The checklist, shown in Table 3, is a guide which should be adapted to suit each canner's requirements. It is important that the scheduled process be developed only by those expert in thermal processing and, only then, when the data upon which recommendations are based are determined in a scientifically sound and acceptable manner. Because of the importance that is attached to correct calculation of thermal processing conditions, it is common to find that in some Countries the regulators overseeing ; canning operations maintain a register of those who are "approved" to establish thermal process schedules.

Once a thermal process schedule has been established it must not be altered without first evaluating the effects of the proposed change on delivery of target Fo values. Also, alterations to product formulation must be evaluated in terms of the possible changes they bring about in the product's heating characteristics. Ideally, specification of the thermal process schedule will be based on data from heat penetration trials with replicate packs, processed under the "worst - case conditions" likely to be encountered in commercial production; however, if this is not possible it is sufficient to refer to those standard texts on canning which recommended process times and conditions for a wide range of canned foods.

In summary therefore, the process schedule provides the specifications which are critical to delivery of an adequate thermal process. The times and temperature of the process schedule is usually contained in the process filing form, an example of which is shown in Figure 3. When completed, the process filing form will also contain additional information which should be specified in the process schedule. It is good practice for the details of the scheduled process to be conveniently located close to the retorts and in a position where it can be seen by the operator.

1.5 Application and Control of the Scheduled Process

Once the process schedule is defined, the manufacturer must implement systems to monitor, control and provide records which confirm, after the event, that all stages in production affecting heat transfer to the SHP of the can were within specification. Records provide the means for a continuous assessment of production and an early warning system with which to initiate corrective action if potential problems arise; also they provide valuable and permanent documentary evidence that delivery of the process was in line with details in the process schedule. The value of permanent records becomes apparent at times of product recalls, when the need may arise for the canner to demonstrate that production techniques complied with good manufacturing practice (GMP) guidelines - without this evidence canners risk facing claims of professional negligence should their product become involved in litigation.

Records should be simple to complete, so as not to discourage their use, and easy to interpret. In some cases it may be appropriate to record data on a quality control chart which shows the change in some variable against time (e.g., fill weight, as in Figure 4). The scales can be chosen to show the change in values about the target value and also include permissible maxima and minima (i.e., tolerances); action levels can be included to alert operators of trends that may cause production to move out of control. Quality control charts are well suited to continuous operations where monitoring takes place throughout production, they are less frequently used when the function being evaluated is a batch operation. Some recording systems are completed by the operator at specified stages of an operation (e.g., the retort log sheet, as in Figure 5) while others are automated and require only minimal operator input (e.g. , retort thermographs, as in Figure 6).

No matter what form of records are adapted, their function is to provide retrospective assurance that the thermal process schedule and those related factors which affect heat transfer to the SHP of the container have been regularly monitored and controlled during production.

Figure 3

Figure 4 Quality control chart for recording container fill weight