IV. THERMAL PROCESSING

As discussed in Chapter IV, thermal processing of fish is a complex operation requiring sophisticated equipment which cannot realistically be scaled down to suit the volumes typical of a small-scale processing operation. In this section, the cost of canning is examined so that comparisons can be drawn with the smaller scale processes already analysed. It is not the intention here to make comparisons between the various methods of thermal processing; however, certain factors relating to the use of flexible pouches, as opposed to cans, are discussed.

Two cost models of a canning factory are analysed (see Edwards et. al, 1981, for a more detailed analysis of the economies of fish canning). In each model, an annual input of 500 tonnes of fresh fish is assumed. Processing is carried out by a single 8-hour shift working 250 days/year and producing a daily output of 10,000 cans. Each can contains 90 g of fish and 35 g of vegetable oil filler (e.g. sunflower oil). The yield of whole fish in can is assumed to be 45%. Model 1 assumes a fully mechanised cannery. Model 2 is identical except that labour is used to replace machinery in certain operations, namely heading, gutting and can labelling.

The different labour requirements of the two models are as follows:

	Model 1	Model 2
Manager	1	1
Engineer	1	1
Supervisor	1	1
Clerical	2	2
Semi-skilled/unskilled labour, of which permanent labour	56 (10)	96 (18)

In other words, an extra 40 semi-skilled and unskilled labourers are employed in Model 2 as follows:

	Model 1	Model 2
Heading and gutting	6	36
Labelling	2	12

The fixed and variable costs (US$) associated with these two models are presented in Tables V.10, V.11 and V.12.

Table V.10.
Production of canned fish; Investment costs (US$, valid for 1981)

Cost items	Model 1	Model 2
Plant and equipment	165,540	140,640
Buildings (400 m2)	60,000	60,000
Land (4,000 m2)	2,000	2,000
Contingencies (10% of above)	22,750	20,260
TOTAL	250,290	222,900
Working capital1	350,000	355,000

Note

¹ Working capital estimated to cover initial supplies of fish, operating costs over first months, and stocks of finished product.

Table V.11.
Production of canned fish: Annual fixed costs
(US$, valid for 1981)

Cost items	Model 1	Model 2
Depreciation1	18,954	16,460
Interest on fixed capital (8%)	20,023	17,830
Interest on working capital (8%)	28,000	28,400
Permanent labour2	19,000	25,800
Insurance (1.5% of investment costs)	3,754	3,340
Maintenance (5% of investment costs)	12,514	11,150
Other overheads	8,000	8,000
TOTAL	110,245	110,980

Notes

¹ Buildings 4%, plant and equipment 10%.

² Manager at $3.500, engineer at $2,800, supervisor at $1,700, clerical at $1,200, semi-skilled/unskilled at $860.

Table V.12.
Production of canned fish Annual variable costs
(US$ valid for 1981)

Cost items	Model 1	Model 2
Fish (at $200/tonne)	100,000	100,000
Filler (87.5 tonnes at $800/tonne)	70,000	70,000
Cans (2.5 million at $0.084 each)1	212,100	212,100
Cartons (25,000 at $0.66 each)	16,500	16,500
Salt (10 tonnes at $120 per tonne)	1,200	1,200
Personnel2	36,800	62,400
Water (8,750 1. at $0.133 per litre)	1,160	1,160
Electricity3	1,670	1,240
Fuel oil (34,000 1. at $0.4/litre)	13,600	13,600
Quality control	10,000	10,000
Sundries	12,000	12,000
TOTAL	475,030	500,200

Notes:

¹ Plus 1% for damaged cans.

² Semi-skilled/unskilled at $3.2 per day.

³ Model 1: 18,500 kWh at $0.09/kWh. Model 2: 13,750 kWh at $0.09/kWh.

With a yield of canned fish at 45% of input, the annual output would be 225 tonnes, which gives the production costs per tonne of finished product shown in Table V.13:

Table V.13.
Production of canned fish; Cost per tonne
(US$)

Cost items	Model 1	Model 2
Fixed costs per tonne	490	493
Variable costs per tonne	2,111	2,223
Total costs per tonne	2,601	2,716

This analysis shows that total production costs per tonne are marginally lower for model 1 than for model 2. In other words, the savings made on machinery were not sufficient to counter-balance the cost of the extra labour (both permanent and direct) required in model 2. Depending on the relationship between labour and capital costs, this situation will generally vary between countries.

Although the canning costs cited above are not directly comparable in all respects with the earlier analyses of salting and smoking, it is very evident that canning is significantly more expensive per tonne of product than the smaller scale processes. In particular, the cost of cans and filler, which in the models discussed, together account for 59% and 56% of annual variable costs, add considerably to processing costs. These items would form an even more significant proportion of costs in any developing country which did not have a can-making plant thereby necessitating the importation of supplies. Furthermore, unlike the small-scale processes, a cannery requires continuous expert maintenance.

The canning plant costed above is a relatively small one. Separate analysis of larger scale plants has not been undertaken. However, economies of scale are known to exist for larger plants. Even at higher levels of throughput, however, canning will still be expensive when compared to small scale processing.

As an alternative to cans, flexible pouches may be used in thermal processing operations. Although claims are made that the use of pouches is generally cheaper than the use of cans, their commercial profitability in the case of processed fish has yet to be proved. The technology is more complex than for cans, and large capital investments are still required when using flexible pouches. For example, a manually operated line - including a retort - to fill, seal, load and unload retorts, dry pouches and cartons at speeds of up to 12 pouches per minute would cost approximately $200,000 if installed in a developing country (Metal Box Limited, personal communication). A line employing more automation and less labour, and running at 25-30 pouches per minute is likely to cost three to four times this amount.

The main advantage which flexible pouches have over cans is their lower weight: 1,000 empty 8 oz cans weigh 109 lb whereas a similar number of pouches weigh only 12.6 lb. This would lead to savings in transport costs, especially in the case of imported supplies. Apart from the pouches, which are likely to be cheaper, variable costs will generally be of similar order to those for canning. Total costs can therefore be expected to be more nearly equated to canning costs than to the small-scale processes discussed earlier.

Besides costs, other considerations need to be borne in mind when comparing thermal processing methods with the smaller-scale processes described earlier. In the first place, the labour requirements of the former are considerably higher: for canning, Model 1 has a total labour requirement of 61 people while Model 2 requires 101 people. This labour input is much lower than the maximum requirement of 14 people - in the case of natural salting/drying, 650 kg throughput - for the smaller scale processes analysed, with only 3 people being required for the traditional smoking process. This variability in labour requirements is likely to affect the location of the respective processing operations. Thermal processing plants are most likely to be based in urban areas where there exists a sufficient pool of skilled and semi-skilled labour. The smaller-scale processes, however, with their much lower labour requirements, are likely to be based at locations as near as possible to the fish supplies. In many cases, this may be in small rural fishing communities. Adoption of such processes can therefore be expected to increase rural employment and arrest the drift of population from rural areas.

The more exacting infrastructural requirements of thermal processing will also tend to favour their location in urban areas. Factories require electricity supplies and the need to ‘buy in’ a high proportion of supplies (i.e. fuel oil, cans, filler, etc.) necessitates reasonable communications. This latter point is also important in that the output of a canning factory may be marketed over a much larger area.

Finally, although no attempt has been made to quantify the foreign exchange requirements, expenditure on the smaller scale processes described is very small (except for mechanical drying where a relatively modest amount of foreign exchange expenditure is required). Canning, on the other hand, has a high foreign exchange component in plant and equipment costs and, where there is a need to import cans, in variable costs.