Z, m, η being integers
5. Packing and Storage
Packing in cans filled with nitrogen gave the best storage properties.
Compression was satisfactory if carried out at a water content of about 10% so that the product was not brittle enough to be pulverized.
Storage life was very dependent on temperature, varying from several years at 10° to 1 5 ° C . (50° to 6 8 ° F . ) to only a few months at 3 7 ° C . ( 9 8 ° F . ) . It did not seem to b e affected by moisture content within the range 2 to 12%, although not all workers are in agreement on this.
E . V A C U U M F R E E Z E - D R Y I N G
In Section 2 , E above, desirable conditions were listed for rapid freeze-drying. From these, it is immediately obvious that on a com
mercial scale an extension of the normal laboratory method of freeze-drying is impracticable because high rates of heat transfer are required.
Clearly, a load of fish supported en masse inside a vacuum chamber would receive practically no heat (except on those surfaces receiving radiation from the walls of the chamber) and most of it would cool by evaporation until no further drying took place.
The simplest method of providing the necessary heat is to support the fish on heated trays, and for most small-scale purposes this is perhaps adequate. However, the heat transfer coefficient between the trays and the fish is not normally great enough to permit the achievement of economically high drying rates.
F. V A C U U M CONTACT DEHYDRATION2
Fish was first dried on a commercial scale in a vacuum by the Danish
"Pressfisk" process (Hay, 1955). The equipment (Fig. 11) consisted of a chamber of rectangular section containing 24 horizontal mild steel heating plates grouped in three banks, one above the other, with 8 plates per bank, each bank thus accommodating 7 trays of material.
In this way, batches of up to 1 ton of raw material could be processed per run. Heat was supplied by hot water circulating in the hollow
2 T h e writer is i n d e b t e d to Mr. J . C . F o r r e s t of the E x p e r i m e n t a l F a c t o r y , Ministry of Agriculture, Fisheries a n d F o o d , A b e r d e e n , S c o t l a n d , for assistance in the p r e p a r a t i o n of this a n d the following subsection.
heating plates. In order to maintain thermal contact between the plates and the fish as the fish shrank during drying, the plates were m a d e to close progressively by a hydraulically actuated lever mechanism. The
F I G . 1 1 . L a r g e v a c u u m c h a m b e r for d r y i n g fish showing a r r a n g e m e n t of h e a t e d plates. ( B y kind permission of Ministry of Agriculture, Fisheries a n d F o o d . )
chamber was evacuated by means of two steam augmenters backed by two water ejectors, and the pressure ranged approximately from 15 down to 2 mm. H g as drying proceeded.
Because the water-vapor pressure in the system was always greater
than the vapor pressure of ice at the initial freezing temperature of fish ( — 1 . 1 ° C . ) freeze-drying could not take place. Although shrinkage occurred during drying, the product was nevertheless somewhat porous.
The time taken to reduce the moisture content of the fish down to 15%
was about 7 hr. At this stage the mat of fillets was removed from the chamber, cut up into a suitable shape, compressed by means of a hydraulic ram, and subsequently dried in vacuo to a moisture content of about 5% (dry weight basis).
The method suffered, among other things, from the disadvantage that reconstitution was very slow compared with that of freeze-dried fish and was by no means complete after many hours of soaking in water.
Even then the texture tended to be tough and stringy (Hanson, 1959).
Despite the apparent disadvantages of the original "Pressfisk"
process, there was promise of improvement when the then British Ministry of Food decided in 1948 to subject the method to intensive investigation. Workers in the Ministry's Experimental Factory in Aber
deen later modified the process (Hay, 1955).
It was found that compression of the dried product led to a marked reduction in the rate of reconstitution so that, despite the implications for storage and transportation, debulking had to b e abandoned. In order to improve both the reconstitution properties and the quality, many variations in the pretreatment and post-treatment of the fish were devised. It was ultimately found possible, by a method consisting essentially of an alkaline dip before drying, to obtain a product which would reabsorb water over a period of an hour or two. However, the fish had by then lost its characteristic texture and would not reconsti
tute to anything like its initial moisture content. It was, therefore, far from having the desirable qualities of freeze-dried fish—retention of nutrients, appearance, flavor and texture, and rapidity of reconstitution.
At this stage, improvements in the ejector system enabled the operating pressure in the chamber to be reduced to about 1 mm. Hg, and it was then found possible partly or completely to freeze-dry the fillets. It had earlier been found that variations in the thickness of layers caused wide discrepancies in the final moisture content at different points in the material, the thinner portions drying more slowly as a result of poor thermal contact. A special flat-cutting machine (Hay, 1955) was developed to cut oversize fillets down to a uniform maximum thickness, and when this was used to prepare fish for the drying plant, freeze-drying could b e achieved consistently, uniformly and more rapidly than was previously possible.
The speed of drying was further increased by cutting fish flesh across the fibers, since the resistance to vapor flow is less in the direction
parallel to the fibers than in a transverse direction. The fish was first frozen and cut into steaks of uniform thickness and then loaded, still frozen, into the drier. By means of this procedure, fish steaks 17 mm.
thick were dried in 1 0 ^ to 11 hr. This process is known as the V C D method.
G . ACCELERATED F R E E Z E - D R Y I N G
The advantages of the V C D method over conventional shelf drying is that heat is conducted into material through top and bottom surfaces simultaneously, and better thermal contact can b e maintained between the heated plates and the fish. In principle, this should lead to a rate of drying considerably in excess of double the rate of shelf drying.
However, the close contact between plates and fish produces a high resistance to vapor flow, and this "stifling" effect tends to offset the advantage of good thermal contact. The mutually conflicting require
ments of maximum heat flow and minimum resistance to vapor flow cannot b e satisfactorily resolved. Excessive heat transfer coupled with a low vapor escape rate can result in thawing, and the fish suffers damage in consequence.
Recently, considerable improvement was effected (Dalgleish and Thompson, 1958) by introducing an innovation which largely removes this difficulty. If a piece of expanded aluminum sheet is placed between prod
uct and heating plate on the top and bottom surfaces ( F i g . 12), a series of channels is formed, separated by the mesh formation. The channels provide an escape path for the water vapor, thus markedly reducing the resistance to flow, while the mesh itself, being in good contact with the material, assures an adequate flow of heat to the surface of the fish. Moreover, since the points of the expanded sheet form indentations on the surface, some penetration of the dry layer of material is achieved. Using this equipment, the drying time for 15-mm.-thick cod steaks is reduced from 1θγ2 or 11 hr. to 6y2 or 7 hr. Figure 13 compares the temperature behavior of fish and heating plates in the V C D process with that of the acelerated freeze-drying ( A F D ) process. T h e tempera
ture deep within the material reflects the approach of the end of drying, and when this temperature reaches that of the water in the heating plates, drying can b e assumed to b e complete.
The application of heat from both sides of the material, coupled with the effect of expanded metal trays on heat transfer and vapor flow, has so increased the initial rate of drying that the removal of water vapor from the vacuum chamber is now a limiting factor at the commence
ment of the process. (It is estimated that the augmenters p u m p water vapor at the rate of about 86,000 liters per second at a pressure of 1 mm.
H g when the initial charge is one ton of fish.) Even so, if some means
F I G . 12. Prefrozen c o d cutlets l o a d e d b e t w e e n e x p a n d e d a l u m i n u m sheets in the "Accelerated F r e e z e - D r y i n g " process.
H. H E A T E D SPIKE F R E E Z E - D R Y I N G
It is of interest to mention that Brynko and Smithies (1958) devised a method of increasing the rate of heat transfer which resulted in drying rates considerably in excess of those of the A F D method. In their method, pieces of meat 19 mm. thick were laid between heater plates fitted with metal spikes. The spikes pierced the meat and conducted heat uniformly and efficiently into the material, and in this way freeze-drying could be completed in 3 to 5 hr. Unfortunately, the practical difficulties involved in impaling the material on the spikes, removing the spikes after drying and keeping them clean have prevented this method from being applied on an industrial scale.
I . V A C U U M F A T DRYING
Piatt and Heard (1946) patented a method in which foodstuffs "in a suitably sized state" are dried by heating the pieces in oil or fat under could be found of maintaining the initial rate of heat transfer to the ice throughout the drying period, such a rate of pumping would result in complete drying of 15-mm.-thick cod steaks in less than 2y2 hr.
Heating water FIG. 13. Temperature behavior on surface and within cod cutlets 17 mm. thick and of water in the heating plates in large-scale vacuum freeze-drying; and corresponding weight loss and plate separation, a: VCD process and b: AFD process. Average absolute pressure, a and b: 1.3 mm. Hg; final moisture content, a: 2.5% and b: 2.3%. ^
reduced pressure at about 8 0 ° C . The fat can later be removed by drainage, or by centrifuging, or by means of a solvent. It is claimed that fish can b e dried in this manner.
Details of the procedure necessary to dry fish are not given, however, nor are drying times. The closest approach to the drying behavior of fish is given by data for raw beef mince which was dried to a water content of 14% after 130 min. No information is given on rates of reconstitution or on the acceptability of the product when reconstituted and cooked.
The method has not been adopted commercially.
J . DEHYDRO-FREEZING
Dehydro-freezing, as its name implies, is a method of food preserva
tion involving the partial drying of the product and subsequent storage at low temperature. The main advantage resulting is that the food is considerably debulked and is, consequently, more economically stored and transported. When required for consumption, it is thawed and reconstituted before preparation.
Although fruit and vegetables have been processed in this way (Talburt and Lindquist, 1954), it is evident from the apparent absence of industrial application that the method is not economically feasible as far as these products are concerned.
The present author has examined the possibility of dehydro-freezing fish. Two groups of cod fillet pieces 10 χ 5 χ 1.5 cm3, were dried in a wind tunnel at 3 0 ° C . and at 30% R.H., under which conditions the drying time-constant was 14.05 hr. The first group was dried to a weight loss of 20.1% (25.1% of the initial moisture content), and the second group was dried to a weight loss of 49.5% (61.9% of the initial moisture content). The fillets were frozen and stored at — 3 0 ° C . for 132 days, thawed and reconstituted in distilled water at 2 0 ° C . The rates of reconstitution were very slow, the water uptake curves being of the form
Wt^W* ( l_e- * / r r )
Wt being the weight of water absorbed at time t after commencement of reconstitution, WM the weight absorbed after immersion for an
"infinite" time, and xr a constant—the "reconstitution time-constant."
The constant xr was 3.47 hr. for the first group and 12.0 hr. for the second group. The former regained 95.6%, and the latter only 76.3%, of their initial water content.
Taste panel assessment of the cooked fish showed both groups to be slightly unpleasant and "unfishlike" to the palate, to have an un
familiar smell, and to be tough and stringy.
The results suggested that the method of dehydro-freezing is un
suitable for the preservation of fish on account of the long reconstitution time, and, more especially, of the inferior and "unfishlike" quality of the cooked, reconstituted product.
IV. A p p r a i s a l of D e h y d r a t i o n a s a Process A. Q U A L I T Y ASPECTS
No dehydrated fish as yet is sufficiently satisfactory to be indistinguish
able from raw fresh fish. However, some products, when incorporated into certain fish dishes, are satisfactory enough. On the other hand, they are not necessarily good enough for use in such standard dishes as fried fish.
Traditionally dried fish, such as stockfish and salt fish, require a considerable time for reconstitution. Unsalted fish is first soaked for many hours in cold water before cooking (the actual time depending on the thickness). Salt fish is usually immersed in cold water overnight, after which the water is decanted and the fish is soaked for a further period of a few hours in fresh water before being cooked. It is often necessary to change the water during the process of cooking in order to reduce the saltiness to a level at which the fish is palatable. Even so, these products are aften excessively salty to most palates accustomed to fresh fish.
All traditional products, when finally ready to eat, are rubbery in texture, difficult to chew, and have a characteristic "cured" flavor.
In many tropical countries, dried fish has such a pronounced flavor that it can only be used as a condiment.
The warm-air-dried product developed in the United Kingdom during World War II can be reconstituted in about 10 min. in cold water. Owing to its granular nature, it is most readily made into fish cakes, kedgerees, or pies, and it is possible to prepare more ambitious and attractive dishes, such as fish souffle. Although acceptable to most consumers, this product is, nevertheless, very different from fresh fish.
Freeze-dried white fish such as is produced by the A F D process is readily reconstituted in cold water in 2 to 3 min. and may be cooked in much the same way as fresh fish. There can be no doubt that this product comes closest in appearance to fresh fish, but in taste and texture it still leaves something to be desired. The taste is "neutral" and not quite "fishlike" and the muscle fibers are somewhat tough to chew (Anonymous, 1957, 1958; Voskresensky, 1959). Nevertheless, when suit
ably flavored and garnished it is generally acceptable.
The reasons for the failure of dehydrated fish to reach perfection are not hard to find. In the first place, owing to the irreversible changes
that take place during drying (denaturation) and the severe damage suffered by the cellular structure, real reconstitution is impossible. The best that can be done is to conserve a porous structure which absorbs and retains sufficient water by capillarity. Compressed products absorb slowly and less completely. With the most successful products, some
thing like 3.5 g. per gram of dried fish can be held. However, as this water is not all "bound," some of it can easily be removed again by pressure or boiled off during cooking in fat.
The manner in which dehydrated fish is prepared for cooking is of considerable importance. Prolonged soaking is undesirable as it is inconvenient and can even be dangerous, since it presents a potential bacteriological hazard. From this point of view, air-dried minced fish and vacuum freeze-dried fish are both very satisfactory. It is evident, too, that all types of dehydrated fish benefit from skilled culinary treatment: unwanted flavors in traditional products are often better masked, while the absence of flavor in the newer products calls for garnishing.
Storage tests (Anonymous 1957, 1958) with nitrogen-packed raw and cooked freeze-dried cod fillets show that there is little or no deteri
oration in texture and taste over a period of many months among samples stored at 0 ° C . and 1 8 ° C . Samples stored at 3 7 ° C . deteriorate rapidly, becoming badly discolored, tough and objectionable to taste after only one month. The discoloration is due to nonenzymic reactions known as "browning" (Jones, 1954a, 1954b, 1957) and the toughening results from cross-links between protein molecules (Connell, 1959).
However, workers at the Experimental Factory, Aberdeen, have found that the storage-life of freeze-dried fish a 3 7 ° C . is very sensitive to moisture content. Above 5 g. water/100 g. nonfatty solids, the fish will be inedible in about a month. At or below 2 g. water/100 g. nonfatty solids, the fish will be palatable after 1 year.
The effect of reducing the moisture content by a given amount is much greater at lower moisture levels (Matheson and Penny, 1961).
B . C O S T S
Finally, as regards cost, the main components are: ( 1 ) raw materials, ( 2 ) labor, ( 3 ) fuel, ( 4 ) depreciation on plant, and ( 5 ) packaging. The raw material cost per pound of dried product depends on the yield of product from the whole fish and the species used. Steam, electricity, and water costs are usually small—only about a cent per pound dry—except for vacuum freeze-drying, where they amount to about 6 cents per pound. Labor and depreciation costs vary greatly with the process. Packaging costs depend largely on the amount of debulking that can be achieved, a likely range being 3 to 20 cents per
pound dry. The actual cost of the drying process itself as itemized under ( 2 ) , ( 3 ) and ( 4 ) above is relatively very cheap, and even the most advanced process, that of accelerated freeze-drying, amounts to only 16 cents per pound dry.
At one extreme, ordinary fish meal is being sold for as little as 3 cents per pound. At the other extreme, vacuum freeze-dried fish could probably not be sold for less than $2.00 per pound. In between, the warm-air-dried minced product might cost 80 cents per pound and salt fish about 20 cents per pound.
For a full treatment of the economics of the A F D method, readers are referred to a publication by Hanson (1961) describing all aspects of the process.
R E F E R E N C E S
Alden, C . ( 1 8 7 7 ) . Processes of T r e a t i n g F i s h . U . S . Pat. N o . 186,093.
Alden, C . ( 1 8 8 0 ) . Preserving F r e s h F i s h . U . S . Pat. N o . 2 3 5 , 1 1 6 .
Alty, T . , a n d M a c K a y , C . A. ( 1 9 3 5 ) . T h e a c c o m m o d a t i o n coefficient a n d the evaporation coefficient of water. Proc. Roy. Soc. A 149, 104-116.
A n o n y m o u s ( 1 9 5 7 ) . Gt. Brit. Dept. Sei. Ind. Research Food Invest. Rept., p . 19. of salt fish muscle. J . Fisheries Research Board Can. 4, 136-140.
C o o p e r , D . le B . ( 1 9 4 2 ) . Artificial drying of salt fish. R e / r i g . and Air Condi
H a n s o n , S. W . F . ( e d . ) ( 1 9 6 1 ) . " T h e A c c e l e r a t e d F r e e z e - D r y i n g ( A F D ) M e t h o d Commission Internationale Technique de Sucrerie, London p p . 170-178.
K a t c h a t u r o v , L . J . ( 1 9 5 6 ) . D e t e r m i n a t i o n of the thermal properties of foodstuffs.
Powell, R . W . ( 1 9 4 0 ) . F u r t h e r experiments on the evaporation of water from saturated surfaces. Trans. Inst. Chem. Engrs. 18, 3 6 - 5 0 .
Preston, J . S. ( 1 9 5 8 ) . Private communication.
Scott, G . C . ( 1 9 2 3 ) . Principles of m o d e r n dehydration a p p l i e d to t h e drying of fish. In " T h e M a r i n e Products of C o m m e r c e " ( D . K . Tressler, e d . ) , p p . 3 7 4 - 3 7 6 . C h e m i c a l C a t a l o g C o . , N e w York.
S i d a w a y , E . P., a n d Y o u n g , O. C . ( 1 9 4 3 ) . A n air-conditioned tunnel for processing fish. Fisheries Research Board Can. Progr. Rept. Pacific Coast. Stas. No. 56, 4 - 5 . Stansby, Μ . E . ( 1 9 4 5 ) . D e h y d r a t i o n of fishery p r o d u c t s . U.S. Fish Wildlife Sew.
Fisheries Leaflet No. 120, 7 p p .
Stephenson, J . L . ( 1 9 5 3 ) . T h e o r y of v a c u u m drying of frozen tissues. Bull. Math.
Biophys. 15, 4 1 1 - 4 2 9 .
T a l b u r t , W . F . , a n d L i n d q u i s t , F . E . ( 1 9 5 4 ) . D e h y d r o - f r e e z i n g of fruits a n d v e g e t a b l e s . Food Sei. Abstr. 26, 3 6 1 .
Tarr, H . L . A. ( 1 9 4 3 ) . C h a n g e s in moisture, micro-organisms a n d volatile b a s e s in d e h y d r a t e d fish d u r i n g p r o c e s s i n g a n d storage. Progr. Rept. Biol. Stas.
Nanaimo and Prince Rupert N o . 57, p p . 16-19.
Tarr, H . L . A. ( 1 9 4 5 ) . W a t e r , m i c r o o r g a n i s m s a n d volatile b a s e s in d e h y d r a t e d fish. J. Fisheries Research Board Can. 6, 3 0 3 - 3 1 0 .
T e m p l e m a n , W . , a n d A n d r e w s , G . L . ( 1 9 5 6 ) . Jellied condition in A m e r i c a n p l a i c e Hippoglossoides platessoides ( F a b r i c u s ) . J. Fisheries Research Board Can.
7, 5 5 8 - 5 9 3 .
T o w n s e n d , C . S . ( 1 9 2 2 ) . I m p r o v e m e n t s Relating to F i s h F o o d s . Brit. P a t . N o . 1 9 7 , 7 2 9 .
Tressler, D . K . ( 1 9 2 3 a ) . T h e preservation of fish b y smoking a n d drying. In
"Marine P r o d u c t s of C o m m e r c e " ( D . K . Tressler, e d . ) , p p . 3 7 0 , 3 7 1 . C h e m i c a l C a t a l o g C o . , N e w York.
Tressler, D . K . ( 1 9 2 3 b ) . T h e preservation of fish b y smoking a n d drying. In
"Marine Products of C o m m e r c e " ( D . K . Tressler, e d . ) , p . 3 7 4 . C h e m i c a l C a t a l o g C o . , N e w York.
Voskresensky, N . A . ( 1 9 5 9 ) . T h e influence of sublimation drying on changes in moisture content of fish tissue. ( I n R u s s i a n . ) Trudy VNIRO 40, 5-20.
Weisselberg, Α., a n d W e i s s e l b e r g , H . ( 1 9 4 4 ) . Better w a y t o d r y salt fish. Food Industs. 16, 9 0 - 9 2 .
W o o d , A. L . ( 1 9 5 1 ) . T h e drying of salt-fish. Fishing News N o . 1 9 8 2 ( A p r . 1 4 ) , p. 12.
Young, L . , a n d L e e , C . F . ( 1 9 4 3 ) . D e h y d r a t i o n of fish p r o d u c t s . Fish Gaz. 60, 64, 7 4 .
Young, O. C , a n d S i d a w a y , E . P. ( 1 9 4 3 ) . T h e dehydration of fish. Fisheries Research Board Can., Progr. Rept. Pacific Coast Stas. No. 56, 7-9.