Three natural-convection solar dryers were constructed and operated, and their performance compared with that of two sun-drying methods. Each of these are described in turn. In order to ensure similar loading and to simplify comparison of performance, drying racks or trays of approximately equivalent dimensions (1.8 m X 0.6 m) were built within each of the solar dryers.
Solar tent dryer
The solar tent dryer was developed by Doe and others, (Doe et al. 1977; Doe, 1979) and initially tested in Bangladesh with fish. TPI has tested this dryer in Africa, SE Asia and Latin America for production of various dried fish products and it has shown considerable promise.
A sketch of the dryer as built at the CDRS is shown in Figure 1, Appendix 1. For this exercise, it differed in one important aspect from the original design of Doe et al., in that both sides of the tent were of clear plastic sheet, whereas for the dryers built in Bangladesh, only the side facing the sun was of clear plastic, the other being of black plastic sheet. The reason for this was that the working site at the CDRS being almost exactly on the equator, the path of the sun was virtually due east to due west for the period of the exercise (April/May). Therefore, it was considered that the slight increase in efficiency of solar collection effected by a side of black plastic would be more than offset by the reduction in the 'greenhouse effect'.
The principle of operation of the tent dryer is that insolation passes through the clear plastic sides and ends of the tent and is absorbed on the black plastic base. Air at the base is thereby heated and rises, thus inducing a draught within the tent. Openings at the base along both sides allow air to be drawn in, and vents in the apex at both ends allow air to exhaust. Some control of the internal temperature, and flow of air through the dryer, can be maintained by adjusting the height of the side openings.
Construction of the dryer was very simple, using a bamboo framework and plastic sheet. Black polyvinyl chloride (PVC) was used for the base of the tent and clear ultra-violet-resistant polyethylene for the sides and ends. Staples were used to attach the plastic sheet to the framework. The drying rack was built along one side of the tent using bamboo and black plastic mesh. Access to the rack was through a movable plastic flap forming half of one end of the tent. The flap could be closed and fastened when not in use. Construction time for the tent dryer was about 6 manhours.
SCD solar dryer
This dryer was developed by Exell and others (Exell and Kornsakoo, 1978; Exell et al., 1979; Exell, 1980) at the Asian Institute of Technology (AIT) in Thailand. It differs principally from the tent dryer in that the solar collector and the drying chamber are distinctly separate as can be seen from the sketch of the dryer in Figure 2, Appendix 1. The dryer was developed for use with paddy but recent work has been conducted at AIT using fish (Exell, private communication).
The solar collector consists of a black plastic base with an inclined transparent plastic cover with a narrow opening across the full width of one end of the collector. Air is heated during its flow through the collector and passes into the drying chamber before exhausting through the chimney. The function of the extended chimney is that the black sides absorb insolation and so heat the air within, thereby enhancing the natural convective flow of air through the dryer.
Identical materials to those used for the solar tent were employed; the base of the collector, the base and back of the drying chamber, and the sides of the chimney were of black PVC, and the collector cover and the top and sides of the drying chamber were of clear polyethylene. Access to the drying rack was provided by plastic flaps on either side of the drying chamber, which also provided a rudimentary means of control of the internal temperature. It should be noted however that for larger models it would be necessary to provide access flaps at the back of the dryer for ease of loading as described by Exell and Kornsakoo, 1978; Exell et al., 1979 and Exell 1980. Construction time for the dryer was about 15 man-hours.
Solar cabinet dryer
This design of dryer was pioneered by Lawand (1966) and the Brace Research Institute (1973) and is probably the most widely-used dryer developed to date, being utilised for a large number of commodities. As its name suggests, and as can be seen from Figure 3, Appendix 1, it is essentially a rectangular cabinet with an inclined transparent cover. The optimum angle of inclination of the cover is dependent upon the geographical latitude (Brace Research Institute, 1973); for the Galapagos this angle was 15°.
Air inlet ports in the base of the cabinet provide an entry for air which is then heated within the cabinet and rises to exhaust through outlet ports in the upper parts of the front and sides. The potential of the cabinet to absorb insolation is enhanced by blackening all interior surfaces. It is normally recommended that the front, sides and base be of double-wall construction with the cavity being filled with a material with good insulation properties, e.g. sawdust. However, the cavity was left unfilled during this exercise to prevent excessive internal temperatures being attained since there is the risk of 'case-hardening' and 'cooking'. Case-hardening is the formation of an impermeable surface layer of the fish, caused by too rapid drying initially, before the moisture in the deeper layers has had an opportunity to diffuse to the surface. The resulting product has a hard, well-dried surface, but the centre remains moist and will spoil.
For this exercise, the cabinet was constructed of plywood on a frame of 50 mm X 50 mm wood; 12-ply thickness sheet was used for the external walls of the sides and front, and top of the base, and 5-ply sheet for the internal walls of the sides and front and the bottom of the base. Two doors of 5-ply sheet formed the back of the cabinet, allowing loading of the trays. All interior surfaces were painted with black matt paint. Plastic hose was used to form ducts through the two walls for the inlet and outlet ports. The cover was a single sheet of clear polyethylene stapled to the frame. The dryer was mounted on legs to facilitate air entry through the base and to reduce the risk of entry of pests into the cabinet. Simple temperature control was achieved by leaving the doors partially open. Construction time for the cabinet dryer was about 25 man-hours.
Drying on rocks
The traditional practice of drying fish in the Galapagos is to spread them on the black lava rocks that abound in the islands. No steps are taken to protect the fish from contact with birds, flies and other pests.
A bed of black rocks was therefore assembled and used in the local manner.
This simple and effective improvement on traditional techniques has been brought into use in recent years in various countries such as Malawi. It consists of an inclined perforated rack supported on a simple framework; it affords some protection of the fish against crawling pests and contamination with dust, but not against flies. The rack built for this exercise was made from black plastic mesh and bamboo poles. Materials, other than plastic mesh, that would allow a virtually unrestricted air movement around the rack could have been used; chicken wire would be one alternative. The drying rack was approximately 1m above the ground.
A total of ten drying experiments were carried out using two species of locally available, freshly caught fish, lisa (Xenomugil thoburni) and pampano (Trachinotus paitensis). The range of preparation methods and pre-treatments was as follows:
Experiments 1 and 2
Dry salted split lisa:The fish were split according to local custom: they were cut along the backbone from head to tail, the guts and gills were removed and a cut made under the backbone to open out the thick part of the flesh and expose a greater surface area for salting and drying. The fish were carefully washed and the flesh scored. The prepared fish were soaked in 10% brine for 30 minutes to make the product firmer and to allow bleeding. They were then packed in dry salt, using one part salt to three parts fish by weight, and left overnight (for a maximum of approximately 21 hours) in a store.
Brined split pampano: The fish were split along the backbone, gutted, washed and all but the last third of the backbone removed. The flesh was scored and the fish totally immersed in saturated brine for 40 minutes.
Brined pampano fillets: Two single fillets were removed from each fish. The fillets were washed, scored and placed in saturated brine for 40 minutes.
Experiments 5 and 7
Brined split lisa: After being prepared in the same manner as for Experiments 1 and 2, the fish were placed in saturated brine for one hour.
Experiments 6 and 8
Unsalted split lisa: The fish were split as described previously but, after washing and scoring, were allowed to drain and were set to dry without salting.
Experiments 9 and 10
Dry salted lisa fillets: Two single fillets were removed from each fish. The fillets were washed and scored prior to salting in the same way as for Experiments 1 and 2 by being soaked in 10% brine for 30 minutes and then packed in dry salt overnight (approximately 12 hours). During this exercise, locally-produced solar salt was used for preparation of brines and for dry salting. After salting, the split fish or fillets were carefully washed to remove excess salt crystals from the surface and were allowed to drain. Each batch of fish was divided into five lots and set to dry in the three solar dryers, on the rack and on the lava rocks simultaneously.
As can be seen from Plate 1, Appendix 6, the rocks, the rack and the three solar dryers were closely grouped in an open sunny aspect. Care was taken in positioning the dryers to prevent any of the taller dryers throwing shade upon another. A series of plates of the dryers is given in Appendix 6.
Over the four-week period of operation the fish were placed in the dryers in the morning (07.30 - 07.45) and removed in the early evening (17.15 - 17.30). Approximately 8 kg of prepared fish could be accommodated in each of the solar dryers; this loading was maintained during ail the drying experiments. The fish were turned regularly, twice a day, to ensure even drying. At night, the fish were press piled in plastic bins which were kept in a store. When it rained, the fish on the rocks and on the rack were covered with plastic sheets.
Measurements were taken at hourly intervals of the following: insolation using a solarimeter and integrator (Lintronic Ltd, 54 - 58 Bartholomew Close, London, EC1A 7HB), ambient and internal dryer temperatures with a temperature recorder (Grant Instruments (Cambridge) Ltd, Barrington, Cambridge, Cambridgeshire), ambient humidity with a whirling hygrometer (C. F. Casella and Co. Ltd, Regent House, Britannia Walk, London, N1 7ND) and wind speed using an anemometer (C. F. Casella and Co. Ltd).* Each batch of fish was weighed at intervals of three hours during the day. When successive weighings indicated little or no change in weight, the fish were considered effectively dry and were removed from the dryers, allowed to cool, and held in plastic bags in the store. During the course of an experiment, a record was kept on the condition of the fish as they dried.
Adjustment to, and control of, the dryers was deliberately kept to a minimum to simulate their operation by a busy fisherman. Action to control the temperatures within the dryers was minimal. Vents and openings were closed in the early morning and late afternoon in order to raise internal temperatures as quickly as possible and maintain them as long as possible, and reasonable care was taken during the hottest periods of the day to prevent the internal temperatures exceeding 55 - 60°C due to the risk of either case-hardening or cooking occurring.
Quadruplicate samples of all of the batches of the final dried salted products were analysed for moisture, salt and acid-insoluble ash. Moisture contents of 10 9 samples were determined using an infra-red moisture balance (Model MB-10, Chyro Balance Corp., Japan). Salt contents were determined by macerating 25 9 samples in water and measuring the chloride concentration in the filtered solution by titration with silver nitrate using potassium chromate as indicator (Pearson, 1970). Acid-insoluble ash was determined after digesting the total ash in 10% hydrochloric acid (AOAC, 1980a and b).