Development of a Model for Thin Layer Solar Drying of Salted Fish Fillets

Abstract A batch type direct passive solar dryer was designed and fabricated to dry salted tilapia fish fillets. The dryer has a capacity a capacity of 1500g/batch. The tilapia fish species were obtained from Domita Farms Nig. Ltd,.Uyo, Akwa Ibom State and transported immediately to Nsukka in a cold storage container. The salted fish fillet was prepared to conform to thin layer drying specification of 3.0mm-5.0mm thickness before drying. The solar drying experiments were conducted at the Department of Agricultural and Bioresources Engineering, University of Nigeria, Nsukka. Nsukka is located on latitude 60 56 N and longitude 70 25 E. and 397m above sea level. The work was carried out as a factorial experiment with two factors in two and three levels respectively arranged in a completely randomized design (CRD) with six treatments and ten replicates. The first factor was fish with two levels, descaled fish and fish with intact scale. The second factor was common salt (sodium chloride) (NaCl) having 0%, 10% and 20% concentrations. The samples when removed from refrigerated storage were allowed to equilibrate with the ambient conditions before being prepared for solar drying. The fish were characterized to obtain its initial moisture content (M,% db),length(m), mass(g), crude protein(CP) (%), crude lipids (CL) (%), equilibrium moisture content (EQ) (%db) and ash (%). A total weight of 1200g/batch drawn from refrigerated storage were cleaned, eviscerated, heads removed, washed and filleted longitudinally to expose the backbone in preparation for solar drying. The fish fillets were treated with salt (Nacl) to prevent spoilage by microorganisms and to dehydrate the fish thereby enhancing the drying process. The experimental drying data was fitted to 19 models and the quality of fit evaluated. A good fit is indicated by a high coefficient of determination (R2 ), low reduced chi- square (χ 2 ), standard error (Se) and mean relative deviation modulus (P). Four out of the 19 models showed reasonable 19 fit. Out of these four models, the Wang and Singh model was fairly good and therefore was further modified and ran on a non linear regression program (NLREG) which gave higher R2 , and lower χ2 , Se and P when compared with the original Wang and Singh model of the form MR = M +At +Bt2 . A is the initial drying constant; B is the final drying constant, MR is the moisture ratio, M is a constant. A fair quality fit to the experimental data when modified yielded the new model. MR = M+ 0.048 t 0.8 + 0.067 t 0.76. For the new model R2 varied between 0.9779 and 0.9989, Se varied between 0.0014067 and 0.042528, χ 2 varied between 1.8495x10-6 and 0.00189 and P varied between 0.0139 and 29.78% respectively. The length of tilapia fish was found to vary between 9cm and 13cm, weight varied between 50g and 100g.The proximate composition of tilapia fish were: initial moisture content (Mo) varied between 254.43 – 293.13%(db), crude protein (CP) varied between 13.0 and 16.25%, crude lipids (CL) varied between 15.5 and 27.83% and ash between 14.74 and 16.16%, equilibrium moisture content (EQ) varied between 15-20% (wb). Drying kinetics of salted fish had moisture ratio (MR) varying between 0.022 – 1.0 at 10 % salt concentration, 0.0408 – 1.0 at 20% salt concentration. Mo was between10.38 - 573.4 and 18.63 – 564.48 for 10% and 20% respectively. Fish fillets without salt treatment had MR varying between 0.020 – 1.0, Mo varied between 12.82 –564.89. Physicochemical quality had the following values microbial count (MC) varying between 0.6x104 -1.1x104 CFU/g (colony forming unit per gram),Fungal count (FC) varied between 0.7x103 – 4.6x104 CFU/g, total volatile basic nitrogen (TVBN) varied between 12.0 –20.7 mg N/100g flesh (milligram nitrogen per 100g flesh). For fish fillets without salt treatment, MC varied between 0.7x104 -1.3x104 ,FC varied between 0.9x103 – 2.8x103 ,TVBN was between 29 -50.TVBN for control was above 30.0. Specific solar radiation available varied between 32.53 -37.70MJ/m2 .