OPTIMIZATION OF THE PROCESS PARAMETERS OF AN EXISTING CASHEW NUT SHELL OIL EXPELLER

Abstract

ABSTRACT The shell of the cashew nut in its natural state is leathery and contains thick vesicant oil called Cashew Nut Shell Liquid (CNSL) within a sponge-like interior. The oil has been known for its innumerable applications and its ability to undergo all the conventional reactions of phenols. This study optimized the existing expeller by monitoring the process parameters used in assessing product quality. The machine consists of a frame, compression chamber, perforated cylinder, hopper and funnels. Design Expert 13 package was used for Analysis of Variance (ANOVA) and response surface optimization. A CCRD of RSM factorial design was used for the process parameters and varied at 2 varieties (Ogbomoso and Brazilian). Also RCBD factorial design of 3 levels of sizes (small, medium and large), 3 levels of moisture content (14.5, 15.9 and 16.4% w.b.), 4 levels of wormshaft speed (30, 45, 60 and 75 rpm) and 2 levels of varieties (Ogbomoso and Brazilian) in triplicates were varied. The machine response varizbles namely machine throughput, machine throughput efficiency, machine processing temperature and oil qualities (free fatty acid, oil volatile impurities at 105 oC, rancidity level and acid value) were studied. The optimum performance of the machine at maximum machine throughput, machine throughput efficiency, machine processing temperature, free fatty acid, oil volatile impurities at 105 oC, rancidity level and acid value were 0.226 and 0.279 Kg/hr, 50.76 and 80.63%, 125.2 and 123.3 oC,1.84 and 2.02%m/m, 14.2 and 22.8%m/m, 5.1 and 7.04 Meq/Kg, 0.3 and 0.14%m/m for Ogbomoso and Brazilian varieties respectively. Meanwhile, the optimum performance of the machine at maximum machine throughput, machine throughput efficiency, wormshaft speed and moisture content were 0.408 and 0.306 Kg/hr, 91.68 and 86.02%, 30 and 35 rpm, 14.5 and 15.3% w.b. for Ogbomoso and Brazilian varieties respectively. The interaction effect of process variables was significant (p<0.05) on oil qualities, machine throughput and machine thrroughput efficency of both Ogbomoso and Brazilian sizes. Thegenerated regressionmodelswere suitable for predicting machine throughput, machine throughput efficiency and machine processing temperature based on highest coefficient of determination (R2) of 0.90-0.99, significant lack of fit (0.050-0.778) and models significance (p<0.05). The optimum condition was found to be heating temperature of 130 and 70 oC, heating duration of 5 and 5 minutes, storage duration of 4 and 1 months, monitored average relative humidity of 58 and 56%, monitored average atmospheric temperature of 32.2 and 31.2 oC, free fatty acid of 1.84 and 2.02%m/m, oil volatile impurities at 105 oC of 14.2 and 22.8%m/m, rancidity level of 5.1 and 7.04 Meq/Kg, acid value of 0.3 and 0.14%m/m with a desirability of 0.695 and 0.728 for Ogbomoso and Brazilian varieties respectively. The study revealed an optimised machine for good quality oil attainable within the process parameters range. The optimum condition of the machine produced up to 0.41 Kg/hr and 90% of machine throughput and machine throughput efficiency respectively for both Ogbomoso and Brazilian varieties without compromising standard quality.