Optimisation of the Cooking Process of Sheanut Kernels (Vitellaria paradoxa Gaertn.) Using the Doehlert Experimental Design

The response surface method employing Doehlert's experimental design was used to optimise the cooking of sheanuts to strike a balance between the advantages and disadvantages usually offered by the process. The independent factors investigated were cooking time, cooking temperature and nut size while the responses were moisture content of the kernels, amount of oil extracted, acid and peroxide values of the butter. Second order polynomial models were generated to describe the process for the responses studied. The validity of the models was tested and it was found that they could be used to explain respectively 83%, 99% and 95% of the variation of moisture content, acid value and peroxide value. The cooking process greatly reduced the free fatty acid values of the butter (< 6%) but increased its peroxide value (up to 20 meq/kg). The cooking process was significantly influenced by all three independent factors investigated. The optimum conditions defined for the cooking process were: cooking time (95-120 min), cooking temperature (75-90 A degrees C) and nut size (40-45 mm). These optimal conditions gave the following responses: moisture content 51.97% w.b., amount of oil extracted 47.47%, acid value (as FFA) 2.76% and peroxide value 3.87 meq/kg. The parameters of the cooking conditions could be set to appropriate values to give butter of either category 1, 2 or 3 in terms of acid and peroxide values.