White kenkey is one of several variants of kenkey, a staple food in Ghana that is made from fermented maize dough. Its production is done by artisanal processors, using non-standardized procedures. These result in wide variations in quality characteristics from among producers, and within batches of kenkey from the same producer.  The determination of optimum process conditions for critical process variables for the production of white kenkey is crucial in achieving a desired product quality in a consistent manner. The objective of this study was to optimize processing parameters of steeping time, fermentation time and steeping temperature on physicochemical properties of white kenkey. A Box-Behnken design was used to study the effects of processing variables of steeping time, steeping temperature, and fermentation time on white kenkey quality parameters of moisture content, pH, titratable acidity, glucose and lactic acid content. Sensory (hedonic) tests were done on the white kenkey using incomplete block designs (k= 5, b= 21, N=105, λ=2, r =7). The data obtained from the sensory tests were fitted to response surface regression models to determine the effects of process variables on the sensory attributes of white kenkey. Contour plots of each of the sensory attributes were generated from the regression models. All the plots were then superimposed on the same axis to obtain the optimum region of the process variables. 

Each of the process variables of steeping time, steeping temperature and fermentation time had significant effects on the physicochemical and sensory characteristics of white kenkey.  The optimum region for  process variables at which the most acceptable kenkey was obtained were steeping time of 30-45h at 30-35ºC temperature and 12 h dough fermentation.  These optimum process conditions were successfully validated, and could be adopted by artisanal processors to assure consistent quality parameters for white kenkey.

Keywords: Box-Behnken, Response surface methodology, optimization, white kenkey, processing variables, contour plots