SIMULATION AND VALIDATION OF ON-FARM IN-BIN DRYING AND STORAGE OF ROUGH RICE

The objectives of this study were: (1) to determine accurate models for predicting equilibrium moisture content (EMC) of rough rice at set conditions of air temperature and relative humidity (RH); and (2) to validate developed mathematical model for predicting moisture content (MC) and temperature profiles during natural air (NA), in-bin drying, and storage of rough rice. Adsorption and desorption isotherms of long-grain hybrid rough rice cultivar, Clearfield (CL) XL745, at temperatures ranging from 15 degrees C to 35 degrees C and RHs 10% to 90% were determined by using a Dynamic Vapor Sorption analysis equipment. Non-linear models were used to determine constants of models for predicting the rough rice adsorption or desorption EMCs. The best model to describe the studied rough rice adsorption isotherms was the modified Halsey equation (RMSE = 0.54% MC dry basis), while the modified Chung-Pfost equation (RMSEs = 0.91% MC dry basis) was best to describe desorption isotherms. The updated rough rice EMC prediction equations were incorporated into an equilibrium-based finite difference model to simulate temperature and MC of rough rice during in-bin drying at selected rice growing locations in Arkansas. The model was validated using field experiments that used modern, on-farm bins equipped with sensors for in-bin air RH and rough rice temperature measurements. The rough rice MC was calculated using the equilibrium models with inputs of measured RH and temperature data. Analyses were conducted to compare sensor-determined rough rice MC and temperature data to that determined from both the simulation model and from laboratory moisture meter-measurements. It was concluded that rough rice MC data determined by field sensors were marginally higher than the meter-measured MCs; the MC offset between in-bin sensor-determined MC data and meter-measured moisture data significantly reduced to nearly 0.5% by the time rice was dried to target levels; and the simulated rough rice MC and temperature profiles reasonably predicted the rough rice temperature and MC data in the field with mean Root Mean Square Errors (RMSE) of rough rice MC and temperature less than 0.57 +/- 0.10% MC and 1.91 +/- 0.21 degrees C, respectively. The study generated useful information for predicting rough rice MC and temperature during NA in-bin field drying and storage.