Accuracy of Grain Moisture Content Prediction Using Temperature and Relative Humidity Sensors

Grain temperature and moisture content (MC) are considered to be principal factors for safe storage of grain.Continuous monitoring of temperatures within grain masses is relatively easy using thermocouples, but monitoring of MCis limited by availability of sensors. However, temperature and relative humidity (RH) can be used to predict grain MC basedon equilibrium moisture content (EMC) equations such as the Modified Henderson, Chung-Pfost, or Oswin. These modelsare limited to quasi-static thermodynamic conditions but do provide a method to predict MC with commercial sensors. Erroranalysis was performed using EMC relationships for wheat to determine the error in grain MC prediction due to sensor error.EMC prediction errors were found to be .0.25% to .0.65% MCdb between the RH ranges of 20% to 70% RH. At higher RHlevels, prediction error increased substantially. Sensor error was set to .2% RH and .0.4.C, for the error analysis. The sensorerror was adopted from a commercial sensor that could be potentially used for a cabled monitoring system. At higher levelsof sensor error (.3% RH, .0.4.C and .4% RH, .0.4.C), prediction error increased from .0.38% to .0.96% MCdb and from.0.65% to .1.29% MCdb , respectively, for the same RH range. Prediction error due to sensor error was found to be of thesame magnitude as the standard errors of regression models developed for wheat. Measurements of sensor accuracy were alsoperformed and accuracy was found to be within or better than rated manufacturer specifications for RH but temperatureaccuracy was less than rated accuracy.