Continuous Monitoring of Moisture Content in Compacted Hay during Drying

Continuous and accurate measurement of moisture content (MC) would represent a significant improvement for on-site monitoring of commercial baled hay drying. A moisture meter based on electrical resistance was used to determine MC of eight hay layers continuously in a laboratory dryer. Two moisture probes and one thermocouple were inserted in individual layers of hay of 135-mm thickness; the eight layers were one over each other in a vertical drying column. The probe MC measurements were compared with the MC obtained by oven drying and mass balance. Three initial levels of MC (20%, 30%, and 40% on a wet basis, w.b.) and two levels of drying air temperature (40°C and 50°C) were repeated three times during the first cut of perennial grass. The difference between oven moisture and moisture probe measurements was influenced by initial MC and air temperature. Linear regression based on moisture probe readings showed that the best prediction model had an R² = 0.919 at an initial MC of 30% and a drying temperature of 40°C. The moisture probes were less accurate to predict oven moisture at a high initial MC of 40% with R² as low as 0.601. When air temperature was included in the prediction model, R² statistics increased and values ranged from 0.685 at high initial MC to 0.949 at low initial MC of 20%. The smallest root mean squared error (RMSE) was observed at low MC (±0.81%); RMSE increased with the initial MC up to ±4.62%.