Energy-efficiency optimization of the biomass pelleting process by using statistical indicators

Biomass pelleting process strongly depends on a number of variables hard to be simultaneously controlled. This paper suggests a method to ensure pellets moisture optimization and process energy saving. An experimental testbed was arranged in order to validate the performance of the proposed strategy. It is based on a closed-loop control system that regulates material moisture and flow rate, but its robustness is affected by the control-loop delay (the actuator delay is about 10 minutes) and by the random arrangement of the pellets inside the cooler that strongly affects product moisture (the measurement errors are not negligible). To overcome those problems, a robust statistical approach was adopted to reach the best tradeoff between estimation accuracy and computational effort. It was derived by the well known Random Close Packing model and statistical estimator. Experimental results prove the effectiveness of the proposed approach that provides moisture errors less than 7.2% with a continuous limitation of energy consumption.