Stable operation of high-rate thermophilic sludge anaerobic reactors is sometime hard to achieve because of the nature of the anaerobic digestion (AD) process itself and the combination of biological and chemical reactions. An interesting and innovative way to handle AD data resides in multivariate statistical approaches since a more accurate analysis can be performed and fault or abnormal schemes detections can be enhanced. In this paper, principal component analysis (PCA) was the basic multivariate tool used to compare single and two stage AD process performances when treating waste activated sludge (WAS), in order to improve their monitoring and control. Two experiments were carried out to perform single and two-stage AD using WAS and fermented WAS as substrates, respectively. Findings from the PCA model agreed with results from the univariate data analysis but additionally showed a higher variability and changes on the stability trend in the AD of WAS. Besides, multivariate statistical process control using Hotelling T2 and Shewhart control charts combined with PCA displayed an out-of-control scheme revealing a transition period, in which the stability pattern of this experiment changed strongly, towards an accumulation of volatile fatty acids
Monitoring and Control Improvement of Single and Two Stage Thermophilic Sludge Digestion Through Multivariate Analysis
Bolzonella, D.
2018-01-01
Abstract
Stable operation of high-rate thermophilic sludge anaerobic reactors is sometime hard to achieve because of the nature of the anaerobic digestion (AD) process itself and the combination of biological and chemical reactions. An interesting and innovative way to handle AD data resides in multivariate statistical approaches since a more accurate analysis can be performed and fault or abnormal schemes detections can be enhanced. In this paper, principal component analysis (PCA) was the basic multivariate tool used to compare single and two stage AD process performances when treating waste activated sludge (WAS), in order to improve their monitoring and control. Two experiments were carried out to perform single and two-stage AD using WAS and fermented WAS as substrates, respectively. Findings from the PCA model agreed with results from the univariate data analysis but additionally showed a higher variability and changes on the stability trend in the AD of WAS. Besides, multivariate statistical process control using Hotelling T2 and Shewhart control charts combined with PCA displayed an out-of-control scheme revealing a transition period, in which the stability pattern of this experiment changed strongly, towards an accumulation of volatile fatty acids
I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
