| Title | Next-generation of instrumental odour monitoring system (IOMS) for the gaseous emissions control in complex industrial plants |
|---|---|
| ID_Doc | 44418 |
| Authors | Oliva, G; Zarra, T; Pittoni, G; Senatore, V; Galang, MG; Castellani, M; Belgiorno, V; Naddeo, V |
| Title | Next-generation of instrumental odour monitoring system (IOMS) for the gaseous emissions control in complex industrial plants |
| Year | 2021 |
| Published | |
| Abstract | Odour emissions from complex industrial plants may cause potential impacts on the surrounding areas. Consequently, the validation of effective tools for the control of the associated environmental pressures, without hindering economic growth, is strongly needed. Nowadays, senso-instrumental methods by using Instrumental Odour Emissions Systems (IOMSs) is among the most attractive tool for the continuous monitoring of environmental odours, allowing the possibility of obtaining real-time information to support the decision-making process and proactive approach. The systems complexity and scarcity of real data limited their wider full-scale employment. The study presents an advanced prototype of IOMS for the continuous classification and quantification of the odours emitted in ambient air by complex industrial plants, to continuously control the plants emissions with backwards approach. The IOMS device was designed and optimized and included the system for the automatic control of the conditions inside the measurement chamber. The designed operational procedures were presented and discussed. Results highlighted the influence of temperature and air flow rate for the measurement repeatability. Accurate prediction model was created and optimized and resulted able to distinguish 3 different industrial odour sources with accuracy approximately equal to 96%. The models were optimized thanks to the software features, which allowed to automatically apply the designed statistical procedures on the identified dataset with different pre-processing approach. The usefulness of having a fully-developed and user-friendly flexible system that allowed to select and automatically compare different settings options, including the different feature extraction methods, was demonstrated in order to identify the best prediction model. (C) 2021 Elsevier Ltd. All rights reserved. |
No similar articles found.