1. Characterization of an innovative waste-derived fertilizer by hyperspectral imaging Serranti S., Trella A., Bonifazi G., Santarelli M., Tirillò J. and Valente M. Department of Chemical Engineering, Materials & Environment, Sapienza - University of Rome, Via Eudossiana, 18, 00184 Rome, Italy. *silvia.serranti@uniroma1.it The MAIN OBJECTIVE of the ongoing project LIFE12 ENV/IT/356 RESAFE: “Innovative fertilizer from urban waste, bio-char and farm residues as substitute of chemical fertilizers” is to demonstrate the production and use of “new” fertilizers characterized by reduced salinity in order to substitute chemicals and mineral fertilizers through a technological route based on urban organic waste, bio-char and farm organic residues “re-utilization” and processing. Following this approach it will be thus possible, for farmers and urban waste managers,, to reduce costs and to obtain environmental and economical incomes that is: − an important material recovery (e.g. reduced quantities of waste sent to landfill disposal and related costs saving); − a further economic benefit due to the obtained fertilizers. SPAIN Agricultural analysis and in field testing CYPRUS Agricultural analysis and in field testing DIPARTIMENTO DI INGEGNERIA CHIMICA MATERIALI AMBIENTE PROJECT BENEFICIARIES CONSEJO SUPERIOR DE INVESTIGACIONES CIENTIFICAS ITALY Waste Expert ITALY Expert on agriculture, agricultural waste handling and field testing ITALY Coordinator of the project. Expert in agricultural waste handling, processing and environmental testing ACTIONS AND TOOLS Actions have been carried out to demonstrate the economic,, social and environmental benefits related to the production and use of reduced salinity fertilizers. Demonstrative plants (i.e. lab and pilot scale plants) have been realized to show as from “different recipes”, based on treated organic waste (i.e. compost and farm organic residues) and bio-char, added with Vegetable Active Principles (VAP), a new class of fertilizers could be obtained. Specific sensing device, based on HyperSpectral Imaging (HSI), are used to follow the transformations of the complex material, during the different stages of the process. Correlations between spectral behaviour and both curing time and measured chemical parameters of the different mixed wastes were investigated. SELECTED MATERIALS FOR PILOT PLANT (Action B.3) A preliminary selection of the different raw materials (Urban Organic Wastes: UOW, Farm Organic Residues: FOR, Bio-char: BC and VAP), constituting the recipe for the fertilizer production, was carried out in Italy and Spain. Mixtures of the raw materials were created. Samples were collected to follow the curing process, at times: t0 and t90 (t in days). The main physical-chemical parameters were measured for the collected samples, that were also acquired by HSI. a b c d Image sequence showing samples, preparation and analysis: a) box containing the raw material mixture; b) sample collection; c) sample preparation; d) acquisition by hyperspectral imaging in the NIR (1000-1650 nm) wavelength range, by NIR Spectral Camera™ (Specim, Finland). HSI SOFTWAREARCHITECTURE SET UP (Action B.3) AND SAMPLES MONITORING (Action C.2) Environmental impact will be also significantly reduced contributing to decrease greenhouse emissions from landfills. All the information about the project objectives, results, dissemination activities are available on-line at http://www.liferesafe.com/ HSI RESULTS Partial Least Squares (PLS) HSI RESULTS Principal Component Analysis (PCA) Demonstration of proper recipes through the characterization of organic waste and final products ACTION B.1 Demonstration of HQ- ORBP (High Quality Organic Matter Based Product) production at laboratory level ACTION B.2 Demonstration of HQ-ORBP production at pilot and semi-industrial scale ACTION B.3 Demonstration of Agriculture Application in Italy ACTION B.4 Demonstration of Agriculture Application in Cyprus ACTION B.6 Demonstration of Agriculture Application in Spain ACTION B.5 CSI 2015 F IGUEIRA DA F OZ, C OIMBRA (P ORTUGAL ), 30 A UGUST - 3 S EPTEMBER 2015 Spectral data have been analysed using the PLS_Toolbox 7.9.2 (Eigenvector Research Inc.) running inside Matlab™ (version 8.4.0) environment. Spectra preprocessing PCA applied for data exploration PLS applied for spectra correlation with different measured parameters The measured parameters utilized to find correlation with the HSI data were: time, soluble nitrogen, soluble carbon and humidity. SpecIm NIR Spectral Camera™ Energizing (lighting) source Optic Waste samples Hyperspectral data display Control Panel Conveyor belt Results clearly outlined as, starting from the collection of NIR- HSI reflectance spectra, good PLS models have been set up for the prediction of soluble nitrogen, soluble carbon and humidity, as well as samples “ageing status”. PCA score plots (PC1 vs. PC2) related to Spanish samples, collected at different curing time (0 and 90 days) PLS results for the different measured parameters (Italian and Spanish samples). Average reflectance spectrum Hypercubic image with selected class Hypercubic image with selected class Average reflectance spectrum PCA score plots (PC1 vs. PC2) related to Italian samples, collected at different curing time (0 and 90 days) Spanish samples Italian samples Time Humidity Soluble Carbon Soluble Nitrogen T0 T90 T0 T90