Agronomic characterization of anaerobic digestates with near-infrared spectroscopy.

Anaerobic digestion is an increasingly widespread process for organic waste treatment and renewable energy production due to the methane content of the biogas. This biological process also produces a digestate (i.e., the remaining content of the waste after treatment) with a high fertilizing potential. The digestate composition is highly variable due to the various organic wastes used as feedstock, the different plant configurations, and the post-treatment processes used. In order to optimize digestate spreading on agricultural soils by optimizing the fertilizer dose and, thus, reducing environmental impacts associated to digestate application, the agronomic characterization of digestate is essential. This study investigates the use of near infrared spectroscopy for predicting the most important agronomic parameters from freeze-dried digestates. A data set of 193 digestates was created to calibrate partial least squares regression models predicting organic matter, total organic carbon, organic nitrogen, phosphorus, and potassium contents. The calibration range of the models were between 249.8 and 878.6 gOM.kgDM-1, 171.9 and 499.5 gC.kgDM-1, 5.3 and 74.1 gN.kgDM-1, 2.7 and 44.9 gP.kgDM-1 and between 0.5 and 171.8 gK.kgDM-1, respectively. The calibrated models reliably predicted organic matter, total organic carbon, and phosphorus contents for the whole diversity of digestates with root mean square errors of prediction of 70.51 gOM.kgDM-1, 34.84 gC.kgDM-1 and 4.08 gP.kgDM-1, respectively. On the other hand, the model prediction of the organic nitrogen content had a root mean square error of 7.55 gN.kgDM-1 and was considered as acceptable. Lastly, the results did not demonstrate the feasibility of predicting the potassium content in digestates with near infrared spectroscopy. These results show that near infrared spectroscopy is a very promising analytical method for the characterization of the fertilizing value of digestates, which could provide large benefits in terms of analysis time and cost.

Relating Near-Infrared Light Path-Length Modifications to the Water Content of Scattering Media in Near-Infrared Spectroscopy: Toward a New Bouguer-Beer-Lambert Law.

In near-infrared spectroscopy (NIRS), the linear relationship between absorbance and an absorbing compound concentration has been strictly defined by the Bouguer-Beer-Lambert law only for the case of transmission measurements of nonscattering media. However, various quantitative calibrations have been successfully built both on reflectance measurements and for scattering media. Although the lack of linearity for scattering media has been observed experimentally, the sound multivariate statistics and signal processing involved in chemometrics have allowed us to overcome this problem in most cases. However, in the case of samples with varying water content, important modifications of scattering levels still make calibrations difficult to build due to nonlinearities. Moreover, even when calibration procedures are successfully developed, many preprocessing methods used do not guarantee correct spectroscopic assignments (in the sense of a pure chemical absorbance). In particular, this may prevent correct modeling and interpretation of the structure of water. In this study, dynamic near-infrared spectra acquired during a drying process allow the study of the physical effects of water content variations, with a focus on the first overtone OH absorbance region. A model sample consisting of aluminum pellets mixed with water allowed us to study this specifically, without any other absorbing interaction terms related to the dry mass-absorbing constituents. A new formulation of the Bouguer-Beer-Lambert law is proposed, by expressing path length as a power function of water content. Through this new formulation, it is shown that a better and simpler prediction model of water content may be developed, with more precise and accurate identification of water absorbance bands.

Predictive models of biohydrogen and biomethane production based on the compositional and structural features of lignocellulosic materials.

In an integrated biorefinery concept, biological hydrogen and methane production from lignocellulosic substrates appears to be one of the most promising alternatives to produce energy from renewable sources. However, lignocellulosic substrates present compositional and structural features that can limit their conversion into biohydrogen and methane. In this study, biohydrogen and methane potentials of 20 lignocellulosic residues were evaluated. Compositional (lignin, cellulose, hemicelluloses, total uronic acids, proteins, and soluble sugars) as well as structural features (crystallinity) were determined for each substrate. Two predictive partial least square (PLS) models were built to determine which compositional and structural parameters affected biohydrogen or methane production from lignocellulosic substrates, among proteins, total uronic acids, soluble sugars, crystalline cellulose, amorphous holocelluloses, and lignin. Only soluble sugars had a significant positive effect on biohydrogen production. Besides, methane potentials correlated negatively to the lignin contents and, to a lower extent, crystalline cellulose showed also a negative impact, whereas soluble sugars, proteins, and amorphous hemicelluloses showed a positive impact. These findings will help to develop further pretreatment strategies for enhancing both biohydrogen and methane production.

Online estimation of VFA, alkalinity and bicarbonate concentrations by electrical conductivity measurement during anaerobic fermentation.

This paper describes the use of electrical conductivity for measurement of volatile fatty acids (VFA), alkalinity and bicarbonate concentrations, during the anaerobic fermentation process. Two anaerobic continuous processes were studied: the first was a laboratory reactor for hydrogen production from molasses and the second was a pilot process for anaerobic digestion (AD) of vinasses producing methane. In the hydrogen production process, the total VFA concentration, but not bicarbonate concentration, was well estimated from the on-line electrical conductivity measurements with a simple linear regression model. In the methane production process, the bicarbonate concentration and the VFA concentration were well estimated from the simultaneous on-line measurements of pH and electrical conductivity by means of non-linear regression with neural network models. Moreover, the total alkalinity concentration was well estimated from electrical conductivity measurements with a simple linear regression model. This demonstrates the use of electrical conductivity for monitoring the AD processes.

Fast and robust NIRS-based characterization of raw organic waste: Using non-linear methods to handle water effects.

Fast characterization of organic waste using near infrared spectroscopy (NIRS) has been successfully developed in the last decade. However, up to now, an on-site use of this technology has been hindered by necessary sample preparation steps (freeze-drying and grinding) to avoid important water effects on NIRS. Recent research studies have shown that these effects are highly non-linear and relate both to the biochemical and physical properties of samples. To account for these complex effects, the current study compares the use of many different types of non-linear methods such as partial least squares regression (PLSR) based methods (global, clustered and local versions of PLSR), machine learning methods (support vector machines, regression trees and ensemble methods) and deep learning methods (artificial and convolutional neural networks). On an independent test data set, non-linear methods showed errors 28% lower than linear methods. The standard errors of prediction obtained for the prediction of total solids content (TS%), chemical oxygen demand (COD) and biochemical methane potential (BMP) were respectively 8%, 160 mg(O2).gTS-1 and 92 mL(CH4).gTS-1. These latter errors are similar to successful NIRS applications developed on freeze-dried samples. These findings hold great promises regarding the development of at-site and online NIRS solutions in anaerobic digestion plants.

Unveiling non-linear water effects in near infrared spectroscopy: A study on organic wastes during drying using chemometrics.

In the context of organic waste management, near infrared spectroscopy (NIRS) is being used to offer a fast, non-destructive, and cost-effective characterization system. However, cumbersome freeze-drying steps of the samples are required to avoid water's interference on near infrared spectra. In order to better understand these effects, spectral variations induced by dry matter content variations were obtained for a wide variety of organic substrates. This was made possible by the development of a customized near infrared acquisition system with dynamic highly-resolved simultaneous scanning of near infrared spectra and estimation of dry matter content during a drying process at ambient temperature. Using principal components analysis, the complex water effects on near infrared spectra are detailed. Water effects are shown to be a combination of both physical and chemical effects, and depend on both the characteristics of the samples (biochemical type and physical structure) and the moisture content level. This results in a non-linear relationship between the measured signal and the analytical characteristic of interest. A typology of substrates with respect to these water effects is provided and could further be efficiently used as a basis for the development of local quantitative calibration models and correction methods accounting for these water effects.

Assessment of pesticides volatilization potential based on their molecular properties using the TyPol tool.

Following treatment, amounts of pesticides can reach the atmosphere because of spray drift, volatilization from soil or plants, and/or wind erosion. Monitoring and risk assessment of air contamination by pesticides is a recent issue and more insights on pesticide transfer to atmosphere are needed. Thus, the objective of this work was to better understand and assess pesticides emission potential to air through volatilization. The TyPol tool was used to explore the relationships between the global, soil and plant volatilization potentials of 178 pesticides, and their molecular properties. The outputs of TyPol were then compared to atmospheric pesticide concentrations monitored in various French regions. TyPol was able to discriminate pesticides that were observed in air from those that were not. Clustering considering parameters driving the emission potential from soil (sorption characteristics) or plant (lipophilic properties), in addition to vapor pressure, allowed better discrimination of the pesticides than clustering considering all parameters for the global emission potential. Pesticides with high volatilization potential have high total energy, and low molecular weight, molecular connectivity indices and polarizability. TyPol helped better understand the volatilization potential of pesticides. It can be used as a first step to assess the risk of air contamination by pesticides.

Fast at-line characterization of solid organic waste: Comparing analytical performance of different compact near infrared spectroscopic systems with different measurement configurations.

Fast characterization of solid organic waste using near infrared spectroscopy has been successfully developed in the last decade. However, its adoption in biogas plants for monitoring the feeding substrates remains limited due to the lack of applicability and high costs. Recent evolutions in the technology have given rise to both more compact and more modular low-cost near infrared systems which could allow a larger scale deployment. The current study investigates the relevance of these new systems by evaluating four different Fourier transform near-infrared spectroscopic systems with different compactness (laboratory, portable, micro spectrometer) but also different measurement configurations (polarized light, at distance, in contact). Though the conventional laboratory spectrometer showed the best performance on the various biochemical parameters tested (carbohydrates, lipids, nitrogen, chemical oxygen demand, biochemical methane potential), the compact systems provided very close results. Prediction of the biochemical methane potential was possible using a low-cost micro spectrometer with an independent validation set error of only 91 NmL(CH4).gTS-1 compared to 60 NmL(CH4).gTS-1 for a laboratory spectrometer. The differences in performance were shown to result mainly from poorer spectral sampling; and not from instrument characteristics such as spectral resolution. Regarding the measurement configurations, none of the evaluated systems allowed a significant gain in robustness. In particular, the polarized light system provided better results when using its multi-scattered signal which brings further evidence of the importance of physical light-scattering properties in the success of models built on solid organic waste.

On-site substrate characterization in the anaerobic digestion context: A dataset of near infrared spectra acquired with four different optical systems on freeze-dried and ground organic waste.

The near infrared spectra of thirty-three freeze-dried and ground organic waste samples of various biochemical composition were collected on four different optical systems, including a laboratory spectrometer, a transportable spectrometer with two measurement configurations (an immersed probe, and a polarized light system) and a micro-spectrometer. The provided data contains one file per spectroscopic system including the reflectance or absorbance spectra with the corresponding sample name and wavelengths. A reference data file containing carbohydrates, lipid and nitrogen content, biochemical methane potential (BMP) and chemical oxygen demand (COD) for each sample is also provided. This data enables the comparison of the optical systems for predictive model calibration based for example on Partial Least Squares Regression (PLS-R) [1], but could be used more broadly to test new chemometrics methods. For example, the data could be used to evaluate different transfer functions between spectroscopic systems [2]. This dataset enabled the research work reported by Mallet et al. 2021 [3].

Selecting the most relevant variables for anaerobic digestion imbalances: two case studies.

In this study, a wrapper approach was applied to objectively select the most important variables related to two different anaerobic digestion imbalances, acidogenic states and foaming. This feature selection method, implemented in artificial neural networks (ANN), was performed using input and output data from a fully instrumented pilot plant (1 m3 upflow fixed bed digester). Results for acidogenic states showed that pH, volatile fatty acids, and inflow rate were the most relevant variables. Results for foaming showed that inflow rate and total organic carbon were among the relevant variables, both of which were related to the feed loading of the digester. Because there is not a complete agreement on the causes of foaming, these results highlight the role of digester feeding patterns in the development of foaming.

Dataset of organic sample near infrared spectra acquired on different spectrometers.

This dataset presents 127 raw near infrared spectra of different organic samples acquired on three different spectrometers in three different labs. An example of data processing is shown to create six spectra transfer models between the three spectrometers (two by two). In order to build and validate these transfer models, the dataset was split into two sets of spectra: a first set was used to compute six spectra transfer models thanks to the Piecewise Direct standardisation function (PDS). A second set of spectra, independent of the first one was used to validate transfer models. Spectrum treatments and models were created on ChemFlow (https://vm-chemflow-francegrille.eu/), a free online chemometric software that includes all the necessary functions.

Modelling hydrolysis: Simultaneous versus sequential biodegradation of the hydrolysable fractions.

Hydrolysis is considered the limiting step during solid waste anaerobic digestion (including co-digestion of sludge and biosolids). Mechanisms of hydrolysis are mechanistically not well understood with detrimental impact on model predictive capability. The common approach to multiple substrates is to consider simultaneous degradation of the substrates. This may not have the capacity to separate the different kinetics. Sequential degradation of substrates is theoretically supported by microbial capacity and the composite nature of substrates (bioaccessibility concept). However, this has not been experimentally assessed. Sequential chemical fractionation has been successfully used to define inputs for an anaerobic digestion model. In this paper, sequential extractions of organic substrates were evaluated in order to compare both models. By removing each fraction (from the most accessible to the least accessible fraction) from three different substrates, anaerobic incubation tests showed that for physically structured substrates, such as activated sludge and wheat straw, sequential approach could better describe experimental results, while this was less important for homogeneous materials such as pulped fruit. Following this, anaerobic incubation tests were performed on five substrates. Cumulative methane production was modelled by the simultaneous and sequential approaches. Results showed that the sequential model could fit the experimental data for all the substrates whereas simultaneous model did not work for some substrates.

Solid-state anaerobic digestion of wheat straw: Impact of S/I ratio and pilot-scale fungal pretreatment.

Solid State Anaerobic Digestion (SSAD) of fungal pretreated wheat straw was evaluated in a leach bed reactor. During a first experiment, the effect of Substrate/Inoculum (S/I) ratios on the start-up phase was investigated. High S/I increased methane productivity but also raised the risk of reactor failure due to Volatile Fatty Acid (VFA) accumulation. With S/I ratios between 1.2 and 3.6 (Volatile Solid (VS) basis), the SSAD start-up using wheat straw was successful. Moreover, reactors were able to recover from acidification when the Total VFA/alkalinity ratio was lower than 2 gHAc_eq/gCaCO3, with VFA concentrations lower than 10 g/L and a pH close to 5.5. The conventional threshold of 0.6 gHAc_eq/gCaCO3 for stable wet AD is therefore not adapted to SSAD. During a second experiment, after the wheat straw was submitted to a fungal pretreatment in a non-sterile pilot-scale reactor, it was digested with an S/I ratio of 2.8-2.9. Under batch SSAD conditions, the biodegradability of pretreated wheat straw was slightly improved in comparison to the control (254 versus 215 NmL/g VS, respectively). Considering mass losses occurring during the pretreatment step, suboptimal pretreatment conditions caused a slightly lower methane production (161 versus 171 NmL/gTSinitial after 60-days anaerobic digestion). Nevertheless, pretreatment improved the start-up phase with lower acidification relative to controls. It would be particularly beneficial to improve the methane production in reactors with short reaction times.

A tool to guide the selection of impact categories for LCA studies by using the representativeness index.

Understanding the environmental profile of a product computed from the Life Cycle Assessment (LCA) framework is sometimes challenging due to the high number of environmental indicators involved. The objective here, in guiding interpretation of LCA results, is to highlight the importance of each impact category for each product alternative studied. For a given product, the proposed methodology identifies the impact categories that are worth focusing on, relatively to a whole set of products from the same cumulated database. The approach extends the analysis of Representativeness Indices (RI) developed by Esnouf et al. (2018). It proposes a new operational tool for calculating RIs at the level of impact categories for a Life Cycle Inventory (LCI) result. Impact categories and LCI results are defined as vectors within a standardized vector space and a procedure is proposed to treat issues coming from the correlation of impact category vectors belonging to the same Life Cycle Impact Assessment (LCIA) method. From the cumulated ecoinvent database, LCI results of the Chinese and the German electricity mixes illustrate the method. Relevant impact categories of the EU-standardized ILCD method are then identified. RI results from all products of a cumulated LCI database were therefore analysed to assess the main tendencies of the impact categories of the ILCD method. This operational approach can then significantly contribute to the interpretation of the LCA results by pointing to the specificities of the inventories analysed and for identifying the main representative impact categories.

A pseudo-stoichiometric dynamic model of anaerobic hydrogen production from molasses.

Despite many mathematical models available in the literature for simulation and optimization of anaerobic digestion processes, only few can accurately account for hydrogen production. In the present study, experiments were performed in a continuous stirred tank reactor with a hydraulic retention time close to 6 h. pH was regulated to 5.5 and agitation was maintained at 300 rpm. Molasses were used as substrate with feeding concentrations varying between 5 and 20 g L(-1). Experimental data were used to estimate the pseudo-stoichiometric coefficients with a constrained nonlinear optimization. The obtained pseudo-stoichiometric matrix is made of two reactions, one being associated with hydrogen production and the other one with acetate production. Finally, a dynamic model is derived and is demonstrated to simulate very accurately the dynamic evolution of hydrogen production, but also biomass and intermediate compounds (i.e., individual volatile fatty acids) concentrations while being very close to the stoichiometric balance. Finally, the best hydrogen production was 15.3L(H)(2)d(-1)L(-1) for a concentration of substrate of 20.09 g L(-1) and a liquid feed flow of 5 L d(-1) (i.e., 1.47 mol-H2 mol-glucose(-1)).

Representativeness of environmental impact assessment methods regarding Life Cycle Inventories.

Life Cycle Assessment (LCA) characterises all the exchanges between human driven activities and the environment, thus representing a powerful approach for tackling the environmental impact of a production system. However, LCA practitioners must still choose the appropriate Life Cycle Impact Assessment (LCIA) method to use and are expected to justify this choice: impacts should be relevant facing the concerns of the study and misrepresentations should be avoided. This work aids practitioners in evaluating the adequacy between the assessed environmental issues and studied production system. Based on a geometrical standpoint of LCA framework, Life Cycle Inventories (LCIs) and LCIA methods were localized in the vector space spanned by elementary flows. A proximity measurement, the Representativeness Index (RI), is proposed to explore the relationship between those datasets (LCIs and LCIA methods) through an angular distance. RIs highlight LCIA methods that measure issues for which the LCI can be particularly harmful. A high RI indicates a close proximity between a LCI and a LCIA method, and highlights a better representation of the elementary flows by the LCIA method. To illustrate the benefits of the proposed approach, representativeness of LCIA methods regarding four electricity mix production LCIs from the ecoinvent database are presented. RIs for 18 LCIA methods (accounting for a total of 232 impact categories) were calculated on these LCIs and the relevance of the methods are discussed. RIs prove to be a criterion for distinguishing the different LCIA methods and could thus be employed by practitioners for deeper interpretations of LCIA results.

Multiplexed chemostat system for quantification of biodiversity and ecosystem functioning in anaerobic digestion.

Continuous cultures in chemostats have proven their value in microbiology, microbial ecology, systems biology and bioprocess engineering, among others. In these systems, microbial growth and ecosystem performance can be quantified under stable and defined environmental conditions. This is essential when linking microbial diversity to ecosystem function. Here, a new system to test this link in anaerobic, methanogenic microbial communities is introduced. Rigorously replicated experiments or a suitable experimental design typically require operating several chemostats in parallel. However, this is labor intensive, especially when measuring biogas production. Commercial solutions for multiplying reactors performing continuous anaerobic digestion exist but are expensive and use comparably large reactor volumes, requiring the preparation of substantial amounts of media. Here, a flexible system of Lab-scale Automated and Multiplexed Anaerobic Chemostat system (LAMACs) with a working volume of 200 mL is introduced. Sterile feeding, biomass wasting and pressure monitoring are automated. One module containing six reactors fits the typical dimensions of a lab bench. Thanks to automation, time required for reactor operation and maintenance are reduced compared to traditional lab-scale systems. Several modules can be used together, and so far the parallel operation of 30 reactors was demonstrated. The chemostats are autoclavable. Parameters like reactor volume, flow rates and operating temperature can be freely set. The robustness of the system was tested in a two-month long experiment in which three inocula in four replicates, i.e., twelve continuous digesters were monitored. Statistically significant differences in the biogas production between inocula were observed. In anaerobic digestion, biogas production and consequently pressure development in a closed environment is a proxy for ecosystem performance. The precision of the pressure measurement is thus crucial. The measured maximum and minimum rates of gas production could be determined at the same precision. The LAMACs is a tool that enables us to put in practice the often-demanded need for replication and rigorous testing in microbial ecology as well as bioprocess engineering.

Clustering pesticides according to their molecular properties, fate, and effects by considering additional ecotoxicological parameters in the TyPol method.

Understanding the fate and ecotoxicological effects of pesticides largely depends on their molecular properties. We recently developed "TyPol" (Typology of Pollutants), a classification method of organic compounds based on statistical analyses. It combines several environmental (sorption coefficient, degradation half-life) and one ecotoxicological (bioconcentration factor) parameters, to structural molecular descriptors (number of atoms in the molecule, molecular surface, dipole moment, energy of orbitals, etc.). The present study attempts to extend TyPol to the ecotoxicological effects of pesticides on non-target organisms, based on data analysis from available literature and databases. It revealed that relevant ecotoxicological endpoints for terrestrial organisms (e.g., soil microorganisms, invertebrates) that support a range of ecosystemic services are lacking as compared to aquatic organisms. The availability of ecotoxicological parameters was also lower for chronic than for acute ecotoxicity endpoints. Consequently, seven parameters were included for acute (EC50, LC50) and chronic (NOEC) ecotoxicological effects for one terrestrial (Eisenia sp.) and three aquatic (Daphnia sp., algae, Lemna sp.) organisms. In this new configuration, we used TyPol to classify 50 pesticides into different clusters that gather molecules with similar environmental behaviors and ecotoxicological effects. The classification results evidenced relationships between molecular descriptors, environmental parameters, and the added ecotoxicological endpoints. This proof-of-concept study also showed that TyPol in silico classification can successfully address new scientific questions and be expanded with other parameters of interest.

Online estimation of assimilable nitrogen by electrical conductivity measurement during alcoholic fermentation in enological conditions.

The monitoring of alcoholic fermentation under enological conditions is currently poor due to the lack of sensors for online measurements. Such monitoring is currently limited to the measurement of CO(2) production or changes in density. In this study, we determined the potential value of measuring electrical conductivity. We showed that this measurement is related to the assimilation of nitrogen, which is typically the limiting nutrient, and directly correlated to ammoniacal nitrogen assimilation at any percentage of ammoniacal nitrogen in the medium. We also used electrical conductivity for the very precise monitoring of the kinetics of nitrogen assimilation after the addition of a pulse of diammonium hydrogen phosphate (DAP) during fermentation. The impact of initial conditions (e.g., must composition, grape variety, pH) remains unclear, but the robustness, precision and low price of the sensor used justify further studies of the potential value of measuring electrical conductivity on the pilot and industrial scales.

Fast ADM1 implementation for the optimization of feeding strategy using near infrared spectroscopy.

Optimization of feeding strategy is an essential issue of anaerobic co-digestion that can be greatly assisted with simulation tools such as the Anaerobic Digestion Model 1. Using this model, a set of parameters, such as the biochemical composition of the waste to be digested, its methane production yield and kinetics, has to be defined for each new substrate. In the recent years, near infrared analyses have been reported as a fast and accurate solution for the estimation of methane production yield and biochemical composition. However, the estimation of methane production kinetics requires time-consuming analysis. Here, a partial least square regression model was developed for a fast and efficient estimation of methane production kinetics using near infrared spectroscopy on 275 bio-waste samples. The development of this characterization reduces the time of analysis from 30 days to a matter of minutes. Then, biochemical composition and methane production yield and kinetics predicted by near infrared spectroscopy were implemented in a modified Anaerobic Digestion Model n°1 in order to simulate the performance of anaerobic digestion processes. This approach was validated using different data sets and was demonstrated to provide a powerful predictive tool for advanced control of anaerobic digestion plants and feeding strategy optimization.