Low-grade mineral deposits lead to a very high tonnage excavation with the adherent economical and environmental problems belong to gas emissions and minerals recovery costs, which, accompanied by the higher operational and equipment costs and the higher demand for the mineral resources, lead to increasing of mineral commodities prices, especially metals. These challenges can be overcome through mine planning optimization. Therefore, an approach for the global optimization of the integrated mining and processing operations is designed by a dynamic and simulation model construction. By applying a case study and through mining selectivity strategy, deeply investigation of the ore parameters (especially mineral liberation grain size and hardness), and proper arrangements for the plant facilities, mineral production is realized, with better quality, lower environmental impacts, lower costs, and higher economic benefits.

Este trabajo tiene como objetivo el desarrollo de la metodología de diseño óptimo de procesos de separación con membranas líquidas selectivas en su aplicación a la separación de mezclas multicomponentes con recuperación selectiva del componente de interés. Para ello se ha investigado la separación selectiva y concentración de cromo(VI) presente inicialmente en un acuífero subterráneo contaminado por acción de la actividad industrial desarrollada a nivel superficial y donde a su vez, coexisten otras especies aniónicas competitivas (sulfato y cloruro mayoritariamente) presentes por las propias características del acuífero y por su localización en una zona litoral. Tras la selección de Alamine 336 (amina terciaria) y NaOH como agentes de extracción y reextracción respectivamente, se caracterizaron experimentalmente las reacciones químicas responsables de la etapa de extracción, así como la cinética del proceso de separación-concentración cuando se emplea la tecnología de pertracción en emulsión (EPT) en contactores de fibra hueca. Finalmente, se desarrolló un modelo matemático multicomponente y se determinaron los parámetros característicos del mismo con el fin de llevar a cabo el diseño óptimo del proceso mediante técnicas de optimización matemática. / This work aims at the development of the methodology for the optimal design of selective liquid membrane processes as efficient alternatives for the separation and selective recovery of raw materials and valuable compounds from multicomponent systems. For this purpose, the proposed methodology has been applied to the remediation of polluted groundwater containing hexavalent chromium as a consequence of effluent leaking from surface deposition of industrial wastes. Furthermore, other competitive anionic species (mainly sulphate and chloride anions) were also present in the groundwaters due to the specific location being close to the shore.After selecting Alamine 336 (tertiary amine) and NaOH as the best extraction and back-extraction agents respectively, a careful experimental design was performed in order to analyse the chemical equilibria involved in the extraction step. Once the emulsion pertraction technology (EPT) was selected as the most suitable separation-concentration process configuration, the experimental kinetic analysis and mathematical modeling of EPT were carried calculating the values of the design parameters. Finally, the proposed multicomponent model was employed to carry out the optimal process design by means of mathematical optimization techniques.

Les capteurs d’images CMOS connaissent une croissance rapide vers des applications à fortes valeurs ajoutées. Certains marchés en devenir, comme les applications d’imagerie médicale,sont axés sur la tenue aux rayonnements ionisants. Des solutions de durcissement par dessin existent actuellement pour limiter les effets de ces dégradations. Cependant, ces dernières peuvent contraindre assez fortement certains paramètres du pixel. Dans ce contexte, cette thèse propose une solution novatrice de durcissement aux effets d’ionisation par les procédés.Elle suggère l’utilisation de pixels intégrant une photodiode pincée à collection de trous pour limiter la dégradation du courant d’obscurité : paramètre le plus sévèrement impacté lors d’irradiations ionisantes. Cette étude est donc premièrement centrée sur la modélisation et l’étude du courant d’obscurité sur des capteurs CMOS standards aussi bien avant qu’après irradiation. Ces dernières assimilées, un démonstrateur d’un capteur intégrant des pixels de1.4 μm à détection de trous est proposé et réalisé. Les résultats en courant d’obscurité, induit par la contribution des interfaces, montrent de belles perspectives avant irradiation. Ce capteur a d’ailleurs été utilisé pour effectuer une comparaison directe sous irradiation entre un capteur à détection de trous et d’électrons à design identique. Ces essais montrent une réduction significative du courant d’obscurité aux fortes doses. Des voies d’amélioration sont proposées pour améliorer l’efficacité quantique du capteur, principal point à optimiser pour des applications aussi bien grand public que médicales. / CMOS image sensors are rapidly gaining momentum in high end applications. Some emerging markets like medical imaging applications are focused on hardening against ionizing radiation. Design solutions currently exist to mitigate the effects of these degradations. However, they may introduce additional limitations on pixel performances. In this context, this thesis proposes an innovative solution of hardening by process against ionization effects. It suggests using hole pinned photodiode pixels to mitigate the dark current degradation: one of the most severely impacted parameter during ionizing radiation. This study is first focused on the modeling and understanding of dark current variation on standard CMOS sensors before and after irradiation. Next, a sensor integrating hole-based 1.4 micron pixels is proposed and demonstrated. Dark current performances induced by interfaces contribution are promising before irradiation. A direct comparison under irradiation between hole and electron based sensors with similar design has been carried out. These experiments show a significant reduction in dark current at high doses. Ways of improvement are proposed to enhance the quantum efficiency of this sensor, the main area for improvement as well consumer as medical applications.

Conselho Nacional de Desenvolvimento Científico e Tecnológico / Lignocellulosic agroindustrial wastes, such as rice husks (RH), are abundant, low cost, renewable sources, available to biotransformation in value-added products. Thus, the main objective of this study was to investigate the use of RH with the purpose of its biotransformation in lactic acid (LacA) starting from hydrolysates of this residual biomass. At the same time, aiming to contribute to the mitigation of the growing environmental concern about the illegal disposal of this waste in the State of Rio Grande do Sul the biggest Brazilian rice producer. For this, a central composite rotational design (CCRD) of the hydrolysis experiments of RH in pressurized conversion system with diluted HCl and H2SO4, revealed that experiments conducted at 175 °C, for 46 min, with acid concentrations of 2.2% v v-1, generate the highest glucose contents. For hydrolysis involving HCl, 17.8 g glucose L-1, with an yield of 106.6 mg glucose g-1 RH; for hydrolysis with H2SO4, 14.1 g glucose L-1, with an yield of 84.5 mg glucose g-1 RH. Previous fermentative experiments were carried out with the aim to select resilient and efficient micro-organisms for the LacA production and, also, to evaluate the need of supplementation of the RH hydrolyzate with specific nutrients. L. rhamnosus, in this study, was the most suitable bacteria for the lactic fermentation, among all the Lactobacillus tested, being able to to produce 5.6 g LacA L-1, with a yield of 33.8 mg LacA g-1 RH (hydrolysed with HCl, without nutrient addition, 96 h fermentation). The acid hydrolysis, in this particular case, was carried out at 160 °C, for 70 min, using 4.0% HCl v v-1, generating 14.7 g glucose L-1. The lactic fermentation via CCRD was batch conducted in 4 mL vials, at 37 °C, pH 6.0, using L. casei subsp. rhamnosus ATCC 10863, specialized in catabolization of both hexoses and pentoses (except arabinose). Experiments with RH hydrolysates containing nutrients added produced up to 18.8 g LacA L-1 (maximum yield of 997,1 mg LacA g-1 sugars) in substrate hydrolyzed with HCl, after 52 h fermentation; and 14.6 g LacA L-1 (maximum yield of 999.8 mg LacA g-1 sugars) in substrate hydrolyzed with H2SO4, after 24 h fermentation. Therefore, in this study, the choices of most influence in the LacA production were fermentation with sulfuric acid and employment of L. rhamnosus. / Resíduos lignocelulósicos agroindustriais, como a casca de arroz (CA), são fontes renováveis, abundantes e de baixo custo, disponíveis para a biotransformação em produtos com valor agregado. O principal objetivo deste trabalho foi investigar o aproveitamento da CA com vistas à sua biotransformação em ácido láctico (ALac), partindo-se de hidrolisados desta biomassa residual. Ao mesmo tempo, objetivou-se contribuir para a mitigação do crescente problema ambiental decorrente da disposição irregular deste resíduo no estado do Rio Grande do Sul, maior produtor brasileiro. Para tanto, o delineamento composto central rotacional (DCCR) do planejamento experimental da hidrólise ácida da CA em sistema de conversão à pressão, com HCl e H2SO4 diluídos, revelou que, experimentos conduzidos a 175 °C, por 46 min, com concentração de 2,2% (v v-1) de ácido, produziram os teores máximos de glicose. Para hidrólises envolvendo HCl, teor de 17,8 g de glicose L-1, com rendimento de 106,6 mg de glicose g-1 CA; para hidrólises com H2SO4, teor de 14,1 g de glicose L-1, com rendimento de 84,5 mg de glicose g-1 CA. Ensaios fermentativos prévios foram feitos com o intuito de selecionar micro-organismos resilientes e eficientes para a produção de ALac e, também, para avaliar a necessidade de suplementação do hidrolisado da CA com nutrientes específicos. L. rhamnosus, neste trabalho, foi a bactéria mais apropriada à fermentação láctica, dentre os Lactobacillus testados, sendo capaz de produzir 5,6 g ALac L-1, com rendimento de 33,8 mg ALac g-1 CA (hidrolisado com HCl, sem adição de nutrientes, 96 h de fermentação). A hidrólise ácida, neste caso específico, foi feita a 160 °C, por 70 min, com 4,0% HCl v v-1, gerando 14,7 g L-1 de glicose. A fermentação láctica via DCCR foi conduzida em vials de 4 mL, em batelada, a 37 °C, pH 6,0, empregando-se L. casei subsp. rhamnosus ATCC 10863, cepa especializada em catabolização de hexoses e pentoses (exceto arabinose). Experimentos com hidrolisados da CA adicionados de nutrientes produziram até 18,8 g ALac L-1 (rendimento máximo de 997,1 mg ALac g-1 açúcares) em substrato hidrolisado com HCl, após 52 h de fermentação; e 14,6 g ALac L-1 (rendimento máximo de 999,8 mg ALac g-1 açúcares) em substrato hidrolisado com H2SO4, após 24 h de fermentação. Portanto, neste trabalho, as escolhas de maior influência na produção de ALac foram a fermentação com hidrolisado sulfúrico e emprego de L. rhamnosus.

Esta tese descreve o desenvolvimento e implementação de um otimizador em tempo real da operação do conversor da unidade de craqueamento catalítico em leito fluidizado (FCC) da refinaria Henrique Lage localizada em São José dos Campos. É desenvolvida uma estratégia de otimização integrada ao controlador preditivo cuja função objetivo e restrições incorporam componentes dinâmicos e estáticos. Os problemas de controle multivariável e otimização econômica são resolvidos simultaneamente num mesmo algoritmo. É utilizado um modelo rigoroso do processo para determinar as condições econômicas ótimas do estado estacionário do conversor. Por intermédio do modelo dinâmico, obtido através de testes em degrau na planta, é determinada a melhor trajetória para conduzir o processo para o seu ponto de maior lucratividade, sem violar as suas restrições operacionais durante as transições. As variáveis controladas e manipuladas possuem restrições tanto no transiente como no estado estacionário. O problema de otimização não linear resultante é resolvido através de um algoritmo SQP. A nova estratégia de otimização forneceu excelentes resultados nas simulações com as condições normais de operação e as perturbações típicas da unidade. Na comparação desta estratégia com a convencional, cujas funções de controle e otimização estão separadas, a mesma mostrou um melhor desempenho por manter o processo mais estável. Na fase de implementação do otimizador, o controlador preditivo multivariável existente no conversor foi adaptado para a nova estrutura de otimização através da inclusão das funções de otimização que envolvem o modelo rigoroso do estado estacionário do processo. O controlador/otimizador foi inserido no software da Petrobras denominado SICON, o qual já era utilizado como plataforma de sintonia e manutenção do controlador multivariável linear do conversor. Esse software tem facilidades para desenvolvimento e implementação de estratégias de controle avançado. Neste trabalho são apresentados os aspectos mais importantes da integração do otimizador na estrutura de automação da refinaria e o seu comissionamento. Estão também relacionadas as principais dificuldades encontradas na implementação do otimizador e a forma que as mesmas foram solucionadas. Os resultados obtidos na planta mostram um bom desempenho do novo algoritmo tanto em relação aos benefícios econômicos como na estabilização da unidade. / This thesis describes the development and implementation of a real time optimizer of the operation of a fluid catalytic cracking (FCC) converter in the Henrique Lage refinery located in São José dos Campos. It is developed an optimization strategy integrated into the predictive controller whose objective function and constraints incorporate dynamic and static components. The multivariable control and the economic optimization problems are solved simultaneously by the same algorithm. A rigorous process model is used to determine the converter steady-state optimum economic operating conditions. By utilizing a dynamic model, obtained through step tests in the plant, calculations are carried out in order to determine the best path to drive the process to its more profitable operating point, without violating operational constraints during the transient modes. All controlled and manipulated variables are constrained, both in the steady-state and in the dynamic states. The resulting optimization problem is nonlinear and is solved by an SQP algorithm. The new optimization strategy has provided excellent results in the simulations performed at normal operating conditions and typical disturbances. Comparing to the conventional strategy, whose control and optimization functions are separated, the new one has shown a better performance by maintaining a smoother operation of the process. In the optimizer implementation phase, the existent predictive linear controller was adapted to the new optimization structure by including the optimization function that involves the rigorous steady-state model of the process. The optimizing controller was inserted in Petrobras SICON software, which was already being used as a tuning and maintenance platform of the linear multivariable controller of the converter. This software has nice features for the development and implementation of advanced process control strategies. In this work it is presented the more important aspects of the optimizer integration into the refinery automation structure and its commissioning. The main difficulties of the optimizer implementation and the way how they were solved are also related. Plant results show a good performance of the new algorithm in terms of economic benefits and unit stabilization.

Diplomová práce se zabývá návrhem procesů oddělení System Activation & Deactivation společnosti IBM Česká Republika, s.r.o. s cílem zlepšení těchto procesů a navýšení efektivity těchto procesů v rámci firmy.

L'orientation service occupe de plus en plus une place importante dans la structuration des systèmes complexes. La conception et le développement d'applications évoluent progressivement d'un modèle traditionnel vers un modèle plus dynamique orienté services où la réutilisation et l'adaptabilité jouent un rôle important. Dans cette thèse, nous proposons une étude portant sur la conception et le développement d'applications par composition de services. Nous décrivons un environnement de partage de services : DyCoSe. Il consiste en un écosystème coopératif où les entreprises membres, organisées en communautés, partagent un consensus global représentant les fonctionnalités métier récurrentes et les propriétés non fonctionnelles communes. La composition d'applications dans DyCoSe repose sur une architecture à trois niveaux combinant à la fois une démarche descendante et une autre ascendante. La démarche descendante permet de décrire l’application à travers une interaction de composants haut niveau et de la raffiner en une ou plusieurs orchestrations de services. La démarche ascendante projette les caractéristiques de l’infrastructure réseau sous-jacente au niveau services. Un processus d’instanciation visant à réaliser une application composite est détaillé. Il formalise le choix des services, selon un ensemble de contraintes données, comme un problème d’optimisation de coûts. Deux solutions au problème d’instanciation sont étudiées. Une solution globale tient compte de l’ensemble des services disponibles dans l’écosystème et une solution locale favorise les services de certaines communautés. Un algorithme génétique est décrit comme implémentation de l’instanciation globale. Une simulation stochastique de l’environnement DyCoSe est proposée. Elle permet d'étudier les possibilités d’instanciation d’une application donnée dans un environnement où la disponibilité des services n’est pas garantie. Elle permet d'étudier aussi, le taux de réussite de l’exécution d’une instance d’une application donnée. / The service oriented computing paradigm plays an increasingly important role in structuring complex systems. Application design and development approaches are witnessing a shift from traditional models towards a more dynamic service oriented model promoting reuse and adaptability. In this thesis, we study an approach for application design and development based on services’ composition. We propose DyCoSe, an environment for sharing services. It consists of an enterprise ecosystem wherein members, organized in communities, share a global agreement describing traditional business functionalities and common non-functional properties. DyCoSe relies on a three level architecture for service based application composition combining both, a top down and a bottom up composition approach. The top down part describes an application using high-level components and refines it to an orchestration of services. The bottom up part projects network characteristics to the services’ level. An instantiation process aiming at realizing an application is described. It formalizes services’ selection as a cost optimization problem considering a set of user given constraints. Two solutions of the instantiation problem are proposed. A global solution considers all the available services in an ecosystem. A local solution gives priority to services of selected communities. A genetic algorithm implements the global instantiation. A stochastic simulation of DyCoSe is proposed. It allows studying the instantiation success rate and the application execution success rate both in a dynamic environment wherein services availability is not guaranteed.

Les protéines extraites des feuilles de luzerne sont une source importante de protéines. La filtration membranaire, technologie de séparation respectueuse de l’environnement avec une productivité élevée et de faible coût a été utilisée pour séparer et concentrer les protéines des feuilles de luzerne à partir de leur jus. Cependant le phénomène du colmatage de la membrane qui réduit sérieusement le flux et la séparation des protéines est un facteur limitant important dans l'application de la filtration membranaire. Pour améliorer la récupération des protéines et amenuiser le phénomène du colmatage, la filtration membranaire associée à fort cisaillement a été utilisée pour la filtration du jus de luzerne. Toujours dans l’objectif d'optimiser le processus de la filtration, "le mode de la filtration" et "les paramètres de fonctionnement" ont été étudiés pour réduire le colmatage de la membrane et améliorer le rendement de la filtration. Puis, l’effluent du jus de luzerne a été filtré par des membranes dans des conditions de fort cisaillement afin de recycler les protéines. En outre, le mécanisme du colmatage a été étudié et a permis d’évaluer les stratégies de contrôle du colmatage. L'optimisation du procédé membranaire, via l’étude du "mode de filtration" et des "paramètres de fonctionnement" a été conduit dans le but d’améliorer la séparation et la concentration des protéines et de réduire le colmatage. Trois types de « mode de filtration » ont été testés : la filtration frontale sur le module de la cellule amicon (DA), la filtration tangentielle dynamique sur le module à disque rotatif (CRDM) et la filtration frontale sur le module à disque rotatif (DRDM)). Les « paramètres de fonctionnement » qui ont été étudiés sont les suivants : le type de membranes (ultrafiltration (UF) et microfiltration (MF)), la vitesse de rotation, la température et la pression transmembranaire (TMP). Le comportement du débit (évolution du flux du perméat au cours de la filtration), les performances de la séparation (taux de clarification et de concentration), l’efficacité du nettoyage de la membrane (récupération de la perméabilité membranaire) et la productivité lors des tests de recyclage et de concentration ont été étudiés dans le but de définir des stratégies dans le contrôle du colmatage. Puis, l’effluent de luzerne a été filtré par UF afin de séparer et purifier les protéines. Le mécanisme du phénomène du colmatage des membranes lors de la filtration du jus de luzerne a été étudié. Le processus du colmatage de la membrane a montré une tendance d’un colmatage multi-site progressif. Le modèle du colmatage multisite progressif selon la loi de Darcy (SMDM) a été proposé afin de mieux décrire et comprendre le processus du colmatage. Les effets de la composition du fluide d’alimentation, du choix de la membrane et des conditions hydraulique ont joué un rôle important dans le processus progressif du colmatage. De plus, les coefficients de résistance et de compressibilité dans les différentes étapes et sites ont été calculés afin d’expliquer le processus complexe du colmatage et d’évaluer l'efficacité des stratégies du contrôle du colmatage. Une série d'essais avec de longues durées de filtration a été réalisée pour étudier le déclin du flux et le colmatage de la membrane à diverses étapes du processus. Ces résultats présentent une utilité pour améliorer la récupération des protéines et contrôler le colmatage dans le processus de la filtration membranaire à fort cisaillement du jus de luzerne. Ces résultats sont aussi utiles pour la conception et la mise en place des technologies membranaires dans les processus industriels. / Alfalfa leaf proteins extracted from plants are an important protein source. As an environmentally friendly separation technology with high productivity and low cost, membrane filtration was used to separate and concentrate leaf protein from alfalfa juice. However membrane fouling seriously reduces flux and protein separation and is an important limitation in the application of membrane filtration. To improve protein recovery and fouling control, dynamic shear-enhanced membrane filtration with high shear rate on membrane surface and excellent anti-fouling capacity was used for alfalfa juice filtration in this work. In order to optimize filtration process, filtration mode and operation parameters were investigated to reduce membrane fouling and improve separation performance. Then, alfalfa wastewater was also treated by dynamic shear-enhanced membrane filtration to recycle proteins. Furthermore, the fouling mechanism was studied and served as a valuable evaluation for fouling strategies. In this study, process optimization including “Filtration mode” and “Operation parameters” was studied to improve protein recovery and fouling control. In “Filtration mode”, three types of filtration modules (dead end filtration using laboratory Amicon cell (DA), dynamic cross filtration using rotating disk module (CRDM) and dead end filtration using rotating disk module (DRDM)) were used to investigate the filtration performance. As for “Operation parameters”, the operation parameters including membranes (ultrafiltration (UF) and microfiltration (MF)), rotating speed, temperature and transmembrane pressure (TMP) were studied to optimize the filtration process. Flux behavior (permeate flux and flux decline), separation performance (clarification and concentration capacity), membrane cleaning efficiency (permeability recovery) and productivity in full recycling tests and concentration tests were utilized to evaluate the various operation strategies. In addition, alfalfa wastewater was treated by UF membrane, while waste proteins were recycled. Fouling mechanism for alfalfa juice filtration was investigated. The fouling process showed significantly stepwise multisite patterns. Based on Darcy’s law, the stepwise multisite Darcy’s law model (SMDM) was proposed to better describe and understand the fouling process. The effects of feed composition, membrane and hydraulic conditions played an important role in stepwise fouling process. Moreover, the resistance coefficient and compressibility for different steps and sites were calculated to explain the complex fouling process and estimate the efficiency of flux decline control strategies. Besides, a series of long tests were utilized to study flux decline and membrane fouling at various fouling step process. These results can be used to understand the protein recovery and fouling control during shear-enhanced membrane filtration process of alfalfa juice. They have important implications for process design of membrane technology in industrial scale.

The problem discussed in this thesis is the optimization of the process of planning, tracking and evaluating tasks in the management area of the department Corporate Technology in Siemens, Brno. Having studied theories of CMMI and principles of ISO 9001, the given process is optimized by automating certain manual tasks. As Corporate Technology uses a SharePoint server to manage tasks, the application is developed to improve his functionality. This application has therefore been tested in the environment of the organization with real-life data and complies to all necessary requirements to be able to save at least 50 % of time in the optimized process. The results of this thesis prove the usability of automating manual repetitive tasks being conducted in the processes of the organization and introduce possibility of increasing process level by preventing human errors.

Im Rahmen dieser Arbeit wurden verschiedene zweiphasige Vergärungssysteme hinsichtlich der mikrobiellen Biomasse und Abbauaktivität charakterisiert, mit dem Ziel, eine mikrobiologische Prozessüberwachung zu ermöglichen. Bei der analytischen Begleitung des Biogasprozesses stellte sich heraus, dass viele Biomasse- und Aktivitätsparameter anlagen- bzw. substratspezifische Werte aufweisen und Prozessstörungen bzw. Verfahrensmodifikationen über diese mikrobiologischen Kenngrößen detektiert und verifiziert werden können. Im Normalbetrieb der Vergärung von Maissilage konnte in der Hydrolysestufe eine starke Vermehrung der mikrobiellen Gesamtzellzahl sowie eine Zunahme des Archaea:Bacteria-Verhältnisses verzeichnet werden. Die Aktivitätsprofile von Hydrolasen (unspezifische Esterase, Polysaccharasen und Proteasen) erlaubten eine Visualisierung des Hydrolysefortschritts. Hierbei erwiesen sich die Esterase-, Cellulase- und Xylanaseaktivität als besonders aussagekräftig. Ähnlich ermöglichte dies die Analyse der Atmungsaktivität, die die mikrobielle Abbauaktivität des Eingangs- bzw. das Restgaspotential des Ausgangsmaterials der Hydrolysestufe wiedergibt. Die phylogenetischen Analysen der ersten Prozessstufe zeigten eine klare Dominanz von cellulolytischen Bakterien der Gattung Clostridium (Ø 35%). Der Anteil der hydrogenotrophen Methanogenen lag in den untersuchten Systemen etwa 50% über dem der Acetoklastischen, was darauf schließen lässt, dass der hydrogenotrophe Methanbildungsweg favorisiert wird. Es wurde aber auch eine erhöhte Abundanz der nicht-methanogenen Crenarchaeota festgestellt, deren Rolle im Biogasprozess noch ungeklärt ist. Im Zuge dieser Arbeit wurde weiterhin die Vergärbarkeit von HCH-belasteter Grassilage bzw. ein potentieller mikrobieller β-HCH-Abbau untersucht. Hier konnte gezeigt werden, dass die Aktivität der Mikroorganismen durch die Schadstoffbelastung des Substrates nicht inhibiert war. Darüber hinaus konnte ein Abbau des β-HCH nach der Hydrolyse der Grassilage nachgewiesen werden. Durch das Auftreten von Prozessstörungen bzw. Verfahrensmodifikationen konnten die Auswirkungen auf die untersuchten mikrobiologischen Kenngrößen näher untersucht und u. a. auch statistisch abgesichert verifiziert werden. Korrelationsanalysen verdeutlichten die mikrobiellen bzw. biochemischen Zusammenhänge im System. Besonders interessant sind dabei die signifikanten Korrelationen zwischen der Gesamtzellzahl, der Esteraseaktivität und dem Chemischen Sauerstoffbedarf, die für jede Verfahrensstufe nachgewiesen werden konnten. Diese Analysen zeigten erstmalig, dass die unspezifische Esteraseaktivität als allgemeiner mikrobieller Aktivitätsparameter in verschiedenen Prozessstufen von Biogasanlagen als Indikator zur Prozesseffizienz und -stabilität einsetzbar ist. Daher wurde die Methodik als Schnelltest weiterentwickelt, der auch vor Ort Anwendung finden kann. Dies ermöglicht eine direkte Analyse des Biogassubstrates, der Effizienz der Substratumsetzung sowie die Detektion von Störungen und eine entsprechende Steuerung und Regelung des Prozesses.