Dissertations / Theses on the topic 'Airlift Reactor'

Bubble columns and airlift reactors were modeled numerically to better understand the hydrodynamics and analyze the mixing characteristics for each configuration. An Eulerian-Eulerian approach was used to model air as the dispersed phase within a continuous phase of water using the commercial software FLUENT. The Schiller-Naumann drag model was employed along with virtual mass and the standard k-e turbulence model. The equations were discretized using the QUICK scheme and solved with the SIMPLE coupling algorithm. The flow regimes of a bubble column were investigated by varying the column diameter and the inlet gas velocity using two-dimensional simulations. The typical characteristics of a homogeneous, slug, and heterogeneous flow were shown by examining gas holdup. The flow field predicted using two-dimensional simulations of the airlift reactor showed a regular oscillation of the gas flow due to recirculation from the downcomer and connectors, whereas the bubble column oscillations were random and resulted in gas flow through the center of the column. The profiles of gas holdup, gas velocity, and liquid velocity showed that the airlift reactor flow was asymmetric and the bubble column flow was symmetric about the vertical axis of the column. The average gas holdup in a 10.2 cm diameter bubble column was calculated and the results for the two-dimensional simulation of varying inlet gas velocities were similar to published experimental results. The average gas holdup in the airlift reactor for the three-dimensional simulations compared well with the experiments, and the two-dimensional simulations underpredicted the average gas holdup.<br>Master of Science

External loop airlift reactors are widely used for biochemical applications such as syngas fermentation and wastewater treatment. To further understand the inherent gas-liquid flow physics within the reactors, computational modeling and simulations of hydrodynamics for air-water external loop airlift reactors were investigated. The gas-liquid flow dynamics in a bubble column were simulated using a FORTRAN code developed by Los Alamos National Laboratory, CFDLib, which employs an Eulerian-Eulerian ensemble averaged method. A two-dimensional Cartesian coordinate system was used to conduct an extensive grid resolution study; it was found that grid cells smaller than the bubble diameter produced unstable solutions. Next, closure models for drag force and turbulent viscosity were investigated for a simple bubble column geometry. The effects of using a bubble pressure model and two drag coefficient models, the White model and the Schiller-Naumann model, were investigated. The bubble pressure model performed best for homogeneous (low velocity) flows and the Schiller-Naumann model was best for all flow regimes. Based on the studies for bubble column flows, an external loop airlift reactor was simulated using both two- and three-dimensional coordinates and results for gas holdup and riser velocity agreed better with experimental data for the 3D simulations. It was concluded that when performing 2D and 3D simulations, care must be taken when specifying the effective bubble diameter size, especially at high flow rates. Population balance models (PBM) for bubble break-up and coalescence were implemented into CFDLib, validated with experiments, and simulated for the external loop airlift reactor at high inlet superficial gas velocities. The PBM predictions for multiple bubble sizes were comparable with the single bubble size simulations; however, the PBM simulations better predicted the formation of the gas bubble in the downcomer. The 3D PBM simulations also gave better predictions for the average bubble diameter size in the riser. It was concluded that a two-dimensional domain is adequate for gas-liquid flow simulations of a simple bubble column geometry, whereas three-dimensional simulations are required for the complex airlift reactor geometry. To conclude, a two-fluid Eulerian-Eulerian model coupled with a PBM is needed for quantitative as well as physical predictions of gas-liquid external loop airlift reactor flows at high inlet superficial gas velocities.<br>Ph. D.

Se evaluó cualitativa y cuantitativamente un sistema multifásico compuesto por un flujo de líquido, aire y mineral dentro de un reactor airlift de recirculación interna con un volumen de trabajo de 18,0 L, empleado en un proceso de lixiviación ácida de minerales oxidados de cobre cuyo tamaño de partícula estuvo en el rango de 55 a 75 µm y con una densidad promedio de 2,94 g/cm3. Se evaluó el efecto de la velocidad superficial del gas de entrada (USG) de 10,55 a 63,59 m/s, el nivel de líquido por encima del tubo interno o draftube (hT) de 5,0 a 20,0 cm y la carga de sólidos en el reactor (sL) de 5,0 a 15,0 % en peso de mineral con respecto al volumen del reactor sobre el comportamiento de los flujos multifásicos y parámetros hidrodinámicos tales como el holdup de gas o fracción volumétrica del gas en la mezcla (εG), las velocidades de líquido en las regiones de flujo ascendente o riser (ULR) y flujo descendente o downcomer en el reactor (ULD), porcentaje de mineral sedimentado (sS) y altura de la columna de espuma formada (hF). La caracterización cualitativa se realizó mediante la visualización de los flujos desarrollados en cada región del reactor airlift. Para todos los valores de velocidad superficial de gas de entrada (USG) testeados, se trabajó con un patrón de flujo de tipo burbuja tanto en la región de flujo ascendente como descendente; caracterizado por burbujas de tamaño pequeño y forma esférica definida. En la zona de desgasificación, el flujo de burbujas es caótico con presencia de burbujas de tamaño y forma heterogénea. Por último, en el fondo del reactor, la presencia de burbujas es prácticamente nula pues transitaban rápidamente desde la región de flujo descendente a la de flujo ascendente. Para los trabajos desarrollados con un flujo bifásico de aire-agua; se encontró que la velocidad superficial del gas de entrada (USG) tiene un mayor impacto sobre los parámetros hidrodinámicos comparado al que ejerce el nivel de líquido por encima del tubo interno (hT). Para el caso del holdup de gas (εG), el valor encontrado en la región de flujo descendente fue aproximadamente un 83,0 % del valor encontrado en la zona de flujo ascendente para todos los valores de USG empleados. Para el caso de la velocidad de líquido, el valor en la región de flujo descendente (ULD) máximo fue de 0,310 m/s y la velocidad de líquido en la zona ascendente (ULR) máxima de 0,305 m/s. En el comportamiento del flujo trifásico (mineral-aire-agua) se observó que la carga de sólidos (sL) impactó directamente en la formación de espuma en el reactor (hF) y el porcentaje de mineral sedimentado (sS), el cual fue como máximo de 60,73 % para una carga de sólidos de 5,0 % y una velocidad superficial del gas de entrada de 10,55 m/s y que se fue reduciendo hasta un 20,24 % para una carga de sólidos de 15,0 % y una velocidad superficial del gas de entrada de 63,59 m/s. Una vez definidas las mejores condiciones de operación del reactor basado en los parámetros hidrodinámicos antes descritos, se realizó la lixiviación de minerales oxidados de cobre usando un nivel de líquido por encima del tubo interno de 15,0 cm, una velocidad superficial de gas de entrada de aire de 10,55 m/s, un porcentaje en peso de 15,0 % de sólidos y con una solución de ácido sulfúrico (0,05 M) como agente lixiviante. Se inició con la solución a pH 1,0 y se registró un incremento hasta un valor de 5,69 a los primeros 5 minutos de iniciado el proceso. Al finalizar el proceso de 4 horas de duración, se extrajo como máximo 0,12 gramos de cobre por kilogramo de mineral, cantidad que fue disminuyendo conforme aumentaba el valor de pH debido a la precipitación del metal a altos valores de pH. Se realizaron, además, dos experimentos de lixiviación a nivel laboratorio en vasos agitados mecánicamente y con iguales condiciones de operación que los empleados en el proceso de lixiviación en el reactor, uno de ellos fue realizado manteniendo el pH en 1,0 durante todo el proceso y el otro a pH variable. Se obtuvo una extracción de 0,21 gramos de cobre por kilogramo de mineral en el vaso con pH controlado y 0,15 gramos de cobre por kilogramo de mineral en el vaso con pH variable, donde se notó el mismo efecto de aumento de pH y disminución en la cantidad de cobre extraído como en el reactor. Se observó que la cantidad de cobre extraído es dependiente de la acidez del sistema, por lo que el control de esta condición abre oportunidades de optimización en el proceso de lixiviación llevado a cabo en estos tipos de reactor.<br>Perú. Concejo Nacional de Ciencia y Tecnología. 002-2016

La lixiviation bactérienne est l'application biotechnologique la plus importante dans l'industrie métallurgique, celle-ci est utilisée pour la récupération des métaux par lessivage, utilisant les capacités que possèdent certains micro-organismes à solubiliser les métaux. Ce travail est divisé en quatre sections: premièrement, nous avons identifié la composition microbiologique de la culture mixte en notre possession. Cette culture mixte est composée de trois populations bactériennes. Deuxièmement, nous avons étudié l'influence de certains paramètres physico-chimiques tels que le pH, la température, la concentration en CO2, le taux de solides et l'ajout de Fe3+ au milieu réactionnel sur la cinétique de la biolixiviation, en culture discontinue. Ces études ont notamment conduit à la détermination d'un pH (1. 75), d'une température (32-37°C) et d'une concentration en CO2 (0,03 et 2% v/v) optimaux pour ce procédé. Troisièmement, nous avons comparé différents systèmes réactionnels (air-lift, lit fluidisé et mécaniquement agité) pour aider au choix d'une future technologie alternative. Notre dernière contribution a porté sur l'étude de la capacité de la culture mixte à oxyder des déchets industriels contenant de l'antimoine, en culture discontinue. Nous avons étudié le rôle important de l'interaction galvanique entre la pyrite et les résidus antimoniés sur la récupération de l'antimoine

Esta tesis doctoral versa sobre la eliminación simultánea de compuestos fenólicos con reactores de biomasa granular trabajando en continuo. El estudio está dividido en dos partes principales; el primer tema trata sobre la modelización de la nitritación, mientras que el otro está dedicado al trabajo experimental sobre la nitritación y eliminación simultánea de compuestos fenólicos. En el estudio de modelización, se desarrolló un modelo matemático de biopelícula para describir la nitritación en reactores de biomasa granular aerobia operando en continuo. El modelo incorpora una estrategia de control del ratio [DO]/[TAN], para mantener un valor deseado de la relación entre las concentraciones de oxígeno disuelto (DO) en el efluente del reactor y nitrógeno amoniacal total [TAN]. El modelo se validó con un gran número de datos experimentales previamente publicados en la bibliografía, así como con datos obtenidos de reactores granulares tratando agua de rechazo a escala laboratorio y piloto. El modelo se utilizó para estudiar el efecto de: a) las consignas de DO y TAN, b) la temperatura de operación, c) las características de la biopelícula (tamaño de partícula, densidad) y d) la concentración de amonio en el afluente, sobre la consecución de la nitritación completa. Los resultados indicaron que la nitritación completa se mantuvo estable y se potenció usando la estrategia de control de la proporción [DO]/[TAN] en la operación del reactor de biomasa granular aerobia. Además, el modelo predijo que gránulos aerobios mayores a 1.5 mm y concentraciones altas de amonio en el afluente potenciaba la obtención de nitritación completa estable, mientras que la densidad de biopelículas tenía poca influencia en este estudio. Además se demostró que era posible la nitritación total a bajas temperaturas con reactores de biomasa granular. Para el trabajo experimental, se utilizó un reactor tipo airlift. Para la puesta en marcha del reactor, se utilizó como inóculo biomasa de un reactor de biomasa granular aerobia que realizaba eliminación de nutrientes. Como alimento del reactor se utilizó un agua residual sintética con un alto contenido de amonio (950 ± 25 mg N L-1). El reactor se operó hasta la obtención de nitritación parcial. Una vez obtenida la nitritación parcial, el reactor se bioaumentó con un lodo activo que contenía biomasa degradadora de p-nitrofenol (PNP) para mejorar el crecimiento de microorganismos degradadores de fenol sobre los gránulos nitrificantes. Acto seguido, mientras el reactor trataba una carga elevada de amonio, se añadieron progresivamente al afluente o-cresol (hasta 100 mg L-1) o PNP (hasta 15 mg L-1), siendo éstos alimentados al reactor con el objetivo de estudiar la nitritación parcial simultánea a la eliminación de o-cresol o de PNP. En el estudio de la nitritación parcial simultánea a la eliminación de o-cresol, se mantuvo el proceso de nitritación parcial estable durante más de 100 días de operación. Además, se obtuvo una biodegradación completa de o-cresol durante todo el periodo experimental. También se realizaron choques de carga de o-cresol, durante los cuales el proceso de nitritación parcial se mantuvo estable y sin verse afectado por esos eventos. Las cargas volumétricas obtenidas de nitrógeno (NLRV) y de o-cresol (CLRV) fueron de 1.1 g N L-1 d-1 y 0.11 g o-cresol L-1 d-1, respectivamente. El análisis de hibridación in situ de fluorescencia (FISH) indicó que en los gránulos había presencia del género Acinetobacter, de bacterias amonio-oxidantes betaproteobacteriales y de Nitrobacter sp. Posteriormente, se continuó con la operación del reactor, y se llevó a cabo un experimento relacionado con el funcionamiento del reactor bajo tres escenarios de alternancia secuencial de contaminantes (SAP). En cada uno de los escenarios SAP se añadieron 15 mg L-1 de compuestos fenólicos secundarios (i.e. PNP, fenol y 2-clorofenol (2CP)) al afluente por un periodo de tiempo corto (entre 20 y 25 años). Los resultados ilustraron que se mantuvo la nitritación parcial y la biodegradación de o-cresol sin mostrar ningún signo de inhibición por la presencia de PNP o de fenol. Sin embargo, en presencia de 2CP en el afluente, se registró durante tres días un 90% de la nitritación parcial y un 25% de la degradación de o-cresol. Estos resultados sugieren que las bacterias amonio oxidantes (AOB) son mas sensibles a la inhibición por 2CP que las heterótrofas (degradadoras de o-cresol). En el estudio de la nitritación simultánea a la eliminación de PNP, se mantuvo la nitritación durante la mayor parte del periodo operacional, obteniéndose un efluente adecuado para la desnitrificación heterotrófica. Sin embargo, durante los primeros 175 días, la biodegradación de PNP fue inestable, observándose diversos episodios de acumulación de PNP. Esta acumulación se determinó que era debida a las condiciones limitantes de DO. El incremento de la concentración de DO en el reactor de 2 a 4 mg O2 L-1 permitió obtener eliminación completa y estable de PNP hasta el fin del periodo experimental. Las NLRV y la carga de PNP obtenidas fueron de 1.0 g N L-1 d-1 y 16 mg PNP L-1 d-1, respectivamente. Además, se evaluó el funcionamiento del reactor realizando dos estudios de hambruna, i) hambruna de PNP y ii) hambruna total (parada del reactor). Los resultados mostraron que 2 días después al fin del periodo de hambruna se obtuvo una recuperación total de la degradación de PNP, mientas que la recuperación total de la nitritación simultánea a la eliminación de PNP se consiguió solo 11 después de volver a poner en marcha el reactor. En conclusión, el uso de reactores de biomasa granular aerobia para realizar nitritación simultánea a la eliminación de compuestos fenólicos es factible. Ésta podría ser considerada la mejor técnica disponible para el tratamiento aguas residuales industriales complejas con contenido de amonio en alta carga y compuestos fenólicos. Se ha probado que la biomasa granular aerobia es resistente a sobrecargas puntuales, a presencia alterna de compuestos recalcitrantes y a periodos de hambruna; estas condiciones, debido a los cambios de planificación de la producción, pueden encontrarse frecuentemente en plantas de tratamiento de aguas residuales industriales. En un futuro próximo, proponemos que la nitritación simultánea a la eliminación de compuestos fenólicos podría combinarse tanto con la desnitrificación heterotrófica o con el proceso anammox para una eliminación sostenible del nitrógeno.<br>The simultaneous nitritation and phenolic compounds removal using aerobic granular reactors in continuous mode were studied in this Ph.D. thesis. The study is divided into two main subjects; the first one is devoted to the modeling of nitritation while the other part is dedicated to the experimental work of simultaneous nitritation and phenolic compounds removal using granular reactors. In the modeling study, a mathematical biofilm model was developed to describe nitritation in aerobic granular reactors operating in continuous mode. The model incorporated a [DO]/[TAN] ratio control strategy to maintain the proportion between the concentrations of dissolved oxygen (DO) and total ammonia nitrogen (TAN) in the reactor effluent to a desired value. The model was validated with a large set of experimental results previously reported in the literature, as well as, data gathered from laboratory scale and pilot plant granular reactors treating reject water. The model was used to study the effect of: a) DO and TAN setpoints, b) operating temperature, c) biofilm characteristics (granules size, density) and d) ammonium concentrations in the influent on the achievement of full nitritation. The results indicated that full nitritation was stably maintained and enhanced by applying the [DO]/[TAN] ratio control strategy in the operation of aerobic granular sludge reactor. Moreover, the model predicted that aerobic granules size larger than 1.5 mm and high ammonium concentrations in the influent enhanced the achievement of stable full nitritation, while poor influence of the biofilm density was found with the simulation study. Furthermore, at low temperature, full nitritation with granular reactors was demonstrated to be possible. In the experimental work, an airlift reactor was employed. In the reactor start-up, granular sludge from a reactor performing biological nutrient removal was used as inoculum. A synthetic wastewater containing high-strength ammonium concentrations (950 ± 25 mg N L- 1) was fed into the airlift reactor. The reactor was operated until partial nitritation was obtained. Once partial nitritation was achieved, the airlift reactor was bioaugmented with pnitrophenol (PNP)-degrading activated sludge to enhance the growth of phenolic-degraders over the nitrifying granules. Immediately, o-cresol (up to 100 mg L-1) or PNP (up to 15 mg L- 1) were progressively added to the high-strength ammonium influent and fed into the reactor with the objective of studying the simultaneous partial nitritation and o-cresol removal and the simultaneous nitritation and (PNP) removal. First, in the study of simultaneous partial nitritation and o-cresol removal, a stably partial nitritation process was maintained for more than 100 days of operation. Moreover, full biodegradation of o-cresol was achieved during the whole experimental period. Also, o-cresol shock load events were applied and the partial nitritation process was kept stable and unaffected during these events. The achievable nitrogen loading rate (NLRv) and o-cresol loading rate (CLRv) were ca. 1.1 g N L-1d-1 and 0.11 g o-cresol L-1d-1, respectively. Analysis of fluorescent in-situ hybridization (FISH) indicated that Acinetobacter genus, betaproteobacterial ammonia-oxidizing bacteria and Nitrobacter sp. were present into the granules. Later, the operation of the reactor was continued, and an experiment devoted to the performance of the reactor under three sequentially alternating pollutant (SAP) scenarios was executed. In each one of the SAP scenarios, 15 mg L-1 of the secondary phenolic compounds (i.e. p-nitrophenol (PNP), phenol and 2-chlorophenol (2CP)) were added in the regular influent for a short period of time (between 20 to 25 days). The results illustrated that partial nitritation and o-cresol biodegradation were maintained without exhibiting any sign of inhibition by the presence of PNP or phenol. However, when 2CP was present in the influent, 90 % of the partial nitritation and 25 % of the o-cresol degradation was inhibited within three days. This finding suggests that the ammonia oxidizing bacteria (AOB) is more sensitive to 2CP inhibition than heterotrophs (o-cresol-degraders). Second, in the study of simultaneous nitritation and PNP removal, nitritation was maintained during most of the operation period producing an effluent suitable for heterotrophic denitrification. However, in the first 175 days, PNP biodegradation was unstable and several accumulation episodes occurred. The oxygen limiting condition was found to be the main explanation of these events. The increase of dissolved oxygen concentration (DO) in the reactor from 2 to 4 mg O2 L-1 permitted to achieve complete and stable PNP removal till the end of the experimental period. The achieved NLRv and PNP loading rate (PNP-LRv) were ca. 1.0 g N L-1d-1 and 16 mg PNP L-1d-1, respectively. Besides, the performance of the reactor was further assessed by performing two starvation studies, i) PNP starvation and ii) total starvation period (reactor shutdown). Results show that full recovery of PNP degradation was achieved within 2 days after the PNP starvation period ended, while full recovery of simultaneous nitritation and PNP removal was accomplished in just 11 days after the restart of the reactor. In conclusion, the use of continuous aerobic granular reactors for the simultaneous nitritation and phenolic compounds removal is feasible. This could be regarded as a best available technique for the treatment of complex industrial wastewaters containing high-strength ammonium concentrations and phenolic compounds. Aerobic granules are proven to be resistant and resilient to the shock loads, to the alternating presence of recalcitrant compounds and to starvation periods; conditions frequently found in industrial wastewater treatment plants due to changes on the industrial production schedules. In the near future, we propose the simultaneous nitritation and phenolic compounds removal should be combined with either heterotrophic denitrification or Anammox for sustainable nitrogen removal.

L’objectif de cette thèse est d’analyser les contributions de Clément Colson à la science économique dans le but de mieux connaître sa pensée, de mieux connaître l’École libérale française au début du XXème siècle, d’étudier l’influence qu’a pu avoir cet auteur sur ses principaux élèves, Divisia, Roy et Rueff, et d’évaluer l’actualité de certaines de ses recommandations.Le premier chapitre montre quels sont les apports théoriques de Colson à la science économique.Pour ses élèves, son principal apport serait la théorie de la détermination conjointe du salaire et du taux d’intérêt. Nous montrerons que cette théorie est proche de la règle de gestion optimale en microéconomie attribuée à Clark ; nous verrons alors que l’on peut parler de découverte multiple.Colson est également intéressant au point de vue de la méthode utilisée. Nous verrons alors qu’il utilise les statistiques et les mathématiques dans ses développements : il est à l’origine d’une évaluation pionnière du revenu de la France, son enseignement impulse le calcul économique, il peut être considéré comme un précurseur de l’économétrie en France. Le deuxième chapitre montre que Colson développe la méthode de tarification des voies de communication exploitées en monopole de Jules Dupuit en proposant des moyens pratiques de révélation des préférences. Nous verrons également que cette théorie est reprise de nos jours avec le Yield Management et par les compagnies aériennes à bas coûts. Le troisième chapitre a pour but de voir comment Colson prend en compte la question sociale. Nous verrons qu’il défend une intervention de l’État plus importante que d’autres économistes libéraux afin d’éviter que les ouvriers ne se tournent vers le socialisme. Le quatrième chapitre étudie l’intervention de l’État préconisée par Colson dans le domaine des chemins de fer. Nous verrons que dans ce domaine où l’État est très présent, l’auteur souhaite le limiter. Il préfère ainsi la concession à la régie et souhaite la construction de nouvelles lignes uniquement si elles sont rentables. Nous verrons qu’à nouveau, la crainte du socialisme n’est pas étrangère à ses positions. Sur chacun des thèmes, nous verrons que Colson accorde à la pratique une place importante. Au niveau théorique tout part de l’observation et se termine par l’observation, au niveau pratique il est marqué par les préoccupations de son époque<br>The object of this dissertation is to analyse the contributions of Clément Colson to the economics in order to be better acquainted with his thought, with the French Liberal School at the beginning of the 20th century, to see the influence he had on his main students, Divisia, Roy and Rueff, and to evaluate the actuality of his recomandations. The first chapter develops the Colson’s theoretical contributions.For his students, his main contribution would be the theory of the joint setting of wage and of the interest rate. We will explain this theory is close to the optimal management rule in microeconomics attributed to Clark; we will see we can speak then about multiple discovery.Colson is interesting too from the point of view of the method used. We will see he uses statistics and mathematics in his developments: he makes one of the first assesments of the French income, his lectures develop economics calculus, he can be seen as a precursor of econometrics in France. The second chapter shows that Colson develops the Jules Dupuit pricing method for means of communications exploited by a monopoly by proposing practical way of preferences revelation. We will show too that this theory is used nowadays with the Yield Management and by airline lowcost companies.The third chapter has for purpose to see how Colson takes into account the social question. We will see he argues for a more important State intervention than other liberal economists in order to avoid workers to turn to socialism. The fourth chapter is devoted to the State intervention recommended by Colson in the field of railways. We will see that in this field in which the State is very present, the author wishes to limit it. So he prefers the concession to the public exploitation and wishes construction of new railway lines just if they are profitable. We will see again that the fear of socialism is not stranger to his positions. On each theme, we will see that Colson gives an important place to the practice. At the theoretical level all starts and finishes with the observation, at the practice level he is influenced by the preoccupations of his time

碩士<br>國立臺灣大學<br>化學工程學研究所<br>88<br>The effect of the superficial gas velocity, volume fraction of solids in gas holdup, pressure fluctuations, power spectral density function and dominant frequency in a concentric-tube airlift reactor was investigated. The outer column is constructed of 15.4cm i.d. and 120cm in height. The draft tube has 11.4cm i.d. and 70cm high. The draft tube is located 10cm above the base of the outer column. The 10cm diameter ring sparger is located at the bottom of the draft tube, with 28 holes drilled in the top of the ring, each of 0.15cm diameter. The solids used in this study is particles of polystyrene and high density polyethylene. The pressure probe and optical fiber probe are positioned inside the outer column and the draft tube for measuring pressure fluctuation signals and gas holdup distribution in the riser and downcomer. The results showed that the gas holdup increased with increasing the superficial gas velocity but the gas holdup decreased with increasing solids loading in the riser and downcomer. The bubble coalescence, pressure fluctuations and standard deviation of pressure fluctuations increased with increasing the superficial gas velocity or solids loading. It was observed that bubbles tended to flow in the central direction from distribution of gas holdup in the radial direction in the riser and downcomer. Power spectral density function and dominant frequency increased with increasing the superficial gas velocity in the riser but decreased in the downcomer. Both Power spectral density function and dominant frequency decreased with increasing solids loading in the riser and downcomer. With higher volume fraction of polystyrene and lower superficial gas velocity, polystyrene piled up the bottom of the reactor.

碩士<br>國立清華大學<br>化學工程學系<br>89<br>Liquid mixing characteristics were studied for the airlift reactor operating with net draft tube. The experimental reactor had a working volume of 26 L, column diameter of 0.13 m, and total height of 2 m. Experiments were carried out using air, water (or CMC solution), and polystyrene particles. Mixing in the liquid phase was assessed by the tracer-response technique and fast conductivity measurements. It was found that the flow pattern in an airlift reactor with net draft tube was similar to that in bubble columns instead of traditional airlift reactor. The mixing time decreased and the axial dispersion coefficient increased with increasing net mesh number. Moreover, the axial dispersion coefficient increased while rising superficial gas velocity except the transition from bubbly to churn-turbulent flow regime. The better diameter ratio of draft tube to column was 0.6, and the reactor with higher draft tube had better mixing performance. However, the top section had slight influence on liquid mixing, and the dispersion coefficient didn’t change with the reactor. Furthermore, adding solids and increasing liquid viscosity would enhance the mixing efficiency. In addition, the correlation between dispersion coefficient and these variables was established and the observed values were in agreement with the calculated values with a deviation of 30 %.

碩士<br>國立清華大學<br>化學工程學系<br>85<br>Because of the swift and continuous progress in biotechnology, it hasbecome a necessary trend to maximize the production or additive value of biochemical products with minimal cost. Consequently a few of modified bioreactors of different types have been proposed to overcome the disadvantages of stirred tanks, such as bubble columns and airlifts. For the aerobic fermentation systems, Tung et al. (1997) proposed an airlift reactor with double concentric net draft tubes to improve the performance of the conventional bubble column. The experimental data showed that under the same operational conditions, both the gas holdup and the oxygen transfer coefficient of the proposed reactor were superior to those of the bubble column. This reactor was also applied to cultivation of Saccharomyces cerevisiae, and successfully shortened the cultivation time by 18%.In order to apply this reactor to the fermentation processes in industrial scales, we chose the gas holdup, liquid mixing, shear stress and gas-liquid mass transfer coefficient as parameters to evaluate the performance on the scale up of the proposed reactor in an air-water system. Axial scale up of the small reactor showed minor effect on the gas holdup and mass transfer, but both the shear stress and mixing time were less than the same scale bubble column. After increasing the inner diameter of reactor from 13 to 19cm, the gas holdup and mass transfer coefficient were still higher than those of the bubble column (id 19cm). Therefore we set a pair of concentric net draft tubes as one module and proposed an airlift reactor with multi-net draft tubes, which was 29cm in diameter and contained 4 modules of net draft tubes. Experimental results showed that there is a similar performance in the large reactor (id 29cm) and the small one (id 13cm) with net draft tubes. The large reactor with net draft tubes was much superior to the same scale bubble column in gas holdup and mass transfer coefficient. The mixing time of the large scale new reactor was also comparable with that of the bubble column. Thus scale-up of the reactor with net draft tubes was successfully achieved.In order to figure out the improvement of the performance of the proposed reactor, we observed the bubble dynamic behavior in the large scale reactor by measuring the bubble size and the bubble number. Experimental results showed that the sizes of bubbles in the proposed reactor and the bubble column were about the same, but in the proposed reactor the bubble number was much higher than that in the bubble column under the same operational conditions. This implied that the net tubes could hold more air and provide more gas-liquid interfacial area, which improved the mass transfer ability.Because of the excellent performance of the new reactor, we hope that it will be applied to fermentation processes in the future, and can improve the conventional bubble columns or the airlift reactors for more energy saving and higher production.

碩士<br>輔英科技大學<br>生物技術系碩士班<br>97<br>Bacterial Cellulose (BC) is an extracellular beta-glucan produced by a species of Gluconacetobacter. This cellulose is different form that found in the cell walls of plants. Gluconacetobacter produce cellulose with special properties such as rigidity, strength, excellent biodegradability, high crystallinity, and large water holding capacity. Because of BC’s particular physical and chemical properties, there are many applications to numerous fields, such as biomedical materials and food products.Thus, researchers and industries are now searching for effective ways to produce BC by Gluconacetobacter. When using traditional reactors, there are challenges to optimal bacterial growth such as high temperatures and shearing stress. For BC production,a airlift reactors has been used for cultivating Gluconacetobacter. Growing bacteria in the Airlift reactor allowed sphere formation, and showed an improvement in bacterial growth, metabolism, and cellulose yield. In this study, different cultivaion conditions have been investigated using the airlift reactors as well as net plates to produce BC.The water holding capacity ,loading force,tension and stress were examined.

碩士<br>淡江大學<br>水資源及環境工程學系<br>89<br>This study is a experiment for treatment efficiency from the fish market wastewater by “biofilm airlift suspension reactor system”. The experiment hydraulic retention time are 2, 4, 6, 8 and 12 hours. It aims to discuss the treatment efficiency, nitrification reaction and biological kinetics constant. According to the results of the experiment: 1.When the hydraulic retention time is 6 hours, the BOD、COD and TKN removal rate can reach 95%、80% and 78%, and the BOD、COD and TKN rate treatment efficiency can reach 1.6 (kgBOD/m3.day)、1.7(kgCOD/m3.day) and 0.5(kgTKN/m3.day). 2.When the hydraulic retention time is 6 hours, this experiment biomass can reach 14000mg/L, sludge retention time reach 60~70 days, and the nitrification rate reach 0.5kg/m3-day。 3.This study can find the “substrate inhibition”, the reason may be the high concentration of substrate. The study applies Haldane mode to compute biological kinetics constants. The maximum rate of BOD utilization (k) is 0.337(kg/kg.day), the half-velocity constant (Ks) of BOD is 0.054(kg/m3), and the inhibition constant of BOD (Ki) is 0.743kg/m3. The maximum rate of COD utilization (k) is 0.452(kg/kg.day), the half-velocity constant (Ks) of COD is 0.164(kg/m3), and the inhibition constant of COD (Ki) is 0.834kg/m3. The maximum rate of TKN utilization (k) is 0.072(kg/kg.day), the half-velocity constant (Ks) of TKN is 0.03(kg/m3), and the inhibition constant of TKN (Ki) is 0.214kg/m3. 4.When COD/TKN is less than 2, it has be found obviously nitrite accumulation. When COD/TKN is greater than 4, the nitrite accumulation has not be found. When COD/TKN is less than 2 and must reach completely nitrification, it may need to increase the amount of carrier. 5.This study keeps dissolved oxygen concentration greater than 3mg/L. The effect of nitrification by dissolved oxygen may argue more in the after study.

博士<br>國立清華大學<br>化學工程研究所<br>82<br>Design and performance analysis of an aerated bioreactor have been investigated in this dissertation. The reactor was applied to an aerobic fermentation process for cultivation of Bacillus thuringiensis. The aim of this dissertation was focused on adaptive computer control of the fermentation system to improve the productivity of thuringiensin. The purpose of the dissertation was also on reactor design for biochemical processes. The airlift reactor with the net draft tube of 6.5 cm in diameter was chosen for fermentation application in thuringiensin production. Cultivations were carried out in agitated reactor, bubble column, and the airlift reactor for comparison. The results exhibited that productions of thuringiensin in a bubble column and an agitated reactor were lower than that in the new reactor, since the new reactor had higher mass transfer rate and lower shear stress in comparison with both bubble column and agitated reactor. The productivity of thuringiensin in the new reactor increased 65% in comparison with the highest production in the literature. Because of the outstanding ability for production of thuringiensin, the new reactor was utilized for fed- batch culture. Different feeding control strategies were applied and the composition of feeding substrate was modified. By combining adaptive control and the well performance of the new reactor, a suitable environment for fermentation was provided. Under such operating conditions, production of thuringiensin could reach 13.73 g/L which was more than tenfold in comparison with the highest production in the literature.

博士<br>國立清華大學<br>化學工程學系<br>85<br>In air-water system, the values of mass transfer coefficient and gas holdup of the proposed reactor (13cm I.D.) were higher than those of bubble column above 100%. Besides, in CMC-air system, the new reactor still exhibited significant improvements on mass transfer and mixing behaviors than in bubble column. Experimental data shown that the static liquid height seemed had little effect on the values of mass transfer coefficient and gas holdup. The empirical equations of the gas holdup and mass transfer coefficient with respected to superficial gas velocity were correlated. The shear stress of the new reactor was lower than the bubble column under the gas velocities between 1.0 cm/s and 7.5 cm/s.Modules of net columns were used for the scale up of the new reactor. One module was constituted of two concentric net columns and one sparger. This idea made the work of scale up easily and formed a common principle. In the first stage of scale up, the diameter of the reactor was extended from 13 cm to 19 cm. The net tubes were mesh 3 and 1 m height. The diameter of inner and outer tube was 8.5 cm and 10 cm, respectly. In the second stage of scale up, the diameter of the reactor was extended to 29 cm.

碩士<br>淡江大學<br>水資源及環境工程學系<br>89<br>The Biofilm Airlift Suspension Reactor（BAS）is used to treat the high strength ammonia-nitrogen wastewater, such as the fish market wastewater in this research. The purpose of this study is to discuss the nutrients removal and BAS system character. Synthetic wastewater which contains high concentration of organic carbon and nitrogen was used as substrate in the experiment. This research applies laboratory scale model to operate the experiments continuously. The hydraulic retention time（HRT） is used as variable parameters to investigate the treatment efficiency and biological kinetics. There were four runs into different hydraulic retention times hold in this experiment. Both of BOD and COD removal were maintained over 90﹪. When the shortest hydraulic retention time is 4 hours, the removal rate are 96.8﹪and 92.7﹪, respectively. In NH4-N removal efficiency, the removal efficiency is increasing with the HRT increasing. The best results were hydraulic retention time of 12 hr, the NH4-N removal rate reached 78.7﹪. When the loading of BOD was 0.22kg BOD/kg MLSS-day and the loading of COD was 0.35 kg COD/kg MLSS-day, the removal rate of BOD and COD could higher than 90﹪. But the removal rate of ammonia-nitrogen was 43.5﹪at the same loading. Possibly because heterotrophic organisms growth faster and inhibited activity of nitrifiers at higher organic loading. When the organic loading depressed, carbon sources decreased, made nitrifiers growth and nitrification increased at the same time. The substrate utilization rate of BOD、COD and NH4-N were highest at HRT=4hr, and decreased with HRT increased. The substrate utilization rate of BOD was 2.68kg/m3-day, of COD was 4.04 kg/m3-day and of NH4-N 0.27 kg/m3-day.The specific utilization rate constants（K）of BOD、COD and NH4-N were KBOD= 0.014 L / mg -d、KCOD= 0.006 L / mg-d and K NH4-N = 0.0004 L / mg -d investigated with Lawrence-McCarty kinetics.

碩士<br>國立清華大學<br>化學工程學系<br>87<br>ABSTRACTSince 1970s, the continuous progress in biotechnology, it has become the important technology in the world. Fermentation engineering of biotechnology deal with the use of living microorganism to produce biotechnological products. In fermentation engineering, bioreactor is the most important part for industrialization of biotechnology. Bioreactor with better liquid mixing and higher oxygen transfer are very important for microorganism. Bubble columns and airlift reactors are the two main kinds of pneumatic bioreactors. Because of many operational advantages, such as simple structure, low shear stress, low operational cost and ease to maintain, they are commonly employed in bioprocess and wastewater treatment. Despite many advantages, the commercial applications of conventional pneumatic bioreactors in biotechnology is still quite limited due to oxygen transfer and liquid mixing in the bioreactors. The purpose of this study was focus on bioreactor design for fermentation engineering. In the present study, we developed a novel bioreactor - airlift reactor with mulit-net draft plates. The novel bioreactor was investigated in the air-water hydrodynamic system and applied to production of chitosan. Other bioreactors including bubble column, airlift reactor, airlift reactor with double net draft tubes[Tung et al., 1997] and agitated reactor are also employed for comparison. For air-water system, the novel bioreactor has properties of both gas-liquid draft flow and gas-liquid dispersion that overcome the disadvantage of poor gas-liquid mixing of conventional bubble column. In liquid mixing, the novel bioreactor shorten the liquid mixing time of bubble column about 40%, about 30% for airlift reactor and about 10% for airlift reactor with double net draft tubes. And the novel bioreactor also has the ability of bubble slicing that slice bubbles into smaller ones and increase the interfacial area of gas bubbles. In oxygen mass transfer, the novel bioreactor had higher volumetric oxygen transfer coefficient ( kLa ) than other pneumatic bioreactors. The kLa in the novel biorector was up to 50% higher than that in the bubble column, 60% higher than airlift reactor and 10% higher than airlift reactor with double net draft tubes. The novel bioreactor has special properties of gas-liquid draft flow and bubble slicing that improve those disadvantages of conventional pneumatic bioreactors. Hence, both oxygen transfer and liquid mixing of the novel bioreactor were better than those of conventional bubble column and airlift reactor. For production of chitosan by fermentation, the novel bioreactor - airlift reactor with 3 net draft plates was chosen for production of chitosan by fermentation. Cultivation of A. coerulea was carried out in agitated reactor, bubble column and airlift reactor with double net draft tubes for comparison. The experiment results showed that shear stress of agitated reactor damaged microorganism. In the pneumatic bioreactors with expulsion shear stress, the ability of oxygen mass transfer of bioreactors had important effect on metabolism of microorganism. We proposed correlation equations that the volumetric oxygen transfer coefficient ( kLa ) of pneumatic bioreactors versus those maximum productivity of biomass and chitosan in the form of : DCW max. = 0. 4899*kLa - 59.166 CHITOSAN max. = 0.0425*kLa - 4.718Above equations correlated the kLa of pneumatic bioreactors with the maximum productivity of biomass and chitosan of microorganism may be useful with regard to information of scale up. The concluded results demonstrated that since the novel bioreactor had higher oxygen mass transfer rate and lower shear stress in comparison with bubble column, airlift reactor with double net draft tubes and agitated reactor, the novel bioreactor had higher productivity of biomass（39.52g/L） and chitosan （3.92g/L）. Furthermore, properties of chitosan including weight-averaged molecular weight （1600~1200 KDa）and degree of deacetylation （95~94%）in the novel bioreactor were higher than that in bubble column, airlift reactor with double net draft tubes and agitated reactor. These experimental facts suggests strongly that the novel bioreactor is excellent candidate for applied to aerobic fermentation engineering.keywords : bioreactor design; chitosan; fermentation; Absidia coerulea

碩士<br>國立清華大學<br>化學工程學系<br>89<br>For developing a fermentation process, selection of a proper fermenter is an important step. There are many types of fermenter in industrial fermentation processes. Stirred tank fermenters and bubble columns are commonly used. For an aerobic fermentaion, especially, oxygen transfer and liquid mixing are two key factors important for the growth and metabolism of microorganisms. Stirred tank fermenter can provide sufficient oxygen transfer by manipulating both aeration and agitation speed. However, the stirred tank fermenter usually have high shear stress which is not suitable for cultivation of some cells. Bubble column has low shear stress, but it has low capability of oxygen transfer and poor mixing. For some fermentation systems, the cells are sensitive to shear stress and have high oxygen consumption rate. Both stirred tank fermenter and bubble column are not suitable for the fermentation systems. The purpose of this study focuses on the design and scale-up of a bioreactor. In the present study, a pilot scale airlift reactor with multiple wire-mesh draft plates was developed. Mixing and mass transfer were investigated in the air-water hydrodynamic system. The experimental results showed that the airlift reactor still had the characteristics of high oxygen transfer rate and well mixing after scaling up. In addition, the proposed reactor also showed that the performance, such as volumetric mass transfer coefficient, gas holdup and mixing time, of the reactor surpassed those of the other bioreactors, including bubble column and airlift reactor with double net draft tubes. The application of the proposed reactor for cultivates Acetobacter xylinum for bacterial cellulose (BC) production has been carried out. The experimental results demonstrated that using the proposed column gave better productions than those of the conventional bubble column and stirred-tank reactor. Further, we have found that BC produced by the proposed reactor formed a unique ellipse pellet (BC-pellet, the average diameter is 10 mm), which is different from the fibrous BC produced by the agitated stirred-tank. Besides, the physical and morpholocigal properties of the BC-pellet also surpassed those of the fibrous BC. Moreover, the BC-pellet suspension was demonstrated to have a higher volumetric oxygen transfer coefficient and mixing capability than those of the fibrous BC suspension in the proposed reactor. The dissolved oxygen during the fermentation process by using the proposed reactor could be maintained above 35% and the growth or metabolism of cells was not inhibited. After 72 hours, the final concentration of BC was 7.72 (g/L) and the productivity was 0.107 (g / h․L), which was three times higher than that of the conventional bubble column and stirred-tank reactor. Consequently, the proposed reactor is quite suitable for cultivating Acetobacter xylinum for bacterial cellulose production. Keywords: bioreactor design, airlift reactor, scale-up, fermentation, bacterial cellulose, Acetobacter xylinum.

碩士<br>國立中正大學<br>化學工程研究所<br>90<br>The goals of this study are to measure and simulate the instantaneous and time-averaged hydrodynamics of an internal-loop airlift reactor by particle image analyzer (PIA) and CFX4.4 software, respectively. Increasing with gas velocity, the macroscopic flow structures of the airlift are similar to those of the bubble column and are divided into two types, dispersed and coalescence bubbling regimes. In the dispersed bubbling regime (Ug< 0.7 cm/s), no bubble exists in the downcomer due to the low liquid circulating velocity. At 0.7 < Ug < 2.2 cm/sec, the 4-region flow region comprises the descending, vortical, fast bubble, and central plume regions in the riser and some small bubbles are dragged to downcomer by the circulating liquid. Descending vortices in the middle zone of downcomer result from force imbalance. When the gas velocity is increased up to 2.2 cm/sec, two fast bubble chains form a larger fast bubble chain in the central region. Circulation liquid velocity relatively increases and drags the bubbles down to the downcomer. Besides, liquid average velocity, gas hold, average bubble size, and Reynolds Stress increases with increasing gas velocity. In simulation aspects, because of applying the sparger modeling the liquid turbulent intensity is enhanced, and leads to that the liquid average velocity is higher and the average bubble size is lower than those of experimental results. However, in general, the trend of the average bubble size distribution in radial location is similar to that of experiment.

碩士<br>國立中正大學<br>化學工程研究所<br>90<br>The goal of the research is using calcium alginate immobilized two broken mycelia of intracellular enzymes respectively, produced high-content fructooligosaccharides（FOS）from sucrose in an internal-loop airlift reactor. The reaction is to covert sucrose to FOS and glucose by the intracellular enzyme of Aspergillus japonicus (A. japonicus), β-fructofuranosidase. Then, glucose was removed by glucose oxidase, which is an intracellular enzyme of Aspergillus niger (A. niger). The research is found that calcium alginate as an immobilized carrier, the optimum entrapped broken mycelia of A. niger is 7%(w/v). As the superficial gas velocity is at 7.32 cm/s in internal-loop airlift reactor, the reaction has a high mass transfer coefficient and well mixing conditions. Comparing with the stirred tank, the internal airlift reactor generates less shear stress and causes the immobilized carriers less broken. If fixed the loading of two kinds calcium alginate immobilized the broken mycelia during producing FOS and the high superficial gas velocity, the DO is enough, the reactions are chemical reaction control. When operated at the low superficial gas velocity, the DO is insufficient, the reactions are mass transfer control, because glucose can not removed by glucose oxidase immediately. If fixed the loading of calcium alginate immobilized the broken mycelia of A. japonicus and changed the loading of calcium alginate immobilized the broken mycelia of A. niger, the superficial gas velocity is operated at the DO is enough, the concentration of FOS is increased with the loading is increased and the optimum loading is between 262.5 grams and 315 grams, that is unobvious increase of FOS concentration if increase the loading of calcium alginate immobilized the broken mycelia of A. niger continually. The result is similar to Duan (2001) but the concentration of FOS is slight low, because the solution exist few fructose by hydrolysis.

碩士<br>國立臺灣大學<br>食品科技研究所<br>93<br>The objectives of this research were to study the optimal condition for cultivating Antrodia cinnamomea in airlift reactors for inhibiting the growth of human hepatocellular carcinoma cell (Hep G2) and human cervical epithelioid carcinoma cell (HeLa cell). The experimental design divided into three parts. In the first part, various airlift reactors including bubble column reactor, airlift reactor with solid draft tube and airlift reactor with net draft tube were investigated for their suitability for cultivating A. cinnamomea. And the stirred tank reactor was used as control. The results indicated that the airlift reactor with solid draft tube operating at aeration rate 0.1 vvm could yield the highest amount of the mycelium (3.13 g/L after 7 days of fermentation). The second part of this research investigated the optimal fermentation conditions for A. cinnamomea for producing the highest amount of triterpenes. Using the airlift reactor with solid draft tube, it was found that when the reactor was operated at 0.1 vvm and 25℃ for 42 days, the mycelium concentration (5.02 g/L) and the crude triterpene content were more than that at 22 and 28℃. The operating scale was increased from 5L to 500L airlift reactor for the third part of this study. It was found that under the optimal fermentation condition, the mycelium in the fermentation broth reached 5.5 g/L in only 28 days, as compared to the 42 days of the 5L reactor. During the second and third part of this study, the inhibition effect on the growth of Hep G2 and HeLa cell by the filtrate and the ethanol extract of the mycelium of A. cinnamomea was investigated. The results indicated that the growth inhibitory effect of the filtrate and the ethanol extract of the mycelium increased with fermentation time. Moreover, the inhibitory effect of the fermentation products cultivated at 25℃ was higher than that at 22 or 28℃. The IC50 of the ethanol extract of the mycelium cultivated at 25℃ for 21 days was 20μg/ml.

碩士<br>長庚大學<br>企業管理研究所<br>92<br>In the efficient market, how an investor respond to the information of firm’s critical accident. This kind of information will affect firm’s securities in stock price and market value that also affect opposite firm and related industry. This paper focuses on the adjustments of stock returns to the 5 times announcements of China Airline accidents. An event study methodology is used in examining the observation and the continuation of abnormal return as well as market model and t test statistics to measure the impact on stock performance. Also this paper researched over reaction effect, transfer effect and spillover effect. The conclusion can support that Taiwan stock market is fitting to semi-strong form theory. The firm’s stock price will respond to the accident information quickly with a negative abnormal return, but the negative abnormal return reduced on the following days that observed no continuous effect. On the event window we cannot find any over-reaction effect on the market. There are transfer effect show on the opposite firm but only at the same market place, and no evidence can prove spillover effect existed due to the accident event

L’objectif de cette thèse est d’analyser les contributions de Clément Colson à la science économique dans le but de mieux connaître sa pensée, de mieux connaître l’École libérale française au début du XXème siècle, d’étudier l’influence qu’a pu avoir cet auteur sur ses principaux élèves, Divisia, Roy et Rueff, et d’évaluer l’actualité de certaines de ses recommandations.Le premier chapitre montre quels sont les apports théoriques de Colson à la science économique.Pour ses élèves, son principal apport serait la théorie de la détermination conjointe du salaire et du taux d’intérêt. Nous montrerons que cette théorie est proche de la règle de gestion optimale en microéconomie attribuée à Clark ; nous verrons alors que l’on peut parler de découverte multiple.Colson est également intéressant au point de vue de la méthode utilisée. Nous verrons alors qu’il utilise les statistiques et les mathématiques dans ses développements : il est à l’origine d’une évaluation pionnière du revenu de la France, son enseignement impulse le calcul économique, il peut être considéré comme un précurseur de l’économétrie en France. Le deuxième chapitre montre que Colson développe la méthode de tarification des voies de communication exploitées en monopole de Jules Dupuit en proposant des moyens pratiques de révélation des préférences. Nous verrons également que cette théorie est reprise de nos jours avec le Yield Management et par les compagnies aériennes à bas coûts. Le troisième chapitre a pour but de voir comment Colson prend en compte la question sociale. Nous verrons qu’il défend une intervention de l’État plus importante que d’autres économistes libéraux afin d’éviter que les ouvriers ne se tournent vers le socialisme. Le quatrième chapitre étudie l’intervention de l’État préconisée par Colson dans le domaine des chemins de fer. Nous verrons que dans ce domaine où l’État est très présent, l’auteur souhaite le limiter. Il préfère ainsi la concession à la régie et souhaite la construction de nouvelles lignes uniquement si elles sont rentables. Nous verrons qu’à nouveau, la crainte du socialisme n’est pas étrangère à ses positions. Sur chacun des thèmes, nous verrons que Colson accorde à la pratique une place importante. Au niveau théorique tout part de l’observation et se termine par l’observation, au niveau pratique il est marqué par les préoccupations de son époque<br>The object of this dissertation is to analyse the contributions of Clément Colson to the economics in order to be better acquainted with his thought, with the French Liberal School at the beginning of the 20th century, to see the influence he had on his main students, Divisia, Roy and Rueff, and to evaluate the actuality of his recomandations. The first chapter develops the Colson’s theoretical contributions.For his students, his main contribution would be the theory of the joint setting of wage and of the interest rate. We will explain this theory is close to the optimal management rule in microeconomics attributed to Clark; we will see we can speak then about multiple discovery.Colson is interesting too from the point of view of the method used. We will see he uses statistics and mathematics in his developments: he makes one of the first assesments of the French income, his lectures develop economics calculus, he can be seen as a precursor of econometrics in France. The second chapter shows that Colson develops the Jules Dupuit pricing method for means of communications exploited by a monopoly by proposing practical way of preferences revelation. We will show too that this theory is used nowadays with the Yield Management and by airline lowcost companies.The third chapter has for purpose to see how Colson takes into account the social question. We will see he argues for a more important State intervention than other liberal economists in order to avoid workers to turn to socialism. The fourth chapter is devoted to the State intervention recommended by Colson in the field of railways. We will see that in this field in which the State is very present, the author wishes to limit it. So he prefers the concession to the public exploitation and wishes construction of new railway lines just if they are profitable. We will see again that the fear of socialism is not stranger to his positions. On each theme, we will see that Colson gives an important place to the practice. At the theoretical level all starts and finishes with the observation, at the practice level he is influenced by the preoccupations of his time