Esse trabalho apresenta um modelo termodinâmico de uma bomba multifásica do tipo duplo parafuso. Para uma dada condição de operação, o modelo calcula a potência consumida, as condições do fluido na descarga e o perfil de pressão ao longo da bomba. Ao invés de simular diretamente o escoamento dentro da bomba, simulou-se os processos que ocorrem dentro das suas câmaras. Para tanto, dividiu-se o processo de bombeamento multifásico em uma seqüência de processos simples, facilitando-se a construção do modelo no simulador de processos Hysys.Process v2.1. Os resultados de potência e temperatura de descarga obtidos com a simulação mostram uma boa concordância com valores experimentais, principalmente para FVG’s baixos. Para FVG’s elevados, o modelo passa a superestimar a potência consumida indicando que as fendas, nesses casos, já não se encontram totalmente preenchidas com líqüido. Dos resultados obtidos para o refluxo, conclui-se que, das equações sugeridas na literatura, aquelas para escoamento turbulento liso são mais adequadas para os números de Reynolds envolvidos. O perfil de pressão e a vazão de refluxo quando o escoamento é multifásico aproxima-se qualitativamente das medições experimentais. Estudou-se a influência de diversos parâmetros na eficiência exergética da bomba. Os resultados mostram que a otimização da eficiência depende das condições de operação da bomba: FVG, tipo de líqüido, diferença de pressão, entre outros. / The goal of this project was to develop a thermodynamic model of a twin-screw multiphase pump. With given operation conditions the model can determine the absorbed power, discharge conditions and the pressure profile along the screw. An alternative approach was suggested to overcome the complex flow problem and the processes inside the pump were simulated instead of direct simulation of the flow. For this purpose, the multiphase pumping process was divided in a sequence of simple processes so the model could be developed in an easier way. The power and temperature values calculated by the model are in good agreement with experimental data, mainly when the gas fraction is low. With higher gas fractions, the model overestimates the absorbed power indicating that screw gaps are not completely filled with liquid anymore. Concerning about the backflow rate, the results show that the equations for turbulent flow in smooth ducts fits better the Reynolds number range in the gaps. The pressure profile and backflow rate for multiphase flow agree qualitatively with experimental results. The influence of several parameters in the exergetic eficiency of the pump were analysed and results show that the efficiency optimization depends on pump operation conditions: gas fraction, liquid type, pressure difference and others.

Orientadores: Everson Alves Miranda, Pedro de Alcantara Pessoa Filho / Tese (doutorado) - Universidade Estadual de Campinas, Faculdade de Engenharia Quimica / Made available in DSpace on 2018-08-09T04:41:28Z (GMT). No. of bitstreams: 1 Tashima_AlexandreKeiji_D.pdf: 2449482 bytes, checksum: a1d1f605f59d627bdb98b47161d1f15d (MD5) Previous issue date: 2007 / Resumo: A técnica de precipitação é freqüentemente utilizada na recuperação de proteínas de soluções aquosas. Uma das formas de se realiza-la é através do processo de precipitação isoelétrica, que consiste no ajuste do pH do meio ao valor em que a proteína tem carga global nula, ou seja, no ponto isoelétrico, pI. Ácidos e bases minerais são normalmente utilizados para ajustar o pH do meio ao pI e provocar a precipitação de proteínas. Contudo, para se obter proteínas de alta pureza e evitar a poluição do meio com resíduos indesejáveis, há a necessidade de um processo adicional para a neutralização da solução e remoção dos sais gerados. Além disso, concentrações locais elevadas dos ácidos podem causar a desnaturação da proteína de interesse. Estudos recentes vêm apontando a utilização de eletrólitos voláteis como uma alternativa promissora aos agentes de precipitação convencionalmente empregados na recuperação de proteínas. Eletrólitos voláteis são obtidos pela dissolução de gases como o dióxido de carbono em solução aquosa; nesta dissolução formam-se íons, cujas concentrações dependem da temperatura e pressão do sistema. Biomoléculas de interesse farmacêutico, como a insulina, por exemplo, podem ser potencialmente recuperadas por processos de precipitação isoelétrica com o CO2. Assim, neste projeto realizou-se um estudo experimental e teórico da precipitação isoelétrica da insulina suína com dióxido de carbono, avaliando-se a influência dos parâmetros de processo como a temperatura, pressão e concentrações de eletrólitos e de proteínas sobre a precipitação. A cinética de variação de pH devida à acidificação das soluções contendo a insulina foi determinada, assim como a cinética de precipitação da proteína. A utilização do bicarbonato de sódio como agente tamponante, em conjunto com o ácido carbônico formado em solução, permitiu que os estudos de equilíbrio fossem realizados sem a necessidade de nenhum componente adicional para o controle do pH do sistema. Dados de solubilidade da insulina suína foram obtidos entre as temperaturas de 5 e 25oC, até a pressão de 16 bar de CO2. Estes dados foram correlacionados por um modelo termodinâmico em que a proteína em solução foi considerada como um eletrólito, o que permitiu uma análise do efeito de forças eletrostáticas sobre a solubilidade da proteína / Abstract: Precipitation is a technique frequently employed in downstream process for the recovery of proteins from aqueous solutions. One example of such technique is the isoelectric precipitation, which consists of the adjustment of solution pH to the value where the protein has zero net charge, the isoelectric point, pI. Commonly, mineral acids and bases are used for pH adjustment and to induce protein precipitation. However, in order to obtain high purity proteins and avoid environmental pollution, an additional process is needed to neutralize and remove residual salts. It is also necessary to take into account the fact that local pH extremes can cause denaturation of the targeted protein. Recent works have pointed the use of volatile electrolytes as promising alternatives to the precipitating agents conventionally utilized in protein recovery. Volatile electrolytes are obtained by dissolution of gases as carbon dioxide in aqueous solutions. In this process, the volatile electrolyte dissociates into ions, whose concentrations depend on system pressure and temperature. Biomolecules of pharmaceutical interest as insulin, for instance, may be potentially recovered by isoelectric precipitation process with carbon dioxide. Thus, in this project an experimental and theoretical study of the isoelectric precipitation of porcine insulin with carbon dioxide was conducted, in order to evaluate the influence of parameters such as temperature, pressure and concentrations of electrolytes and protein over the precipitation process. Kinetics of pH variation due to acidification of the insulin solutions was determined, as well as protein precipitation kinetics. The use of sodium bicarbonate as buffer agent with carbonic acid produced in solution allowed the equilibrium studies be conducted without the need of any additional component to control system pH. Solubility data of porcine insulin were obtained in the temperature range of 5 to 25oC and up to CO2 pressures of 16 bar. These data were correlated by a thermodynamic model in which the protein in solution was considered as an electrolyte, which allowed an analysis of electrostatic force effects over protein solubility / Doutorado / Engenharia de Processos / Doutor em Engenharia Química

The master’s thesis is focused on creating a thermodynamic model of the diesel engine for heavy commercial vehicle. After the model was created, new intake and exhaust pipes are designed based on simulations. Those proposed pipes variants are compared with the series design. After evaluating individual variants, the final version of the new pipes is selected and 3D model for this version is created.

The energy demands in the world is rapidly increasing and with this,  a supply nuclear  power  is  of  much  interest.    Nuclear  fuel  is  relatively  efficient  when comparing to power sources like wind­ and hydropower plants. Pellets are used as fuel by many plants however, its main concern is to find maximize cost efficiency and  minimize  fuel­waste.   Studying  how  to  get  the  pellets  to  be  as  optimal  as possible is of massive importance and in huge focus in order to match the worlds power demand. These  pellets  are  sintered  in  a  furnace  type  known  as  ”pusher­type”  furnaces that functions continuously and is incredibly efficient when it comes to its heat transfer capacity and high­performance output.  In this sintering process, a gas­ flow from the opposite side from the pellets interacts with the solid pellets in order to get the desired reaction. However, the turbulence and the nature of the multi­ phase flow problem causes many unknown interactions and the main focus is do create a theoretical model based on the process parameters to understand what is happening in the furnace. In this study, a simplified model of the inside of the furnace chamber was created in order to observe where and when in the furnace a dissociation from CO2 to CO + O2 would occur. Data given by Westinghouse was put into a mathematical model created in MATLAB and parameters given by the thermodynamic model was in turn put in to ANSYS, a program based on Computational Fluid Dynamics for a simulation. The simulation was considered a success when the gas­mix goes from 3% CO2 to 0.4%. The CFD of the model estimates this to happen at 250 seconds, where as the thermodynamic model predicts the exchange time to be about 200 seconds.   This study is a major first step in understanding the dynamics of the furnace. / Energibehovet i världen ökar snabbt och då blir ett stadigt tillförsel av kärnenergi mycket intressant.  Kärnbränsle är relativt effektivt jämfört med kraftkällor som vind­  och  vattenkraftverk.    Pellets  används  som  bränsle  av många  kraftverk och då blir det ett upphov att hitta maximal kostnadseffektivitet och minimera bränsleavfall.  Att forska fram till hur man gör pellets så optimala som möjligt är av enorm betydelse och i stort fokus för att matcha världens energi behov. Dessa pellets sintras i en ugnstyp som kallas ”pusher­type” ugnar som fungerar kontinuerligt och är otroligt effektiva när det gäller dess värmeöverförings-kapacitet och högpresterande effekt.  I denna sintringsprocess startar ett gasflöde från motsatt sida från pelletsen med de fasta pelletsen för att få den önskade reaktionen.  Det blir ett flerfasigt flödesproblem och orsakar många okända interaktioner och huvudfokus är att skapa en teoretisk modell baserad på processparametrarna för att förstå vad som händer i ugnen. I  denna  studie  gjordes  en  förenklad  modell  av  ugnskammarens  insida för  att observera var och när i ugnen en dissociation från CO2 till CO + O2 skulle inträffa. Data från Westinghouse placerades i en matematisk modell skapad i MATLAB och  parametrar  som  gavs  av  den  termodynamiska modellen  lades  i  sin  tur  till ANSYS, ett program baserat på Computational Fluid Dynamics för en simulering. Simuleringen ansågs vara färdig när gasblandningen går från 3% CO2 till 0,4%. CFD:n  för  modellen  uppskattar att  detta  händer  vid  250  sekunder,  där  den termodynamiska modellen förutspår utbytestiden till cirka 200 sekunder. Denna studie är ett stort första steg för att förstå ugnens dynamik.

Die vorliegende Studie präsentiert eine Analyse der Integration der AhR und Wnt/beta-catenin Signalwege in den CYP1A1 Promotor sowie den regulatorischen Einfluss der Promotorlogik auf die Genexpression. Experimentell wurde diese Analyse mithilfe 29 mutagener Reporterkonstrukte des humanen CYP1A1 Promotors durchgeführt. Ein mathematisches Modell, welches eine Repräsentation des Crosstalks der Signaltransduktionswege mit einer statistisch mechanischen Beschreibung der kombinatorischen Promotorbelegung kombiniert, komplementierte den experimentellen Ansatz. Unter zusätzlicher Zuhilfenahme von gut kontrollierbaren synthetischen Promotorkonstrukten fand ich heraus, dass nur jenes Dioxin-responsive Element das sich am nächsten am Transkriptionsstartpunkt befindet, die Promotorbelegung an die RNA Polymerase kommuniziert. Außerdem beobachtete ich, dass Transkriptionsfaktoren alleine mit Transkriptionsfaktoren interagieren die mit benachbarten Bindestellen assoziieren, d.h. Interaktionen überbrücken keine größeren Entfernungen. Der Modellierungsansatz ermöglichte zudem die erfolgreiche Vorhersage einer UND-Gatter-ähnlichen Integration der beiden Signalwege in den Promotor. Für die genomische Architektur des CYP1A1 Promotors konnte ich die Signifikanz der Zielbindestelle des Wnt/beta-catenin Signalwegs innerhalb des cis-regulatorischen Region demonstrieren. Mithilfe des Modells fand ich heraus, dass diese Bindestelle am stärksten und vielfältigsten mit den restlichen Transkriptionsfaktoren interagiert. Zusätzliche konnte, im Vergleich zu dem alles-oder-nichts UND-Gatter der synthetischen Konstrukte, eine sehr viel graduellere Antwort auf die Integration der beiden Signalwege aufgezeigt werden. Abschließend wurde das physiologisch zu beobachtende Expressionsmuster von dem Modell vorhergesagt und experimentell validiert. / The study at hand presents an analysis of the integration of the AhR and the Wnt/beta-catenin signaling pathways into the CYP1A1 promoter as well as the regulatory influence of the promoter logic on gene expression. Experimentally, this analysis was conducted with the help of 29 mutant constructs of the human CYP1A1 promoter. I complemented this experimental approach with a set of mathematical models that combined a representation of the signaling crosstalk with a statistical mechanics description of the combinatorial promoter occupancy. With the help of well controllable synthetic promoter constructs I found that only the dioxin responsive element closest to the transcription start site communicates the promoter occupancy to the RNA polymerase. Furthermore, transcription factors only interact with transcription factors that associate with nearby binding sites, i.e., no long-distance binding was observed. The modeling approach subsequently enabled the successful prediction of an AND-gate-like integration of the two signaling pathways into the promoter. For the genomic architecture of the CYP1A1 promoter, I could demonstrate the importance of the Wnt/beta-catenin pathway target binding site within the cis-regulatory region. The model uncovered that this binding site is the strongest and most promiscuous interaction partner of the remaining transcription factors. In addition, a less switch-like response to the integration of the two signaling pathways as compared to the all-or-none AND-gate within the synthetic constructs could be demonstrated. And lastly, the physiological expression pattern in liver lobules could be successfully predicted by the model and experimentally verified.

Phase behaviour information is essential for the development and optimization of hydrocarbon resource production, transport and refining technologies. Experimental data sets for mixtures containing heavy oil and bitumen are sparse as phase behaviour data are difficult to obtain and cost remains prohibitive for most applications. A computational tool that predicts phase behaviours reliably for mixtures containing such ill-defined components, over broad temperature, pressure and composition ranges would play a central role in the advancement of bitumen production and refining process knowledge and would have favourable impacts on the economics and environmental effects linked to the exploitation of such ill-defined hydrocarbon resources. Prior to this work, predictive computational methods were reliable for dilute mixtures of ill-defined constituents. To include a much wider range of conditions, three major challenges were addressed. The challenges include: creation of a robust and accurate numerical approach, implementation of a reliable thermodynamic model, and speciation of ill-defined constituents like Athabasca Bitumen Vacuum Residue (AVR). The first challenge was addressed by creating a novel computational approach based on a global minimization method for phase equilibrium calculations. The second challenge was tackled by proposing a thermodynamic model that combines the Peng-Robinson equation of state with group contribution and related parameter prediction methods. The speciation challenge was addressed by another research group at the University of Alberta. Pseudo components they proposed were used to assign groups and estimate thermodynamic properties. The new phase equilibrium computational tool was validated by comparing simulated phase diagrams with experimental data for mixtures containing AVR and n-alkanes. There is good qualitative and quantitative agreement between computed and experimental phase diagrams over industrially relevant ranges of compositions, pressures and temperatures. Mismatch was only observed over a limited range of compositions, temperatures and pressures. This computational breakthrough provides, for the first time, a platform for reliable phase behaviour computations with broad potential for application in the hydrocarbon resource sector. The specific computational results can be applied directly to solvent assisted recovery, paraffinic deasphalting, and distillation and refining processes for Athabasca bitumen a strategic resource for Canada. / Chemical Engineering

No description available.

Capturer le CO2 est devenu un domaine de recherche important en raison principalement des forts effets de serre dont il est jugé responsable. La formation d'hydrate de gaz comme technique de séparation montre un potentiel considérable, d'une part pour sa faisabilité physique et d'autre part pour une consommation énergétique réduite. En bref, les hydrates de gaz (clathrates) sont des composés ″cages″ non-stoechiométriques, cristallins comme la glace et formés par une combinaison de molécules d'eau et de molécules hôtes convenables, à basses températures et pressions élevées. Puisque la pression exigée pour la formation d'hydrate de gaz est généralement forte, il est judicieux d'ajouter du bromure tétra-n-butylique d'ammonium (TBAB) comme promoteur de formation d'hydrate de gaz. En effet, le TBAB permet généralement de réduire la pression exigée et/ou d'augmenter la température de formation aussi que de modifier la sélectivité des cages d'hydrates au profit des molécules de CO2. TBAB participe à la formation des cages par liaisons ″hydrogène″. De tels hydrates sont nommés "semi-clathrate hydrates". Évidemment, des données d'équilibres de phase fiables et précises, des modèles thermodynamiques acceptables, et d'autres études thermodynamiques sont requises pour concevoir des procédés de séparation efficaces utilisant la technologie mentionnée ci-dessus. Dans ce but, des équilibres de phase de clathrate/semi-clathrate hydrates de de divers mélanges avec des gaz contenant CO2 (CO2 + CH4/N2/H2) ont été mesurés, ici, en présence d'eau pure et de solutions aqueuses de TBAB. La partie théorique de la thèse présente un modèle thermodynamique développé avec succès sur la base de la théorie des solutions solides de van der Waals et Platteeuw (vdW-P) associée aux équations modifiées de la détermination des constantes de Langmuir des promoteurs d'hydrates pour la représentation/prédiction des équilibres en présence de ″semi-clathrate hydrates″ de CO2, CH4, et N2. Plusieurs tests de cohérence thermodynamique basés soit sur l'équation de Gibbs-Duhem, soit sur une approche statistique ont été appliqués aux données d'équilibre de phase des systèmes de ″clathrate hydrates″ simples/mélanges afin de statuer sur leur qualité. / CO2 capture has become an important area of research mainly due to its drastic green-house effects. Gas hydrate formation as a separation technique shows tremendous potential, both from a physical feasibility as well as an envisaged lower energy utilization criterion. Briefly, gas (clathrate) hydrates are non-stoichiometric, ice-like crystalline compounds formed through a combination of water and suitably sized guest molecule(s) under low-temperatures and elevated pressures. As the pressure required for gas hydrate formation is generally high, therefore, aqueous solution of tetra-n-butyl ammonium bromide (TBAB) is added to the system as a gas hydrate promoter. TBAB generally reduces the required hydrate formation pressure and/or increases the formation temperature as well as modifies the selectivity of hydrate cages to capture CO2 molecules. TBAB also takes part in the hydrogen-bonded cages. Such hydrates are called "semi-clathrate" hydrates. Evidently, reliable and accurate phase equilibrium data, acceptable thermodynamic models, and other thermodynamic studies should be provided to design efficient separation processes using the aforementioned technology. For this purpose, phase equilibria of clathrate/semi-clathrate hydrates of various gas mixtures containing CO2 (CO2 + CH4/N2/H2) in the presence of pure water and aqueous solutions of TBAB have been measured in this thesis. In the theoretical section of the thesis, a thermodynamic model on the basis of the van der Waals and Platteeuw (vdW-P) solid solution theory along with the modified equations for determination of the Langmuir constants of the hydrate formers has been successfully developed to represent/predict equilibrium conditions of semi-clathrate hydrates of CO2, CH4, and N2. Later, several thermodynamic consistency tests on the basis of Gibbs-Duhem equation as well as a statistical approach have been applied on the phase equilibrium data of the systems of mixed/simple clathrate hydrates to conclude about their quality.

Le développement de simulateurs de procédés alimentaires est limité par le manque d’outils permettant de prédire les propriétés physicochimiques (pH, aw, Eh) dans les aliments. Les aliments sont généralement des milieux très complexes contenant un nombre élevé de constituants (eau, solvants organiques, solides dissous, gaz dissous, espèces ioniques, macromolécules), et ces propriétés sont essentielles pour caractériser leur qualité technologique, sanitaire et organoleptique. Cependant, elles évoluent au cours d’un procédé de transformation ou de conservation. La prédiction de ces propriétés physicochimiques passe par la détermination de potentiels chimiques. Une approche de modélisation basée sur la thermodynamique est utilisée pour prédire l’évolution du pH et de l’aw de produits carné et produits laitiers dans certaines conditions rencontrées en industrie alimentaire. En s’appuyant sur le modèle ULPDHS développé par Achard en 1992 pour les milieux liquides, ce travail a nécessité la création de molécules fictives à ayant le même comportement que les aliments modélisés. La cohérence entre les données expérimentales et les données prédites est satisfaisante, le modèle thermodynamique permet de prédire correctement le pH et l’aw des différents milieux étudiés. Cependant l’intégration du modèle thermodynamique dans un simulateur de procédé nécessite la création d’un réseau de neurone mimétique du modèle thermodynamique. En conséquent a été développé un réseau de neurones dont les sorties sont identiques aux sortie du modèle thermodynamique. Ainsi il est possible d’intégrer les résultats du modèles thermodynamique dans un simulateur de procédé sans trop rallonger les temps de simulations. Les résultats obtenus dans ce travail peuvent être une aide importante à la formulation de nouveaux produits. Le modèle étant entièrement prédictif, il est possible de déterminer l’effet d’une modification de composition de l’aliment sur le pH et l’aw. / The development of food process simulators is actually limited by absence of tools that can predict de evolution of the physical-chemical properties (pH, aw, Eh) in food. Food are generallymulticomponent medias (including water, organic solvents, dissolved solids, dissolved gases, ionic species, macromolecules), and these properties are essential to characterize technological, sanitary and organoleptic qualities. However, these physical-chemical properties change during transformation or preservation process. The prediction of these properties requires the determination of chemical potentials. Thermodynamic approach is used to predict pH and water activity of meat and dairy products in different condition encountered in the food industry. Based on the model ULPDHS developed by Achard in 1992 for liquid medias, this work required the creation of fictive molecule who have the same behavior as modeled food. Consistency between the experimental data and the predicted data is very satisfactory, the thermodynamic model correctly predict the pH and aw of different products studied. The integration of the thermodynamic model in a process simulator requires the creation of a mimetic neural network. Therefore, was developed a neural network whose outputs are identical to the output of the thermodynamic model. So it is possible to integrate the results of thermodynamic models in a process simulator without too lengthen the time simulations. The results obtained in this work can be an important aid to the formulation of new products. The model is completely predictive and it is possible to determine the effect of a change in the composition of the feed on the pH and aw.

Le travail vise à développer un procédé de dessalement de l’eau de mer par congélation sur parois froides. L’étude thermodynamique a permis de quantifier l’effet de la composition et de la salinité de l’eau sur la température de congélation et la température de précipitation de Na2SO4,10 H2O. Les résultats sont bien décrits par le code de calcul de Frezchem, dérivé du modèle de Pitzer. Les essais de dessalement par congélation ont été conduits avec des solutions eau/NaCl de différentes concentrations, ainsi qu’avec de l’eau de mer de Rabat, Nice et Marseille. Le montage expérimental mis au point se compose d’un doigt de gant plongé dans une cuve double enveloppée contenant l’eau à traiter. Le procédé complet de dessalement est conduit en deux étapes: l’étape de congélation produisant un dépôt de glace sur le doigt de gant et l’étape de ressuage, effectuée après vidange de la saumure, qui consiste à purifier en profondeur la glace en opérant une fusion des zones impures. Une caméra filme la couche de glace et fournit la cinétique de croissance de la couche. La congélation a été effectuée dans une solution stagnante ou dans une solution agitée par injection d’air. Elle est conduite en appliquant deux rampes de refroidissement, respectivement dans le doigt de gant et dans la double enveloppe, qui peuvent être ou non identiques. Les essais ont montré la nécessité d’ensemencer la couche de glace sur le tube et d’ajuster finement la température initiale de la rampe de refroidissement. L’étude systématique de l’influence des paramètres opératoires a mis en évidence les rôles importants de la rampe de refroidissement et de la salinité de la solution sur la pureté de la glace produite. En l’absence d’agitation, la température de la double enveloppe a également un effet notoire sur la pureté de la glace à cause des gradients de température, et par suite, des courants de convection qu’elle peut engendrer au sein de la solution. En régime agité, la solution semble être à chaque instant en équilibre avec la glace. Une même salinité finale de la glace peut être obtenue avec des rampes de refroidissement beaucoup plus rapides qu’en statique. Quelles que soient les performances de la congélation, le ressuage est ensuite indispensable pour atteindre la norme de potabilité. La recherche de conditions opératoires optimales a permis de réduire la durée du procédé global à 8h (5h de cristallisation et 3h de ressuage). Les résultats de ce travail montrent la faisabilité de la technique et donnent une bonne idée des conditions de fonctionnement qui peuvent être employées pour produire l'eau potable / The work aims to develop a process for freezing desalination of seawater on cold walls. The thermodynamic study has quantified the effect of composition and of salinity on the freezing temperature and the precipitation temperature of Na2SO4,10 H2O. The results are well described by the calculation code Frezchem, derived from the model of Pitzer. The experiments were performed with water/NaCl solutions of different concentrations and with samples of sea water from Nice, Rabat and Marseille. The pilot crystallizer consists of a cooled tube immersed in a cylindrical double jacketed tank containing water to be treated. The complete process of desalination is conducted in two steps: the freezing step, leading to the crystallization of the ice layer and the sweating step, which consists of purifying in depth the ice layer by melting the impure zones. A camera films the ice and provides the growth kinetics of the layer. Freezing was performed in a stagnant solution to a stirred solution or by injecting air. It is conducted by applying two cooling ramps, respectively in the tube and the double jacket that may or may not be identical. Tests have shown the need to seed the ice on the tube and fine-tune the initial temperature of the cooling ramp. The systematic study of the influence of operating parameters has highlighted the important role of the cooling ramp and salinity of the solution on the purity of the ice produced. In the absence of stirring, the temperature of the double jacket also has a noticeable effect on the purity of the ice due to temperature gradients, and consequently, convection currents that may result in the solution. Steady stirring, the solution seems to be at all times in equilibrium with ice. Same final salinity of the ice can be obtained with ramps cooling much faster than static. Whatever the performance of the freezing, the sweating is then necessary to meet the standard for drinking water. The search for optimum operating conditions has reduced the duration of the overall process to 8 hours (5 hours the crystallization and 3 hours of sweating). The results of this work show the feasibility of the technique and give a good indication of operating conditions that can be used to produce drinking water