HAL-MINES ParisTech
Not a member yet
27348 research outputs found
Sort by
Gestion prévisionnelle optimisée sous incertitudes jointes
No full textThe expansion of renewable energy sources (RES) leads to the growth of uncertainty in the power distribution network operation. The inherent variability and intermittency of RES present significant challenges to the efficient and reliable operation of power systems. To address these challenges, operational planning performed by distribution system operators should evolve, in particular, to allow the efficient utilization of different flexibility levers, such as active power modulation and reactive power management. Decisions on lever activation are based on the resolution of an alternating current optimal power flow problem (AC-OPF). This thesis develops algorithms for handling two stochastic AC-OPF models. These optimization problems are simultaneously nonconvex, nonsmooth, and discrete. The thesis aims to grasp these complexities accurately, by addressing the AC power flow equations without relying on convexification and by handling interdependent uncertainties either through a joint probability constraint or via scenario decomposition to cope with the discrete levers.More specifically, the first proposed methodology addresses a continuous version of the joint chance-constrained AC-OPF. A first contribution of this work is the design of a numerical procedure (oracle) that enables the representation of the probability constraint as a difference of two convex functions. This step is followed by applying a known Difference-of-Convex (DoC) bundle method to the resulting continuous optimization problem. A second contribution concerns a new bundle algorithm with stronger convergence guarantees under weaker assumptions. For the chance-constrained AC-OPF, this algorithm provides a critical (generalized KKT) point. The work builds upon the employed DoC bundle and proposes a different master program and an original rule to update proximal parameter. The algorithm is capable of handling a broad class of nonsmooth and nonconvex optimization problems beyond the stochastic AC-OPF framework, provided the objective and constraint functions can be represented as differences of convex and weakly convex (CwC) functions. The practical performance of the algorithm is illustrated through numerical experiments on some nonconvex stochastic problems and is compared to the DoC bundle method for the chance-constrained AC-OPF in a 33-bus distribution network.The second proposed methodology addresses operational planning rules for power modulation and curtailment, like priority and fairness, which result in logical and discrete formulations. The numerical results demonstrate the limitations of the bundle method for integrating integer variables. As an alternative, an optimization model is proposed that assigns a binary variable to each scenario and maximizes the number of satisfied scenarios within a limited budget. Applying penalization and block coordination allows separating those discrete considerations from the stochastic AC-OPF component, which is then decomposed into an individual deterministic AC-OPF for each scenario. Although it lacks theoretical convergence guarantees, the relevance of this approach is validated in practice.L'expansion des sources d'énergie renouvelable accroît le degré d'incertitude dans l'exploitation des réseaux de distribution d'électricité. La variabilité et l'intermittence inhérentes à ces énergies posent aussi d'importants défis aux gestionnaires de réseaux au niveau opérationnel. La gestion prévisionnelle doit ainsi évoluer pour intégrer des leviers de flexibilité, telles la modulation de puissance active et la gestion de puissance réactive. La décision relative à l'activation de ces leviers se traduit par un problème d'Optimal Power Flow. Cette thèse développe des algorithmes de résolution pour deux modèles stochastiques en courant alternatif (AC-OPF). Ces problèmes d'optimisation sont, à la fois, non-convexes, non-lisses et discrets. Cette thèse vise à appréhender ces complexités, sans recourir à la convexification des équations de flux de puissance,et en considérant l'interdépendance des incertitudes, via une contrainte probabiliste jointe ou une décomposition par scénarios dans le cas de leviers discrets.Précisément, la première méthodologie proposée s'applique à une version continue de l'AC-OPF sous contrainte probabiliste jointe. Une contribution de ce travail porte sur la conception d'une procédure numérique (oracle) traitant la contrainte probabiliste comme la différence de deux fonctions convexes. L'oracle est alors associé à une méthode de faisceaux pour les problèmes DoC (différence de convexes). Une seconde contribution porte sur le développement d'un nouvel algorithme de faisceaux offrant des garanties de convergence plus fortes sous des hypothèses plus faibles. Il produit ainsi un point critique (satisfaisant des conditions KKT généralisées) de l'AC-OPF probabiliste. Basé sur la méthode DoC précédente, cet algorithme exploite un programme maître différent, ainsi qu'une règle originale de mise à jour du paramètre proximal. Il s'applique à la classe générale des problèmes d'optimisation non-convexes et non-lisses dont objectif et contraintes sont modélisables comme différence de fonctions convexes et faiblement convexes (CwC). L'évaluation empirique de l'algorithme est menée sur différents problèmes non-convexes et stochastiques. Ses performances pratiques sont comparées à celles de la méthode DoC sur un cas d'étude de l'AC-OPF probabiliste dans un réseau de distribution à 33 nœuds.La seconde méthodologie proposée considère des règles discrètes en gestion prévisionnelle, telles que des règles de priorité et d'équité pour la modulation de puissance.L'expérimentation montre les limites de la méthode des faisceaux pour intégrer des variables entières. Comme alternative, il est proposé un modèle d'optimisation attachant une variable binaire par scénario, et maximisant le nombre de scénarios réalisés dans un budget limité. La dualisation des contraintes couplantes et la coordination par blocs permettent de séparer les règles discrètes de l'AC-OPF stochastique, qui se décompose, à son tour, en AC-OPF déterministes individuels par scénario. Si la convergence théorique n'est plus garantie par cette séparation, la pertinence pratique de l'approche est illustrée numériquement
Operating policies for robotic cellular warehousing systems
No full textInternational audienceRobotic Cellular Warehousing Systems provide an innovative robot-to-goods picking approach designed to improve robot transportation efficiency, where robots move to pick items and transport the picked items to workstations. In this study, we investigate the optimal operating policies for such a system by comparing two picking strategies (pick-while-sort and pick-then-sort) and three robot-to-workstation assignment rules (random, closest, and dedicated). Specifically, we develop dedicated closed queuing networks to model robot-to-goods picking and estimate warehouse throughput under different policies through single-class and multi-class models. The effectiveness of these analytical models is validated through numerical simulations, with an average gap of 5.53% between simulation and analytical results. Additionally, we conduct a series of numerical experiments to examine the impact of various factors on warehouse performance, including the numbers of robots and workstations, robot capacity, order size, and sorting efficiency. Based on the experimental findings, we provide managerial implications that offer insights into optimizing resource allocation and system configuration. These insights enable warehouse managers to improve operational efficiency and overall performance
Combinaison de techniques de bio-traitement passives ou semi-passives applicables aux drainages miniers
No full textInternational audienceLe projet COMPAS (APR-GESIPOL-2017-COMPAs) a visé à déterminer la combinaison optimale de traitements passifs ou semipassifs pour réduire l'impact environnemental des drainages miniers acides riches en fer, métaux lourds et arsenic. Les deux principales étapes de traitement biologique testées étaient la sulfato-réduction, permettant d'éliminer les métaux sous forme de sulfures et augmenter le pH de l'eau à une valeur proche de la neutralité et la bio-oxydation induisant la coprécipitation du fer (Fe) et de l'arsenic (As). Un traitement physico-chimique complémentaire utilisant des matériaux calcaires a également été évalué. Le projet a été conduit principalement sur le site minier de Carnoulès, connu pour ses importantes contaminations en Fe, As et zinc (Zn) avec des concentrations typiques respectivement de 1000, 100 et 20 mg/L. Une modélisation biogéochimique a été réalisée pour optimiser les conditions expérimentales. Les résultats des essais en laboratoire et sur site ont montré une efficacité prometteuse, mais des ajustements sont nécessaires pour une application industrielle. En parallèle, une évaluation économique et une analyse du cycle de vie (ACV) ont été menées pour estimer les coûts et l'impact environnemental des différents procédés de traitement proposés. Les conclusions indiquent que des essais à l'échelle démonstrateur sont indispensables pour valider les performances et affiner les paramètres de traitement.</div
Contributions à des méthodes de bases réduites pour l'inférence bayésienne non-paramétrique en géosciences
No full textThis thesis proposes advanced nonparametric Bayesian inference methods for inverse problems in geosciences. These methods enable uncertainty quantification when estimating fields from indirect observations. The standard Bayesian approach to such problems involves choosing a finite-dimensional field parametrization. The posterior distribution of this field is obtained by sampling the parameters, typically via a Markov Chain Monte Carlo method. The two main contributions of this thesis are the development of a general field parameterization and the construction of an adaptive dimension reduction method. The first contribution relies on Karhunen--Loève decompositions with uncertain hyperparameters. The inference problem is reformulated within a hierarchical Bayesian framework thanks to a change of measure, facilitating the sampling of the joint posterior distribution of KL coordinates and hyperparameters. This approach enables obtaining a more reliable estimation of posterior distribution uncertainties by exploring an enlarged prior set. The second contribution develops a gradient-free dimension reduction approach by identifying the field features that are informed by the observations. Two indicators determine these features: the posterior-to-prior covariance ratio and the correlation between the log-likelihood and parameters. The former is optimal in the linear Gaussian case and successfully applies to some nonlinear cases, while the latter becomes essential in more complex settings. These developments facilitate uncertainty quantification in high-dimensional and computationally intensive inverse problems. The efficiency of these methods is illustrated on different problems in geosciences, especially on traveltime tomography and groundwater problems.Cette thèse propose des méthodes avancées d'inférence bayésienne non-paramétrique pour les problèmes inverses en géosciences. Ces méthodes permettent une quantification des incertitudes lors de l'estimation de champs à partir d'observations indirectes. L'approche bayésienne classique nécessite de choisir une paramétrisation de dimension finie du champ. La distribution a posteriori de ce champ est obtenue en échantillonnant les paramètres, généralement par méthodes de Monte-Carlo par chaînes de Markov. Les deux contributions principales de cette thèse sont le développement d'une paramétrisation générale du champ et la construction d'une méthode adaptative de réduction de dimension. La première contribution s'appuie sur des décompositions de Karhunen-Loève (KL) ayant des hyperparamètres incertains. Le problème d'inférence est reformulé dans un cadre bayésien hiérarchique grâce à un changement de mesure, ce qui facilite l'échantillonnage de la distribution a posteriori conjointe des coordonnées KL et des hyperparamètres. Cette approche permet d'obtenir une estimation plus robuste des incertitudes en explorant un espace a priori plus riche. La seconde contribution développe une approche de réduction de dimension sans gradient en identifiant les caractéristiques du champ informées par les observations. Deux indicateurs déterminent ces caractéristiques : le rapport des covariances a posteriori et a priori, et la corrélation entre la log-vraisemblance et les paramètres. Le premier indicateur est optimal dans le cas linéaire gaussien et s'applique avec succès dans certains cas non linéaires, tandis que le second devient essentiel dans des configurations plus complexes. Ces développements facilitent la quantification des incertitudes dans les problèmes inverses de haute dimension coûteux en temps de calcul. L'efficacité des méthodes est illustrée sur différents problèmes en géosciences, notamment sur des applications en tomographie des temps de trajet et pour des écoulements en milieu poreux
Effect of Dissolved Oxygen and Temperature on Oxidation and Stress Corrosion Cracking of 316L Stainless Steel in Nuclear Primary Environment
No full textInternational audienceThe effects of dissolved oxygen and temperature on oxidation and stress corrosion cracking (SCC) were investigated by exposing 316L coupons and Slow Strain Rate Tensile Tests (SSRT) specimens in simulated primary water at 290°C and 320°C in both hydrogenated and oxygenated environments. A duplex oxide layer has been observed on all specimens, with a continuous inner layer and an outer layer composed of crystallites. Oxide thickness on coupons is maximum at 320°C in the range [290°C, 340°C] in hydrogenated environment, and is minimum at the same temperature in oxygenated environment. Intergranular oxidation penetrations were numerous in hydrogenated environment and are rare or absent in oxygenated environment. In SSRT, the material is less susceptible to SCC in oxygenated environment at 320°C. No SCC cracks were found at 290°C in hydrogenated environment. The susceptibility to both oxidation and SCC in this environment increases within the range [290°C, 340°C]
Légitimer, produire : le travail pour rendre tangible les nombres du CO2e
No full textInternational audienceSince March 2019, a group of French research staff has been engaged in efforts to quantify the greenhouse gas emissions of public research laboratories in France. The group’s efforts are based on a specific metric: the equivalent carbon dioxide produced (CO2e) or carbon footprint, which is quantified using the group's own software. CO2e quantification was not present, or only marginally, in public laboratories until the end of the 2010s. Two processes explain its diffusion. The first is “legitimization”, through a process of involvement in a scientific and political arena, of demarcation from existing methodologies by asserting an approach in terms of order of magnitude, and of negotiation, both with the CNRS and at the laboratory level. The second is “production”, combining the phases of conventions and measurement, winning allies, and adapting the calculator. These operations make CO2e tangible in the laboratories.Depuis mars 2019, un groupe de personnels de recherche français se mobilise pour quantifier les émissions de gaz à effet de serre des laboratoires publics de recherche en France. Ce groupe appuie son action sur un type de nombre : le dioxyde de carbone équivalent (CO2e) ou empreinte carbone, quantifié à l’aide de son propre logiciel. Or, la quantification en CO2e n’était pas présente, ou seulement de manière marginale, dans les laboratoires publics avant la fin des années 2010. Deux opérations expliquent sa diffusion. La première est la légitimation par un travail d’inscription dans un espace scientifique et politique, de démarcation méthodologique en revendiquant une approche en termes d’ordre de grandeur, et de négociation, aussi bien avec le CNRS qu’à l’échelle des laboratoires. La seconde est la production de ces nombres par un travail d’association des phases de conventionnement et de mesure, d’enrôlement d’allié·es et d’adaptation du calculateur. Ce sont ces opérations qui rendent tangible le CO2e dans les laboratoires
Investigating Delayed Rupture of Flow Diverter-Treated Giant Aneurysm Using Simulated Fluid–Structure Interactions
No full textInternational audienceGiant intracranial aneurysms are frequently treated shortly after discovery due to their increased risk of rupture and commonly symptomatic nature. Among available treatments, flow diverters are often the sole viable option, though they carry a rare but serious risk of delayed post-operative rupture. The underlying mechanisms of these ruptures remain unknown, due to the biomechanical complexity of giant aneurysms and challenges in replicating in vivo hemodynamic conditions within numerical simulation frameworks. This study presents a novel fluid–structure interaction simulation of a giant intracranial aneurysm treated with a flow diverter, based on high-resolution rotational angiography imaging. The resulting hemodynamics are compared to three established delayed-rupture hypotheses involving pressure rises, chaotic flow and autolysis. When considering wall compliance, the analysis reveals a consistent phase shift, dampening in pressure cycles, and an increased aneurysmal flow. These findings highlight the need for revisiting existing hypotheses and provide a foundation for advancing both computational modelling and clinical management strategies for giant intracranial aneurysms
The phase equilibrium behavior of icy bodies' surface mixtures: The simulation of the methane + acetonitrile system on Titan
No full textInternational audienc
Optimizing the Integration of Heat Pumps in Solvent Based Post-Combustion CO2 Capture Plants
No full textInternational audienceAlthough the chemical absorption method currently stands as the predominant approach for capturing CO2 in industrial processes, its broader implementation is hindered by the process’s energy penalties. This energy penalty mainly stems from the reboiler that uses steam, often produced from fossil fuels, to regenerate the solvent. This leads to additional CO2 generation, which will either decrease the capture rate or will require to increase the capture plant capacity to process the additional flue gas. In this context, technologies allowing to produce decarbonized steam suitable for solvent regeneration looks particularly appealing. Heat pumps are key technologies in this perspective, which allows to recover waste heat and upgrade it to useful heat in many applications. Recently, with the expansion of their operating temperature ranges and their technological enhancements, the integration of high-temperature heat pumps poses a promising solution that can address the energy challenges in carbon capture plants. In fact, this application has been investigated in the literature. For instance, Zhang et al. explored the efficiency of a Monoethanolamine (MEA)-based carbon capture plant by recovering low-grade waste heat using different thermodynamic units, including an absorption heat pump (Zhang et al., 2023). The technologies were mainly utilized through cascade variations models which led to a notable decrease in energy consumption. Another promising study conducted by Alabdulkarem et al. investigated seven heat pump configurations along with forty-one different working fluids yielding an overall enhancement in the plant’s efficiency (Alabdulkarem et al., 2015). The aforementioned studies, along with various others in the literature, achieved promising results; nevertheless, they highlighted the necessity for further optimization and the need for guidance in proper heat pump integration, laying the groundwork for future research endeavors.For carbon capture in an industrial environment, each use-case exhibits a unique combination of a CO2 capture unit characteristics with a given reboiler duty and temperature, a set of potentially valorizable heat sources within the CO2 capture scope, a CO2 conditioning section, and finally a set of waste heat sources from other nearby processes that could be advantageously utilized. This paper aims at illustrating, through the comparison of several heat pump implementation options, our current research aiming at developing a systematic methodology for highlighting themost relevant way to implement heat pumps for a given use-case.In this study, the focus is made on mechanical compression type of heat pumps. Both closed-cycle compression heat pumps and open-cycle steam compression heat pumps (also called mechanical vapor recompression, MVR) are examined by the authors. The study strives to go beyond the basic heat pump cycle consisting of a single-stage compressor, a condenser, an expansion valve, and an evaporator, in order to improve the industrial relevancy of the proposed solutions. In the course of this research, the CESAR1 Aspen Plus model, developed by the Carbon Capture Simulation Initiative (CCSI) (Morgan et al., 2022), was upscaled and used. The reboiler has a heat demand of 55 MW (135°C/3.1 bara steam), and three waste heat sources candidate for valorization have been selected: the flue gas pretreatment unit, the regenerator’s overhead condenser, and the CO₂ compression chain. Within the study, various heat pump architectures, including the pure refrigerant basic configuration, cascade heat pump configuration, and economizer heat pump, were modelled using Python. Different refrigerants, namely, Hydrofluro-Olefins (HFO), Hydrocarbons, and water (in MVRs) were considered. The use of water-based heat pumps is advantageous due to the benefit of enabling direct steam delivery to the reboiler. Finally, the possibility to implement a waterloop to harvest, transport and centralize the waste heat to a suitable location for implementing a large heat pump system has been also considered.The study on the use-case examined three key aspects. First, it investigated the optimal heat pump architecture by comparing MVR-only configurations to MVR coupled with an advanced HFO or hydrocarbon cycle. Additionally, five different refrigerants in the HFO/hydrocarbon cycles were evaluated to determine which minimized exergy destruction. The results demonstrated that coupling an MVR with an economizer heat pump (containing an internal heat exchanger) was the most efficient configuration and that cyclopentane was the most suited refrigerant.The second aspect of the study focused on maximizing the waste heat recovery potential of heat sources within the carbon capture plant. In particular, the CO₂ compression line was analyzed and assessed through various compression strategies, including wet and dry CO₂ compression, adjustments to the number of compression stages, and variations in intercooling. The specific conditions under which each strategy was most beneficial were identified.The final part of the study integrated all previous results: the cycles were optimized and integrated into the carbon capture process.The preliminary results obtained for a conventional CO2 compression train (dry compression) are summarized below:•Full Heat Coverage with MVR Alone:To cover 100% of the heat demand using only an MVR, a flash tank temperature as low as 65°C (0.25 bara) is required. While the achieved COP is promising (>4.4), this option seems to raise many challenges for an industrial implementation.•MVR with a 1 bara Flash Tank Limit:If a pressure limit of 1 bara is imposed on the flash tank, to avoid vacuum conditions in the MVR suction line, only 33% of the heat requirement can be supplied, with a COP > 8.0.•MVR Coupled with an Economizer Heat Pump:In this configuration, the MVR is coupled with closed-loop heat pump to supply 100% of the heat demand while avoiding an excessive vacuum operation. A waterloop is considered for simulating a case with a centralized heat pump. An economizer heat pump produces additional hot water to the flash tank, which operates at an optimized temperature of 85°C (0.6 bara) (see Figure 1). This setup results in an overall COP > 3.0.These initial results illustrate the numerous possible tradeoffs in the implementation of heat pumps to efficiently generate decarbonized steam for amine-based CO2 capture plants. It has been demonstrated that fulfilling the entire reboiler duty with heat pump systems, with the considered waste heat sources, is possible but raises some challenges. Also, the high COP of the second case suggest that a techno-economic optimum could be to only produce a portion of the reboiler duty with a MVR. The developed methodology is a powerful tool to systematically assess technicalsolutions for a given use-case and its available waste heat sources and provide valuable first insights about the impact of the selection of a given technological solution. Enhanced dry CO2 compression scheme, tailored to generate heat with heat pump systems, wet CO2 compression schemes enabling the optimal valorization of water latent heat as well as the availability of a waste heat source from a nearby process unit will be explored to provide additional insights about the potential of heat pumps to improve amine-based CO2 capture units
