Auswahl der wissenschaftlichen Literatur zum Thema „Single medium thermocline“

Efficient and reliable thermal storage is an important requirement for substituting conventional power systems with solar thermal facilities. Storage will synchronize the intermittent supply of solar radiation with the usually constant demand of technical thermal processes. The distributed collector system of the IEA/SSPS Project in Almeri´a (Spain) uses two different storage systems for the 100 to 295°C temperature range: a single thermocline vessel and a dual medium storage tank (DMST). In the first tank, thermal oil is used as the energy carrier as well as for energy storage; in the dual medium tank, the storage medium is cast iron, and the oil acts primarily as a heat transfer fluid. At the SSPS, this concept’s potential for future process heat applications has been assessed. Performance and operational restrictions of the DMST were systematically studied over a wide range of temperatures, and an existing simulation model was verified and adapted at the same time. The thermodynamic model of the DMST is presented and compared with the first results of the 1985 test program.

A molten-salt thermocline tank is a low-cost option for thermal energy storage (TES) in concentrating solar power (CSP) plants. Typical dual-media thermocline (DMT) tanks contain molten salt and a filler material that provides sensible heat capacity at reduced cost. However, conventional quartzite rock filler introduces the potential for thermomechanical failure by successive thermal ratcheting of the tank wall under cyclical operation. To avoid this potential mode of failure, the tank may be operated as a single-medium thermocline (SMT) tank containing solely molten salt. However, in the absence of filler material to dampen tank-scale convection eddies, internal mixing can reduce the quality of the stored thermal energy. To assess the relative merits of these two approaches, the operation of DMT and SMT tanks is simulated under different periodic charge/discharge cycles and tank wall boundary conditions to compare the performance with and without a filler material. For all conditions assessed, both thermocline tank designs have excellent thermal storage performance, although marginally higher first- and second-law efficiencies are predicted for the SMT tank. While heat loss through the tank wall to the ambient induces internal flow nonuniformities in the SMT design over the scale of the entire tank, strong stratification maintains separation of the hot and cold regions by a narrow thermocline; thermocline growth is limited by the low thermal diffusivity of the molten salt. Heat transport and flow phenomena inside the DMT tank, on the other hand, are governed to a great extent by thermal diffusion, which causes elongation of the thermocline. Both tanks are highly resistant to performance loss over periods of static operation, and the deleterious effects of dwell time are limited in both tank designs.

This paper presents the modeling theory and results of an innovative thermal energy storage (TES) facility, ideated, realized, and tested by ENEA (Italy). This prototype enabled the thermocline storage with molten salts in a novel geometry ideated for small-medium scale decentralized solutions, which includes two vertical channels to force the circulation through two heat exchangers, respectively, and realized for charging and discharging phases (in a single tank). A thermophysical model was built and tested properly for this particular geometry in order to analyze the temperature distribution along the radius. The numerical results well reproduced the experimental values. Furthermore, the analytical solution provided a short-cut methodology able to evaluate the thermocline distribution (along the vertical axis) depending on both the time and the radius values. Hence, the influence of the radial position (r) on the thermocline degradation was studied finding that, at the edges (r → 1), the thermocline remains unchanged for longer (around ten times more) than at the center of the tank (r → 0). The obtained numerical modeling and the analytical correlation can be useful for the process analysis to scale-up the thermal storage system and to evaluate the system reliability for industrial plants.

The growing interest in large-scale solar power production has led to a renewed exploration of thermal storage technologies. In a thermocline storage system, heat transfer fluid (HTF) from the collection field is simultaneously stored at both excited and dead thermal states inside a single tank by exploiting buoyancy forces. A granulated porous medium included in the tank provides additional thermal mass for storage and reduces the volume of HTF required. While the thermocline tank offers a low-cost storage option, thermal ratcheting of the tank wall (generated by reorientation of the granular material from continuous thermal cycling) poses a significant design concern. A comprehensive simulation of the 170 MWht thermocline tank used in conjunction with the Solar One pilot plant is performed with a multidimensional two-temperature computational fluid dynamics model to investigate ratcheting potential. In operation from 1982 to 1986, this tank was subject to extensive instrumentation, including multiple strain gages along the tank wall to monitor hoop stress. Temperature profiles along the wall material are extracted from the simulation results to compute hoop stress via finite element models and compared with the original gage data. While the strain gages experienced large uncertainty, the maximum predicted hoop stress agrees to within 6.8% of the maximum stress recorded by the most reliable strain gages.

The overall efficiency of a concentrating solar power (CSP) plant depends on the effectiveness of thermal energy storage (TES) system (Kearney and Herrmann, 2002, “Assessment of a Molten Salt Heat Transfer Fluid,” ASME). A single tank TES system consists of a thermocline region which produces the temperature gradient between hot and cold storage fluid by density difference (Energy Efficiency and Renewable Energy, http://www.eere.energy.gov/basics/renewable_energy/thermal_storage.html). Preservation of this thermocline region in the tank during charging and discharging cycles depends on the uniformity of the velocity profile at any horizontal plane. Our objective is to maximize the uniformity of the velocity distribution using a pipe-network distributor by varying the number of holes, distance between the holes, position of the holes and number of distributor pipes. For simplicity, we consider that the diameter of the inlet, main pipe, the distributor pipes and the height and the width of the tank are constant. We use Hitec® molten salt as the storage medium and the commercial software Gambit 2.4.6 and Fluent 6.3 for the computational analysis. We analyze the standard deviation in the velocity field and compare the deviations at different positions of the tank height for different configurations. Since the distance of the holes from the inlet and their respective arrangements affects the flow distribution throughout the tank; we investigate the impacts of rearranging the holes position on flow distribution. Impact of the number of holes and distributor pipes are also analyzed. We analyze our findings to determine a configuration for the best case scenario.

Le stockage d'énergie est essentiel à la transition énergétique car il permet de découpler la production de l'énergie de sa consommation. Le stockage de chaleur thermocline en eau, utilisé dans les réseaux de chaleur à moyenne ou basse température, repose sur la stratification thermique dans une cuve. Les performances de ce type de stockage sont fortement liées à la bonne stratification du fluide qui peut être perturbée par l'injection et le soutirage du liquide, des aspects peu explorés dans la littérature.L'objectif de cette thèse est de modéliser un tel stockage de manière fiable pour analyser la distribution du fluide. En effet, le but est de mieux appréhender les phénomènes physiques gouvernant la thermocline pendant les cycles de fonctionnement et d'accroître ses performances énergétiques par un design ou un pilotage amélioré. Pour ce faire, des études numériques utilisant la CFD (Computational Fluid Dynamics) ont été réalisées et comparées à des données expérimentales disponibles dans la littérature et obtenues via une nouvelle section d'essais.Dans un premier temps, un modèle CFD a été développé basé sur un cas expérimental existant de la littérature. Dans un stockage thermocline en eau, il y a bien souvent coexistence entre une région laminaire dans la cuve et turbulente à proximité des distributeurs. Cette coexistence est un enjeu majeur de la modélisation car la plupart des modèles de turbulence ne sont pas capables de représenter fiablement la transition d'un écoulement turbulent vers laminaire. Pour ces travaux, une méthode statistique RANS (Reynolds Average Numerical Simulation) est adoptée et le modèle k-omega-SST est sélectionnée car il permet de représenter les écoulements en proche paroi. Concernant la flottabilité, il existe deux méthodes pour considérer ses effets : utiliser une masse volumique variable dans l'ensemble des équations, ou constante sauf dans le terme de flottabilité . Cette dernière est connue sous le nom de l'approximation de Boussinesq mais n'est valable que sur une faible gamme de ΔT. La précision de l'approximation de Boussinesq a été remise en question et une approche au second ordre de ce modèle est employée. Celle-ci permet d'obtenir le même terme de flottabilité qu'un modèle à masse volumique variable mais avec un temps de calcul réduit de moitié. La comparaison avec des données expérimentales a permis de souligner l'impact de l'état initial en température (stockage stratifié ou homogène). Une étude exploratoire de l'impact d'une injection progressive selon une rampe en débit a montré son impact sur la réduction de l'épaisseur de la thermocline au moment de sa création.Dans une démarche de validation du modèle et de vérification des observations numériques, un nouveau dispositif expérimental a été conçu. Celui-ci mesure la température grâce à 300 thermocouples disposés dans la cuve et permet un contrôle précis des conditions opératoires. Des études en phase statiques pour évaluer les pertes thermiques ont été réalisées. Des études dynamiques ont permis de faire varier les paramètres opératoires pertinents : la vitesse de propagation axiale, l'écart de température, le dispositif de soutirage ou encore l'injection progressive. Pour ce système, les résultats montrent qu'il est possible d'obtenir une stratification à forte vitesse (> 2 mm/s) tant que le ΔT est suffisamment élevé.Enfin, l'écoulement dans la section d'essais a été étudié numériquement avec un modèle CFD actualisé. Les champs de variables ont montré que les résultats numériques et expérimentaux sont cohérents, en particulier lors de la formation de la thermocline. Toutefois, un excès de diffusion lors de la propagation du gradient thermique à faible débit est notable. Pour tous les essais réalisés les écarts expérimentaux et numériques ont été quantifiés: à l'exception des conditions critiques, l'écart sur l'épaisseur de thermocline est de ±50% et se situe entre 0 et -10% pour le taux de restitution
Energy storage is essential to the energy transition as it allows decoupling energy production from its consumption. Water-based thermocline heat storage, used in medium or low-temperature heating networks, relies on thermal stratification in a tank. The performance of this type of storage is strongly linked to the proper stratification of the fluid, which can be disrupted by the injection and extraction of the liquid, aspects that are scarcely explored in the literature.The objective of this thesis is to reliably model such storage to analyze the fluid distribution. The aim is to better understand the physical phenomena governing the thermocline during operating cycles and to enhance its energy performance through improved design or control. To achieve this, numerical studies using CFD (Computational Fluid Dynamics) were conducted and compared with experimental data available in the literature and obtained via a new experimental setup.Initially, a CFD model was developed based on an existing experimental case from the literature. In water thermocline storage, there is often coexistence between a laminar region in the tank and a turbulent region near the distributors. This coexistence is a major challenge in modeling because most turbulence models cannot reliably represent the transition from turbulent to laminar flow. For this work, a RANS (Reynolds Average Numerical Simulation) statistical method is adopted, and the k-omega-SST model is selected as it can represent near-wall flows. Regarding buoyancy, there are two methods to consider its effects: using a variable density in all equations or a constant density except in the buoyancy term. The latter is known as the Boussinesq approximation but is only valid over a narrow range of ΔT. The accuracy of the Boussinesq approximation has been questioned, and a second-order approach of this model is employed. This allows obtaining the same buoyancy term as a variable density model but with a calculation time reduced by half. Comparison with experimental data highlighted the impact of the initial temperature state (stratified or homogeneous storage). An exploratory study of the impact of progressive injection according to a flow ramp showed its effect on reducing the thermocline thickness at the time of its creation.As part of the model validation and verification of numerical observations, a new experimental setup was designed. It measures the temperature using 300 thermocouples placed in the tank and allows precise control of operating conditions. Static phase studies to evaluate thermal losses were conducted. Dynamic studies allowed varying relevant operating parameters: axial propagation speed, temperature difference, extraction device, and progressive injection. For this system, the results show that it is possible to obtain stratification at high speed (> 2 mm/s) as long as the ΔT is sufficiently high.Finally, the flow in the test section was numerically studied with an updated CFD model. The variable fields showed that the numerical and experimental results are consistent, especially during the formation of the thermocline. However, excessive diffusion during the propagation of the thermal gradient at low flow is notable. For all the tests carried out, the experimental and numerical discrepancies were quantified: except for critical conditions, the discrepancy in thermocline thickness is ±50% and ranges from 0 to -10% for the restitution rate

Conversion of direct solar energy, in particular the Concentrated Solar Power (CSP) technologies, has a significant role on conventional energy cost and efficiency. A single tank thermocline Thermal Energy Storage (TES) system is accountable for the overall efficiency of this conversion system. A single tank TES system has a thermocline region that produces the temperature gradient between hot and cold storage fluid by density difference. The overall energy storage capacity depends on sustaining of this region at uniform manner. This paper analyzes how the difference in the percentage of porous medium influences the effectiveness of the flow-distribution and hence, the overall performance of the TES system. The effectiveness is assessed by the optimal flow distribution. The optimal distribution is obtained by examining the velocity profile at any horizontal plane. This plane should be uniform for sustaining the thermocline region during the operation period. To achieve a uniform velocity distribution, two symmetric perforated plate flow distributors were placed in the tank. The distributors were positioned near the inlet and outlet, and checked the performance by varying the percentage of porous medium since the distribution is influenced by the porosity. Porous distributors with hexagonal shape pore were considered and Hitec® molten salt was used as a heat transfer fluid. These respective percentages of porosity affect the flow distribution throughout the tank during the flow distribution. The standard deviations of the velocity field at different positions along z-plane and thermal diffusivity were analyzed. The analyses of our cases were done to distinguish a configuration for the minimum thermal diffusivity and velocity deviation from the mean flow. A finite volume based computational fluid dynamics software was used to execute the computational analysis.

The growing interest in large-scale solar power production has led to a renewed exploration of thermal storage technologies. In a thermocline storage system, heat transfer fluid (HTF) from the collection field is simultaneously stored at both excited and dead thermal states inside a single tank. A granulated porous medium included in the tank provides thermal mass for storage and reduces the amount of HTF volume required. While the thermocline offers a low-cost storage option, thermal ratcheting of the tank wall (generated by filler material reorientation from continuous thermal cycling) poses a significant design concern. A comprehensive simulation of the 170 MWht thermocline tank used in conjunction with the Solar One pilot plant is performed with a multi-dimensional two-temperature computational fluid dynamics model. In operation from 1982 to 1986, this tank was subject to extensive instrumentation, including multiple strain gages along the tank wall to monitor hoop stress. Temperature profiles along the wall material are extracted from the simulation results to compute hoop stress via finite element models and compared with the original gage data. While the strain gages experienced large uncertainty, the stresses computed from the simulation agree reasonably well with the experimental measurements. The maximum predicted hoop stress agrees to within 6.8% of the maximum stress recorded by the most reliable strain gages.

The overall efficiency of a Concentrating Solar Power (CSP) plant depends on the effectiveness of Thermal Energy Storage (TES) system [1]. A single tank TES system consists of a thermocline region which produces the temperature gradient between hot and cold storage fluid by density difference [2]. Preservation of this thermocline region in the tank during charging and discharging cycles depends on the uniformity of the velocity profile at any horizontal plane. Our objective is to maximize the uniformity of the velocity distribution using a pipe-network distributor by varying the number of holes, distance between the holes, position of the holes and number of distributor pipes. For simplicity, we consider that the diameter of the inlet, main pipe, the distributor pipes and the height and the width of the tank are constant. We use Hitec® molten salt as the storage medium and the commercial software Gambit 2.4.6 and Fluent 6.3 for the computational analysis. We analyze the standard deviation in the velocity field and compare the deviations at different positions of the tank height for different configurations. Since, the distance of the holes from the inlet and their respective arrangements affects the flow distribution throughout the tank; we investigate the impacts of rearranging the holes position on flow distribution. Impact of the number of holes and distributor pipes are also analyzed. We analyze our findings to determine a configuration for the best case scenario.

Due to their relatively high capital and environmental cost of two-tank molten salt thermal storage systems, a significant amount of research has gone into looking for sensible and latent thermal energy storage alternatives suitable for concentrated solar thermal (CST) plants. Despite a large number of developments in the last decade, comparative studies among promising options have been lacking. In particular, only a few comparative studies are available in which thermal energy storage (TES) systems are integrated as an active subcomponent of CST plant. Therefore, this study compares selected sensible heat thermal energy storage systems based on their integrated performance with other CST components (e.g. a tower -based CST plant with a Rankine cycle) over a year of operation. In the present study, annual performances of single-medium thermocline (SMT), double-medium thermocline (DMT), and shell-and-tube (ST) system were compared with that of a conventional two-tank molten salt storage system. Concrete with porosity of 0.2 (concrete occupies 80% of the system) was selected as a low cost filler material in the DMT and ST systems. The systems were sized for 15 hours of storage capacity and integrated into a validated 19.9 MWe Gemasolar power plant model with solar multiple of 2.5. Before performing annual integrated simulations, an optimum design of each storage system was selected based on a performance analysis of the storage system over a constant 15 hours discharge. A CST plant with a two-tank molten salt system enables the highest amount of electricity generation in a year followed by the SMT and DMT systems, which resulted in 7% and 9% less electricity generation, respectively. For the CST plant with ST system, 20% less electricity was generated over a year. Overall, this study provides a methodology for the comparison of the TES alternatives, and it gives insight the most promising alternative for replacing two-tank molten salt systems.