Thesis title: Thermal-hydraulic study and optimization of the DEMO Water Cooled Lithium-Lead Breeding Blanket
One of the key components of a nuclear fusion power plant is the Breeding Blanket (BB), in charge of ensuring the essential functions of Tritium production, shield the Vacuum Vessel (VV) and remove the heat generated in the toroidal chamber. Two conceptual designs are currently being studied for the implementation in the DEMOnstration Fusion Reactor (DEMO) in the framework of R&D activities under the coordination of the EUROfusion Consortium. One of these two BB is the Water-Cooled Lithium-Lead (WCLL), which relies on two different fluid: the water, necessary to remove the generated heat in the tokamak and to shield the vacuum vessel from neutrons, and the Lithium-Lead (PbLi) eutectic alloy, adopted as breeder and neutron multiplier, necessary for the Tritium production in order to make the fusion self-sustaining. The first function is fulfilled by two independent cooling systems: the First Wall (FW), that facing the plasma removes the heat flux raised from it, and the Breeding Zone (BZ), that removes the deposited power due to neutron and photon interaction inside the breeder. To guarantee good energy conversion efficiencies, these two systems must operate under certain conditions, and pressurized water at the typical pressure of the nuclear Pressurized Water Reactor (PWR) is adopted.
The Ph.D. work has been developed in collaboration with ENEA Brasimone Research Center, under the coordination of the EUROfusion Consortium in the task of the Work Package Breeding Blanket.
The aim of this Ph.D. thesis is to contribute to the development of the conceptual design of the WCLL breeding blanket, in order to design an efficient and reliable system, demonstrating the capability to fully withstands the DEMO requirements in normal and off-normal conditions. The activity has been focused on the thermal-hydraulic of the system; specifically, the analyses were performed on one single elementary cell, that compose the WCLL due to its periodicity. To perform realistic analyses, multiple factors have taken into account: engineering aspects, neutronic, thermo-mechanic and magneto-hydrodynamic. This has been pursued through the engineering approach and with the application of the numerical CFD code to represent the behaviors of the different analyzed models.
The first part of this study (Chapter 3) starts from the WCLL design review, which is described in Section 3.1. This section concerns the previously studied and developed configurations of the WCLL. Four different configurations (T01.A, T01.B, T02 and T03) have been studied in a comparative analysis, evaluating the main advantages and issues, that have led to the development of the WCLL 2018 V0.2 configuration. The WCLL 2018 V0.2 is the starting point of this research activity. The elementary cell is fully described in Section 3.2, where all the components have been expounded, concluding the introductory part of this Ph.D. work.
Subsequently, in Chapter 4, different thermal-hydraulic analyses have been performed through numerical simulations, where a complete three-dimensional finite volume model of the WCLL elementary cell has been set-up in each analysis, using the commercial CFD code ANSYS CFX v18.2. Several steady-state analyses have been performed in order to optimize the BZ tubes layout, the FW cooling system, the BZ manifold layout and to evaluate the impact of the heat transfer modelling approach through the PbLi modelling and its properties. Once the CAD has been defined, the thermal power and the related cooling systems, FW and BZ, have been set through an analytical approach, in order to guarantee compliance with the main DEMO requirements design. The numerical model includes fluid and solid domains, representing in detail the WCLL elementary cell with its different structures and fluids. The Section 4.3 has the aim of optimizing the arrangement of the BZ pipes, guaranteeing a Eurofer temperature below the imposed limit of 550°C, and water at certain conditions in compliance with the thermodynamic cycle assumed for the electricity production. Several configurations have been analyzed to identify a promising BZ coolant system layout, which satisfies the DEMO requirements. The CFD analyses have been carried out investigating the temperature field of the solid structures, Eurofer and Tungsten, and also the thermal-hydraulic performances of the water-cooling systems and PbLi. These optimization analyses led to the V0.6 configuration, which has set the minimum number of BZ tubes to 22 and ensuring a symmetric temperature field in the toroidal direction in the BZ and FW systems and concerning the FW also in poloidal direction, a maximum temperature of the Eurofer structures of around 500°C, which not exceeds the imposed limit of 550°C.
In the next paragraph, the Section 4.4, a FW water channels optimization has been achieved, reducing the channels number from 10 up to 4. This channels reduction has been pursued thanks to the fact that the FW temperature was considerably below the imposed limit of 550°C, and to a DEMO's thermal load review that has led to a modification on the imposed heat flux on the FW from 0.5 MW/m2 to 0.32 MW/m2. The optimization has returned a greater homogeneity of the temperature field between BZ and FW systems, reducing the passive heat removal of the FW system from the BZ system. Although the pressure drops strongly increase, it has resulted in a decrease in the volume of water present in the first centimeters of the cell which positively affects the Tritium Breeding Ratio (TBR), the fundamental parameter for the Tritium production. The V0.6 configuration with 4 water channels (named V0.6_FW4), fully withstands the DEMO thermal loads not exceeding the Eurofer temperature limit of 550°C. Another important parameter that has been affected by this reduction is the FW water velocity that has been strongly increased, enhancing the water thermal-hydraulic performances.
In Section 4.5, the results from the analysis of the V0.6_FW4_R model have highlighted that the recirculation manifold have to be adopted to guarantee large and safety margins from a thermal crisis. The recirculation ensures a lower wall temperature in all the BZ tubes, especially in the tubes near the FW which are subjected to greater power deposition. In addition, a greater flow of water flow rate, guarantee a large and safety margin from the possibility of thermal crisis.
The second part of the activity, reported in Section 4.6, concerns the assessment of the assumption made in the numerical model development. Since the PbLi thermal conductivity deeply affects the elementary cell temperature field, due to its huge thermal inertia, a study was also conducted on it. Two different PbLi thermal conductivity have been chosen to evaluate the temperature field of the cell, the suggested Mogahed correlation and the conservative IAEA correlation. The analyses have also been performed with the PbLi domain set as liquid and solid, evaluating the impact of this assumption on the numerical prediction. It has been evaluated that the PbLi modelling approach does not affect the results, increasing the temperature field by only a few degrees in the model with solid PbLi. In addition, simulations with liquid PbLi, have demonstrated that the convective contribution to the heat exchange is almost irrelevant. The PbLi reaches, in both cases with forced convection and in the absence of buoyancy effects, velocities in the range of v0.1-0.15 mm/s, resulting in a laminar flow. Moreover, the Prandtl number has been analytically evaluated, returning a value close to 0.02. This means that the thermal diffusivity, which is related to the rate of heat transfer by conduction, unambiguously dominates prevailing on convection, that can be neglected. Instead, what significantly affects the model is the adopted PbLi thermal conductivity correlation, that with the IAEA causes the Eurofer temperature limit to be exceeded and the hot-spots onset. Due to the choice of setting as reference property, for the PbLi thermal conductivity, the IAEA correlation, the analyses on the updated configuration returns an Eurofer temperature field which exceeds the imposed Eurofer temperature limits in different areas.
The analyses, reported in Section 4.7, have led to a further set of simulations to find out a BZ tubes layout in order to guarantee the respects of the requirements, even in the worst conservative conditions. The second BZ optimization has carried out an alternative BZ tubes disposal, which has resulted in a configuration named WCLL 2018 V0.6_B. These modifications have not led to major changes in the BZ water domain, keeping almost unchanged the relevant thermal-hydraulic parameters, as velocity, pressure drops and temperature. However, they have involved the PbLi and structures temperature field reaching a Eurofer temperature slightly below the imposed limit of 550°C, extinguishing the structures hot-spots. The V0.6_B has been chosen as final configuration, able to cope with the operative DEMO thermal loads, guaranteeing the respect of the DEMO Eurofer temperature limit and its related requirements.
The last part of this work (Chapter 5) has been focused on the operational phases of the WCLL breeding blanket. One of the main functions of the BB is the conversions of the thermal energy from the fusion reaction in energy suitable for the power generation. The pulsed nature of the DEMO fusion reactor divides the operative phase in two main phases: pulse where the Deuterium and Tritium are burnt and energy is produced for 120 min, and dwell estimated to be 10 min where the central solenoid is recharging, and decay heat is removed. This research activity has been pursued to verify if the selected V0.6_B design is able to face the DEMO operational phases, investigating the response of the systems from the thermal-hydraulic point of view. In Section 5.1, a three-dimensional model has been reproduced according to the main outcomes of the previous Chapter. Different CFD transient analyses have been performed to verify if the selected design is suitable with the DEMO constrains and requirements, investigating different thermal behaviors. Transient thermal-hydraulic analyses have been set-up to simulate the burning phase, composed by ramp-up and ramp-down, after which steady state conditions of full power and dwell are reached respectively. Moreover, power fluctuation analyses have been performed to investigate the plasma instabilities caused by the pellets injection during the normal operation, this causes peaks of over or under power. In addition, an artificial analysis focused on the evaluation of the effect of the PbLi thermal inertia has been performed.
Unfortunately, the transient analyses require a significantly higher computational cost, and this implies that in order to obtain results in a reasonable time, it is mandatory to reduce the number of elements in the numerical model. Although it is interesting to obtain extremely accurate output values of the model, the main goal of these analyses is to investigate the global performances of the cooling systems, therefore, in Section 5.2 a mesh sensitivity has been performed in order to reduce the high number of elements without losing accuracy in the obtained results.
To perform the transient analyses different time-dependent thermal loads have been considered. The operational phases of DEMO have been characterized by different power contributions which vary in space and time. These thermal loads have been widely discussed in Section 5.3, identifying the power contribution curves to adopt for the model. In particular, in Section 5.4, the model solver settings and boundary conditions have been set-up per each run.
In Section 5.5, the results have been described and differentiated by steady-state and transient analyses. The first part (Section 5.5.1) concerns the steady-state analyses performed to impose the initial conditions for the transient, and in the second part (Section 5.5.2), the transient analyses have been widely discussed. In particular, in sub-sections 5.5.2.1 and 5.5.2.2, the fast ramp-up and ramp-down phases are respectively analyzed. Both show how the PbLi thermal inertia plays a key role in the rise and fall of the temperature of the structures, ensuring to the systems a slowly gradual trend. The FW system promptly reacts to the power variation, showing a temperature trend similar to the power one. Instead, the BZ system slowly reacts to power variations. It depends from the PbLi thermal inertia, which strongly slows down the temperature trend, causing a considerable delay in reaching steady-state conditions. In both simulations, the Eurofer temperature is below the limit and the operative water constrains are respected and guaranteed.
The study of the power fluctuations has been reported in sub-sections 5.5.2.3 and 5.5.2.4, where the V0.6_B layout has been subjected to different power oscillation. These analyses have been performed for the purpose of studying if the PbLi thermal inertia can mitigate these oscillations, continuing to operate within the required requirements. The thermal-hydraulic goal is that the elementary cell continues to supply water to the Primary Heat Transfer System (PHTS) at the required conditions for the electricity production. In addition, it has been highlighted that, in the remote case of an overpower peaks series, the huge PbLi thermal inertia absorbs these oscillations not provoking a temperature build-up. The same cannot be stated regarding the FW systems where a slight temperature build-up has been evidenced.
In sub-section 5.5.2.5, an artificial study has been performed to evaluate the PbLi thermal inertia. The system has been subjected to a step-down ramp, passing from the nominal power to a zero-power condition. The analysis is aimed at identifying the characteristic time trend of both systems. A strong cooling time differentiation, between FW and BZ system, has been evinced, where the cooling time of the PbLi has different order of magnitude compared with the FW.
In conclusions, different numerical simulations have been performed to fully evaluate the thermal behavior of the WCLL elementary cell, within steady-state and transient analyses. The first part of the Ph.D. research activity has focused in the development of a WCLL elementary cell design that satisfies the DEMO requirements, optimizing the BZ tubes and the FW water channels. The analyses have led to choose a final design the V0.6_B layout, since it withstands the normal operative thermal loads. The second part, instead, has demonstrated that the selected design is able to withstand all the operational phases of the DEMO fusion reactor, confirming the choice made in the first part. The final design has been selected to comply with the DEMO requirements, withstanding the thermal loads, and guaranteeing adequate water conditions for the electricity production during all the reactor phases.