Search

The energy and power demands implied by start-up and attainment of operating conditions for a pulsed DEMO have been explored. The information gained has been used to establish the requirements both for transformer recharge and for an energy storage system designed to ensure the plant effectively maintains an uninterrupted 1 GW supply to the grid. It was determined that a minimum of approximately 150 MW of auxiliary heating would be required to access the desired operating conditions for the burn phase, and the current-drive potential of a heating system of this capacity was assessed to determine how the pulse length could be extended beyond purely inductive operation. The overall peak power and energy demand associated with the start-up phase of each pulse of output power was determined for a range of dwell times. These two quantities determine the costs of the subsystems required for the power provision during the transformer recharge phase and the energy storage needed for constant plant output. We are therefore able to produce a simple estimate of such costs as a function of dwell time. For falling dwell times, costs are found to rise sharply below around 500 s, but they remain relatively flat as dwell times increase beyond around 1000 s, over the range considered.

Following earlier publication of a model of crack flow and filtering based on a 2-D crack representation, the model has now been improved and submitted to validation by comparison with data from a number of other theoretical and experimental studies. This process has led to improved understanding of gas flow and filtering in real crack geometry and reinforced expectations that the model predictions of barrier filtering capacity will be conservative for realistic containment conditions. Our analysis also indicates that, because of drag effects related to the existence of reduced aperture sites, the viscosity limited flow cannot be approximated by the plane Poiseuille model. As a result, theoretical models employing this assumption over-predict measured flow rates by around an order of magnitude. The narrowing of the flow stream at these reduced aperture sites is also found to be a factor which significantly boosts the filtering of particles. As a result of these investigations, there is a strong possibility that we have now identified all essential elements required for a detailed description of aerosol transport through cracks via the laminar flow of a non-condensible carrier gas.

Although fusion power is being developed because of its large resource base, low environmental impact and high levels of intrinsic safety, it is also important to investigate the economics of a future fusion power plant in order to assess the potential market for the electricity produced. As part of the PPCS (Power Plant Conceptual Study) in Europe, published in 2005, an assessment was made of the likely economic performance of the range of fusion power plant concepts studied. Since that time, new work has been carried out, within the fusion programme, and particularly in the EU DEMO study, that changes a number of the important assumptions made in the PPCS. These changes allow either reduced cost versions of the PPCS plant models or, alternatively, plants with less ambitious technical assumptions at constant cost. The impact of the new results, emerging from the EU DEMO studies, on the role of fusion in the future energy market is described. A new energy economics model is employed to analyse the potential market performance of fusion power in a range of future energy scenarios and this shows that there can be a significant role for fusion in a future energy market.

A model for the prediction of the particulate filtering capacity of containment barriers has been developed, and equations are derived which describe gas flow in cracks, as well as the inertial impaction of particles onto crack walls. Some specific containment options being considered for use with fusion power plants have been investigated, using these methods, providing insights into the impact of design options on the filtering performance of containment barriers. In particular, the modelling indicates that, for some hypothetical scenarios, a bare concrete barrier is effective in filtering particulates, but also that the addition of a steel liner, to reduce the gas flow rate, may also reduce the instantaneous barrier filtering efficiency.

An analysis has been made of the IFMIF lithium loop activation using the time-dependent, activation, transport and deposition code, TRACT (TRansport of ACTivation). Corresponding operator dose rates have also been derived using the neutron and photon transport Monte Carlo code MCNP-4C, enabling an assessment of operator exposure, both during IFMIF operation and for times up to 1 year after operation has halted. The dose rates were obtained for specific configurations of loop geometry. Where comparisons can be made, these results agree very well with those of an earlier preliminary investigation.

A ‘worst case’ assessment of the radiologically significant environmental source term, based on bounding accident analysis of a single blanket cooling loop loss of cooling accident (LOCA), has been performed for Fusion Plant Models A and B of the Power Plant Conceptual Study (PPCS). This paper describes the calculations of ensuing material mobilisation, transport in the containment, release to the environment and consequential dose to the public. Also reported is a comparison with more detailed dose calculations by Forschungszentrum Karlsruhe (FZK), for the same scenario, using real weather data. The results show that the calculated doses are comparable with, or slightly above, typical annual doses from natural causes.

Cryogenic pellet injection has been used in START and it is under commissioning for MAST for fuelling and diagnostics. In START NBI, heated plasmas fuelled with deuterium pellets, high normalized density (Greenwald number NG≤1.1) with good energy confinement time (τE/τEelmfITER97≤1.3) both at high beta (βT≤25%, βN≤4.9) have been obtained for time scales close to the energy confinement time (∼τE). In addition, high fuelling efficiency (∼100%) have also been simultaneously observed. A complete pellet injection programme is currently under development in MAST, including the commissioning of a eight pellet injector, technical aspects of launching pellets through guide tubes, dedicated diagnostics, and modelling.

Bounding calculations of thermal transients following a hypothetical (Category V) loss of coolant accident in two international thermonuclear experimental reactor (ITER) design options, EU-I and ITER-fusion energy advanced tokamak (FEAT), have been carried out for a variety of assumptions concerning heat-transfer to and heat rejection from, the cryostat. It is shown that, even when unrealistically conservative assumptions are made in illustrative calculations, the temperatures do not approach melting point for any part of the plant structure. In all cases, maximum temperatures are reached within 1 year. It is also concluded that for the baseline scenario for ITER-FEAT in which helium gas is assumed to be present in the cryostat, convectively transferring heat between components, the outboard first-wall temperature never exceeds 530°C. A significant improvement could still be made if the surface emissivity of the thermal shields could be increased by some means during the post-accident period.

We report here improvements to a code, FUSCON, developed for the SEAFP-2 (Safety and Environmental Aspects of Fusion Power) study to model containment transport and release to the environment. The previous version of FUSCON did not include heat transfer to the wall, accounted for condensation for only a limited set of conditions, and hence placed restrictions on some of the containment parameters investigated. The improvements described involve modelling of post-accident convective transport in the containment system. This is a good basis for enhancement of the model to include mass transfer effects associated with condensation. Three-dimensional, computational-fluid-dynamics, finite-element calculations have guided the derivation of a computationally economic model of convective containment transport. The efficiency of the model will ensure that FUSCON is still sufficiently fast-running to be used to study confinement system optimisation and selection.

Analyses have been performed of the potential consequences to the public of hypothetical loss-of-coolant accidents in conceptual fusion power plant designs. In order to establish upper bounds to the consequences of such events, a case has been studied in which total loss of all active cooling has been assumed, with no remedial intervention for the duration of the accident sequence. The analyses are based on three conceptual power plant designs, two of them similar to those assumed in the earlier safety and environmental assessment of fusion power (SEAFP) study (Raeder et al., 1995), with updating of assumed structural materials. The three models studied provide a broad range of design options. In all cases the decay-heat driven temperature transients are well below the level at which structural melting would begin. Based on conservative assumptions, mobilisation, release and dose calculations show that potential maximum doses to the public are very far below the levels at which evacuation might be considered.

It is found for conditions relative to fusion safety studies that there is a range of initial aerosol masses (up to ≈ 1000 kg) within which the aerosol particle growth dynamics are dominated by Brownian agglomeration. Above this range, the dominant growth mechanism is gravitational agglomeration. Previously, monodisperse aerosol modelling has been reported (W.E. Han, 19th Symposium on Fusion Teechnology, Lisbon, Portugal, 15–20 September, 1996) which enables the accurate prediction of aerosol behaviour up to initial masses of ≈ 1000 kg. Here monodisperse modelling is extended to deal with initial aerosol masses above 1000 kg, enabling the prediction of aerosol transport and the release to the environment in the gravitationally dominated agglomeration regime for hypothetical accidents involving mobilised radionuclides. It is shown that increasing the initial aerosol mass in the Brownian dominated agglomeration regime results in increased release to the environment. However, increases in initial mass in the gravitationally dominated regime soon lead to a maximum in release to the environment; as a result, further increases in initial mass produce smaller releases to the environment.

Calculations have been made of the temperature evolution of the structures in the outboard zone and of the divertor of the standard and low-activation versions of the EEF (Environmental, Safety-Related, and Economic Potential of Fusion Power) Reference Reactor, following hypothesised loss-of-coolant events. For the standard version, temperatures of the first wall and blanket are found to rise only very slowly (after an initial adjustment) and for the low-activation version never exceed about 850 K. Although the results are specific to the particular materials and reactor power density used in this study, they augur well for the development of a purely passive approach to fusion safety.