Experiments for the hydrogen combustion aspects of ITER LOVA scenarios

https://doi.org/10.1016/j.fusengdes.2011.01.050Get rights and content

Abstract

This work addresses the hydrogen safety issue of an International Thermonuclear Experimental Reactor (ITER) in case of a loss of vacuum accident (LOVA) scenario. In this scenario, accidentally generated hydrogen mixed with injected air could be ignited at reduced, sub-atmospheric pressures. The general question is whether the induced combustion pressure could exceed the ITER vacuum vessel design pressure. The paper presents the results of large scale dynamic experiments on hydrogen ignition and combustion at reduced pressures in the presence of a turbulent air jet injected into the hydrogen atmosphere. Experiments have been performed in a cylindrical vessel with a volume of 8.8 m3 filled with hydrogen at an initial pressure of 200 mbar. The orifice size of 6 mm i.d. was chosen to model in the same time scale the real leak through a 100 mm × 100 mm area into the ITER vacuum vessel with a volume of about 3000 m3. The structure and dynamics of the air jet into the hydrogen atmosphere at different initial pressures were investigated using microscopic liquid droplets as markers. During the combustion tests, it was found that more distant ignition positions and stronger ignition energies lead to maximum combustion pressures that are lower than the design pressure of the ITER vacuum vessel.

Introduction

One of the ITER scenarios with water leakage and the following interaction with hot beryllium dust in the vacuum vessel (VV) after an air ingress may lead to the formation of a combustible hydrogen–air–steam mixture [1]. The GASFLOW analysis showed the possibility of formation of hydrogen–air mixture, which is capable for effective flame acceleration, and even detonation. In order to prevent such scenarios, a combustion mitigation system based on nitrogen injection into the vacuum vessel (VV) and a hydrogen recombination system was proposed [1], [2]. It was shown [3] that such mitigation systems will significantly reduce the risks of a hydrogen detonation and fast deflagration in the wet bypass scenario. However, the atmosphere in the ITER sub-volumes with partial pressure of 200 mbar of hydrogen cannot be completely inerted, especially at the early stage of the scenario. Slow or fast deflagrations could still generate relatively high pressures above 2 bar and threaten the integrity of the VV.

Contrary to the nitrogen-dilution mitigation system, an alternative method based on pulsating spark igniters that prevent a gathering of significant amounts of burnable hydrogen–air mixture was currently proposed. Because the maximum combustion pressure depends on the mixture reactivity and proportional to the initial pressure, the “early” ignition of a hydrogen–air mixture in the VV was organized in the presence of an air jet in hydrogen atmosphere. It reproduces the real mixing procedure and level of turbulence. Since the mixture properties during the air injection were changed from 200 mbar of pure hydrogen (100% H2) to 1 bar of 20% of hydrogen in air, the mixture reactivity and ignition conditions were changed from the Upper Flammability Limit (UFL) with lower mixture reactivity to a practically stoichiometric hydrogen–air mixture with very high reactivity. The idea was to ignite the mixture in the VV as early as possible at the low reactivity level to prevent catastrophic pressure loads.

The main objective of this experimental study was to investigate the operation of the alternative mitigation system and to perform measurements of maximum pressure loads and flame velocities of the combustion process as function of the ignition energy and ignition position in order to evaluate an optimum regime for the igniters leading to lower pressure loads and slower flame velocity. The dynamics of air ingress, effect of scale and combustion of hydrogen–air mixtures at sub-atmospheric pressures in VV were also experimentally investigated.

Section snippets

Flammability limits

The normal operating pressure of the ITER VV is sub-atmospheric. This was the reason why accident scenarios consider formation of hydrogen–air mixtures at sub-atmospheric pressures. The flammability properties of such mixtures are well-known at atmospheric and elevated pressures as well [4], [5], [6], [7]. Only one paper [8] is related to the flammability of hydrogen–air mixtures at sub-atmospheric pressures and ambient temperature. The lowest ignition pressure of 150 mbar was achieved in this

Experimental details

A schematic diagram of the experimental facility for the large scale hydrogen combustion experiments in the presence of a turbulent air jet is shown in Fig. 3. It consists of a thick wall (100 mm) high pressure chamber with a volume of 8.8 m3, an internal diameter of 1.8 m, and a length of 3.7 m. The volume has a nozzle (d = 4, 6, 10 mm i.d.) located at the top in the middle position. The volume is equipped with pressure sensors for pressure measurements and several windows for the optical observation

Conclusions

Air injection into hydrogen atmosphere, flammability limits and flame velocities of hydrogen–air mixtures at ITER relevant sub-atmospheric pressures were studied.

It was shown that an early ignition could not be realized at reduced pressure due to the highly turbulent air jet and the very fast mixing procedure. The ignition takes place at the initial pressure when average hydrogen concentration is below the Upper Flammability Limit (75% H2). This corresponds to a pressure of more than 275 mbar,

Acknowledgments

This work was done in the frame of the Fusion for Energy (F4E) project and funded by the grant: F4E-2010-GRT-01-01 (ES-SF).

References (11)

  • C.J. Sun et al.

    Dynamics of weakly stretched flames: quantitative description and extraction of global flame parameters

    Combust. Flame

    (1999)
  • L. Topilski, Main parameters and assumptions that should be taken into account in the model of inert gas injection...
  • J. Xiao, J. Travis, W. Breitung, T. Jordan, Feasibility study of possible prevention of hydrogen combustion in ITER by...
  • J.R. Travis, J.W. Spore, P. Royl, K.L. Lam, T.L. Wilson, C. Müller, et al., GASFLOW: a computational fluid dynamics...
  • R.K. Kumar

    Flammability limits of hydrogen–oxygen–diluent mixtures

    J. Fire Sci.

    (1985)
There are more references available in the full text version of this article.

Cited by (3)

  • Evaluation of different models for turbulent combustion of hydrogen-air mixtures. Large Eddy Simulation of a LOVA sequence with hydrogen deflagration in ITER Vacuum Vessel

    2020, Fusion Engineering and Design
    Citation Excerpt :

    According to the theoretical estimations by Kuznetsov et al. [45], maximum adiabatic combustion pressure may exceed the design pressure of ITER VV (i.e., 2 bar) with inventories of 4 kg of H2, when the mixture ignites at the pressure over 400 mbar. Indeed, Kuznetsov et al. [46] had performed large-scale hydrogen ignition experiments in an 8.8 m3 vessel, as well as combustion experiences at reduced pressures, in the presence of a turbulent air jet injected into a hydrogen atmosphere. They found that for mixtures with a partial hydrogen pressure of 0.2 bars at 293 K and initial ignition pressures over 0.33 bars, final combustion pressures would reach values over 2 bars.

  • New insights into the peculiar behavior of laminar burning velocities of hydrogen-air flames according to pressure and equivalence ratio

    2014, Combustion and Flame
    Citation Excerpt :

    Furthermore, whereas recent efforts concerned high-pressure flames, only a few data have been obtained under reduced pressure, as presented in Table 1, where nonmonotonic variation of flame speed with pressure has been reported [13]. The subatmospheric conditions, specifically around 200 mbar, are of particular importance to guarantee safe operation of the International Thermonuclear Experimental Reactor (ITER) [14]. Updates of hydrogen kinetic reaction mechanisms have been presented recently by Hong et al. [15], Burke et al. [12], and Keromnès et al. [16].

View full text