Experiment and simulation on Zr₂Fe bed for tritium capturing

Abstract

Zr based alloys are widely used in hydrogen storage and purification systems. For hydrogen isotope capturing and recycling, it is possible to design the tritium extraction system of the glove box taking into account the operating conditions and technical constraints. Zr₂Fe alloy bed was adopted in the 2.5%H₂/Ar capturing experiments as well as numerical simulation was also constructed. In this work, the breakthrough curve under conditions of 300 °C@200 sccm for the carrier gas penetration was obtained by establishing a single flowing bed. Then a new numerical model of the bed was established for simulating the penetration process, and a numerical solution to the mass conservation equation of carrier gas passing through the bed was obtained concerning the actual experimental parameters. Calculation results have shown the two-dimensional distribution of H₂ concentration flowing through a one-dimensional reactor. It can be seen clearly that the H₂ concentration distributed along the bed axial at different time. This distribution profiles can be used to globally illuminate the concentration variation of hydrogen in the reactor within the whole reaction time scale. Comparing the experimental result of 1.27 days in the breakthrough curve, the results of numerical simulation can predict the curve of about 1.5 days began to breakthrough, although there were certain deviation (∼18%). The results can provide a feasible numerical model support for optimizing the tritium capturing bed design.

---

1. Introduction

Zr based alloys are widely used in hydrogen storage and purification systems.^1–5 A glove box with a tritium extraction system is an important safety containment system for tritium processing system in fusion reactor tritium factories. The flow bed filled with Zr₂Fe alloy powders has a satisfying ability for hydrogen capture from an inert gas carrier situation, and its absorption efficiency is sufficiently high to fulfill the tritium recovery material requirements.^6–8 According to different engineering requirements such as in the International Thermonuclear Experimental Reactor (ITER) and China Fusion Engineering Test Reactor (CFETR), it is necessary to design different parameters for the tritium capturing flow bed to meet certain operation and lifetime requirements. Generally, it is common practice to do a shrinkage ratio model test prior to the amplified reactor for meeting the actual technical requirements. Without numerous tests, this method cannot adapt to different working situations, complicated and changeable parameters, especially for the lifetime estimation of the engineered bed. Establishing a numerical model can speed up and optimize the reactor design.⁹ Therefore, this work utilized Zr₂Fe flow bed to study the capture of hydrogen isotopes from inert gas carrier by both experiment and numerical simulation, in order to optimize the tritium capturing process at the given conditions.

2. Experimental

Zr₂Fe alloy powders were purchased from GRINM Beijing. The particle size distribution is among from 100 to 200 mesh. Hydrogen breakthrough curves were measured in a setup (Fig. 1a) mainly consisting of a vacuum-pumping system, gas-mixing system, reactor with φ14 stainless steel tube in the furnace and micro chromatography with a 1 ppm resolution (Micro GC, Inficon 3000). Before measuring, all pipelines were purged by carrier gas and the reactor tube was preheated for removing contamination. The measurement processes can be described as: (i) Zr₂Fe alloy powders and fillers were filled into the reactor tube; (ii) pumping down the whole system except gas-mixing system into a ∼5 Pa vacuum level; (iii) activating Zr₂Fe powders at 500 °C for 2 hours in the vacuum state then cooling down to the setpoint temperature; (iv) shutting up the pump system, then filling the reactor with required mixing gas by the gas-mixing system and meanwhile using the Micro GC to analyze the H₂ concentration in the exhaust gas.

Fig. 1 (a) Hydrogen penetration measurement setup of Zr₂Fe reactor, (b) numerical simulation model for Zr₂Fe bed.

For the hydriding kinetic measurements were taken in a SETARAM PCT-Pro equipment. A sample of 1.5 g Zr₂Fe powder used in the flow reactor was taken to measure the hydriding kinetics at the required temperature. Before measuring, an activation process was also employed, which was consistent with penetration reaction condition (Table 1).

Table 1 The parameters used in simulation

Symbol	Numerical value	Parameter
C		The concentration of hydrogen in carrier gas (mol m^−3)
ρ	6900	Packing density of Zr2Fe in the bed (kg m^−3)
β	2.5%	Hydrogen concentration in the carrier gas
x		The length of reaction bed (m)
u	0.02	The areal velocity (m s^−1)
Dax	1.63 × 10^−4	The axial dispersion coefficient of constituent in carrier gas (m^2 s^−1)
ε	0.5	The porosity of fixed bed
ϕ	0.014	The diameter of the reaction bed cross section (m)
[small gamma, Greek, vector]		The reaction kinetics at cross section unit facing the gas flow
Dm	2.23 × 10^−4	Molecular diffusion coefficient (m^2 s^−1@573 K)
dp	1.5 × 10^−4	The average diameter of Zr2Fe particle (m)
ξ		The reacted fraction of hydriding in the hydrogen absorption kinetics fitting parameters of Zr2Fe^9,13

---

3. Numerical simulations

According to the law of mass conservation and the hydrodynamics, the carrier gas penetration in a porous column can be represented by a plug flow model including axial dispersion as shown in Fig. 1b. A carrier gas with 2.5% H₂ in Ar was flowed into the reactor tube, then the concentration distribution C(x,t) was calculated in the reactor. The mass balance equation in fluid phase for hydrogen concentration can be expressed as eqn (1).

With boundary conditions: in the section of (0,b)

Initial condition:

t = 0, C(x,0) = 0

Estimation of parameters

(i) Reaction rate [small gamma, Greek, vector]

Former kinetics reports^9–13 have used the grain scale reaction kinetics to estimate the particle scale reaction kinetics. However, it seems too simplified for practical engineering applications due to the idealized assumptions. This work used practical measured kinetics of Zr₂Fe powders to represent the reaction rate in the flow bed reactor. According to the plug flow model, the reaction rate of sectional area in the reactor can be expressed as eqn (2).

(ii) Axial dispersion coefficient D_ax

Considering the Zr₂Fe particle size and flow speed of carrier gas, the axial dispersion coefficient can be represented eqn (3) as followed by Edwards et. al.¹³

4. Results and discussion

The hydriding kinetic measurements were taken for deriving the reaction rate [small gamma, Greek, vector] used in the numerical simulation. The Zr₂Fe powder of 1.5 g was used to measure the hydrogen absorption properties in SETARAM PCT-Pro setup under 300 °C and 1 bar conditions. A formula suitable for alloy powder ξ_H2 = 1 − (1 − 1/2√(t/t_0.875))³ was used to fit our isothermal hydrogen absorption curve.^9–12 The fitting result was generally acceptable with a deviation R² = 0.989 as shown in Fig. 2. In addition, the reaction rate [small gamma, Greek, vector] can be derived from eqn (2).

Fig. 2 Hydrogen absorption properties of Zr₂Fe particles, the blue line is measurement results and green star line is simulation results.

For the hydrogen breakthrough curve measurement, an 8 cm length bed was filled with Zr₂Fe powder into the tube reactor. A 200 sccm mixture gas of 2.5% H₂/Ar was supplied into the reactor controlled by an accurate gas-mixing system. The hydrogen breakthrough curve is shown in Fig. 3, which indicates that hydrogen began to penetrate the bed at 1.25 days and then its concentration rises rapidly to the feed mixture gas concentration of 1.12 mol m⁻³, meaning breakthrough completely.

Fig. 3 H₂ breakthrough curve at 300 °C@200 sccm.

A numerical model was built up to simulate the hydrogen penetration process in the flow bed reactor. A program was coded to solve the mass conservation equation in this flow bed model. All the parameters used in the simulation were adopted from the practical experimental process, as well as the boundary and initial conditions. The simulation results are demonstrated in the Fig. 4. A 2D concentration profile concerning the axial and time coordinates is shown in the Fig. 4a, which can generally illuminate the concentration variation of hydrogen in the reactor within the whole reaction time scale. Furthermore, the concentration profile along the reactor axial at different time could be easily derived from the model as seen in the Fig. 4b. By using this, it can be used to discover what happened to the concentration profile in the reactor when the hydrogen was about to penetrate, as well as in the post-penetration process. Fig. 4c is the simulation result for the hydrogen breakthrough curve in the practical experiment. It shows that the breakthrough point was about 1.5 days, which has a diversion of ∼18% compared with the measuring result of 1.27 days. This can be attributed to the simplified assumption for the powder size of 150 μm, but in fact, it has a broad distribution of 100–200 mesh.

Fig. 4 Numerical simulations. (a) The 2D H₂ concentration along axial direction at different location and time; (b) the concentration profile of 0.8, 1.6 and 2.2 days for showing before-/after-breakthrough profiles; (c) H₂ breakthrough curve for Zr₂Fe bed.

5. Conclusions

Incorporating the numerical simulation with experimental methods on the hydrogen penetration behavior through the Zr₂Fe flow bed, this work established a new model that can globally illuminated the concentration variation of hydrogen in the reactor within the whole reaction time scale. The results could be beneficial to improving the design of tritium capture bed as follows:

(1) This model can predict an acceptable hydrogen breakthrough happening for the tritium capturing bed;

(2) The reaction rate [small gamma, Greek, vector] used in the numerical model is derived directly from hydriding kinetics experiment, rather than the calculation from grain scale kinetics. This methods including the powder size dispersion effect will be more suitable to the practical engineering.

Conflicts of interest

There are no conflicts to declare.

Acknowledgements

Project supported by National Magnetic Confinement Fusion Science Program (Grant No. 2017YFE0301502, 2013GB108002 and 2014GB112005).

Notes and references

1. Y. Yang, X. Song and C. Zhang, Int. J. Hydrogen Energy, 2016, 47, 22206–22213.
2. Z. Cao, L. Ouyang and H. Wang, Int. J. Hydrogen Energy, 2016, 41, 11242–11253.
3. D. Conić, A. Gradišek and J. Radaković, et al., Int. J. Hydrogen Energy, 2015, 40, 5677–5682.
4. A. Dagher, H. G. Salem and T. M. Moustafa, et al., Int. J. Hydrogen Energy, 2014, 39, 17740–17746.
5. L. Mihailov, T. Spassov and M. Bojinov, Int. J. Hydrogen Energy, 2012, 37, 10499–10506.
6. Z. Y. Huang, C. X. Liu and J. F. Song, et al., Chem. Eng. J., 2010, 38(10), 205.
7. S. Fukada, Y. Toyoshima and M. Nishikawa, Fusion Eng. Des., 2000, 49, 805–809.
8. S. Fukada, K. Tokunaga and M. Nishikawa, Fusion Eng. Des., 1997, 36(4), 471–478.
9. K. C. Chou, Q. Li, Q. Lin, L. J. Jiang and K. D. Xu, Int. J. Hydrogen Energy, 2005, 30, 301–309.
10. M. F. Edwards and J. F. Richardson, Chem. Eng. Sci., 1968, 23, 109–123.
11. J. M. P. Q. Delgado, Heat Mass Transfer, 2006, 42, 279–331.
12. K. Ligera, X. Lefebvrea, A. Ciampichettib, A. Aiello and I. Ricapito, Fusion Eng. Des., 2011, 86, 1859–1862.
13. M. F. Edwards and J. F. Richardson, Chem. Eng. Sci., 1968, 23, 109–l23.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c8ra08784a

---