Professional Documents
Culture Documents
Evaluation of Reaction Kinetics For The Catalyst Used in PCRD and Study of Channel Affect On The Same
Evaluation of Reaction Kinetics For The Catalyst Used in PCRD and Study of Channel Affect On The Same
ScienceDirect
highlights
Kinetics of indigenous PCRD catalyst for Indian NPP has been experimentally studied.
Behavior of catalyst plates stacked in parallel inside PCRD has also been evaluated.
Variation in migration mechanism of reactants to catalyst surface affects behavior.
Overall affect has been empirically approximated as single step Arrhenius equation.
This is significant in modelling of PCRDs for faster containment analysis using CFD.
Article history: Hydrogen mitigation strategies have gained importance for Nuclear reactors owing to
Received 5 November 2019 damage caused to integrity of reactor containment by hydrogen fire in three major nuclear
Received in revised form accidents of Three Miles Island, Chernobyl and Fukushima. One promising technology for
8 January 2020 hydrogen mitigation is deploying Passive Catalytic Recombiner Devices (PCRDs). Principle
Accepted 13 January 2020 involved here is recombining hydrogen released during accident with oxygen from
Available online 12 February 2020 ambient air inside reactor containment on catalyst surface to form steam. Present work
focuses on experimental evaluation of reaction kinetics associated with hydrogen-oxygen
Keywords: recombination on surface of indigenous PCRD catalyst developed for Indian Nuclear Power
Nuclear power plants Plants. Behavior of catalyst plates stacked in parallel inside PCRD has also been evaluated.
Severe accidents This effect is due to difference in migration mechanisms of reactants and products to and
Hydrogen mitigation from the catalyst surface. Overall affect has been empirically approximated as single step
Noble metal catalyst Arrhenius equation. This is significant in modelling of PCRDs for faster containment
Hydrogen-oxygen recombination analysis using CFD.
Reaction kinetics © 2020 Hydrogen Energy Publications LLC. Published by Elsevier Ltd. All rights reserved.
which should always be less than the self-ignition magnetronic and electrochemical technique. The detailed
temperature of the gas mixture. experimental data are not reported in open literature.
(iii) They are simple in construction and easy in application As part of the technology development programme for
as they consist of rectangular box having a set of par- hydrogen management in Indian Pressurised Heavy Water
allel catalytic plates having flow channels between Reactors (IPHWRs), experimental as well as analytical studies
them. Thus their handling, installation and mainte- are underway. The first type of catalyst device developed in
nance are relatively easy. BARC, India was based on a chemically coated polyester cur-
tain. This curtain was proposed to be kept in a rolled position
The development of catalyst for hydrogen oxidation dates and would roll down in the event of an accident. However, this
back to the work of Johann Wolfgang Dobereiner, who was the device could not perform beyond 150 C as the polyester cloth
first to study the catalytic features of platinum at the University showed its tendency to melt above this temperature. Hence,
of Gena in Germany almost more than a century back. The first its use was limited for very low hydrogen concentrations only.
application of the principle of catalytic recombination for miti- Subsequently, metal substrate based catalyst devices have
gating hydrogen in NPPs was developed in Germany in 1986 as been developed. Three different techniques for deposition of
reported by Chakraborty et al. [8]. It was a sandwich type the PtePd catalyst on polyester [9], cordierite honeycomb [10]
recombiner comprising of stainless steel carrier plates, over and stainless steel gauze [11] have been developed. Low sur-
which thin layers of platinum, vanadium and palladium were face area catalysts viz. Pt, Pt þ Pd (simultaneous) and Pt þ Pd
vacuum deposited. These plates were stacked in an inert at- (sequentially) deposited over stainless steel wire gauze,
mospheric condition maintained in a box to avoid contamina- employing single bath electroless reduction with formalde-
tion during plant operation and these were automatically hyde as reducing agent has been developed and effect of
unfolded and lowered during an accident. After the develop- preparation route on overall morphology of noble metal
ment of sandwich type recombiner, other type of catalytic coating has been evaluated in detail [12]. This catalyst having
recombiners developed in Germany are (i) Box-Type Recombiner PtePd deposited simultaneously over stainless steel wire
- SIEMENS Design and (ii) Box Type Recombiner - NIS Design. gauze has been employed in PCRDs in Indian NPPs.
Subsequently, other countries also started their own R&D On the analytical front also many researchers have pub-
for recombiner development. Reinecke published a paper on lished their studies in recent literature. ‘Modular Accident
Studies on innovative hydrogen recombiners as safety devices Analysis Program version 5.04 (MAAP5.04)’ software was used
in the containments of light water reactors [4]. Their study in Ref. [13] to calculate the flammable gas generation, migra-
was based on experiments carried at REKO facilities at For- tion, and combustion risk within the containment to calculate
schungs Zentrum, Juelich (FZJ). The samples of sandwich type the number of PARs need in a large Dry PWR containment in
plates were tested in laboratory scale test setup. In order to order to prevent significant flammable gas combustion
demonstrate the functioning of the catalytic plates in large events. CFD has been used by Ref. [14] to study the dynamic
volumes, two series of experiments, described as demon- behavior of a plate-type Areva FR-380 recombiner in a quies-
stration test and convection test, were performed in HDR test cent environment. The model was able to identify the phe-
facility. Development of catalytic recombiners in Canada has nomenon of downward gas flow through the PAR metal casing
resulted in two types of models based on natural convection in the initial phase of the start-up period and to correlate the
and on forced convection. The catalyst developed by AECL, magnitude and duration of this effect with the rate of increase
Canada consists of small platinum crystallites on a hydro- of hydrogen concentration in the environment [15]. Con-
phobic high surface area support, bonded to a stainless-steel structed and justified, a simple engineering thermo-hydraulic
screen. The catalyst element for the forced convection test model of hydrogen removal in the operation of a passive
was an open cylindrical module formed by wrapping up of autocatalytic recombiner based on the available experimental
alternate layers of flat and corrugated screen with cylindrical data. The paper presented the application results of the model
passages in between the catalyst layers. For the natural con- as a part of contour industry codes RELAP, TRACE and COR-
vection test, the catalyst elements were in the form of loose SAR, intended, among other things, for carrying out multi-
thin plates. The concept of using catalytic recombination of factor and full-scale calculations of the dynamics of
hydrogen was also investigated at the Kurchatov Institute of emergency processes with the release of hydrogen into nu-
Technology, Russia, both experimentally and analytically. clear power plant premises. A passive containment cooling
Experiments were performed with flat plate type recombiners system (PCCS) combined with hydrogen mitigation system is
comprising of catalytically activated high porosity cell mate- suggested in Ref. [16] that will passively remove containment
rial (HPCM) in 1.2 m3 cylindrical shaped reaction vessel. Ex- heat load as well as mitigate hydrogen gas generated and
periments with both dry and wet air indicated effective accumulated in the containment atmosphere during pro-
hydrogen combination with sufficiently small volumes of longed station blackout. They make use of atmospheric air as
HPCM. The HPCM samples made of stainless steel and ultimate heat sink to reject the PCCS heat, which in turn
nichrome are the typical volumetrically regular structures of transfers heat from the containment. In the proposed design a
periodically repeating cells, in the form of polyhedron with a finned vertical tube type PCCS system submersed in pool
characteristic cell scale of approximately 1 mm. Material of water is considered with supply of pool water lasting three
the walls of cells is generally small compared to overall vol- days following which airflow cools PCCS tubes for long term.
ume of cells thereby causing highly porous (90%) structure The PCCS inlet flow path has a catalytic recombiner that is
with low specific gravity. The walls of cell are covered with a effective during first few days should there be any hydrogen
coating of aluminium oxide and platinum layer deposited by generation and accumulation in the containment.
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 5 ( 2 0 2 0 ) 8 5 8 4 e8 5 9 4 8587
As a part of the analytical studies at BARC, India, Compu- reactor and the cylinder is pressurised to 1.1 bars with air.
tational Fluid Dynamics (CFD) model [17] for simulating the During the start of each experiment, a known quantity of
performance of PCRD in reactor containment has been hydrogen is injected into the reactor to attain a pre-decided
developed and validated against data from open literature. value of ~0.8e4.5% initial v/v hydrogen concentration for
The reaction kinetics data used for this model is specific to the various experiments.
particular catalyst reported in open literature [18]. The present The hydrogen concentration inside reactor is monitored
study is aimed at detailed evaluation of the reaction kinetics using thermal conductivity detector (TCD) based online
of the catalyst device developed at BARC, so that the analytical hydrogen monitor. The reaction being exothermic, the tem-
model [17] can be fine-tuned for the India specific PCRD. perature rise on the catalyst surface is recorded with the help
of two K-type thermocouples placed on the catalyst. The
change in concentration of hydrogen and the catalyst surface
Experimental temperature are recorded as functions of time.
From a series of experiments, data points have been
Catalyst preparation generated pertaining to variation of hydrogen concentration
and temperature with time [19]. The derivative plot of [H2] vs.
The PtePd over stainless steel wire gauze catalyst has been time provides for d [H2]/dt data for time ‘t’, at temperature ‘T’.
prepared by electroless deposition method [12]. This method Substituting the values of [H2] and d [H2]/dt in Eq. (6) later, kT is
involves chemical deposition of noble metals (Platinum and calculated for the respective temperature T.
Palladium) over the stainless-steel wire gauze in presence of a For a typical experiment, where initial hydrogen concen-
reducing agent. The precursors used for Pt and Pd are Chlor- tration was taken as 3.4% volume by volume, Fig. 3 depicts the
oplatinic acid (Hindustan Platinum Private Limited, 39.8% Pt) data used for reaction kinetics and activation energy calcu-
and palladium chloride (Hindustan Platinum Private Limited, lations. The scatter plots in Fig. 3(a) represents change in
59.96% Pd) and reducing agent is formaldehyde (Thomas hydrogen concentration in moles/litre, Fig. 3(b) depicts the
Baker, 37e41%). change in temperature with respect to time and the derivative
plot -d [H2]/dt in moles/litre/sec. corresponding to rate of re-
Catalyst characterization action, with time t is depicted in Fig. 3(c). For each set of data,
rate constant kT is calculated for that temperature T,
Powder X-ray Diffraction has been used to assess the crys- employing Eq. (6). The activation energy calculated from ln
talline phase purity of deposited noble metals over the wire (kT) vs 1/T plot is shown in Fig. 3(d).
gauge surface. The XRD patterns were recorded with Philips
analytical instrument with Ni filtered CueK-a radiation in the
scanning range of 10e70 two theta value. Surface Results and discussion
morphology and particle size of deposited noble metal nano-
particles has been studied using Scanning Electron Micro- Catalysts preparation
scopy (JEOLJSM-6360).
The catalyst is prepared using formaldehyde (Thomas Baker,
Catalytic activity evaluation 37e41%) as reducing agent with chloroplatinic acid (Hindustan
Platinum Private Limited, 39.8% Pt) and palladium chloride
For generation of kinetics data for the prepared catalyst, the (Hindustan Platinum Private Limited, 59.96% Pd) as noble metal
H2 and O2 recombination reaction was performed under static precursors and SS-304 wire gauze of wire diameter of 0.125 mm
air condition at room temperature in a 40 L stainless steel and of 80 mesh number as a substrate for deposition [12,20].
reactor Fig. 2[12]. The catalyst is hung inside the cylindrical The coating process employed is based on the following
Fig. 3 e Hydrogen concentration (a), catalyst temperature (b), derivative plot of [H2] (c) with respect to time and ln (kT) vs 1/T
plot (d).
Catalyst characterization
Fig. 4 e XRD pattern of SS wire gauge support and PtePt
X-ray diffraction catalyst.
Fig. 4 shows the XRD pattern of wire gauge as well as the
catalyst. The XRD pattern of the catalyst when compared to 0802) and palladium (JCPDS 46-1043). The average crystallite
the SS wire gauge have additional peaks at ~40 , ~46 and ~67 , size of noble metal particles determined by Scherrer equation
2Q values. These peaks correspond to Platinum (JCPDS No.04- using 40 peak and it was found to be around 21 nm.
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 5 ( 2 0 2 0 ) 8 5 8 4 e8 5 9 4 8589
H2 þ ½ O2 4H2 O (3)
d½H2
Rate ¼ ¼ kT :½H2 x :½O2 y (4)
dt
where, kT is rate constant at temperature T (Kelvin) and x and
y are the order of the reaction with respect to H2 and O2
respectively.
For 4% initial hydrogen concentration in the reactor, the
stoechiometric oxygen concentration required is 2%.
Compared to this, the oxygen concentration in air is 20%,
which is 10 times more than the stoechiometric require-
ment. Hence, it can be assumed that the rate is indepen-
dent of the oxygen concentration in the reactor. Thus,
the rate equation can be reduced to the form shown in
Eq. (5):
d½H2
¼ kT :½H2 x (5)
dt
This equation is rearranged to Eq. (6) for Arrhenius plot
1 d½H2
kT ¼ : (6)
½H2 x dt
reaction rate subsides with decreasing hydrogen concentra- 109e150 kJ/mol [27]. Compared to this surface assisted het-
tion in the reactor vessel. This clearly indicates the usefulness erogeneous reaction has been reported to take place through
of stainless steel wire gauze support in terms of effective considerably lower activation energy barrier. The extent of
dissipation of generated reaction heat. ln (kT) vs 1/T data ob- this barrier depends on the pathway followed.
tained from various set of experiments is plotted in Fig. 7. The Two of the prominent approaches are based on reaction
activation energy obtained from the above experiments is between adsorbed oxygen with either the gaseous hydrogen
found be in the range of 18.1 kJ/mol while the intercept or the dissociated hydrogen atom. The first mechanism as
signifying pre-exponential factor is 0.76. discussed by Clyne [28] involving two step reaction based on
The kinetics for oxidation of hydrogen has been reported in [O] þ H2 4 OH þ H and H2 þ OH 4 H2O þ H, exhibiting acti-
detail by various researchers. The gas phase oxidation process vation energy requirements of 25 kJ/mol. Since the activation
is hindered by large activation energy in the range of energy for these alternate paths exceeds the activation energy
data obtained from our experiments by large margin, the
feasibility for their occurrence is very low. Compared with
this, the reaction mechanisms based on dissociative adsorp-
a tion of hydrogen and oxygen exhibit activation energy for
various reaction steps in the range of 11.5 and 17.5 kJ/mol.
Hence it can be implied that the H2 þ O2 4 H2O reaction on the
present set of catalysts follows the similar pathway, based on
dissociative adsorption of both hydrogen and oxygen, fol-
lowed by combination of adsorbed species.
initial hydrogen concentrations. Fig. 9 shows catalytic ac- the recombination reaction. The derivative plot of [H2] vs.
tivity data obtained for catalyst for initial hydrogen concen- time provides for d [H2]/dt data for time ‘t’ and at tempera-
tration ranging from around 2.0 to 4.5% in air. The ture ‘T’. Substituting the values of [H2] and d [H2]/dt in Eq. (6),
temperatures at different heights of central channel were kT is calculated for the respective temperature T. ln (kT) vs. 1/
averaged suitably to get the temperature value resulting from T values obtained from various set of experiments are plotted
Acknowledgement
references
Fig. 12 e ln (kT) vs 1/T plot for parallel channels
arrangement between four plates with corner plates
blocked from outside. [1] Gharari R, Kazeminejad H, Kojouri NM, Hedayat A. A review
on hydrogen generation, explosion and mitigation during
severe accidents in light water nuclear reactors. Int J
Hydrogen Energy 2018;43:1939e65.
earlier case to get the temperature value resulting from the [2] Institut de Radioprotection et de Surete Nucleaire (Irsn).
recombination reaction. The derivative plot of [H2] was also Research and development with regard to severe accidents in
plotted similarly. pressurized water reactors: summary and outlook. Technical
Similar to the above procedure, the values of [H2] and dH2/ Report IRSN. France: IRSN Fontenay-aux-Roses; 2007. 083.
dt were substituted in Eq. (6), to calculate the kT respective [3] Bentaib A, Meynet N, Bleyer A. Overview on hydrogen risk
research and development activities: methodology and open
temperature T. Fig. 12 shows the ln (kT) vs 1/T values obtained
issues. Nucl Eng Tech 2015;47:26e32.
from various set of experiments. The activation energy for the
[4] Reinecke EA, Tragsdorf IM, Gierling K. Studies on innovative
parallel channels arrangement by blocking the outer surface hydrogen recombiners as safety devices in the containments
of the two corner plates with aluminium foil is found out to be of light water reactors. Nucl Eng Des 2004;230:49e59.
20.6 kJ/mol, while the intercept signifying pre-exponential [5] Yu W, Yu X, Tu ST, Tian P. Super hydrophobic Pt-Pd/Al2O3
factor is 1.81. catalyst coating for hydrogen mitigation system of nuclear
power plant. Int J Hydrogen Energy 2017;42:14829e40.
[6] Lalik E, Kosydar R, Sobieraj RT, Witko M. Humidity induced
deactivation of Al2O3 and SiO2 supported Pd, Pt, Pd-Pt
Conclusion catalysts in H2þ O2 recombination reaction: the catalytic,
micro calorimetric and DFT studies. Appl Catal Gen
Reaction kinetics for the indigenous catalyst developed for 2015;501:27e40.
[7] Sandeep KC, Bhattacharyya R, Warghat C, Bhanja K,
PCRDs to be deployed in Indian NPPs has been studied by
Mohan S. Experimental investigation on the kinetics of
carrying out various lab scale experiments on the same. The catalytic recombination of hydrogen with oxygen in air. Int J
behaviour of these catalyst plates, when stacked in parallel Hydrogen Energy 2014;39:17906e12.
inside the recombiner box, has also been evaluated by car- [8] Chakraborty AK, Klatt KH, Konrad RD. Erste Me Beerebnisse
rying out semi-engineering scale experiments. The activation von Experimentzur Besetigung von H2 ausbrennbaren
energy calculated for the recombination reaction on the single Gasgemisch Unter Nutzungwirkender Metallfollen
catalyst plate is found to be in range 18.1 kJ/mol, while the (Zwischenbericht) GRS) Mai. 1986. p. 1222. A e.
[9] Dewit WA, Koroll GW, Sitar JL, Graham WRC, Knight C.
intercept signifying pre exponential factor is 0.76. This value is
Hydrogen recombiner development at AECL. In: Proc. OECD/
in accordance with the values reported for similar catalyst in NEA/CSNI, AECL-11762, NEA/CSNI/R (96) 8. Pinawa,
open literature. For the parallel channels arrangement where Manitoba. White Shell Laboratories; 1996.
the sides are open, the activation energy is found be in the [10] Belapurkar AD, Gupta NM. A hydrophobic catalyst for
similar range 18.7 kJ/mol, and pre exponential factor is 0.8. recombining H2/D2 and O2 in nuclear reactors. Power Plant
This shows that for the central channel, there is an extra Chem 2000;2:452e7.
[11] Belapurkar AD, Varma S, Shirole A, Sharma J. Cordierite
resistance for the diffusion of hydrogen towards the central
supported platinum catalyst for hydrogen-oxygen
plates. This is due to part of the hydrogen getting consumed
recombination for use in nuclear reactors under LOCA.
on the outer surface of the corner plates due to lateral diffu- Power Plant Chem 2008;10:461e7.
sion from sides. This deprives the central channel with free [12] Sanap KK, Varma S, Dalavi D, Patil PS, Waghmode SB,
availability of hydrogen causing the overall activation energy Bharadwaj SR. Variation in noble metal morphology and its
to be slightly more. When the outer surface of the two corner impact on functioning of hydrogen mitigation catalyst. Int J
plates in the parallel channels arrangement is blocked, the Hydrogen Energy 2011;36:10455e67.
lateral diffusion of hydrogen from the sides is prohibited. This
8594 i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 5 ( 2 0 2 0 ) 8 5 8 4 e8 5 9 4
[13] McMinn P. Passive autocatalytic recombiners for flammable [21] Barker BD. Electroless deposition of metals. Surf Technol
gas control. July: Fauske & Associates, LLC; 2017. 1981;15:77e88.
_ n
[14] Roze Antoni. Simulation of start-up behaviour of a passive [22] Morfin F, Sabroux JC, Renouprez A. Catalytic combustion of
autocatalytic hydrogen recombiner. Nukleonika 2018;63/ hydrogen for mitigating hydrogen risk in case of a severe
2:27e41. accident in a nuclear power plant: study of catalysts
[15] Avdeenkov AV, Sergeev Vl V, Stepanov AV, Malakhov AA, poisoning in a representative atmosphere. Appl Catal B
Koshmanov DY, Soloviev SL, Bessarabov DG. Math Environ 2004;47:47e58.
hydrogen catalytic recombiner: engineering model for [23] Acres GJK. The reaction between hydrogen and oxygen on
dynamic full-scale calculations. Int J Hydrogen Energy platinum. Platin Met Rev 1966;10:60e4.
2018;43:23523e37. [24] Kim KR, Paek SW, Chung H. Catalytic recombination of
[16] Revankar ST. Passive containment cooling system with hydrogen and oxygen in air stream. J Ind Eng Chem
hydrogen recombiner for prolonged station blackout. Int J 2001;7:116e20.
Nucl Saf Secur 2019;1/1:1e17. [25] Warnatz J, Allendorf MD, Kee RJ, Coltrin ME. A model of
[17] Shukla V, Ganju S, Varma S, Sengupta S, Maheshwari NK. elementary chemistry & fluid mechanics in combustion of
Affect of recombiner location on its performance in closed hydrogen on platinum surfaces. Combust Flame
containment under dry and steam conditions. Int J Hydrogen 1994;96:393e406.
Energy 2019;44/47:25957e73. [26] Hellsing B, Kasemo B, Zhdanov VP. Kinetics of the
[18] Schefer RW. Catalyzed combustion of H2/air mixtures in a hydrogen-oxygen reaction on platinum. J Catal
flat plate boundary layer: II, numerical model. Combust 1991;132:210e28.
Flame 1982;45:171e90. [27] Seshadri K, Peters N, Williams FA. Asymptotic analyses of
[19] Sanap KK, Varma S, Waghmode SB, Bharadwaj SR. stoechiometric and lean hydrogen-air flames. Combust
Supported Pt nano particles for the hydrogen mitigation Flame 1994;96:407e27.
application. Int J Hydrogen Energy 2014;39:15142e55. [28] Clyne MAA, Thrush BA. Reaction of oxygen atom with
[20] Sanap KK, Varma S, Waghmode SB, Bharadwaj SR. Wire gauze hydrogen. Nature 1961;189:135e6.
and cordierite supported noble metal catalysts for passive
autocatalytic recombiner. Nucl Eng Des 2015;14:226e32.