595
A publication of
CHEMICAL ENGINEERING TRANSACTIONS
VOL. 77, 2019
The Italian Association
of Chemical Engineering
Online at www.cetjournal.it
Guest Editors: Genserik Reniers, Bruno Fabiano
Copyright © 2019, AIDIC Servizi S.r.l.
ISBN 978-88-95608-74-7; ISSN 2283-9216
DOI: 10.3303/CET1977100
Bursting Test of a Silicon Carbide Micro Reactor
Roland Guidata, Olivier Lobeta, Claire Villemurb,*, Jérôme Daubechb, Patricia Vicotb
a
Corning SAS 7 bis avenue de Valvins F-77215 Avon Cedex, France
INERIS, Parc Technologique Alata, BP2, 60550 Verneuil-en-Halatte,France
claire.villemur@ineris.fr
b
This document describes and presents the results of a pneumatic bursting test of a CORNING AFR® G4
industrial silicon carbon reactor performed at INERIS in France. The reactor burst beyond 150 bar, well above
its design pressure of 18 bar. The kinetic energy of fragments ejected during the bursting and the pressure
effects were measured. The observed effects are below the minimum levels of significant injury for human
body.
1. Introduction
The micro reactor technology has notably gained more and more interest over the past years, and its
penetration in production plants is increasing rapidly. Micro reactors are devices operating at flow rates up to
several thousands of millilitres per minute, with a characteristic dimension (diameter) ranging from several
micro to several millimetres (the latter being sometimes named “milli “or “meso” reactors) (Jensen, 2017).
This kind of equipment offers a much better heat transfer than conventional equipment; the higher heat
release management allowing to operate at a higher temperature, which leads to faster kinetics. They also
provide a better mass transfer that can increase the reaction rate. Therefore, they allow producing similar
throughput as batch processes with a much lower quantity processed at the same time, leading to a drastic
reduction of damages resulting from a failure or a chemical runaway. In that respect, micro reactor technology
is clearly part of IST (Inherently Safer Technology).
However, operating at higher temperatures means operating at more severe condition. A sudden shut down of
the feeding pump or cooling circulation may cause a temperature increase and the formation of hot spots
leading to a reaction runaway accompanied by gas or vapor production. A micro reactor being generally full of
liquid (except for gas-liquid reactions), a local pressure increase would not be damped by a free volume as in
a batch reactor, and the pressure could rise very rapidly into the reactor. The rate of pressure rise would
depend on several factors, such as the capacity of the material constituting the reactor wall to partially absorb
the pressure wave (elasticity), the head loss in the reactor and the rate of gas generation by the reaction. As
the pressure would rise very rapidly, a conventional mechanical protection device (PSV) should not operate as
fast as requested due to its mechanical inertia. Bursting disk could be suitable providing that it would be
installed precisely at the point where the pressure increase would be initiated. Ultimately, the micro-reactor
itself would acts as a bursting disk.This phenomenon so far has not been studied from a theoretical point of
view, and modelling may not be so relevant as many parameters are involved. End users may therefore
question about the effects of a pressure rise resulting from a reaction runaway in a micro reactor, and what
could be the consequences on the direct surrounding of the reactor, in particular on workers. In order to
determine the consequences of a rapid pressure rise in a micro reactor, a pneumatic bursting test was carried
®
out at INERIS on a CORNING Advanced Flow™ G4 industrial reactor.
2. CORNING® Advanced Flow™ Reactor (AFR)
Corning is involved in micro-reactor development and sales since 2002. The first reactors were made of glass
and dedicated mainly to lab development and production for high added value products (pharmaceutical
industry). High throughput micro reactors for industrial production (several thousands of tons per year) have
been introduced on the market for more than seven years, and several industrial units are in operation in
Paper Received: 25 October 2018; Revised: 15 April 2019; Accepted: 25 June 2019
Please cite this article as: Guidat R., Lobet O., Villemur C., Daubech J., Vicot P., 2019, BURSTING TEST OF A SILICON CARBIDE MICRO
REACTOR, Chemical Engineering Transactions, 77, 595-600 DOI:10.3303/CET1977100
596
Europe, India and China. In the micro reactor for industrial production (G4 AFR), the process flow is contained
in plates made of sintered silicon carbide (SiC). This material has very good thermal properties, is not subject
to thermal shock and has a large thermal inertia, another key point for safety in case of sudden shut-down of
the cooling system. On top of that, it has been chosen because it is virtually not subject to corrosion by any
medium (including all acids, caustic and hydrofluoric acid) at the design temperature (ranging from -40 to
+200°C). The G4 AFR looks like a plate and frame heat exchanger with a stack of SiC plate for process,
interposed between metal plates for heating or cooling medium.
In order to protect people against any unexpected liquid projection or bursting, it is surrounded by a stainlesssteel casing (see Picture 1). Even though G4 AFR has a small volume (typically a few liters), it is considered
as a pressure vessel and complies with European and international regulation (PED in Europe, SELO in
China, KGS in Korea….).
Picture 1. G4 AFR with heat exchange piping without stainless-steel protective casing (a), G4 AFR with
protective casing (b).
3. Bursting test and results
6
60
5
50
4
40
Surpression (mbar)
Altitude (m)
The bursting test of a G4 AFR model was performed at INERIS, the French National Institute for Industrial
Environment and Risks. The mission of the institute is to evaluate and prevent accidental or chronic risks to
human and the environment related to industrial installations, chemical substances and underground
operations. INERIS has experimental facilities for conducting large-scale tests.
3
30
2
20
1
10
0
0
0
5
10
15
20
Distance (m)
25
30
35
0
50
100
150
200
250
300
350
400
Distance (m)
Figure 1. Maximum projection distance (a) and overpressure (b) as a function of distance to the reactor
calculated with PROJEX EPHEDRA software.
597
3.1 Preliminary study
A predictive analysis of the maximum potential effects of the bursting was conducted before the test to
determine safety distances for operators. The pressure effects and the projection distance of the fragments
were calculated using the PROJEX EPHEDRA software (Heudier, 2015). The maximum projection distance
was estimated at 30m from the reactor and an overpressure of 50 mbar was estimated at 5m (see Figure 1).
3.2 Test process
The reactor model tested was made of two SiC plates interposed between three metal plates surrounded by
the stainless-steel casing. The test consisted in causing the rupture of one of the SiC plates by rapidly
introducing compressed air from a pressurized tank to simulate a quick pressure increase in the reactor. The
test was conducted by carrying out several successive sudden loads of the plate until bursting, using remotely
controlled pneumatic valves.
In order to record information on overpressure and projection of fragments that could be produced by the
bursting, the following instruments were installed on or close to the reactor (see picture 2):
• A piezoresistive pressure transducer on the outlet of the pressurized SiC plate,
• Two piezoresistive pressure transducers placed on lenticular supports and positioned respectively at 1m
and 2.5m from the center of the reactor, in the lateral axis of the plates,
• A high-speed camera (1000 pictures/sec), also placed in the lateral axis of the plates,
• In addition, two conventional cameras were also placed to monitor the test.
Picture 2: Test installation with atmospheric pressure sensors and pressure sensor on the SiC plate (middle of
the picture). The cameras are not visible on the picture.
3.3 Test results
The pneumatic rupture of the plate occurred beyond 150 bar, well above the design pressure of the reactor
(18 bar).
Reactor structure damages
After the bursting, the global structure of the reactor did not show significant damages (see picture 3).
The two SiC plates were broken during the rupture and the two heat transfer fluid plates surrounding the
pressurized plate were disengaged. In addition, the lower part of the protective casing was partly torn off (see
picture 4).
Pressure effects
The release of the pressure contained in an enclosure during its pneumatic rupture results in the external
propagation of an atmospheric pressure wave.
The pressure recorded close to the reactor during the bursting did not show a significant pressure increase.
The peak of overpressure measured at 1m from the center of the reactor was 6 mbar, and 2 mbar at 2.5m
598
(see Figure 2). It is also to notice that the pressure release in the SiC plate after breakage lasted 15
milliseconds.
Picture 3. AFR before test (a) and after test (b).
Picture 4. The protective casing was partially torn off.
Projection effects
24 small SiC pieces were projected by the bursting around the reactor through little openings of the casing
(see picture 5). Their weight was less than 25 grams, and the maximum distance of projection was 6 meters.
The projection distances and the intensity of the pressure wave observed during the test are well below the
estimates of the preliminary study. It is likely that the presence of the casing and the formation of multiple
cracks in the SiC plate rolled the flow of pressurized air, thus limiting the intensity of the pressure effects. The
projection effects were also limited by the protective casing which prevented the ejection of a number of plate
fragments outside the reactor structure and by the fixing of the plates to the chassis.
599
Figure 2: Pressure evolution after bursting: top dark line represents the pressure evolution inside the
pressurized SiC plate, the green line is the atmospheric pressure measured at 1m from the centre of the
reactor, the yellow line is the pressure measured at 2.5 m from the centre of the reactor.
3.4 Estimation of impacts on people and environment
Estimation of potential consequences of projection effects
A missile is likely to produce two types of impacts on human or on structures: shocks, that can cause fractures
in the human body or lead to the deformation of structures, and penetration, which is mainly the case for small
fragments. The consequences of the impact of projectiles formed by the bursting of the plate were estimated
by calculating their kinetic energy and comparing it to the thresholds related to the effects of penetration of
small fragments (less than 1 kg) defined in the French pyrotechnic regulation (FR, 2007).
The kinetic energy of the projectiles was calculated from their velocity assessed from the recording of the
high-speed camera (see Figure 3). The maximum kinetic energy calculated was less than 1 Joule, far less
5
than the threshold of occurrence of significant injuries (8 J) and damages on structures (3,7.10 J).
Figure 3. Extracts from the recording of the high-speed camera showing the path of a SiC fragment that
allowed the estimation of the kinetic energy of projection.
600
Estimation of potential consequences of pressure effects
A pressure wave is likely to induce bending or shear forces in the structures and compression in the human
body. A pressure wave can also propel projectiles or cause the fall of an individual. In order to estimate the
potential consequence of pressure effects of the bursting, the intensity of the pressure wave measured around
the reactor was compared to the thresholds of pressure effects on the structures and on the human body
proposed in the Decree of 29 September 2005 relative to the assessment of dangerous phenomena that can
occur in Classified Installations (FR, 2005). The intensity of the pressure wave observed after the pneumatic
rupture (6 mbar) is well below the threshold of appearance of effects on individuals and structures (20 mbar).
4. Conclusion
The burst test performed at INERIS on a Corning G4 micro reactor partially made of silicon carbide has shown
that under test conditions, it was unlikely that any people, even staying close to the reactor, would have been
significantly injured. In addition, consequences for the close environment would be very limited.
The bursting pressure (beyond 150 bar) is significantly higher than the design pressure (18 bar) which
constitute a high safety margin in case of mis-operation.The mechanism of rupture of SiC leads to a multiple
opening of the SiC plate, which limits the effect of pressure and kinetic energy of fragments. The mechanical
design of the reactor, as well as the stainless-steel casing have also a positive impact in the limitation of the
consequences of the bursting.The consequences both in term of pressure wave and projection of fragments
are significantly lower than the potential maximum effect calculated by the PROJECT EPHEDRA software.
Although the test was conducted with compressed air, it gives a good trend of what could happen after a
severe chemical reaction runaway leading to a rapid pressure increase by gas or vapor generation.
References
FR, 2007, Circular DPPR/SEI2/IH-07-0111 of 20 April 2007 on the application of the decree laying down the
rules on risk assessment and accident prevention in pyrotechnic establishments.
FR, 2005, Decree of 29 September 2005 on the evaluation and consideration of the probability of occurrence,
kinetics, intensity of effects and severity of consequences of potential accidents in hazard studies of
classified installations subject to authorization.
Heudier L., 2015, Les éclatements de capacité, phénoménologie et modélisation des effets, Oméga 15,
INERIS, Verneuil-en-Halatte, France.
Jensen K.F., 2017, Flow Chemistry-Microreaction Technology comes of Age, AIChE Journal, March 2017, Vol
63, N°3, Cambridge, USA.