Chemical Quenching
Alexei Pylilo
Chemical Engineering and Biotechnology
Submission date: July 2012
Edd Anders Blekkan, IKP
Torbjørn Gjervan, SINTEF
Norwegian University of Science and Technology
Department of Chemical Engineering
The present master thesis is an extension of specialization project and has been carried out at the
Department of Chemical Engineering at the Norwegian University of Science and Technology
The work with the project started after personal meeting with my supervisor Professor Edd. A.
Blekkan who proposed to investigate the opportunity of chemical quenching of synthesis gas by
light hydrocarbons. The project idea has seemed very attractive and challenging because it promised
a lot of interesting work.
Catalysis group at the Department of Chemical Engineering disposes experimental set-up
Pyrolyserigg that has been moved to a new location at the Chemistry Hall D. Pyrolyserigg was
placed previously at Varmeteknisk Lab,(NTNU). In connection with relocation, it was necessary to
re-establish apparatus and perform a series of methane pyrolysis experiments in order to confirm
that the Pyrolyserigg works properly. Verification tests were finished in the middle of February.
In order to perform quenching experiments, it was necessary to design and install a new quencher
part for the cooling/quenching system. First, in the middle of March the modified quencher was
installed and it was possible to continue with experiments. Further, a series of experimental
problems (gas leakages, welding of connection ceram tube-quencher, problems with gas flows,
problems with thermocouples, PC – problems, some problems with gas chromatograph) has
occurred that slowed experimental work. Finally, one of the heating elements of high temperature
furnace was broken.
Experiments with a synthesis gas were planned, but were not performed as a result of the events
mentioned above. However, some practical knowledge were obtained that can form the basis for the
future work if it will be decided to continue with the topic.
I wish to thank a lot my supervisor Professor Edd A. Blekkan and co-supervisor Torbjørn Gjervan
for their participation in my work and supervision.
I am very thankful to all technical staff at the Department of Chemical Engineering that helped me
with solving of practical tasks that have arisen during my project work.
Alexei Pylilo
Trondheim, June 2012
The main objective of the present work was to investigate the opportunity for the use of light
hydrocarbons as quenching agents to quench high temperature gas streams. When light
hydrocarbon, for example propane, is introduced into a hot gas, for example synthesis gas, the
cooling of hot gas will occur both by dilution and by endothermic reactions of hydrocarbon. Thus,
the hot gas may be quenched to a lower temperature. At the same time, thermal energy present in
the hot gas may be recovered in the form of desired products that are produced during hydrocarbon
cracking. This type of quenching is called for chemical quenching. A proof-of-concept study would
be performed, but was not completed because of problems with a realization of experimental
conditions and problems with equipment.
Methane pyrolysis experiments were performed at high temperatures 1400 and 1450 °C. Short
residence times and high dilution with hydrogen were applied in order to minimize coke formation.
Methane conversions, selectivities and yields of products for different experimental conditions were
calculated from the gas chromatographic analysis. The aim of pyrolysis experiments was to validate
that the experimental set-up Pyrolyserigg works properly. This was done by comparison of
experimental results with results that were obtained before on the same set-up and by comparison
with a literature data.
After the work of Pyrolyserigg was verified, it was possible to modify existing cooling/quenching
system. A new quencher part that gives opportunity to introduce a cold gas into hot gas, heated by
high temperature furnace, was designed. The quencher part was designed in a way that the
temperature of hot gas entering the quencher and temperature of cooled gas mixture leaving the
quencher could be measured. Thus, quenching effect (temperature drop) could be measured.
Three types of experiments were performed after modification of cooling system: experiments with
only hot inert gas (N2), experiments with a hot nitrogen gas «quenched» by methane, and hot
nitrogen gas quenched by propane. Hot gas temperature and gas flows were varying in order to
study the modified system's behaviour.
Experiments with a hot nitrogen showed that there is a limit for maximum obtainable temperature
inside the quencher, 1195 – 795 °C, for the given quencher design. A high temperature gradients
between measurements points were registered that may be an indication of high heat losses. The
temperature gradient increases with increasing gas temperature and decreases with increasing gas
Quenching effect (temperature drop) was measured in quenching experiments with propane, and
expected temperature drop was calculated for the applied experimental conditions. Low measured
temperature drop in quenching experiments indicates poor mixing of hot and cold gases. Quenching
effect of propane introduction increases with increasing hot gas temperature because heat
consumption by endothermic reactions increases, that is a consequence of increasing propane
A product mixture from quenching experiments with propane was analysed, and conversion of
propane, yield and selectivities to products were calculated. Ethylene and propylene are considered
to be most valuable products. Ethylene yield increases with conversion and is around 37 % at 100 %
conversion. Propylene yield goes through a maximum at 55 % conversion and approaches zero at
100 % conversion.
The main identified problems for the proof of concept study are a poor gas mixing, high
temperature gradients through the quencher and high heat loss from the quencher part. In addition,
problems with connection ceram tube – quencher at high temperature have occurred.
Front page
Table of content
1. Introduction
1.1 Increase in energy demand and use of natural resources
1.2 Trends for more energy-efficient solutions (processes)
1.3 Synthesis gas production
1.4 New way to energy recovery
1.5 Background for the project: concept of chemical quenching
1.6 Scope of the work
2. Theory and Literature
2.1 General about quenching and quenching methods
2.2.1 Gas mixing
2.2.2 Turbulent vs. Laminar flow
2.2.3 Velocity profiles in cylindrical tube
2.2.4 Mass and energy balance over control volume, CV
2.2.5 Calculation of heat loss over CV
2.3 General about hydrocarbons pyrolysis
2.3.1 Pyrolysis of methane
2.3.2 Thermal cracking of propane
2.4 Synthesis gas: Reactions and thermodynamic data
2.5 Temperature measurements at high temperatures
3. Apparatus and Procedures
3.1 Pyrolysis equipment: Pyrolyserigg
3.1.1 Electrical furnace
3.1.2 Temperature regulation
3.1.3 Gas lines
3.1.4 Cooling system (Quencher/Cold Finger)
3.2 Analysis equipment
3.2.1 Gas Chromatograph (GC)
3.2.2 Temperature measurements
3.3 Pyrolysis of methane
3.3.1 Procedure: methane pyrolysis
3.3.2 Procedure: reactor regeneration
3.4 Modification of cooling system
3.4.1 Volumes of different sections of the cooling system
3.4.2 Detailed sketch of the modified cooling system
3.5 Experiments with a hot inert gas – nitrogen, «quenched»
by methane or propane
4. Results
4.1 Pyrolysis of methane: conversions, selectivities and yields
4.2 Experiments with hot inert – nitrogen. Practical knowledge about
quenching system. Problems.
4.2.1 Limitation of maximum obtainable temperature in the quencher
4.2.2 Connection Quencher – Ceram Tube
4.2.3 Coke formation and plugging of the quencher
4.3 Experiments with hot nitrogen quenched by propane
4.3.1 Conversions, selectivities and yields
4.3.2 Quenching effect: measured and expected temperature drop.
5. Discussion
5.1 Pyrolysis of Methane: verification of previous experiments
5.2 Experiments with hot nitrogen quenched by propane
5.2.1 Conversions, selectivities and yields
5.2.2 Quenching effect: measured and expected temperature drop.
5.3 Quencher Design
5.3.1 High temperature gradients and high heat losses
5.3.2 Bad mixing of hot and cold gases
6. Conclusions and Recommendations
6.1 Conclusions
6.2 Recommendations
7. References
Appendix A1: Example of calculation of Reynolds number NRe for the quencher and for
the ceram tube
Appendix A2: Calculations of conversions, selectivities and yields from the methane
Appendix B: GC – method. Calibration of GC. Example of retention times from the
quenching experiments with propane
Appendix C: Calibration_Curves(Pyrolysis).xls
(Calculation of heat loss and
expected temperature drop)
(summary of results with propane)
Quench_Propane(1200 C).openDocument-regneark
Quench_Propane(1300 C).openDocument-regneark
Quench_Propane(1450 C).openDocument-regneark
Results_Methane Pyrolysis.xls
(summary of results from methane
Set i)_Methane_Pyrolysis.xls
Set ii)_Methane_Pyrolysis.xls
Set ii)_Methane_Pyrolysis.xls
Tg vs. Tw.xls
(Calculation of true gas temperature as a function of
wall temperature)
Appendix D: Technical drawing of furnace
1.1 Increase in energy demand and use of natural resources
Economic growth and expansion of world's population are the two most important driving forces
for the increasing use of natural resources and for the increasing world's energy demand. As amount
of people with a more income increases, the production and consumption of energy will also
increase [1]. The International Energy Outlook 2011 (IEO) prepared by the U.S. Energy
Information Administration predicts that the world's energy consumption will increases by 53 %
from 2008 to 2035. This corresponds to 1.6 % increase per year. Developing countries will
contribute for the main part of the rise in energy demand.[2]
Figure 1.1 World energy consumption, 1990-2035 (quadrillion Btu)[2]
OECD – Organization for Economic Cooperation and Development (members)
According to IEO the consumption of natural gas, oil, coal and renewable resources will continue to
increase in order to cover the world's energy demand. To meet the challenges of the future, it is an
absolute necessity to develop new technologies and improve existing technologies for energy
production and processing of natural resources.
1.2 Trends for more energy-efficient solutions
Saved energy is an earned energy. Improving energy efficiency of existing processes is one way to
save or reduce energy consumption. Making a process more energy-efficient may often result in
reduction of emissions and wastes from the process, reduction of operating cost and may give
financial gains for the company. Many companies have a policies directed to reduce their emissions
and save energy consumption. Depending on particular process, different opportunities exist to
make it more energy effective. In chemical industry improvement of processes producing chemicals
can be reached by development and use of novel materials and catalysts. Energy efficiency may be
increased by properly process design and process integration that maximize heat recovery and
minimize energy consumption. Development of more sophisticated processes using advanced
technologies is a trend and need of the future.
An example of successful energy-efficient solution is Combined Heat and Power plants (CHP).
CHP plants produce electricity and heat in the form of steam that can be used as a process steam or
used for district heating. CHP plants may reach electrical efficiencies of more than 50 % and total
efficiency may be as high as 90 % [3].
Just to mention, when efficiency of a process is discussed, the term energy efficiency may be
somewhat misleading if energy quality is not taken into account. Thus, when two or more process
alternatives are compared, the exergy analysis should be performed to get an answer about which
alternative is the best.
1.3 Synthesis gas production
Synthesis gas is a general name used to describe gas mixtures that contain various amounts of
hydrogen and carbon monoxide. Other important components of synthesis gas mixture may be
carbon dioxide, water and methane. Synthesis gas is a key intermediate for preparation of base
chemicals like hydrogen, ammonia, methanol and Fischer-Tropsch fuels. In addition, it is used
directly as a fuel gas in power generation by the gas turbine. Three main processes for the syngas
production are steam reforming of natural gas, partial oxidation of heavy hydrocarbons with steam
and oxygen, and gasification of coal.[4] Biomass gasification is also an important rout for synthesis
gas production that is in constant development.
Depending on the process configuration, the exit temperature of raw syngas out of reactor may vary
significantly. Typical exit temperatures for different syngas production methods are given in Table
Table 1.1 Typical syngas outlet temperatures from reactor unit
Texit [°C]
Steam reforming of
methane [4]
Higher temperature is
preferred, but material
constraints is limiting factor
Partial oxidation of heavy
hydrocarbons [5],[6]
1300 - 1400
Contains particles of residual
carbon and ash
Gasification of coal*[4],[7] 1300 - 1500
Ash (or slag) is in liquid
*Entrained flow gasifier
As shown in Table 1.1 syngas is produced at high temperatures and contains a lot of energy in the
form of heat. The heat should be removed in order to meet downstream applications. For the case
when syngas is produced in entrained flow gasifier, it is possible to recover 5 – 25 % of the energy
in the feed, relying on the applied technology [8].
Different configurations of syngas cooling and heat recovery system exist. The radiant and/or
convective heat exchangers may be used in combination with direct quench systems, in which water
or cool recycle gas are introduced into the hot raw synthesis gas. Afterwards, a heat recovery at
lower temperatures occurs typically through a series of heat exchangers and different quality steam
is produced.[7],[8]
In the case of coal gasification the outlet temperature of syngas is very high, and ash or (slag) is in
liquid form. Fouling of downstream process equipment can be a serious problem, thus a quench is
necessary to solidify the slag. Four main options is identified for quenching [7]:
2. Radiant syngas cooling
3. Water quench
4. Gas recycle quench
5. Chemical quench
Normally, a first step of syngas treatment is a quenching by one of methods mentioned above from
1500 °C to around 900 °C with a subsequent heat recovery in syngas coolers (fire tube boilers or
water tube boilers) by steam production. During the water quench and gas recycle quench a thermal
energy is degraded to a lower level and energy recovery is limited.
1.4 New way to energy recovery
In principle, the syngas or any other hot process stream may be quenched at the same time as
thermal energy is used to produce other valuable products. For example, if light alkane is used as a
quenching medium, the quenching will occur by dilution with cold alkane and by endothermic
reactions of alkane. The valuable products will be a simple olefines and hydrogen. This may give a
more effective heat utilization than the production of steam. The energy will be recovered in the
form of valuable products. The quench process where quenching medium undergo chemical
reactions may be referred as chemical quench. Chemical quench may be applied in combination
with the heat recovery by steam and the total efficiency of the energy recovery may be increased.
On the global level, the improvement of efficiency of syngas quenching may be an important step
on the way to reduction of energy consumption
1.5 Background for the project: concept of chemical quenching
Background for the project was idea about chemical quenching. The concept of chemical quenching
is not very widespread in industry and not so much literature was found about the topic. L. Dessau
and H.-J. Spangenberg have performed studies of a plasma acetylene process. In this process the
reactive mixture of H2, CH4, C2H4 and Cn compounds at the temperatures of about 2200 K was
quenched by injection of liquid hydrocarbons such as gasoline or discrete n-alkanes. The process
uses high enthalpy of hot gas mixture to convert quench-hydrocarbons to simple olefines. At the
same time the reactive gas mixture is cooled very fast to avoid the consecutive reactions, such as
coke formation. The quenching process with liquids can be divided in physical (formation, heating,
evaporation of droplet) and chemical parts (cracking).[9]
One of the recent developments is a concept of chemical quenching of hot synthesis gas coming
from an entrained-flow gasifier. The temperature of synthesis gas out from a such gasifier is around
1500 °C and synthesis gas should be cooled down to temperatures around 900 °C. This is because
raw synthesis gas, produced from coal based feedstock, contains ash and slag on liquid form. These
contaminants should be removed to protect downstream process equipment from fouling. By
introducing a chemical quench or a second non-slagging gasification stage, the temperature out
from gasifier can be reduced sufficiently, and ash or slag become sticky and easy to separate. At the
same time the energy of the hot synthesis gas is used in endothermic reactions to gasify a secondstage feed. This results in increased cold gas efficiency. The quenching medium may be a dry feed
or slurry feed as in E-Gas process. In the case the feed is the coal-water slurry, the energy is used to
heat up the quenching medium, to evaporate the water and for the pyrolysis of coal.[7],[10]
A similar principle has been demonstrated by CHOREN in so called Carbo-V® process. The charcoal
produced in the first gasification stage of biomass is used as a quenching medium in the second
gasification stage [11].
1.6 Scope of the work
Primary aim of the project work was to investigate if light hydrocarbons, for example propane, can
be used as a quenching medium to quench very hot synthesis gas. Initial tests of the concept should
be performed and quenching effect (temperature drop), conversion of used hydrocarbon, obtained
products and other aspects should be studied.
Previous to this, it was necessary to perform a series of methane pyrolysis experiments in order to
validate that the experimental set-up (Pyrolyserigg) is working properly, and to verify earlier
experimental results from the methane pyrolysis. In order to perform quenching experiments the
cooling system/cold finger was modified and new quencher part with opportunity for hydrocarbon
injection was installed. A series of experiments with N2 as a hot inert gas was performed with the
purpose of study the behaviour of modified quencher. A series of experiments with hot N2
quenched by propane was done to study the quencher performance.
Experiments with a hot syngas quenched by propane were planned, but were not done because one
of the heating elements of the high temperature furnace was broken.
2.1 General about quenching and quenching methods
Many processes for the preparation of chemicals are carried out at high temperatures. In some cases
it is necessary with a rapid cooling of chemical substances or process flows. In this context the rapid
cooling refers to a quenching. Depending on particular process the main objectives of quenching are
heat recovery from process streams, temperature control in reactors, prevention of consecutive
reactions, retention of product composition and increase in a process efficiency.[4]
Quenching can be direct or indirect, dependent on the way it is done. Indirect quenching is often
used for the generation of high-pressure steam by special designed heat exchangers. In the case with
indirect quenching there is no direct contact between heating medium and quenching medium.
Direct quenching is associated with injection of solids, liquid or gas into the heating medium. Direct
quenching can be very efficient and high cooling rates can be achieved.[4] The cooling rate is often
a measure of performance of quenching process[12].
Sundstrom and DeMichiell have investigated the following quenching techniques: mixing with a
cold gas, injection into a fluidized bed, contact with cold surface and evaporation of liquid spray.
They concluded that all techniques are able to provide cooling rates greater than 106 °R/sec, but
factors such as product recovery, energy recovery and scale-up should be considered, in selecting a
technique. Gas mixing technique gave the fastest quenching rate. It is difficult to compare
quenching methods because of the specific parameters associated with each method. In the Figure
2.1 a temperature decay curves are illustrated at typical quenching conditions.[13]
Figure 2.1 Temperature decay curves at typical quenching
conditions [13]
In the case of coaxial mixing of hot and cold gases Figure 2.2, the most important factors for
cooling rate were the diameter of the hot jet and the ratio of coaxial to jet velocity. A smaller hot jet
diameter and higher velocity ratios increased the cooling rate by reduction the time necessary to
establish turbulent mixing processes.[13]
Figure 2.2 Schematic diagram of quenching by coaxial gas mixing [13]
Energy recovery is generally preferred at high temperatures. The high temperature conditions result
in high driving forces for the heat exchange processes. At very high temperatures the material
constraints can be a problem. During quenching by gas mixing or liquid spray the thermal energy
present in the process stream is degraded to a lower level. In addition, dilution of the process stream
by cooling medium occurs. The consequence of dilution may be an increase in separation costs of
final gas stream. This problem can be avoided by using cold product stream as cooling medium by
sending it in the return to the quencher. In principle the cooling medium can be a reactive
component or mixture of reactive components that undergo endothermic reactions. In this case the
quenching may result in reduction of temperature and production of desirable products. [13]
2.2.1 Gas mixing
Transport phenomena such as mass, heat and momentum transport should be considered when
flows of hot and cold gases are mixed together. To obtain a high quenching effect, it is necessary
with a rapid and completely mixing of hot and cold gases. The mixing of gases will occur mainly by
to mechanisms: molecular diffusion and convective transport.
The molecular diffusion can be explained by the kinetic theory of gases. The gas molecules are in
rapid random movement and often collide with each other because of their kinetic energy. Because
molecules move randomly in all directions, there are fluxes in all directions. When concentration
gradients occur in the bulk gas phase, there is a net flux of molecules from a region with a high
concentration to a region with a low concentration. The concentration gradient is a driving force for
diffusion transport.[14] In the case when two gas flows in a tube in laminar regime and have the
same velocity the mixing will occur mainly by molecular diffusion through the boundary layer.
On the other hand, the convective transport of gas occurs by the forced movement of the gas where
the ensembles of molecules move in concerted and collective motion. The convective transport may
occur at all scales that are larger than a few atoms. The cause for the convective mixing of the gas
flow inside the tube or reactor is viscous forces and shear stresses. When gas velocity is sufficient
high, the flow pattern is unstable and eddies or small packets of fluid particles are presented. These
are moving in all direction and at all angles to the normal line of flow. When high degree of
fluctuations exists, the gas flow is said to be in turbulent flow regime. In the turbulent flow regime
the mixing of gases will occur mainly by convective transport.[14]
The mixing of hot and cold gas flows will depend on flow conditions such as temperature, gas
velocities, viscosities, gas densities and geometry of mixing device. Turbulent flow regime will give
a better gas mixing than laminar flow in which there is no lateral mixing of the gas.
2.2.2 Turbulent vs. laminar flow
To predict where a gas flow is in laminar or turbulent regime, the Reynolds number can be used,
which is dimensionless. For the tube geometry the Reynolds number defined as:
N Re 
D – inner diameter of the tube
υ – average velocity of the fluid
: defined as volumetric rate of flow divided by the cross – sectional
area of the pipe
ρ – fluid density
μ – fluid viscosity
«For a straight circular pipe, when the value of Reynolds number is less than 2100, the flow is
always laminar. When the value is over 4000, the flow will be turbulent, except in very special
cases. In between – called the transition region – the flow can be viscous or turbulent, depending
upon the apparatus detail, which can not be predicted». [14]
2.2.3 Velocity profiles in cylindrical tube
The knowledge about velocity profile of gas flow inside the tube (Quencher) can be important to
explain the temperature measurements. For a simple tube geometry and laminar flow at steady state,
the expression for velocity profile can be derived by making a shell momentum balance and using
equation for the definition of viscosity. The following equations can be obtained [14]:
  r 2 
 x  2 x ,av 1    
  R  
 x ,max  2 x ,av
υx – flow velocity in x direction [m/s]
υx,av – average flow velocity for a cross section [m/s]
r – distance from the tube center in radial direction [m/s]
R – tube inside radius [m]
υx,max – maximum flow velocity in the tube center [m/s]
For the fluid flow the velocity profile is parabolic as represented in Figure 2.21
Figure 2.2.1 Velocity and momentum flux profiles for laminar flow in a tube [14]
It has been shown by experiments that fluid moving in the center of the tube is moving faster than
the fluid near the walls. For laminar flow, the velocity profile is true parabola. The velocity at wall
is zero. For turbulent flow the velocity profile is somewhat flattened in the center as shown in
Figure below.[14]
Figure 2.2.2 Velocity distribution of fluid across a tube [14]
A certain length of tube is necessary to establish a fully developed velocity profile at the entrance
region of the tube. This length is called for entry length, Le, and can be approximately calculated as
(for laminar flow)
Le  0, 0575DN Re
For turbulent flow an approximation is that the entry length is nearly independent of the Reynolds
number and is around 50 times of the tube diameter.[14]
2.2.4 Mass and energy balance over control volume, CV[14]
For heat-transfer systems, a steady state condition means that there is no temperature change at any
given point and heat fluxes are constant over the time. With assumption about steady state, a simple
energy or heat balance over control volume (CV) or over a system can be written on the form
in ,i
  H j ,reac  Q   H out ,i
Hin,i – enthalpy of flow of component i into the CV [J/s]
ΔHj,reac – heat produced or consumed i reaction j
Q – net heat added or removed from CV
Hout,i – enthalpy of flow of component i out of CV [J/s]
A sketch over the region of quencher considered as CV is illustrated in the Figure 2.2.3
Figure 2.2.3 Control Volume (CV) considered in the heat and mass balances. The
boundaries in axial direction are at temperature measurement points
The enthalpy of flows of components defined as:
H in,i  Fin ,i Cpi (Tin  T 0 )
H out ,i  Fout ,i Cpi (Tout  T 0 )
Fin,i – molar flow of component i into CV
Cpi – heat capacity of component i
Tin – flow temperature into CV
Tout – flow temperature out of CV
T0 – reference temperature at
standard conditions
Fin,i is calculated, using ideal gas low, from the volumetric flow of component i, given by MFC:
Fin,i 
P – standard pressure
Vin,i – volumetric flow of component I
T – temperature at standard conditions
R – universal gas constant
The flow of component i out of CV is calculated using experimental data for conversion of
component i:
NB! The component i will to some extent react outside of CV. This means that the obtainable conversion inside the CV
is lower than conversion predicted by experiments. This can be a source of error for heat balance over CV.
Fout ,i  Fin ,i (1  X i )
Where Xi – total conversion of component i
2.2.5 Calculation of heat loss over CV
When approximately steady state condition for cooling system is reached, the heat loose over CV to
the surroundings can be calculated. The temperature is measured for the gas flows in and out of CV.
There is no reaction, since only inert hot gas is introduced during the heating period. The heat loose
is a difference between enthalpies of incoming and out coming gas flows:
Qlose   Fout ,i Cpi (Tout  T 0 )   Fin ,i Cpi (Tin  T 0 )
The calculated heat loss is a heat loss at actual experimental conditions and will be dependent on
amount of gas passing CV, the gas temperature and isolation layer around the quencher.
When reacting or inert cold gas are introduced into the quencher, the established temperature profile
through the cooling system and CV, will change. The heat loss will also be changed. It will probably
be reduced because of reduction of temperature. However, this change is considered to be negligible
and heat loss is assumed to be constant.
When heat loss over CV and conversions to products (product flows) are known, it is possible to
calculate expected temperature out of CV. This is a way to calculate expected quenching effect
(temperature drop) of cold gas injection.
By rearranging equation (8), the expected temperature out of CV calculated as
in ,i
Tout  T 
Cpi (Tin  T 0 )   H j , reac  Qlose
out ,i
 F Cp
out ,i
out ,i
2.3 General about hydrocarbons pyrolysis
Pyrolysis or thermal cracking of hydrocarbons is a widely applied process. When hydrocarbons
from natural gases, refinery gases or petroleum are heated up to a sufficient high temperature,
thermal cracking take place. When pyrolysis of hydrocarbons occurs in the presence of steam,
steam acts as a diluents, the process called for steam cracking. The higher alkanes are converted to
olefines in a high yields. The final product mixture is complex, but ethene is considered to be the
main product. Depending on feedstock and working conditions many other olefines and aromatics
are produced: propene, butenes, butadiene, benzene, toluene, etc. In addition, it can be formed a less
valuable products such as methane, heavy pyrolytic naphthas and coke. The coke formation should
normally be avoided because of associated fouling problems.[4],[15]
Steam cracking of hydrocarbons is a high temperature process, depending on feedstock a
temperatures higher than 600 °C are applied [15]. Because the overall process is endothermic,
considerable heat input at high temperature level is necessary. The primary reactions in hydrocarbon
pyrolysis are dehydrogenation of alkanes and cracking of long chain alkanes to lower alkenes and
(dehydrogenation reaction, ΔHr >0)
C2nH2n+2↔C(n-m)H2(n-m)+CmH2m+2 (cracking reaction, ΔHr >0)
The primary products of pyrolysis may undergo secondary reactions like further pyrolysis,
dehydrogenation, condensation and coke formation.[4] Thermal decomposition of hydrocarbons can
be described by free-radical mechanism and using a set of elementary reactions. For example, it was
used 35 reversible and 1 irreversible elementary reactions to describe pyrolysis of methane in
temperature range 1200 – 1500 °C [16]. L. Dessau has used 38 equation in reaction kinetic
modelling of n-hexane, when it was used as a quenching agent for high temperature conversion of
methane [9].
In general, smaller alkanes are more stable than higher alkanes. Thus, the pyrolysis of smaller
alkanes requires a higher temperature and higher heat input to obtain a given conversion. Both
dehydrogenation reactions and cracking reactions produce two molecules for every converting
alkane molecule. As a consequence, the conversion of alkanes is dependent on the partial pressures
of reactant and products.[4] The Figure 2.3.1 illustrates how the equilibrium conversion for some
alkanes is affected by temperature and total pressure.
Figure 2.3.1 Equilibrium conversion of smaller alkanes [20]
Surface effects
The experimental results of methane pyrolysis and pyrolysis of other hydrocarbons may be affected
by the surface effects like heterogeneous reactions, occurring on the reactor wall. To minimize
these surface effects, the surface-to-volume ratio S/V of reactor should be kept low.[16]
Makarov et al. studied methane pyrolysis in a flow system in the temperature range 815-1100 °C.
They concluded that at S/V ratios less than 20 cm-1 only homogeneous reactions were present. This
means no surface effect. At S/V ratios greater than 104 cm-1 homogeneous reactions were
The surface effects will also be dependent on catalytic activity of the surface, and therefore, on the
type of material used in reactor. The reaction rates are higher in iron than in quartz or gold reactors .
The tendency to coking is higher in iron, monel or cobalt than in quartz, porcelain, silver or gold
reactors. The pyrolysis of n-alkanes in a stainless steel reactor favour formation of ethene versus
methane and propene, than pyrolysis in quartz reactor working at comparable conditions.[15]
The deposition of coke on the reactor surface may be s serious problem in the pyrolysis studies
because the surface property is changed continuously as coke is formed [15]. In addition, an
uncontrolled coke formation may cause the plugging of the system.
Other possibility to minimize the surface effects is the passivation of inner surface with surfaceactive agents [15]. [17] has used CS2 to deactivate chromium steel when thermal cracking of
propane was studied. [21] has used mixture of N2, H2 and CH4 to passivate the ceram reactor after
the coke burn-off.
2.3.1 Pyrolysis of methane
Ola Olsvik has studied the conversion of methane into C2 compounds at the temperature region
(1000 – 1500 °C). At temperatures above 600 °C thermodynamic equilibrium favour formation of C
and H2, thus the residence time inside reaction zone become an important parameter to consider.
The quenching step of methane pyrolysis is also important for product distribution. “The pyrolysis
of methane may be described as stepwise dehydrogenation at high temperature”[16]:
2CH4 → C2H6 → C2H4 → C2H2 → 2C
+ H2
The main products of methane pyrolysis are ethyne, ethene, benzene, hydrogen and carbon.
Ethane is the primary product. Ethene, ethyne and benzene are secondary products, and at high
residence times tar and coke are also produced. The dilution by inert gas and applied short residence
times, reduce the opportunity for condensation reactions and formation of heavier compounds. [16]
The pyrolysis of methane is explained by free radical mechanisms. The formation of methyl radical
and hydrogen radical is an initiation step. The reaction has high activation energy and supposed to
be rate-determining [16]:
CH4 → CH3· + H·
The overall reaction rate of methane can be considered as a 1. order reaction [16],[22]:
r = k · CCH4
The methane conversion and the ratio of ethyne to ethene have been shown to increase with
increasing temperature. The Figure 2.3.2 shows how conversion vary as a function of residence
time for different temperatures.
Figure 2.3.2 Conversion vs. Residence time for different temperatures.
Ptot= 1bar, di =9 mm, except at 1500 °C where di = 4 mm, H2:CH4 = 2:1 [16]
The Figure 2.3.3 shows how the maximum yields of products: ethyne, ethene and benzene vary with
increasing temperature.
Figure 2.3.3 Yields as a function of temperature, H2:CH4 = 2:1[16]
“The rate of pyrolysis decreases markedly in the presence of hydrogen”[22]. Hydrogen dilution has
an important effect on the kinetics of the methane pyrolysis. The effect of hydrogen dilution
becomes more important at higher reaction temperatures. Hydrogen strongly depresses carbon
formation and increases the yield of ethyne [16],[22]. The Figure 2.3.4 illustrates how the
conversion depends on hydrogen dilution for different residence times.
Figure 2.3.4 Conversion of methane as a function of residence time for
different hydrogen dilutions. T = 1300 °C[16]
2.3.2 Thermal cracking of propane
G. Buekens and F. Froment [17] have performed studies of pyrolysis of propane for temperature
region between 625 and 850 °C, and near the atmospheric pressure. They concluded that the main
products of thermal cracking of propane are methane, ethylene, propylene and hydrogen. “Ethane is
also partly a primary product, but it is produced in much smaller quantities”. In addition, butenes,
butadiene and aromatics are formed.[17]
Cracking of propane can be explained by two parallel decomposition reactions [17]:
C3H8 ↔ C2H4 + CH4 (cracking)
ΔHr0 = 83 [kJ/mol][19]
ΔHr(700 °C) = 78,2 [kJ/mol][18]
C3H8 ↔ C3H6 + H2
ΔHr0 = 125 [kJ/mol][19]
ΔHr(700 °C) = 129,3 [kJ/mol][18]
Table 2.1 illustrate slectivities to primary products at zero conversion for several temperature
Table 2.1 Primary product distribution at Zero Conversion [17]
As conversion of propane is increasing, the selectivity for ethylene and methane is also increasing,
but selectivity toward propane is decreasing. The Figure 2.3.5 illustrates how selectivities to
primary products change with increasing conversion for temperature region 725 – 750 °C. [17]
Figure 2.3.5 Selectivity diagram. (Temperature 725 – 750 °C) [17]
V.R. Choundary and V.H. Rane [18] have investigated thermal cracking and non catalytic oxidative
conversion of propane in the presence of steam at various process conditions. They concluded that
the thermal cracking of propane happens to a considerable level only at high temperatures (>700
°C). At lower temperatures the conversion was very small (< 7 %). At temperatures higher than 700
°C, the propane conversion and selectivity for propylene increased markedly. The selectivity for
ethylene and methane decreased with increasing temperature, Figure 2.3.6. The coke formation on
the reactor walls and/or a tarlike product on the cooler parts of reactor was observed.[18]
Figure 2.3.6 Influence of temperature on the thermal cracking of propane [18]
2.4 Synthesis gas: Reactions and thermodynamic data
Table 2.2 presents the main reactions for the production of syngas by steam reforming. Almost the
same reactions may be used to describe the coal and biomass gasification. Because different
feedstock have a different composition, the various ratios of H2O/CH4(C) and O2/CH4(C) are
applied for syngas production. The reactions equations in table 2.2 are not independent.[4] The
synthesis gas system may be described by only three of independent equations [20].
Table 2.2 Reactions during methane conversion with steam and/or oxygen [4]
The Figure 2.4.1 shows the thermodynamic equilibrium for the steam reforming of methane and for
the partial oxidation of methane in the case with stoichiometric feed and 1 bar pressure.
Figur 2.4.1 Equilibrium gas composition at 1 bar as a function of temperature (a) steam
reforming of methane; H2O/CH4 = 1 mol/mol; (b) partial oxidation of methane;
O2/CH4= 0.5 mol/mol [4]
From the figure above can be concluded that the synthesis gas mixtures can be quenched to
approximately 1000 K without considerable change in equilibrium gas composition.
2.5 Temperature measurements at high temperature
Temperature measurement at high temperature can be a difficult task, and significant errors can
occur. Temperature gradients are the main driving force for the heat transfer. Heat transfer may
occur by conduction, convection or radiation. In many cases more than one of heat transfer
mechanisms are involved.[14] Processes performed at high temperatures can give a rise to a high
temperature gradients in measurement field/points. The consequence may be a wrong temperature
measurement or wrong interpretation of the measurement.
In the case, when temperature measurement of flowing gas in a tube is taken by temperature sensor
(thermocouple), both the convective heat transfer and heat transfer by radiation should be taken into
account. When hot gas is passing the sensor, the convective heat exchange will occur between the
sensor and the gas, and heat exchange by radiation will take place between the sensor and the wall
of the tube. The sensor will indicate a temperature between true gas and wall surface temperatures.
The Figure 2.5.1 shows how heat is transferred, where the wall temperature Tw is lower than the
true gas temperature Tg, (Tw < Tg).Tp is a measured temperature.[14]
Figure 2.5.1Temperature measurement of a gas showing radiative and
convective heat transfer for a sensor. [14]
At steady state conditions a net heat balance for the sensor can be written as[14]
qc =qr
qc – rate of convective heat transfer to the sensor from the gas [J/s]
qr – net radiation heat from the sensor to the wall [J/s]
The convective heat transfer can be expressed as
qc = hcAp(Tg – Tp)
and heat transfer by radiation can be expressed as
qr = ɛσAp(T4p – T4w)
hc – convective heat transfer coefficient
Ap – surface area of the sensor
ɛ – emissivity of the sensor surface, < 1
σ – constant
(17) can also be written on the form
Tg  Tp 
 (Tp4  Tw4 )
The difference between true gas temperature and measured temperature can be calculated by trial
and error, when all parameters are known.
However, the coefficient hc is a function of the system geometry, fluid properties, temperature
difference between gas and sensor, and flow velocity. True gas temperature Tg will also be
dependent on flow conditions like gas velocity profile and will vary in radial direction when
temperature gradient is present. For turbulent flow the temperature is more uniform in the bulk gas
flow because of turbulent mixing, occurring on some distance from the tube wall. For laminar flow
conditions, the temperature gradient will be more definite in radial direction.[14]
Emissivity ɛ is the ratio of the emissive power of a surface to that of perfect black body. The surface
emits a radiation, depending on its temperature. For all real materials the emissivity is lower than
one, ɛ < 1. Emissivity for polished metal surface is low and high for oxidized surface.[14]
During continuously temperature measurements, the surface of the sensor may be exposed to
conditions that will change its emissivity. For example, the surface may be oxidized or coke
deposits may be formed on the surface.
Both convective heat transfer to the sensor and radiant heat exchange with the wall, and as
consequence the measured temperature may vary significantly with a flow conditions. High
temperature «gives» high temperature gradients. That all can affect the temperature measurements.
3.1 Pyrolysis equipment: Pyrolyserigg
Pyrolyserigg is a platform that gives opportunity to perform experiments with gases at high
temperature reaction conditions, up to 1500 °C. The main parts of the platform include electrical
furnace, power supply-and control system, gas lines, gas storage cylinders, quenching system of hot
gas mixture, equipment for analysis of product gases (GC-gas chromatograph) and gas alarm
system. The platform is connected to central ventilation system in order to get the product gases out
of the facility and to avoid accumulation of the gases inside the platform by eventual leakage. A
schematic diagram of experimental apparatus is shown in Figure 3.1.1, and picture of the
pyrolyserigg is shown in Figure 3.1.2
Figure 3.1.1 A schematic diagram of experimental apparatus with some modifications [21]
Figure 3.1.2 Pyrolyserigg and temperature control box to the write
3.1.2 Electrical furnace
The electrical furnace (Kanthal) has a vertical length of 60 cm and gives a total power output of 7
kW. The furnace is placed inside the rig in such a way that a ventilation aperture located above the
furnace. This is done with regard to ensure a safe operation, since the probability for leakage in
connections with the reactor tube is high. The furnace has two small apertures on the side. These are
used to set thermocouples S-type (Rh-Pt) for temperature measurement and control. Schematic
drawing of the furnace is in Appendix D
Figure 3.1.3: Electrical Furnace
3.1.2 Temperature regulation
Power supply and temperature control are provided by the «regulation box», which was delivered
by Siemens. It was found out that temperature control is best at high temperatures. At temperatures
lower than approx. 1000 °C the temperature inside the furnace is unstable and swings with
amplitude of approx. 30 °C in the measurement point. This is probably due to programming of
«regulation box».
3.1.3 Gas lines
As shown in Figure 3.1.1, the gas supply to the reactor may occur by four gas lines. During
experiments the gases were metered from storage cylinders, placed outside the rig. Two of gas lines
(H2, CH4) are connected to central gas supply system with opportunity to provide gases from this
system. At the moment of experimental work, the central gas supply system was not in use, with
exception of technical air which was supplied to GC. If necessary, the line for introduction of liquid
hydrocarbons to the reactor may be easy connected.
In experiments after modification of the quenching system, the “nitrogen line” was connected
directly to the modified quencher and was used to introduce quenching gas.
3.1.4 Cooling system (Quencher/Cold finger)
The main path of the cooling system is the Quencher/Cold finger. The gas mixture from the reactor
outlet passes through the quencher where a rapid cooling of the gas mixture happens. The quencher
is of indirect type and water is used as a quenching medium. The quencher is connected to the
central cooling system through which the cooling water is circulating. After that methane pyrolysis
experiments were completed, the cooling system was modified. The sketch of quencher used in
methane pyrolysis experiments is shown in the Figure 3.1.4.
Figure 3.1.4 Quencher [21]
3.2 Analysis equipment
3.2.1 Gas Chromatograph (GC)
Gas chromatograph is a device for analysis of product gases. The GC is of the type (HP 5890 Series
2) equipped with a Thermal Conductivity Detector (TCD) (Carbosieve S-2) and a Flame Ionization
Detector (FID) (GS-Q). The GC is the old device and some problems have occurred during its
exploitation. Some of valves are not working, but it is still possible to use GC.
Figure 3.2.1 GC- Gas Chromatograph
3.2.2 Temperature measurements
One thermocouple S-type (Rh-Pt) was placed in upper aperture on the furnace, sticking
approximately 5 mm from the inside wall. The thermocouple was connected to the “regulation box”
which controlled temperature inside the furnace. It was observed that at high temperatures small
displacement in thermocouple's position gave impact on measured temperature. Another S-type
thermocouple was placed in lower aperture and was connected to the temperature Logger.
Temperature measurements inside modified quencher were taken by two K - type thermocouples,
placed approximately in the middle of the quencher.
Temperature measurements during experiments were saved on PC with exception of Runs 4-12 for
methane pyrolysis. The PC was out of stand.
3.3 Pyrolysis of methane
One of suggestions in the end of specialization project [23], which form the basis for the present
work, was to continue with methane pyrolysis. The pyrolysis experiments have been repeated
because of some inconsistencies during experiments in the specialization project. The experimental
conditions were chosen in the way to verify the results that was obtained before[21], and are listed
in Table 3.1.
Table 3.1: Experimental runs. Reaction temperature, feed composition by MFC, inner diameter of
the reactor tube, residence time (at normal conditions), reactor pressure (during GC-analysis) for
different runs.
Temp. Temp.
Pinlet** τ***
[Nml/min] [Nml/min] [Nml/min] [mm]
[barg] [ms]
Prøve [º C]* (logger)
2,2-3,2 49
1,6-2,4 49
* Temperature is given as reference temperature at upper measurement point at furnace wall
(thermocouple sticking approximately 5 mm from the inside wall of furnace)
** Reactor pressure varied during GC-analyses for Runs 1 and 2. The results from these runs are
not taken into account.
***Residence time is calculated, based on estimation of reaction zone from specialization project
3.3.1 Procedure: methane pyrolysis
Pyrolyserigg is equipment that was used for the methane pyrolysis. Gas flows of CH4, H2 and N2
were adjusted by mass flow controllers (MFC) accordingly to planed Runs and in a way to give the
same residence time. High dilution with hydrogen and short residence time were applied in order to
avoid formation of coke. Two GC – analyses of feed gas were taken for each gas mixture
composition. The furnace was heated up to setpoint temperature. After 5-10 minutes 1. GC –
analyse was initiated. Three GC – analyses of product gas mixture were taken for each Run.
Temperature inside furnace, pressure to the reactor inlet and outlet were registered for each GC –
Experiments were started with a lowest methane concentration and nitrogen flow was held constant.
The alsint ceramic tube with inner diameter 3 mm was used as a reactor. The product gases were
quenched at the reactor outlet by annular water – cooled quencher. After passing the quencher,
product gases were filtered for coke particles in a box filled with a glass wool and afterwards, in a
metal sinter (15 microns).
When pyrolysis experiment was finished a small flow of N2 (50 Nml/min) was used to purge the
gas lines in approx. 30 min.
3.3.2 Procedure: reactor regeneration
In Runs 1-2 pressure build-up at reactor inlet was observed. This is an indication of extensive coke
formation and plugging of the reactor tube. Reactor was regenerated by burning off coke in air.
The methane storage cylinder was replaced by the air storage cylinder. The gas lines were emptied
of gas previously. After run 1 the reactor regeneration was performed by procedure given by [21]:
(20 Nml/min) of air diluted in N2 (250 Nml/min) at 1000 °C. Decoking was very slow, no marked
pressure decrease after four hours with regeneration.
It was decided to increase flow of air to (80 Nml/min) and temperature to 1100 °C. Cold water was
disconnected during decoking. Decoking was finished when pressure at reactor inlet was 0,2 barg.
The last regeneration procedure was applied after Runs 2 and 3.
[21] has mentioned that the Grafoil sealant may be exposed to oxidation when high flow of air is
3.4 Modification of cooling system
After the methane pyrolysis experiments were completed and the work of the Pyrolyserigg was
verified, the cooling system could be modified in order to perform quenching experiments with
hydrocarbons. The quencher with opportunity to introduce hydrocarbons into hot gas has been
designed. It was mentioned in [23] that the quencher design is of great importance in order to study
the quenching concept. The quencher should be designed in a way that gives a good gas mixing
properties, minimal heat loses to the room (operate in adiabatic regime), opportunity for
temperature measurement of outcoming gases and opportunity to remove coke deposits.
In the absence of practical knowledge about how such quencher could be constructed, a start point
was to make a simplest possible design. A sketch of the proposed quencher design is shown in the
figure below.
Figure 3.4.1 A sketch of proposed quencher design
This type of design should give opportunity to introduce cold hydrocarbon into hot gas that is
heated by high temperature furnace. At the same time, the temperature of the injected hot gas and
temperature at the outlet of the quencher could be measured. It was also proposed to install a nozzle
at the hydrocarbon inlet or use some packing material inside the quencher to improve gas mixing.
This idea was dropped because of expected coke formation that would result in plugging of the
quencher and/or hydrocarbon inlet.
The following aspects were considered for the choice of material of construction: tolerance for high
temperature, surface effects and easy-to-handle. Quartz or steel were proposed as materials. Quartz
material tolerates high temperatures and surface effects are reduced in comparison to steel. The
surface effects were considered to be of minor importance for the study of quenching concept at this
stage. Stainless steel was chosen as material of construction because it is easy-to-handle and its
tolerance to high temperatures. A path of Conax fitting was welded to the one end of the steel tube
to provide connection with a ceramic tube. On the other end of the tube a Swagelok fitting was
welded to connect the new quencher to the «cold finger». Ports for thermocouples and hydrocarbon
inlet were also welded to the tube as shown on the sketch above.
It was also necessary to modify the «cold finger». A piece of tube with a Swagelok fitting was
welded to the upper path of «cold finger». The inner part of the old quencher – actually, the cold
finger was removed. This was done because of limitation in vertical length that could be applied.
Otherwise, the furnace should be lifted up.
After removal of the inner part, the cooling of gases inside the «cold finger» would occur only by
outer shell, and cold wall quenching would be less effective. However, it was expected that the hot
gas would be quenched by hydrocarbon in the upper section, previously. A detailed sketch of the
modified cooling system (new quencher part + modified «cold finger») is shown in Figure 3.4.2.
3.4.1 Volumes of different sections of the cooling system
The cooling system can be divided into three sections:
1. section is a CV defined as in chapter 2.2.4. Ideally, the direct quenching of hot gas by
hydrocarbon should be performed inside this section.
2. section is a volume between CV and «cold finger». This section can be considered as «died
volume» because temperature measured above the section. The section should be minimal. Initially,
it was not planned to have this section: lower point for temperature measurement should be placed
write above «cold finger».
3. section is volume where the gas mixture is cooled indirect by water. The gas mixture should be at
approximately room temperature when it leaves this zone.
Table 3.2: Calculated volumes and residence times for different
sections inside cooling system
Volume [ml]
Residence time [ms] at
gas flow 6250 [Nml/min]
Section 1
Section 2
Section 3
3.4.2 Detailed sketch of the modified cooling system
Figure 3.4.2 A detailed sketch of modified cooling system.
3.5 Experiments with a hot inert gas – nitrogen, «quenched» by methane or propane
After modification of cooling system was finished and new quencher was installed, a set of
experiments with a hot inert gas (N2) was performed. Three types of experiments were made:
experiments with only nitrogen gas, experiments with nitrogen gas «quenched» by methane and
nitrogen gas quenched by propane. Experiments, where synthesis gas should be quenched by
propane were planed, but were not performed, because the furnace has been out of stand.
N2 gas was heated up in alsint tube, (di= 3 mm) in the furnace at different setpoint
temperatures, ranged from 1000 to 1450 °C, and was send through the cooling system. Different gas
flows were applied, 100 – 7500 [Nml/min].Temperature variations in two measurement points
inside the quencher were measured and registered by temperature Logger. The pressures into reactor
system and out of reactor system were registered. The main purpose of experiments with inert
nitrogen was to study the system's behaviour.
In experiment with methane, a cooling effect of introduction of cold inert was investigated.
It was expected that the methane acts as inert at applied conditions, and that the cooling of hot
nitrogen flow will occur mainly by dilution with cold methane.
The furnace was heated up to setpoint temperature 1200 °C. The N2 flow was setting to 6250
[Nml/min]. After one hour, approximately a steady state condition was reached, T1= 908 °C T2 =
605 °C. Than, CH4 at room temperature, 506 [Nml/min] was introduced to the quencher. Two GC –
analyses were taken afterwards. GC – analyses confirmed that the methane acts as inert.
(T1 - temperature at upper measurement point; T2 – temperature at lower measurement point
inside Quencher)
30 min after the methane was introduced, T1= 903 °C and T2=623 °C. Its mean that temperature
drop at upper point is 5 °C and temperature increase at lower point is 18 °C. The methane flow was
The N2 flow was reduced to appr. 2500 [Nml/min]. After 20 minutes T1 = 810 °C and T2 = 490 °C.
Again, 506 [Nml/min] methane was introduced and no cooling effect was observed.
Since experiments with methane did not give expected cooling effect, it was proposed to use
propane as a quenching gas. The quenching would occur both by dilution and by endothermic
reactions of propane.
The MFC, used for methane gas, was calibrated for propane. A maximum obtainable propane flow
with this MFC was 246 [Nml/min]. As in the case of experiment with methane, the furnace was
heated up to 1200 °C and nitrogen flow was setting to 6250 [Nml/min]. After 1,5 hour appr. steady
state was reached, T1=897 °C and T2=608 °C. Then, 246 [Nml/min] of propane gas was introduced
to the quencher.
90 minutes later, T1=895 °C and T2=621 °C. Again, no cooling effect was observed. GC - analyses
were failed at this experiment: problems with sample injection.
It was decided to increase the temperature of hot nitrogen and use higher propane flows in later
experiments. It was expected that cooling effect would increase because of higher dilution with cold
gas and higher conversions of propane at higher temperature. It was also necessary to change MFC
for propane with a MFC that could give higher flows. Experimental conditions in experiments
where hot nitrogen was «quenched» by propane and GC – analyses of product gases were taken, are
summarized in Table 3.3
Table 3.3: Experimental conditions in Nitrogen/Propane experiments (alsint tube used for heating
up N2, di = 6 mm)
Temperatu T1 [°C]ii) T2 [°C]ii)
τ [ms]iii)
re [°C]
[Nml/min] [Nml/min]
3. Setpoint temperature to the furnace
4. Measured temperature in upper and lower points inside quencher, given on the form (xxx-yyy), where
xxx is a temperature at «steady state»: before propane is introduced.
yyy is a temperature registered during the last GC – analysis for the run, after propane introduction.
5. Residence time for the gas mixture inside CV - defined as in chapter 2.2.4
* Run 22, only 1 GC – analysis is taken; problem with propane flow (probably because of plugging of inlet
or MFC failed)
4.1 Pyrolysis of methane: Conversions, Selectivities and Yields
The main objectives of the pyrolysis experiments were to verify the results, obtained in earlier
experiments, and to validate that the Pyrolyserigg work properly. Methane conversions, selectivities
and yields of products for different experimental conditions were calculated from gas
chromatographic analysis. Methane concentration present in the figures below is calculated from
TCD – measurements of feed gas mixture.
Three Sets of Runs accordingly to Table 3.1 were done:
i) Runs 1,2,3
ii) Runs 4,5,6,7,13
iii) Runs 8,9,10,11,12
The Figures 4.1.1 and 4.1.2 show how methane conversion increases with an increase of methane
concentration at temperatures 1400 ºC and 1450 ºC. The figures show estimated methane
conversion, based on both FID and TCD – measurements.
Figure 4.1.1 Methane conversions given by FID and TCD,
T = 1400 ºC, Set iii).
Figure 4.1.2 Methane conversions given by FID and TCD.
T = 1450 ºC, Set ii)
In the Figure 4.1.3 the methane conversion (given by FID) is plotted against methane concentration
for different Sets: i) ii) iii). In addition a set from initial tests [23] and Reference Set is plotted in the
figure. The Reference Set shows conversions of methane that were achieved at Pyrolyserigg by [21].
The reported experimental conditions for the Reference Set are T = 1450 º C, di = 3 mm, τ = 41 ms.
CH4 Conversion [%]
CH4 Conversion vs. Concentration (by FID)
ii) T=1450 C
iii) T =1400 C
Reference Set
T=1450 C)
i) T=1450 C
Initial tests
T=1450 C
CH4 Concentration [vol %]
Figure 4.1.3 Methane conversion as a function of methane concentration
for different experimental Sets
Ethylene and acetylene are main products of methane pyrolysis at applied experimental conditions.
However, C3 C4 C5 – hydrocarbons, benzene and coke were formed in some experiments. The
selectivities and yields of products were calculated based on FID measurements. It is also assumed
that coke is not produced. This should be acceptable assumption since low methane concentrations
and short residence time are applied.
In the Figure 4.1.4 the selectivities to ethylene and acetylene are plotted as a function of methane
conversion for temperatures 1400 and 1450 °C.
Figure 4.1.4 Selectivity to ethylene and acetylene as function of
methane conversion
At temperature 1400 °C and applied experimental conditions very small amounts of hydrocarbons
containing more than two carbon atoms were formed. The data from experiments at 1450 °C are
more representative, and selectivities for C3-C5 hydrocarbons and benzene are plotted in the Figure
4.1.5, as a function of methane conversion.
Se le cttivity vs . Conversion T=1450 C
C3 -carbons
C4 - carbons
M e thane Conversion [%]
Figure 4.1.5 Selectivity to C3-C5 hydrocarbons and benzene
The selectivities to ethene and ethyne at 1450 °C is plotted as a function of methane concentration
in the Figure 4.1.6
Figure 4.1.6 Selectivity to ethene and ethyne against methane concentration
The Figure 4.1.7 represents yields of ethene and ethyne that were achieved in experiments. The
yields are plotted against methane conversion for the temperatures 1400 and 1450 °C.
Figure 4.1.7 Yields of ethene and ethyne as a function of methane
4.2 Experiments with hot inert – nitrogen. Practical knowledge about quenching
system. Problems.
After the modified quenching/cooling system was installed, a set of experiments with a hot nitrogen
gas was performed. The intention of these experiments was to study quenching system's behaviour:
heat loses, which temperatures could be obtained inside the quencher, which gas flows could be
applied, pressure variations and other practical knowledge.
Before experiments were started, it was assumed that temperature about 1300 – 1400 °C could be
obtained inside the quencher. The measurement of such high temperature could be performed by S –
type thermocouples which are very expensive and breakable. The first problems have occurred
during leakage testing of the system. The gas leakage was identified at inlet, where thermocouple
was placed, as shown in Figure 4.2.1. During the leakage was making tight, thermocouple was
broken. Other S – type thermocouple was broken later, after an unlucky event. Then, it was decided
to use K – type thermocouples for the temperature measurements. These are more durable, but
temperature limitation for K – type thermocouples is around 1200 °C.
Figure 4.2.1 A sketch shows how S – type thermocouple was placed at inlet of quencher
A teflon sealing was used in junction thermocouple/tube inlet. When high temperatures are applied,
it is a risk that sealing can be destroyed and gas leakage can occur. Experiments have shown that a
20 cm long ¼ - tube connected to the quencher results in sufficient cooling of junction at upper
measurement point. At lower measurement point the temperature was much lower and 10 – 15 cm
tube was enough to prevent destruction of the sealing.
In the first experiment with only nitrogen, the furnace was heated up to 1450 °C, in the same
manner as it was done in methane pyrolysis experiments. The N2 flow was 580 [Nml/min] and
quencher was not isolated. The maximum temperature that was obtained in the upper measurement
point T1= 513 °C, the temperature measurement in lower point failed (problem with K –
The experiment was repeated but quencher was isolated with quartz wool. The maximum
temperature was T1= 775 °C at this time. The experiments showed that the heat losses are very high.
Suggestion was to apply higher flows of nitrogen. For this purpose, nitrogen storage cylinder was
connected to «hydrogen-line-2» that has high-flow MFC.
The last experiment was repeated with higher nitrogen flows. Pressure increase in the system was
observed when high gas flows were applied, Pin 3,6 barg and Pout 3,4 barg at 6250 Nml/min. The
sources of pressure increase were identified: a metal sinter and a needle valve. These were removed
from the gas lines to avoid build-up of pressure. The maximum obtained temperatures were T1 =
1160 °C and T2 = 667 °C at 6250 Nml/min nitrogen flow and setpoint T = 1450 °C.
Table 4.2.1 represent approximately «steady state» temperatures that were measured inside the
quencher for different nitrogen flows and setpoint temperatures at the furnace. Temperatures T1 and
T2 are dependent on heat losses and thus on isolation around quencher. The amount of isolation was
different in some experiments.
Table 4.2.1: Measured “steady state” temperatures at upper and lower points for different setpoint
temperatures and nitrogen flows
Flow N2
2500 3760 6250 6250 6250
Setpoint T [°C]
1450 1450
T1 [°C] upper point
1062 1130
T2 [°C] lower point
Before experiments with hot nitrogen were performed, it was assumed that quencher would operate
adiabatic (minimal heat losses to the surroundings). At least, it was expected that the heat loss
between upper and lower measurement points would be minimal. As shown in the Table 4.2.1 a
temperature gradient is very high between measurement points, around 30-50 °C/cm. The high
temperature gradient between measurement points may be explained by flow pattern of gas inside
the quencher and temperature gradient in radial direction. Later, it was suggested to investigate
temperature gradient in radial direction but experiment was not done because the furnace was
destroyed. Another explanation may be that the thermocouples showed wrong temperature.
However, thermocouples were tested in temperature calibrator device and showed write temperature
up to 500 °C that is a limit temperature for calibrator. ΔT for thermocouples were 15 °C at this
4.2.1 Limitation of maximum obtainable temperature in the quencher
As a consequence of heat losses and temperature gradients, the maximum registered temperatures
were 1195 °C at upper point and 764 °C at lower point. So, the achieved temperatures were much
lower than it was supposed and high heat loss was a problem. An extra layer of insulation around
the quencher has been used, without significant temperature increase was marked. An increase of
gas flow to around 7500 Nml/min did not give subsequent temperature increase. Vice versa, the
temperature started to drop at upper point, as shown in the Figure 4.2.2.
Figure 4.2.2 Sketch of how temperature drop with time at upper measurement point
when high gas flow is applied, 7500 [Nml/min]. Setpoint temperature, 1200 °C.
The reason for the temperature drop may be that the heat transport from the ceramic tube to the
passing gas begins to be a limitation. As amount of gas passing the tube increases to some level, the
tube begins to cool down to some extent and wall temperature becomes lower. This results in
reduction of gas temperature. The effect of temperature decrease can be compensated by increase of
setpoint temperature on the furnace, but furnace has operating temperature limitation 1500 °C. It
seems that the reduction of heat loss is an only way to go up in temperature if necessary.
4.2.2 Connection Quencher – Ceram Tube
Conax fitting with grafoil sealing was used to connect ceram tube and modified quencher. In
methane pyrolysis experiments ceram tube was connected directly to the cold finger. This gave a
relative good cooling of the connection to the ceram tube, and it was possible to disconnect fitting
after experiments in usual way by using spanners. For modified quencher, the cooling of the
connection is less effective, and fitting are exposed to much higher temperatures. This results in
welding of the connection. A metal sawback was used to cut the connection, and then a new Conax
fitting could be welded to the quencher. This procedure is somewhat cumbersome and time
consuming. The Conax fitting was replaced two times because of identified gas leakage in the
The gas leakage in Conax fitting was detected during leakage tests, after experiments with setpoint
temperature 1450 °C were performed. Presumably, grafoil sealing was destructed as a consequence
of very high temperature conditions. Other problem was that the ceram tube was easy broken (in
connection) while the quencher was disconnected from the tube. This indicates that the tube
material and fitting are exposed to both mechanical and thermal stresses, which cause the weakness
of the tube material and deformation of the carbon sealing. The leakages were not detected at lower
setpoint temperatures. A deformed sealing and a piece of ceram tube that remains in the fitting are
shown in the Figure 4.2.3
Figure 4.2.3 Picture of connection quencher to ceram tube
after use. Red arrow points to a cavity/hole in grafoil sealing.
4.2.3 Coke formation and plugging of the quencher
It was mentioned in section 3.4 that a simple quencher design was chosen, because it was expected
that the coke formation and plugging of the quencher can be a serious problem during quenching
experiments. Experiments, where propane was used as a quenching medium, showed that the coke
formation is not a big problem. Coke was produced, but it was produced in small quantities. In
experiment with a setpoint temperature 1450 °C and propane flow 140 Nml/min, a temperature
increase was registered at some time. At the same moment a control box for MFC showed that the
propane flow decreases by itself. This could be an indication of plugging of hydrocarbon inlet. The
experiment was stopped and inlet was tested for plugging with a metal rod. The plugging was not
observed. The experiment was continued but higher propane flow was used. No more propane flow
decrease was observed. Presumably, the problem lay in used MFC.
4.3 Experiments with hot nitrogen quenched by propane
The main objective of experiments, where propane at room temperature was introduced to the hot
nitrogen, was to study a quenching effect of propane introduction by measure temperature drop. On
the basis of conversion obtained experimental, the heat consumed in endothermic reactions and
expected temperature drop were calculated.
4.3.1 Conversions, selectivities and yields
Conversion of propane, selectivities and yields to typical products were calculated from GC –
analyses (by FID) . The calculations do not take into account the coke formation and formation of
hydrocarbons with carbon content higher than six atoms. The main identified hydrocarbon products
are methane, ethylene and propylene. Ethane, acetylene and C4 – hydrocarbons are produced in
lower quantities. C5 – hydrocarbons, benzene, coke and tarlike substances were indicated at higher
conversions of propane.
Figure 4.3.1 shows conversions of propane when it was used as a quenching medium at different
setpoint temperatures. Setpoint temperature is used as reference. The actual temperature conditions
are given in Table 3.3. The nitrogen flow was held constant and propane flow was varied.
Volumetric ratio calculated from gas flows given by MFC.
Conversion vs. ratio N2/Propane
Conversion [%]
T=1200 C
T=1300 C
T=1450 C
Volumetric ratio N2/Propane
Figure 4.3.1 Total conversion of propane as a function of volumetric
ratio N2/Propane. Flow N2 6250 [Nml/min].
Figure 4.3.2 shows calculated selectivities to typical products as a function of propane conversion.
The selectivities are influenced of that the conversion is obtained at different temperatures,
residence times and N2/Propane ratios.
Selectivities vs. Conversion
Selectivity [%]
90 100
Conversion of propane [%]
Figure 4.3.2 Selectivity to typical products as function of propane
conversion. Temperature range 620-1180 °C.
In the Figure 4.3.3 yields of typical products are plotted as a function of propane conversion.
Yields vs. Conversion
Conversion of Propane
Figure 4.3.3 Yields of typical products as a function of conversion.
Experimental conditions are given in Table 3.3
4.3.2 Quenching effect: measured and expected temperature drop
In the Figure 4.3.4 a measured temperature drop at upper measurement point is plotted against
volumetric ratio (N2/Propane) for applied setpoint temperatures. Actual measured temperatures are
given in Table 3.3. Temperature drop calculated as a difference between temperature at «steady
state» (before propane is introduced) and temperature registered during the last GC – analysis for
the run, after propane introduction.
Tem perature drop at upper point [C]
Tem perature drop vs. volum etric ratio ( N2/Propane)
Setpoint T =1200 C
Setpoint T=1300
Setpoint T=1450 C
Volum etric ratio N2/Propane
Figure 4.3.4 Temperature drop at upper measurement point as a function
of volumetric ratio N2/Propane gas flows. Nitrogen flow 6250 Nml/min is
In the Figure 4.3.5 a measured temperature drop at lower measurement point is plotted against
volumetric ratio (N2/Propane) for applied setpoint temperatures. Temperature drop given in the
same manner as for upper point.
Temperature drop vs. volumetric ratio (N2/Propane)
Temperature drop at lower point [C]
Setpoint T = 1200 C
Setpoint T = 1300 C
Setpoint T = 1450 C
-20 0
Volumetric ratio (N2/Propane)
Figure 4.3.5 Temperature drop at lower measurement point as a function
of volumetric ratio N2/Propane gas flows. Nitrogen flow 6250 Nml/min.
Quenching effect or expected temperature drop is based on calculations from combined energy and
mass balance, calculated heat loss and experimental conversions of propane. Calculations are in
Appendix C Combined_Energy_Mass_Balance.xls and in chapter 2.2.4
Figure 4.3.6 shows expected temperature drop at lower measurement point if propane would act as
inert (quenching only by dilution with a cold gas) vs. volumetric ratio N2/Propane. Calculation of
expected temperature drop based on assumptions: constant heat loss over CV for given temperature
regime and constant heat capacities of gases.
Expected temperature drop at lower point only by dilution with
propane vs. volumetric ratio
Temperature drop [C]
Setpoint T=1200 C
Setpoint T =1300
Setpoint T=1450 C
Volumetric raito N2/Propane
Figure 4.3.6 Expected temperature drop at lower measurement point assumed
that quenching occurs only by dilution with propane vs. volumetric ratio
Figure 4.3.7 shows expected temperature drop at lower measurement point in the case quenching
happens both by dilution with cold propane and by endothermic reactions of propane:
dehydrogenation and cracking to methane/ethylene. Calculated heat removal by endothermic
reactions is based on conversions obtained experimentally. It is assumed that conversion in each
reaction is a half of the total experimental conversion and heat of reactions is constant.
Expected temperature drop at lower point including heat consumed in
endotermic reactions vs. volumetric ratio
Temperature drop [C]
Setpoint T =1200 C
Setpoint T =1300 C
Setpoint T= 1450 C
Volumetric ratio N2/Propane
Figure 4.3.7 Expected temperature drop at lower measurement point assumed that
quenching occurs both by dilution with propane and by heat consumed in
endothermic reactions vs. volumetric ratio N2/Propane.
5.1 Pyrolysis of Methane: verification of previous experiments
The conversion was chosen as a useful parameter in order to compare and verify the results from
the earlier methane pyrolysis experiments. During initial tests of apparatus [23] the obtained
conversion was higher than expected. For that reason, it was necessary to repeat experiments. The
Set i) was performed. The Runs 1 and 2 were unlikely. Measurements through these runs were
accompanied with pressure increase at the reactor inlet. Pressure increase indicates plugging of
reactor and extensive coke formation inside the reactor tube. No good reason were found to describe
the phenomenon. In later methane pyrolysis experiments plugging of reactor and coke formation
were not observed.
Run 3 has shown an unexpected high conversion again. Than, it was suggested to perform a series
of experiments at lower setpoint temperature, T =1400 °C. At this temperature the methane
conversion is in accordance with the results used from the Reference Set that was performed at 1450
°C, as shown in Figure 4.1.3.
The following explanations may be given for the deviation of obtained methane conversion from
the conversion used as reference: the temperature measurements can be taken at the furnace wall at
two different points. The measured temperature at the bottom path of furnace was around 1500 °C,
while it was 1450 °C (set point) at the upper path. The temperature gradient between two
measurement points can explain why conversion in initial test and in experimental sets i)-ii) was
higher than in the Reference Set. In addition, at high temperatures there are probably a high
temperature gradients in radial direction inside the furnace and measured temperature may be very
sensitive to thermocouple's position. A small deviations from the reference position (5 mm from the
inside wall) may result in different temperatures "reading" by "regulation box", and as a
consequence, different effective temperature in the reactor tube. It can be concluded that the
temperature control and measurement are essential to interpret the results from the pyrolysis
As shown in Figures 4.1.1 - 4.1.3, the methane conversion increases almost linearly with an
increase in methane concentration. This indicates that the overall consumption rate of methane is
the 1. order reaction, that is in agreement with [16] and [22]. Figure 4.1.3 shows that the higher
temperatures give higher conversion that is a consequence of increasing reaction rate. The same
conclusion may be drawn looking on the Figure 2.3.2
In Figures 4.1.1 and 4.1.2 the conversions obtained by TCD and by FID are compared. Ideally, the
measurements by TCD and FID would give the same conversion, if coke is not formed in the
experiment. If coke is formed, the conversion given by TCD would be higher than conversion given
by FID. Some points in the figures show that the conversion calculated by TCD is lower than
conversion calculated by FID. This indicates that it is an inaccuracy between to measurement
methods. An explanation of inaccuracy can be a bad or wrong calibration of TCD.
The Figure 4.1.3 shows that the methane conversions from sets i) ii) and initial test, plotted against
methane concentration with a reference temperature 1450 °C, follow the same “line”. This is an
indication of reproducibility of pyrolysis experiments.
In the Figure 4.1.4 the selectivities to ethylene and acetylene is plotted as function of methane
conversion for the temperatures 1400 °C and 1450 °C. The results show that the selectivity to
acetylene increases and selectivity to ethylene decreases with an increase in temperature from 1400
to 1450 °C. This is in agreement with results obtained by [16].
However, both experimental and simulated results from [16] show that the selectivity to ethylene
decreases with increase in methane conversion at temperatures 1400 and 1500 °C, and selectivity to
acetylene has a maximum at 20 – 40 % conversion. As shown in the figure 4.1.4, ethylene
selectivity increases slightly with methane conversion, and selectivity to acetylene decreases
without identifiable maximum. Deviation of the selectivity curves can be explained by the choice of
experimental parameters which were varied to obtain a given conversion. In the first case, the
H2/CH4 = 2 was held constant and residence time was varied. In the second case, the residence time
was held constant and methane concentration was varied, and high hydrogen dilutions were applied.
As mentioned in chapter 2.3.1, hydrogen dilution has an important effect on kinetics of methane
Figure 4.1.5 represent selectivities to C3-C5 hydrocarbons and benzene that were obtained in
experiments (gas phase products/ coke is not considered). As shown in the figure, the selectivities
are lower than 1,5 % in most cases. Low selectivity to benzene and other higher hydrocarbons can
be explained by short residence times and high hydrogen dilution that prevents the condensation
reactions. The accuracy of these results is not very good. Some problems with identification of
compounds happened and integration curves from GC – analyses were long from perfect for some
In the figure 4.1.6 selectivities to ethene and ethyne are plotted as function of methane
concentration. The shapes of the selectivity curves are similar to that for the case where these are
plotted as function of conversion. This is probably because conversion varies linearly with methane
The Figure 4.1.7 shows yields of ethylene and acetylene. The yields for acetylene are higher than
for ethylene, and ratio acetylene to ethylene increases with methane conversion. Yields of both
ethylene and acetylene increase linearly, and temperature difference about 50 °C have low influence
on Yield – Conversion relation.
5.2 Experiments with hot nitrogen quenched by propane
5.2.1 Conversions, selectivities and yields
As shown in the Figure 4.3.1, conversion of propane decreases with decreasing N2/Propane ratio
and conversion increases with increasing temperature. At lower N2/Propane ratios (high propane
flows) conversion is low that means that quenching of hot gas occurs mainly by dilution and heat
removal by endothermic reactions is low. As temperature of hot gas increases, the propane
conversion increases and heat removal by endothermic reactions also increases. At setpoint
temperature 1450 °C, measured temperature at «steady state» inside quencher was 1193 – 745 °C
(high gradient and high heat loss). If heat loss could be eliminated, a higher conversion could be
obtained although at low N2/Propane ratios.
From Figures 2.3.1 and 2.3.6 can be concluded that high propane conversion can be obtained at
temperatures higher than 800 °C at atmospheric pressure. This means that the hot gas could be
quenched effective by propane from, let say, 1200 °C to 800 °C. These conditions were not realized
experimentally because of presence of temperature gradient and heat loss.
The equilibrium conversion of propane is a function of total pressure in the system, Figure 2.3.1 As
a consequence, the quenching efficiency will also be dependent on the pressure of the system. Heat
consumption by endothermic reactions will decrease with increasing pressure. Experiments were
performed at Ptot = 0,6 barg or 1,6 bara that give something lower conversion than at atmospheric
pressure. It can also be mentioned, because synthesis gas is produced often at elevated pressures,
the opportunity for the quenching with alkanes may be reduced.
Figure 4.3.2 shows selectivities to typical products as a function of propane conversion. The results
are in accordance with a Figure 2.3.5 in that the selectivity to propylene decreases with increasing
conversion, selectivity to methane and ethylene increase with increasing conversion and selectivity
to ethane is low, < 9 %. However, obtained selectivities of methane is much lower than presented in
Figure 2.3.5, and selectivity to ethylene and propylene is also sufficiently lower. These deviations
can be explained by the differences in experimental conditions. Deviations at higher conversion can
also be explained by the fact that the hydrocarbons with more than 6 carbon atoms and coke are not
included in calculations. The results at higher propane conversion are more uncertain. The indirect
quenching at lower path of cooling system was less effective after removal of cold finger. This can
also affect the results.
Yields of typical products as a function of conversion are presented in the Figure 4.3.3. Yield of
ethylene increases almost linearly with propane conversion. Yield of propylene goes through a
maximum at 55 % conversion and approach a zero at 100 % conversion. Yield of acetylene is low,
< 5 % for the most runs but reaches 25 % at run that gave 100 % conversion. This run is very
uncertain because of problems with propane flow.
The idea of chemical quenching experiments was to investigate the opportunity of heat utilization of
high energy streams in a desirable way. Effective chemical quenching means that the temperature
reduction of the hot gas would occur by heat consumption in endothermic reactions to highest
possible extent. Thus conversion of quenching medium needs to be highest possible. An other
important aspect is which valuable products could be obtained at these high conversions. In the case
when propane is used as quenching medium, ethylene and propylene are considered to be the most
valuable products. It seems that the yield of propylene is very low at high conversions. Ethylene
yield is much higher. [17] reports ethylene to propylene ratio is 6 at 95 % conversion. Hydrogen and
other hydrocarbons are also produced during propane cracking. It can be mentioned that if the
synthesis gas would be quenched by propane the selectivities and yields of products can be
different. As in the case with methane pyrolysis, hydrogen present in synthesis gas will probably
inhibit propane cracking. The experiments with synthesis gas were not performed because the
furnace has been out of stand.
5.2.2 Quenching effect: measured and expected temperature drop
Temperature was measured at upper and lower points inside the quencher. Figures 4.3.4 and 4.3.5
shows temperature drop at these points after propane was introduced. Temperature drop is plotted as
a function of volumetric N2/Propane ratio for three setpoint temperatures, used as reference
temperatures. Actual measured temperatures are given in Table 3.3.
Initially, the temperature drop at upper point was not expected at all, because thermocouple was
placed above inlet to hydrocarbon and temperature of hot gas alone should be measured. However,
the temperature drop increased with decreasing N2/ Propane ratio (increasing propane flow).
Assumed that the propane do not reach the sensor at upper point, the temperature drop may be
explained by reduction of wall temperature. At higher propane flows, cold gas reaches probably the
sensor that result in lower temperature measurement. As seen in Figure 4.3.4 a setpoint temperature
has low effect on temperature drop, except for the case Tsetpoint =1450 °C and low N2/Propane ratio.
At lower point, setpoint temperature 1200 °C and high N2/Propane ratio the temperature increase
(negative temperature drop on the figure) was observed after propane was introduced. This is
probably because the «steady state» was not reached previous to propane introduction. As seen in
the Figure 4.3.5, temperature drop increases both with decreasing N2/Propane ratio and with
increasing temperature of hot gas. A higher temperature drop at higher hot gas temperatures for the
same N2/Propane ratio indicates that the quenching is more effective at higher temperatures.
Nevertheless, the obtained quenching effect and measured temperature drop are long from what can
be expected. In the figure 4.3.6 expected temperature drop was calculated for the case when
quenching occurs only by dilution with a cold gas, propane acts as inert. It was assumed that the
heat loss from CV is constant for a given setpoint temperature and do not vary with N2/Propane
ratio. Heat loss calculation is based on measured temperatures at upper and lower points. When
results from Figures 4.3.6 and 4.3.5 are compared, one can conclude that, even if quenching would
occur only by dilution, the expected temperature drop would be higher than measured. The
explanation of deviation is a poor mixing of hot and cold gases inside the quencher.
In Figure 4.3.7 heat consumption by endothermic reactions of propane (dehydrogenation and
cracking to methane/ethylene) is included in calculation of expected temperature drop at lower
point. Calculated heat consumption is based on propane conversions obtained in experiments. The
calculated temperature drop are much higher than measured.
Temperature drop for the two cases, including endothermic reactions and not including, was
calculated to give a sense of importance of heat consumption in endothermic reactions in
quenching process. At setpoint temperature 1450 °C and N2/Propane ratio 6,3 expected temperature
drop 1,6 times higher in the case when endothermic reactions are included. As N2/Propane ratio
increases to 44, the temperature drop is 3,5 times higher for the same case.
5.3 Quencher Design
As was mentioned in chapter 3.4, a quencher with a simple design was constructed and used to test
the concept of chemical quenching. After the experimental work was performed the following can
be said about quencher design:
Very high temperature gradients between measurement points registered. The temperature
gradient increases with increasing gas temperature and decreases with increasing gas flow.
Heat losses are high and maximum obtainable temperatures inside the quencher are low,
1195 – 795 °C.
Heat transfer from the ceram tube to the hot gas limits obtainable temperature in quencher
when gas flows higher than approximately 7000 Nml/min are applied
Low measured temperature drop in quenching experiments indicates poor mixing of hot and
cold gases.
At setpoint temperature 1450 °C problems with connection quencher – ceram tube occurs:
welding of the fitting and destruction of the grafoil sealing that cause the gas leakage.
K – type thermocouples are more durable and easy-to-handle than S – type thermocouples
with respect to make connections tightly. K – type thermocouples have a temperature limit
around 1200 °C. If experiments at higher temperature will be performed, the S – type should
be used.
Removal of cold finger reduces efficiency of cold wall quenching in lower part of cooling
system. This can influence the results of experiments with propane. Temperature of
outcoming gases from the cooling system was not measured but it was much higher than
room temperature when high setpoint temperatures were applied.
As shown in the Figure 3.4.2 it is a quencher section between lower temperature
measurement point and «cold finger». The volume of this section is higher than the control
volume, CV. The presence of the section affects the conversion, selectivities and yields in
the quenching experiments because of increasing residence time. The section should be
reduced or extra thermocouple inlet should be installed write above «cold finger».
5.3.1 High temperature gradients and high heat losses
In chapter 4.2 was mentioned that high temperature gradients between measurement points can be
explained by high heat losses or by high temperature gradient in radial direction. If high
temperature gradient in radial direction is a really case, especially at lower point, the calculations of
the heat loss may be wrong, thus calculated expected temperature drop may also be wrong. The
fact, that the use of extra insulation layer around quencher has not resulted in significant
temperature increase, can indicate that the «cold finger» is an important source of heat loss and
temperature gradient for the quencher-part.
Heat transfer by conduction through the metal material downwards the cooling system, from the
point with a high temperature (connection to the ceram tube) to the point with a low temperature
(«cold finger»), will occur. Figure 5.3.1 shows how heat is transferred through the quencher wall.
Figure 5.3.1 Heat transfer through the wall of quencher part
If this conductive heat transfer is high compared to heat transfer from the gas phase to the tube
wall, the consequence may be the high temperature gradient through the wall of the quencher. Thus
the wall temperature at lower measurement point may be much lower than the wall temperature at
upper measurement point. Low wall temperature may result in that the measured temperature is
much lower than the true gas temperature.
Neither wall temperature nor true gas temperature are known. Equation (20) from chapter 2.5 was
used to calculate how true gas temperature Tg may vary as a function of wall temperature Tw , for
the same measured temperatures Tp , 610 °C and 909 °C, taken from Table 4.2.1. The results are
shown in figures 5.3.2 and 5.3.3 below.
True gas temperature vs. wall temperature
Temperature of hot gas [ C ]
Tg - true gas
Tp- measured
Tw - wall temperature [ C]
Figure 5.3.2 True gas temperature vs. wall temperature. Tp = 610 °C
Parameters used in calculations: hc = 40 W/m2·K; ε = 0,6; σ = 5,676·10^-8
[W/m2·K4] [14]
True gas temperature vs. wall temperature
Temperature of hot gas [C]
Tg - true gas
Tp - measured
Tw - wall temperature [C]
Figure 5.3.3 True gas temperature vs. wall temperature. Tp = 909 °C
Parameters used in calculations: hc = 40 W/m2·K; ε = 0,6; σ = 5,676·10^-8
[W/m2·K4] [14]
The calculations show that the difference between true gas temperature and measured temperature
increases with decreasing wall temperature.
As shown in the Figure 5.3.1, heat is transferred through the wall both in axial and radial directions.
Assuming that the heat transfer in radial direction is negligible, the one dimensional model for the
heat transfer at steady state can be used to describe the temperature distribution through the wall.
Assume also that Tw = 1000 °C in connection to ceram tube and Tw = 70 °C in connection to «cold
finger» and L = 27,2 cm is a distance in between. Then, the temperature distribution will be as
presented in the Figure 5.3.4.
Temperature drop through the quencher wall
Tw - wall temperature
Tw = -34,2L + 1000
L [cm]
Figure 5.3.4 Temperature distribution through the quencher wall.
Boundary conditions: Tw(0)=1000 °C Tw(L)=70 °C. L = 27,2 cm
Heat transport in radial direction is not taken into account.
A lower measurement point is at L=13,5 cm that give Tw = 538 °C according to model above. From
the Figure 5.3.2, true gas temperature is Tg = 760 °C and the difference Tg-Tp = 150 °C.
Due to numerous assumptions, it is likely that the temperature difference is overestimated.
However, the example above shows that the conductive heat transfer through the quencher wall may
have a significant impact on temperature measurements, and «cold finger» may be a source of heat
losses for the quencher – part. If this is a really case, the material with a lower thermal conductivity
should be chosen for the quencher.
5.3.2 Bad mixing of hot and cold gases
Quenching experiments with propane have shown an unexpected low temperature drop that
indicates a poor mixing of hot and cold gases inside the quencher. As was discussed in chapter
2.2.1, a gas mixing occurs usually by diffusion or by convective transport or by both transport
types, depending on flow conditions. In turbulent flow regime gas mixing is mainly by convective
transport and mixing is much better than in laminar flow.
The Reynolds number for the typical hot gas flow in experiments was calculated as NRe = 457 for
the quencher – part and NRe = 1085 for the ceram tube, calculation is in Appendix A. The low
Reynolds number indicates that the hot gas flows in laminar regime and there is no effective
turbulent mixing. According to equation (4) in chapter 2.2.3 an entry length Le is calculated to be 34
cm that means the fully developed velocity profile was not established.
The cold propane flow is introduced perpendicular to the hot nitrogen flow and the gas flow pattern
is not really known. Since temperature drop at lower point is low and it is assumed that the reason
for that is a bad gas mixing, one can think of situation when cold propane flow push hot nitrogen to
the quencher wall. Then, both gases flow in some kind laminar flow with a boundary layer in
between. The Figure 5.3.5 shows how the gas flow pattern may look out. The consequence may be
that the mixing of gases is dominated by the diffusion transport that is relative slow.
Figure 5.3.5 Gas flow pattern. To the left: cold gas is
introduced. To the right: only hot gas entering quencher.
The gas flow pattern inside quencher was discussed with Hugo A. Jakobsen, Professor at Reactor
Technology Group, NTNU [24]. He has mentioned that the gas flow is probably much higher in the
middle of the quencher than the gas flow at the quencher wall. He meant also that the temperature
difference between upper and lower measurement points is to high, when isolation around quencher
is applied. The temperature gradient in radial direction and positions of thermocouples are most
likely the explanation of the high measured temperature difference between upper and lower points.
On the basis of estimated Reynolds number and measured temperature drop, it can be concluded
that the mixing of gases is poor. Thus, the used quencher design do not realize the conditions of
completely mixing of hot and cold gases that makes testing of quenching concept difficult.
In chapter 2.1.1 was mentioned that in the case of coaxial mixing of hot and cold gases the diameter
of the hot jet and the ratio of coaxial to jet velocity are important factors for cooling rate. A smaller
hot jet diameter and higher velocity ratios reduce the time necessary to establish turbulent mixing
processes. In quenching experiments with N2 and propane the turbulent mixing was not established,
and gas flows are perpendicular to each other. However, conclusions from chapter 2.1.1 may
probably be used to improve the gas mixing. Ceram tube used to heat up nitrogen had inner
diameter 6 mm. If ceram tube with a smaller diameter will be used, the hot jet will be smaller and
gas velocity out of tube will be higher.
A nozzle should be installed at cold gas inlet or diameter of the inlet should be reduced to increase
the cold gas velocity. The cold gas inlet may also be placed at an angel to hot gas flow as shown in
Figure 5.3.6. When hot and cold gas flows will meet together at higher velocities a forced
convective mixing would occur to an larger extent. These small changes in quencher design may
hopefully improve gas mixing sufficiently.
Figure 5.3.6 Sketch of proposed changes in quencher design
to improve gas mixing
6.1 Conclusions
Concept of chemical quenching was demonstrated by experiments with a hot nitrogen gas
quenched by propane: temperature drop was indicated and propane was converted to the
products. However, the observed quenching effect was low that can be explained by poor
mixing of gases.
Quenching effect increases with a hot gas temperature that can be explained by increasing
propane conversion.
A high conversion of quenching medium is a requirement because thermal energy should be
recovered in the form of valuable products to an largest extent. At high conversion of
propane the ethylene is a product that was obtained in the highest yield.
If synthesis gas will be quenched by propane, the conversion, yields and selctivities will be
different. The synthesis gas produces often at elevated pressures and high partial hydrogen
pressure. This may result in lower conversion of propane and as a consequence, the
quenching effect may be reduced.
Conversion of hydrocarbon (propane) decreases with decreasing N2/Propane ratio. As the
N2/Propane ratio decreases (propane flow increases), the importance of endothermic heat
consumption decreases and quenching occurs by dilution to a higher degree.
Quencher design is poor and do not realize expected experimental conditions: completely
gas mixing and adiabatic quench that makes the proof-of-concept study difficult. The
obtained results (conversions,yields and selectivities) are not very useful.
The quenching of syngas by light hydrocarbons may result in production of different
amounts of products that will be necessary to separate later. Even if, quenching will be
effective and high yields of valuable products will be obtained, the separation costs of the
products will reduce the total efficiency of the process.
A hypothetical situation may be that the quenching process is optimized in such a way that
the quenching hydrocarbon reacts to CO and H2 in the presence of O2 or water. Thus, the
final product is synthesis gas, and problem with a products separation may be avoided.
A better quencher design is necessary in order to investigate the concept.
6.2 Recommendations
1. A nozzle should be designed and incorporated into hydrocarbon inlet, to increase gas
velocity. Gases at higher velocity will have higher kinetic energy and as result – better
mixing. If coke will plug the nozzle, the inlet can be disconnected and the needle or metal
rod can be used to remove the plug.
2. A hydrocarbon inlet should be placed at at angel as shown in the Figure 5.3.6.
3. The alsint tube with a lower diameter should be used for heating up the hot gas, di = 3 mm
or smaller if possible, to improve mixing. The gas velocity out of tube will increase.
Different tube diameters may affect the heat transfer from the tube to the gas. If it will be
possible to apply higher gas flows, the higher temperature inside the quencher can be
obtained and temperature gradient between measurement points can be reduced.
4. 2. section of the quencher, defined in chapter 3.2, should be reduced/cut. 1. section may
retain its dimensions. If high gas velocities will be applied, it will be enough space to mix
5. First of all, it is recommended to investigate if temperature gradients in radial direction
exist. This may be an explanation of measured temperature gradient between upper and
lower points. If radial temperature gradient is not significant, the gradient in axial direction
may be explained by conductive heat transport through the quencher wall.
If gradient in axial direction and/or heat losses can not be eliminated, it is proposed to made
a model describing heat transport and temperature distribution through the quencher.
6. The cold finger should be moved back into the outer shell of water quencher. It will be
necessary to lift up the furnace.
7. If higher temperatures in the quencher will not be reached (problems with Conax fitting at
8. It is not recommended to go higher than T = 1300 °C for the setpoint temperature on the
furnace, for given quencher configuration because problems with connection to the ceram
9. It is proposed to make a mathematical model, if possible, that will describe the quenching of
synthesis gas by alkanes. The model should include reactions kinetic, thermodynamics, mass
and heat transfer phenomena. To make such model may be a difficult task in itself, so it is
recommended to cooperate with Reactor Technology Group at NTNU. In combination with
experimental work a more useful results may be obtained.
BP Energy Outlook 2030,
London, © BP 2011
20. June 2012
International Energy Outlook 2011, U.S. Energy Information Administration
20. June 2012
Combined Heat and Power Plants, © Siemens AG 2006
21. June
Chemical Process Technology, Jacob. A. Moulijn, Michiel Makkee, Annelies Van Diepen,
23. June 2012
Perspectives and Experience with Partial Oxidation of Heavy Residues, C.Higman, Paris,
(28. June 1994)
23. June 2012
An Overview of Coal based Integrated Gasification Combined Cycle (IGCC) Tehnology, Ola
Maurstad, (2005)
http://sequestration.mit.edu/pdf/LFEE_2005-002_WP.pdf 23. June 2012
The National Energy Technology Laboratory (NETL)
23. June 2012
Chemical quenching of methane-acetylene conversion by injection of liquid hydrocarbons
into a plasma jet, L. Dessau, H.-J.Spangenberg, R. Kleffe 30. June 2012
C. Higman, M. Van der Burgt, «Gasification», Elsevier, 2003
Sustainable SunFuel from CHOREN's Carbo-V® Process, Presentation, T. Blades, M.
Rudloff, O. Schulze, © CHOREN, San-Diego (2005)
http://www.eri.ucr.edu/ISAFXVCD/ISAFXVAF/SSFCCVP.pdf 23. June 2102
Modelling of the quenching in cylindrical and annular reactors, M. Ferrer, J. Lédé, Solar
Energy Vol. 66, No. 2, p.161, (1999)
Quenching Processes for High Temperature Chemical Reactions, D.W. Sundstrom, R.L.
DeMichiell, Ind. Eng. Process Des. Develop., Vol. 10, No 1, (1971)
Transport Processes and Separation Process Principles, 4th ed., C. J. Geankoplis, (2008)
Steam Cracking of Hydrocarbons. 1. Pyrolysis of Heptane, Martin Bajus, Václav Veselý,
Piet A. Leciercq, Jacques A. Rijks, Ind. Eng. Chem. Prod. Res. Dev., Vol. 18, No. 1, (1979),
pp. 30-37
Thermal Coupling of Methane, Ola Olsvik, Ph.D. thesis, University of Trondheim, (1993)
Thermal Cracking of Propane, Alfons G. Buekens, Gilbert F. Froment, Ind. Eng. Chem.
Process Des. Dev., 1968, 7 (3), pp 435–447
Simultaneous Thermal Cracking and Oxidation of Propane to Propylene and Ethylene, V.R.
Choudhary, V.H. Rane, A.M. Rajput, AIChe Journal, Vol.44, No. 10, (1998), pp. 2293-2301
SI Chemical Data, G.H. Aylward, Tristan J.V. Findlay, (2007)
Lecture Notes, TKP 4150, NTNU, (2011)
Personal communication with Torbjørn Gjervan, Research Manager, SINTEF, Trondheim
Pyrolysis of methane in the temperature range 100 – 1700 K, V.S. Arutyunov, V.I. Vedeneev,
Russian Chemical Reviews 60 (12) (1991), pp. 1384 - 1397
Specialization project, TKP4510, Alexei Pylilo, NTNU (2011)
Personal communication with Hugo A. Jacobsen, Professor at Reactor Technology Group,
NTNU, May 2012
Appendix A 1
Example of calculation of Reynolds number NRe for the quencher and for the ceram tube
A gas flow given by MFC (volume flow at standard conditions)
VN2  7
 Nm3 
Nl 0, 007 Nm3
 1,16667 104 
 s 
1 1
 Rn  2 2
By ideal gas low
 V1
For the approximately temperature inside the quencher T2 = 900 °C =1173 K
assume constant pressure
V2 
 m3 
V (T  900 oC )  5 104  
 s 
The cross sectional area of the quencher calculated as
Across 
 Di2
  (0.02m)2
 3,142 104 m 2  av 
Reynolds number is calculated as
N Re 
 2,3  m / s 
- average gas velocity
Diav 
Di – inner diameter of the tube/quencher [m]
ρ – gas density [kg/m3]
μ – gas viscosity [Pa˖s]
ρ900 °C = 0,45914
μ900 °C = 46,191˖10 [Pa˖s]*
(P = 1,6 bara)
=> NRe = 457 (Reynolds number for the quencher)
For the ceram tube
Assume an average gas temperature for the ceram tube Tav = 1100 °C,
then the volume flow of nitrogen
 m3 
V (T  1100 oC )  5,87 104  
 s 
 Di2   (0.006m) 2
 2,827 105 m 2  av   20,8  m / s 
Across 
Use viscosity and density data at 999 °C
ρ999 °C = 0,4234
μ999°C = 48,7˖10-6
=> NRe = 1085 ( Reynolds number for the ceram
* http://www.peacesoftware.de/einigewerte/stickstoff_e.html
25.Juni 2012
Appendix A 2
Pyrolyse experiments. The method of analysis.
The analysis of the product gas mixture is done by gas chromatograph (GC) equipped with Thermal
Conductivity Detector (TCD) and Flame Ionization Detector (FID). The sample of the product gas
is introduced to the GC where separation of mixture's components occurs inside a two columns:
Carbosieve S-2) and GS-Q. Each component separates according to its retention time. The detectors
register the output signals in a chromatogram. Areas corresponding to each component are obtained
after integration. These areas are used to calculate the relative amount of components in the sample
and product gas mixture composition.
Calculations for the product gases
In the following it assumed that integrated areas from chromatogram are proportional to the
amounts of components in the analysed sample. In addition the areas given by FID are proportional
to the number of carbon atoms corresponding to the hydrocarbon presented in the sample. The
nitrogen gas is used as internal standard since nitrogen is not consumed during the reactions.
The gas mixture containing hydrogen, nitrogen, methane, ethane, propane and n-butane are used to
calibrate the GC and to identify retention times for the listed components.
Calibration gas used from the bottle No 10181806.
The combination of outputs from TCD and FID is used in calculation of mass balances for the
Mass balance by TCD
Using the calibration gas mixture correction/response factors fi are estimated for H2, N2 and CH4 by
X cP
fi  i c
100 Ai
Xci – mole fraction of component i in calibration gas
Aci – measured area corresponding to component i in calibration gas
P – the atmospheric pressure
i - H2, N2 or CH4
The calculated response factors for nitrogen and methane are assumed to be constant, but response
factor for hydrogen will vary since helium is used as carrier gas for GC. The variation in response
factor for hydrogen will not affect the calculations of methane conversion because nitrogen used as
internal standard. The methane flow in product gas is related to nitrogen gas flow. However, the
estimated flow of hydrogen in product gas can be something uncertain.
The ratio
Ki 
f N2
shows how response factor for specie i is related to response factor for nitrogen.
Molar ratio of species (hydrogen or methane) to nitrogen is estimated as
ri  K i
 i
AN2 X N2
Than, molar flow of component i can be calculated as
Fi  FN 2 ri
FN2 = F0N2 – nitrogen gas flow is given by mass flow controller
Conversion of CH4 given by TCD measurements:
Methane conversion estimated as
 FCH 4
X CH 4 
F0CH4 - average methane flow given by two measurements of feed gas mixture
FCH4 – the flow given by measurement at experimental conditions
The final calculated methane conversion is an average conversion of three measurements at
experimental conditions.
Mass balance by FID
Molar ratios of carbon species present in product gas mixture registered by FID calculated as
rj 
X N2
jACH 4
 rCH 4
rj – molar ratio of hydrocarbon specie to nitrogen
Xj – mole fraction of carbon specie in product gas
XN2 – mole fraction of nitrogen in product gas
Aj – measured area corresponding to identified carbon specie in product gas
ACH4 – area of methane (given by FID)
rCH4 – molar ratio of methane to nitrogen (given by TCD - measurement)
j – number of carbon atoms in specie j
Than, molar flow of hydrocarbon specie estimated as
F j  FN 2 rj
And coke flow rate can be estimated as C1
Fcoke  FCH 4   jFj
Conversion of CH4 (gas phase/ by FID):
Conversion of CH4 estimated only by FID measurements. Conversion given as a difference between
total carbon atoms registered by FID and C – atoms present as methane, divided by total carbon
atoms present in the product mixture. This estimation method does not take into consideration the
coke formation.
X CH 4 
mol %(total _ C  atoms )  mol %(CH 4 )
mol %(total _ C  atoms )
The mole % of carbon compound present in the product gas mixture is calculated as
mol %Cn 
n nn
n – number of carbon atoms in the compound
An – area measured by FID for specie with n carbon atoms
When mole % of carbons species is known, the carbon distribution in gas phase is calculated as
(% C – atoms as Cn) =
%Cn  n
n n n
The selectivity to hydrocarbons is estimated as percent of carbon atoms presented as hydrocarbon
relative to the total amount of carbon atoms, divided by the percent of converted methane:
Sn % 
(% C _ atoms _ as _ C n )
100%  (% C _ atoms _ as _ C 1)
The Yield of hydrocarbons calculated as conversion of methane multiplied by selectivity to
Yn %  X CH 4S n
NB! The same formulas were used for calculation of results from quenching experiments with
propane. Propane is considered to be reactant. Analyses were taken only by FID.
Appendix B
GC - method
The GC – method used for the experimental analysis is a method that was used in earlier methane
pyrolysis experiments. The name of the method is ODDL1.M.
Oven temperature versus time
Initial temperature: 40 °C
Initial time: 6 min
Heating rate: 25 °C/min
Final temperature: 200 °C
Valve on: 0,1 min
Valve off: 3,0 min
Calibration of GC:
Carrying gas (He):
H2 gas:
4,5 [ml/min]
30 [ml/min]
248 [ml/min]
313 [ml/min]
Carrying gas (He)
Refer. flow
Total flow
Retention times for the calibration gas (average of three analysis):
Rt = 2,179 min
Rt = 4,417 min
Rt = 9,818 min
Rt = 12,364 min
Rt = 1,210
Rt = 5,884
Rt = 11,427
27 [ml/min]
46 [ml/min]
73 [ml/min]
Example of retention times from the quenching experiments with propane:
Tset = 1300 °C, Vpropan = 143 [Nml/min] VN2 = 6250 [Nml/min]
Rt = 2,210
Rt = 3,647
Rt = 4,349
Rt = 4,478
Rt = 9,643
Rt = 9,838
Rt = 12,259; 12,420
Rt = 17,607
Retention times for Acetylene and Ethane, and for Propylene and Propane are very
close to each other. The separation of components in the column is not very good.
The areas on chromatogram overlap that result in some uncertainty in
calculations. The GC-method should be changed to get a better separation of
Appendix D
Was this manual useful for you? yes no
Thank you for your participation!

* Your assessment is very important for improving the work of artificial intelligence, which forms the content of this project

Download PDF