Combustion engineering and gas utilization Third Edition. Edition Cornforth instant download
Combustion engineering and gas utilization Third Edition. Edition Cornforth instant download
https://ebookgate.com/product/combustion-engineering-and-gas-
utilization-third-edition-edition-cornforth/
https://ebookgate.com/product/synthesis-gas-combustion-fundamentals-
and-applications-1st-edition-lieuwen/
ebookgate.com
https://ebookgate.com/product/combustion-science-and-engineering-1st-
edition-kalyan-annamalai-author/
ebookgate.com
https://ebookgate.com/product/combustion-engineering-issues-for-solid-
fuel-systems-1st-edition-bruce-g-miller/
ebookgate.com
https://ebookgate.com/product/handbook-of-natural-gas-transmission-
and-processing-third-edition-principles-and-practices-saeid-mokhatab/
ebookgate.com
Spacecraft Systems Engineering Third Edition Peter
Fortescue
https://ebookgate.com/product/spacecraft-systems-engineering-third-
edition-peter-fortescue/
ebookgate.com
https://ebookgate.com/product/plastics-third-3ed-microstructure-and-
engineering-applications-3rd-edition-nigel-mills/
ebookgate.com
https://ebookgate.com/product/engineering-management-principles-and-
economics-etext-third-edition-niall-m-fraser/
ebookgate.com
https://ebookgate.com/product/sunflower-chemistry-production-
processing-and-utilization-1st-edition-dunford/
ebookgate.com
https://ebookgate.com/product/electric-power-transformer-engineering-
third-edition-james-h-harlow/
ebookgate.com
Combustion Engineering and Gas Utilisation
This page intentionally left blank
ii
Combustion Engineering
and Gas Utilisation
BRITISH GAS SCHOOL OF FUEL MANAGEMENT
British Gas-
~~o~;~~n~~~up
LONDON AND NEW YORK
Published by British Gas pic
Rivermill House, 152 Grosvenor Road, London SW1V 3JL
iv
Foreword
This is the third edition of 'Combustion Engineering and Gas Utilisation' and I am pleased to have
the opportunity to contribute a foreword to a publication that has clearly established itself as an
essential work for study and reference. That it is produced by the British Gas School of Fuel
Management is not only a tribute to the expertise and commitment of the School's staff but should
re-assure those concerned with gas utilisation that the School continues its world-leading work in
energy efficiency. This has become so much more important as we recognise the fundamental role
of energy efficiency in meeting environmental challenges.
This book offers a comprehensive source of reference on all aspects of gas utilisation in the
industrial, commercial and administrative market sectors. It covers process, space-heating and hot-
water equipment in detail, Codes of Practice and Standards and the important issue of fuel
efficiency.
In commending it to a wide readership, I am conscious of the success of the two previous editions,
which attracted a world-wide usage among gas undertakings and academic institutions. I am certain
that it will continue as a valued work for reference and education in the gas industry.
Mr R Evans
Chairman and Chief Executive, British Gas pic
v
This page intentionally left blank
vi
Contents
Foreword v
Preface ix
Contributors xi
vii
Chapter Ten: IGNITION AND COMBUSTION SAFEGUARDS 627
Start flames and pilot burners, Basic ignition procedure for industrial gas equipment,
Shut down, Combustion safeguards - general, Manual, semi-automatic and fully automatic burner light up,
monitoring and shut down, References.
Index 877
viii
Preface
Since the second edition of this handbook was published in 1983, a whole range of new efficient gas
fired space, water and process heating equipment has appeared on the industrial and commercial
market. Also 8ritish Gas has been privatised and many new Standards, Codes of Practice and
Regulations have been introduced. This publication attempts to bring engineers up to date with
equipment currently available and sound combustion engineering principles and practice.
In addition, the importance of correct installation, commissioning, servicing and maintenance of gas
utilisation equipment for safety and efficiency cannot be stressed too strongly. It is important that
only competent and trained staff are employed in such activities. All operators should be aware of
the safe light up, shut down and emergency isolation procedures.
The installation of pipework should conform to the Gas Safety (Installation and Use) Regulations
1984 and also to British Gas Publication IM/16, Guidance Notes on the Installation of Pipework in
Customer's Premises. Before gas fired equipment is brought into production for the first time it
should be properly commissioned, as in the Guide to Commissioning Procedures for Gas Fired Plant
document which is being prepared by the Institution of Gas Engineers.
Gas fired plant requires regular attention throughout its life in order to ensure safe and efficient
operation. Planned maintenance is therefore essential in order to extend the useful life of compo-
nents and prevent where possible breakdowns during production periods.
Additionally, wear and tear on mechanically moving parts means that over a period of time pressure
settings, air gas ratios and !Iue gas analyses may tend to drift from their original setting resu Iting in
loss of efficiency and perhaps a reduction in the safety margin employed. For this reason then, it
will usually be more economic as well as good practice to regularly inspect and maintain gas fired
equipment. Appliances must not be used if they are unsafe in any respect.
The frequency of servicing will be determined by the type of equipment employed, period of use,
the environment in which it is installed and the degree of reliability expected. In practice whilst
many tests can be carried out annually, there will be some checking which should be done
quarterly, weekly or even daily, in order to maintain optimum efficiency and safety. All new
equipment should be supplied with a service manual giving an extensive guide to servicing
requirements but older equipment should be serviced using sound engineering practice and
judgement.
J. R. CORNFORTH
ix
This page intentionally left blank
x
Contributors
The initial script for a large proportion of this textbook was prepared by FEC Consultants Ltd.
However, during its preparation there have been many contributors and much advice given by
friends and colleagues from within the gas industry and also by equipment suppliers and manufac-
turers. We hope they will all accept our sincere expression of thanks. In addition, most of the
original artwork has had to be redrawn and the whole work standardised and unified. We would like
to thank the College of Fuel Technology for their help and guidance here.
The main contributors and the chapters involved are listed below.
xi
This page intentionally left blank
xii
Combustion 1
1.1 INTRODUCTION 3
REFERENCES 35
3
CHAPTER 1
COMBUSTION
1.1 INTRODUCTION
This section deals with the essential properties of natural gas and liquefied petroleum gases when
used in heating processes. In view of the widespread use of dual-fuel burners, fuel oil combustion
properties are also covered. The basic combustion reactions are discussed but the detailed mechanism
of combustion is not treated in any depth.
The main properties of natural gas relating to flame structure and development are described in
detail with particular emphasis on burner design.
To determine combustion performance, the heat losses in flue gases are calculated and presented
in a series of graphs. For condensing flue systems a method of calculation is described for assessing
combustion performance based on hygrometric charts.
Techniques for the determination of flue gas loss on actual plant are also highlighted.
The commodity which is distributed throughout Britain as a gaseous fuel is natural gas. The term
natural gas is generally used to describe the flammable gas which is derived from the earth's crust,
usually from deep boreholes and frequently in association with crude petroleum. The main com-
bustible constituent is methane but there are others, notably ethane, propane, butane, pentane
and some higher hydrocarbons, in decreasing order of proportion. There may also be other, non-
combustible, gases such as nitrogen and carbon dioxide.
The exact composition varies according to the source of the fuel and a detailed analysis of natural
gas types is shown in Table 1.1.
Methane forms the major constituent of natural gas and it is from methane that most of the heat of
combustion is obtained.
Propane and butane are readily liquefiable gases associated with petroleum distillation which are
handled as liquids but find application as gaseous fuels. They are generally refered to as liquefied
petroleum gases and are marketed as fuels in their own right. Although described as 'propane'
and 'butane', they are in fact mixtures, with the titular constituent comprising about 90%. The
combustion properties of methane, ethane, propane and butane are broadly similar although ethane
generally occurs only as a constituent of commercially distributed natural gas. Composition analysis
of liquefied petroleum gases is shown in Table 1.2.
4 Chapter One
Composition % Volume
Nitrogen N2 2.72 1.78 1.56 2.53 0.47 7.97
Helium He - 0.05 - - - -
Carbon Dioxide CO 2 0.15 0.13 0.54 0.52 0.32 0.55
Methane CH 4 92.21 93.63 93.53 91.43 95.28 85.26
Ethane C2 H6 3.6 3.25 3.36 4.10 3.71 4.45
Propane C3 HS 0.9 0.69 0.70 0.99 0.18 1.02
8utanes C4 H 10 0.25 0.27 0.24 0.33 0.03 0.52
Pentanes CsH 12 0.07 0.09 0.07 0.10 0.01
Higher Hydrocarbons 0.1 0.11 - - - 10 .23
Composition % Mass
Nitrogen N2 4.38 2.90 2.55 4.05 0.79 12.10
Helium He - 0.003 - - - -
Carbon Dioxide CO 2 0.38 0.33 1.39 1.31 0.84 1.31
Methane CH 4 84.89 87.37 87.28 83.62 91.10 74.08
Ethane C2 H6 6.21 5.69 5.88 7.03 6.65 7.25
Propane C3 HS 2.28 1.59 1.80 2.49 0.47 2.44
8utanes C4 H lO 0.83 0.91 0.81 1.09 0.11 1.64
Pentanes CS H12 0.29 0.38 0.29 0.41 0.04
Higher Hydrocarbons 0.74 0.827 - - - 11.18
Ultimate Composition % Mass
Carbon 72.16 73.23 72.93 71.98 74.38 66.07
Hydrogen 23.18 23.62 23.51 23.02 24.21 20.88
Nitrogen 4.38 2.91 2.55 4.05 0.79 12.10
Oxygen 0.28 0.24 1.01 0.95 0.62 0.95
Composition % volume
Methane CH 4 - 0.1
Ethane C2H6 1.5 0.5
Propane C3 HS 91.0 7.2
8utane C4 H lO 2.5 87.1
Propylene C3 H6 5.0 4.2
Composition %mass
Methane CH 4 - 0.03
Ethane C2H6 1.02 0.27
Propane C3 H S 90.91 5.70
8utane C4 H lO 3.30 90.83
Propylene C 3 H6 4.77 3.17
The main properties of gaseous fuels related to combustion are given in Table 1.3. Calorific Value
(specific enthalpy) is the most significant property of a fuel, and for gases is defined as the quantity
of heat released by the complete combustion of a standard volume in air and the subsequent cooling
of the products of combustion back to standard reference conditions. British Gas standard conditions
are 2B8.7 K and 1 013.7405 mbar absolute and specific enthalpy units are MJ/m 3 . (Consult Chapter
12 for further information on gas standard conditions).
Notes: 1. Reference conditions 288 K and 1013 mbar absolute unless otherwise stated.
2. Where fuel is liquid at reference conditions, quoted gaseous properties are for
ideal-gas volumes
3. Calorific value (saturated) = 0.9826 x calorific value (dry).
Cooling the products to 288 K (15°C) implies the inclusion of the latent heat of condensation of
the water vapour from combustion of the hydrogen content and the calorific value is thus termed
the gross value. This latent heat is frequently not available in practice as most heating processes
reject flue gases at temperatures above the water dewpoint; the nett calorific value, which exludes
the water vapour latent heat, is a more realistic expression of the heat potentially available. For
natural gas, the nett calorific value is about 10% lower than the gross; for any fuel, the difference
6 Chapter One
between gross and nett depends on the fuel's carbon: hydrogen ratio. I n the UK, the gross calorific
value is generally used, but in most other countries the nett value is quoted and care must always be
exercised when comparing or collating information from various sources.
The relative density of gases is generally quoted as the ratio of the mass of a unit volume of the gas
to the mass of the same volume of air. The relative density of natural gas is a little over half that of
air so that natural gas if freely released in air rises and thus more readily disperses. This is important
since leaks in reasonably well ventilated spaces clear quickly and gas released from vents and pres-
sure reliefs will disperse with minimum hazard.
The liquefied petroleum gases have densities greater than air and if released into the atmosphere
sink to the ground and accumulate in ducts etc. This is a serious disadvantage to their use and repre-
sents a potentially hazardous situation. Great care must be taken to avoid leaks in L.P.G. systems
and freely discharging relief valves are not acceptable except under very carefully chosen conditions.
Combustion is a chemical reaction involving the combination of fuel and oxygen to produce heat
and combustion products. The most convenient source of oxygen is that which forms almost 21 %
of atmospheric air. Hence the combustion of natural gas is in most cases carried out in this way
although some specialised applications demand the use of pure oxygen.
The combustion reaction can be represented by way of a chemical equation for each constituent
present in the fuel. This enables the theoretical air to gas ratio to be determined together with the
volumes of combustion products produced. For example the reaction for methane gas is:
However since two volumes of oxygen are taken from 9.52 volumes of air the complete reaction
using ambient air is:
It should be noted that this equation is an over-simplification and represents the terminal condition
only. Combustion with oxygen of a fuel containing carbon and hydrogen is initiated by a chain
reaction with the intermediate transient existence of many unstable compounds. 'Nitrogen' in air
includes very small amounts of other inert gases (e.g. carbon dioxide) but this may be ignored for all
except the most exacting calculations. Similarly, the small quantity of water vapour in air is only
taken into account when preparing very detailed heat or material balances.
Stoichiometric combustion conditions are those where the relative fuel and air quantities are the
theoretical minimum needed to give complete combustion, and can be readily calculated knowing
the analysis of the gaseous fuel by volume or indeed any fuel by ultimate analysis.
For mean Bacton natural gas the calculation procedure is tabulated in Table 1.4.
Column 2 contains the required information from the combustion equation concerning stoichio-
metric oxygen demand. Thus for methane:
Propane C3 Ha + 50 2 ~ 3C0 2 + 4H 2O
Butane C4 H u1 + 6.50 2 ~ 4C0 2 + 5H 2O
Pentane CS H12 + 80 2 ~ 5C02 + 6H 2O
Higher hydrocarbons C9 H20 + 140 2 ~ 9C0 2 + 10H2 O
(assumed C9 H 20 )
CO 2 H 2O N2
1 2 3 4 5
Below the total oxygen requirement in Column 2 is the associated amount of nitrogen, on the
basis that ambient air is oomprised of 3.76 volumes of N2 to 1 volume of O 2 .
Column 3 gives the amount of CO 2 produced on stoichiometric combustion for each of the consti-
tuent gases, with columns 4 and 5 giving the same information for water and nitrogen.
Combustion in practice is hardly ever carried out in stoichiometric conditions. Some industrial
burners may operate at air/gas ratios which are extremely close to the theoretically correct value
but the majority of burners require a measure of air in excess of the stoichiometric quantity to
ensure that combustion is carried to completion.
The main reason for the requirement of excess air is failure of the gas and air streams to mix
completely in the distance over which combustion is required to take place. The occurrence of
combustion is dependent upon the collision of molecules of fuel with molecules of oxygen. If there
are deficiencies in the mixing of the two fluids then an excess of oxygen must be provided to
increase the incidence of molecular collisions.
A convenient method of relating the proportion of air actually present in a combustion system to
the theoretically required amount is as a percentage above requirements or as a factor.
The air factor is expressed as the ratio of the air actually used (vol./vol.fuel) to the stoichiometric
air demand (vol./vol.fuel):
If the flow of fuel and air to a burner system could always be reliably metered the determination of
air factor could be readily calculated. This is however seldom practicable particularly as regards the
supply of air and thus the air factor must be determined from other measurable parameters. The
presence of excess air dilutes the stoichiometric flue gases and thus reduces the % CO 2 and intro-
duces a % O 2 , Thus excess air can be determined knowing either the carbon dioxide or oxygen
content of the flue products using one of the following equations:
or AF - 1 = (V O /Ao)[02/(21 -0 2 )] 1.3
The ratio Vo/Ao varies from 0.89 to 0.98 for various fuels with a value of 0.898 for natural gas. For
routine purposes as an approximation it can be taken as unity.
For sub-stoichiometric conditions unburned fuel will be present in the form of products of incom-
plete combustion, with carbon monoxide as the major constituent. Clearly the quantity of carbon
in the flue gases from the combustion of a given fuel is the same whether or not the theoretical air
volume is used. Thus to allow for any carbon monoxide (CO) present the excess air equation is
modified:
AF - 1 = (Vo/Ao) [(COzst)/(COzact + CO) - 1] 1.4
where CO = measured CO in dry combustion products % vol.
Combustion 9
The variation in combustion products with air factor for natural gas is shown in Figure 1.1.
carbon dioxide
10
QJ
E
::J
o>
?fi.
oxygen
carbon monoxide
theoretical
real burner
(typical)
The ideal graph of combustion products in Figure 1.1 shows that carbon monoxide is completely
eliminated at the stoichiometric air/gas ratio and that no free oxygen exists at air factors less than
unity. In practice because of inadequate mixing, too small a combustion chamber, and other factors,
this idealised situation will rarely be achieved. A typical real burner performance is also shown in
Figure 1.1.
This emphasises the need to monitor CO as well as O 2 (or CO 2 ) when adjusting burners. If oxygen
concentration alone is monitored then it is possible to reach a burner setting that apparently has a
small amount of free oxygen present but which in fact produces a high level of carbon monoxide.
Products of combustion from a fossil fuel will always include water vapour. The amount of water
vapour is dependent on the initial hydrogen content of the fuel. The proportion of water vapour in
the flue gases will vary from about 19% with natural gas to 7% with bituminous coal and less than
10 Chapter One
12% with oil. It is therefore of considerable importance to define whether an analysis is made on a
wet or dry basis. The relationship between wet and dry analysis for natural gas is shown in Figure 1.2.
12 12
10 10
O2 wet
8 8
CO2 wet
.,
E
">
"0 6 6
>
.c
~
~
u
"
"C
0
.,C. 4 4
"
;::
;ji
2 2
0
0
0 20 40 60 80 100
% excess air
Fig. 1.2 Relationship between wet and dry flue gas analysis
Most instruments used for the analysis of flue gases measure on the dry basis because a sample is
extracted and cooled; water vapour is removed either in the instrument or in an intermediate drying
column and knock-out pot. An important exception is the zirconia oxygen analyser where the
measuring device is inserted directly in the flue: this type of instrumentation measures on the wet
basis.
An understanding of the flame properties of natural gas is important as a basis for burner design and
in order to assess the suitability of burner types for particular applications.
These are the lower and upper percentages of fuel in fuel-air mixtures outside which self-sustaining
flames will not occur. Clearly any burner system must be capable of giving a reliable air-to-fuel ratio
within these limits if a stable flame is to be produced. This is especially important where the estab-
lishment of pilot flames is concerned.
For common gases the accepted limits are given in Table 1.5 although they may vary slightly with
ambient temperature and pressure conditions. The limits are conventionally expressed as %vol/vol
fuel in air, Zo and Zu for lower and upper respectively.
Combustion 11
Combustible Gas Zo Zu
These limits may be expressed in terms of air factor. For natural gas they are:
It is important to note that these limits apply to the combustion zone; in practice, combustion
outside these limits will be found in the furnace or flue due to such factors as hot surroundings
or the use of preheated air.
The limits of flammability of gases have important safety connotations. Leaks, either into the
general atmosphere or into a combustion chamber, are more easily rendered harmless if the gas in
question has narrow flammability limits. A leak of natural gas is more easily diluted, compared
with acetylene or hydrogen, to below its lower level of flammability. If the leak is into a closed
combustion chamber then the concentration will quickly build up until it has passed through the
upper flammability limit.
Flammability limits can be calculated for mixed industrial fuel gases using Le Chatelier's principle:
Combustion is initiated by ignition, the purpose of which is to raise the system represented by the
fuel and air mixture to a level at which the combustion reaction becomes self-propagating-. The
concept of ignition temperature postulates a minimum temperature below which ignition cannot
occur. Depending on external conditions, development of enough heat to compensate for heat
dissipated will permit such a temperature to be reached.
The amount of energy required to initiate the combustion reaction can be measured and is expressed
as the minimum ignition energy. For natural gas the value is 0.30 mJ and applies to an air-to-gas ratio
12 Chapter One
of 1.5: 1 or approximately 17% excess air. The amount of energy required increases for other ratios
as shown in Figure 1.3 and approaches infinity at the limits of flammability. Similarly an increase in
proportion of inerts increases the minimum ignition energy. Minimum ignition energy is also related
to the laminar burning velocity:
The minimum ignition energies at stoichiometric conditions for propane and butane are 0.30 mJ
and 0.38 mJ respectively, compared with 0.341T).J for natural gas.
Where heat is used to initiate combustion, the minimum ignition temperature is appropriate, which
for natural gas is approximately 650 DC.
2.0
1.0
--, 0.9
E
> 0.8
Cl
~ 0.7
c:
11>
c: 0.6
o
.+='
c: 0.5
.~
E
E 0.4
·c
·E
0.3
0.2
0 5 10 15 20
fuel in air/fuel mixture % vol.
Fig. 1.3 Minimum ignition energy for the methane/air mixture
The burning velocity of a gaseous fuel is the rate at which the flame front travels through a com-
pletely self-burning mixture of gas and air. The rates obtained from experimental work are influenced
by the experimental conditions and any quoted values must be taken only as a guide, since dis-
crepancies may be apparent in values taken from different sources for the same gas.
Combustion 13
Burning velocity is important in the design of premix burners since it is the parameter which deter-
mines how fast the mixture of gas and air may be pushed out of the burner head before the flame
lifts away from the burner and goes out. A gas flame represents a balance between the velocity at
which the gas and air mixture is ejected from the burner and the speed with which the flame front
travels back towards the mixer. Low burning velocities give low maximum flame port loadings
corresponding to a low amount of heat generated in unit time per unit area of burner port. High
burning velocities mean relatively high minimum rates of firing to avoid light-back.
For natural gas the maximum value of burning velocity is 0.43 m/s. This maximum value occurs
only at mixtures close to stoichiometric while as the level of excess air or air deficiency increases
the burning velocity decreases as shown in Figure 1.4. The differences in burning velocity between
gaseous fuels is generally attributable to the differences in hydrogen content. The maximum value
of burning velocity is 0.47 m/s for propane and 0.45 m/s for butane, while that for hydrogen is
3.5 m/s.
0.45
0.40
~
M
en
N
cill
E
~
C"CI 0.35
~
C"CI
--
'"
E
>-
·u
~
E 0.30
III
>
Cl
t:
E
::l
.c
0.25
The rate at which heat may be generated in a burner depends upon two factors:
The volume flow rate of fuel gas through the burner port, qv.
The specific enthalpy of the fuel gas, h.
Thus the heat generated, 0, is proportional to the product of the enthalpy value and the flow rate.
14 Chapter One
Now for a given pressure of fuel gas at the burner the flow rate is inversely proportional to the
square root of the relative density of the gas, d. Thus the heat released can be expressed as:
Q ex h/dO. S 1.7
where h/d OS
•
•
IS
th e W0 bbe N um ber = Calorific value of gas
V Specific gravity
For the same pressure at the burner, gases of equal Wobbe Number will generate heat at equal
rates per unit of burner port area. The burner port area is the area of the orifice through which
fuel gas or fuel gas and air mixture flows from a burner to the atmosphere and upon which the
flame forms. In a multiport burner it is the sum of the areas of all the ports.
In addition to Wobbe Number other factors influence the rates at which gas may be burned on a
burner of given port area, the most important of which is the burning velocity.
For simple atmospheric burners without any form of flame retention device the limiting burning
rate before lift-off occurs is 550 to 660 W per m 2 of burner port for natural gas.
Once having ignited a flow of mixture, a flame is initiated. The development of the flame itself
follows very much the same pattern of chemical events as that relative to ignition, and its progress
is almost entirely proportional to mixing rates, since the reaction time between fuel particles and
oxygen is relatively so small as to be negligible. The precise structure ofthe visible flame will depend
on two factors:
The aerodynamic situation relative to introduction of air and gas to the point of mixing,
and to the intensity with which mixing occurs.
The thermodynamic and aerodynamic situation of the flame relative to its surroundings.
Industrial application of flames depends for success almost entirely on the ability to produce stable
flames. Stability here is the property of location of the concentration of flame reactions in a
constant and predictable position. Normally, this condition is best satisfied when the flame appears
to be securely seated on the port from which gas, or mixture, is issuing. Departure from this condi-
tion can occur with a number of results. The flame can wander away from the issue port, and may
do so to an extent at which extinction finally occurs. With premix burners, the flame can transfer
itself from the mixture issue location to a point further back where gas alone is issuing from an
orifice: this fault is prevalent in ill-desi~ned or badly operated burners using high burning-velocity
fuels. Even incipient instability will contribute to difficulties of reliable ignition and monitoring
of the flame.
Flame structure
Development of flame structure is the result of mixing between combustible gas (or vapour) and the
air (or other oxygen source) which supports combustion. A flame can exist, once a flammable
mixture is ignited, providing the balance of energy distribution of the process producing the flame
is favourable. In general terms this will entail that a supply of energy is continuously available at
the point where it is desired to initiate the flame. This can be provided by several means, including
for example:
In the simplest event, a boundary can be imagined between a body of combustible gas and the air
which is supporting its combustion.
When combustion is occurring, the boundary layer constitutes a flame front, usually visible. In a
flowing system, such as for example in a stream of neat gas containing no added air and issuing
into air, the rate of mixing may be relatively slow, since the mixing process is one comprising
the diffusion of particles from one stream into the other, with a very slow transfer rate indeed.
The boundary between the two gas streams can be imagined as almost intact over relatively long
distances: that is, bulk penetration of one stream into the other, as opposed to particle diffusion,
is relatively slight. When the relative velocity of one stream compared with the other is increased,
the boundary becomes more and more disturbed, and bulk penetration of one stream into the other
becomes an important factor in assisting the eventual particle collision which must occur before
combustion can take place, largely by increasing the area of any boundary between the component
gas streams. These conditions would describe a laminar diffusion flame system, the term laminar
being used in its strict sense as a measure of the Reynolds number of the gas flow condition, related
to a geometrical dimension of the gas port.
The emissions of radiation from flames arises from molecules and particles within the flame gases.
A yellow flame, due to the presence of a high level of suspended particles, is called a luminous
flame, while a blue flame, arising from the presence of the reacting gases only, is described as a
non-luminous flame.
Non-luminous flames
Non-luminous flames are produced, especially from gaseous combustibles, in conditions where
combustion air is supplied in such a way as to prevent subsequent pyrolysis, or cracking, of hydro-
carbons in the combustible gas.
In gas burners, this condition is met when gas and air are mixed prior to discharge from the burner
port on which the flame is to be established. The design characteristics of this type of burner are
discussed more fully in Chapter 2. The flame which such systems produce is characterised by the
familiar double-cone structure, indicating a primary reaction of mixture with secondary air, and a
secondary stage in which the products of the first reaction react with further secondary air. The
result is a blue flame, whose size characteristics (overall length, height of inner cone, volume)
depend on the port diameter on which the flame is seated, and the rate at which mixture is supplied
to the port.
When air and gas are supplied under pressure to a burner system in which they issue from their
respective discharge nozzles in close proximity (in the so-called nozzle-mix configuration) there is
rapid mixing between gas and air streams. Non-luminous blue flames are thus produced so long as
air mixture occurs before unmixed gas can be raised to a temperature high enough for cracking to
occur. Flame length is consequently reduced, and combustion intensity increased. Nozzle mix
burners can be used for heating furnaces and simi lar installations, in temperature ranges where heat
transmission is required to be mainly convective with a minor contribution from radiation from
the flame. The furnace lining is directly involved as a recipient of convected heat, which is partly
re-radiated to the furnace charge and is itself working at a temperature low enough to tolerate high
coefficients of heat transfer. Compactness of furnace, and hence of flame is highly desirable in such
situations.
16 Chapter One
Luminous flames
The principle mode of heat transmission from this type of flame is by radiation from suspended
carbon particles, carried in the stream of flame gases. The carbon particles stem from pyrolytic
decomposition of carbon compounds in the fuel, e.g. methane, during the period at which fuel gas
temperature is high enough for this to occur, but prior to access of combustion air. Control over
the pattern of mixing in such a system is therefore a very important factor, and thus the burner
configuration is decisive to flame performance.
Diffusion (or neat gas burner) flames and pre-mixed slow burning flames produce larger particles
whilst pre-mixed fast burning flames produce smaller and fewer particles. A lesser but still important
amount of heat is transmitted from such flames by radiation from tri-atomic gases, principally
CO 2 and H2 O. The amount of heat transmitted by this route will depend on both the thickness
of the gas stream and its temperature.
The principles of gas and multi-fuel burners are discussed at greater length in Chapters 2 and 3.
One of the most important aspects of the application of combustion theory is the determination of
the efficiency of the heating process. If stoichiometric combustion could be achieved reliably and
the flue gases reduced to room temperature before being exhausted then the maximum use of the
heat released is made. In most practical cases, air factors other than 1.0 are encountered and flue
gases must be exhausted at above the process temperature. This induces two sources of inefficiency.
Firstly, if the air factor measured is less than 1.0 unburned fuel will be present in the form of
products of incomplete combustion with carbon monoxide being the major product. In this event
some of the calorific value of the fuel is discharged into the flue without ever being released in the
form of heat. On the other hand where excess air is present, heat is being absorbed to increase the
temperature of additional air as it passes from burner to flue.
Secondly, flue products leaving the working chamber of equipment at above ambient temperatures
take out useful heat which is lost unless recovered in a recuperator, waste heat boiler or other
means of heat recovery.
In all applications it is necessary to accept that the full gross calorific value of the fuel cannot be
usefully employed and some heat will be lost both from the flue and the heated surfaces of the
plant. However, there are obvious advantages in reducing the heat loss to a minimum. Combustion
performance is an expression used to determine the amount of heat liberated in the process as a
proportion of the calorific value, and is usually derived from measuring the flue gas heat loss:
combustion performance % = (100 - % flue gas loss). (It is often termed 'combustion efficiency',
which is technically incorrect, and is not the same as thermal efficiency since no account is taken of
structural or other losses.)
The heat lost in the products of combustion from fossil fuel fired plant is a combination of several
parameters:
The total enthalpy of the various component gases at the flue gas temperature.
The enthalpy due to water vapour of the flue gas temperature.
Combustion 17
For complete combustion, the percentage gross stack losses can be evaluated using the formula:
where H = total sensible heat content of the combustion products, including the sensible
heat of water vapour (kJ/kg).
hg,hn = gross and net calorific values of the fuel respectively (kJ/kg).
Thus the amount of heat lost can be calculated from the constituent flue gases and graphs can be
=
drawn showing flue losses plotted against excess air for different flue temperatures. Figure 1.5
shows two such graphs for natural gas.
20 100
0c
5c
0c
0c
c
0c
5c
c
0c
0c
0c
0
45
42
40
37
35
50
60
70
55
47
65
15
c
17.5 90
5
32
en
yg
c
ox
0
30
12.5 50
15 80
c
0c 5
40 c 27
5
ca
3 7
rbo
c c
350 12.5 0 70
25
nd
10 c
325
iox
c 40
300 5c
ide
c 22
c 10 60
250 0c
7.5 c 20
225 30
c
200
5 c 5c
1 7 7.5 17 50
c
150
125
c 20 15 0c
2.5 100 c
5 40
mperature c
flue gas te 125
carb
on d
ioxi c
10 de 100
en
2.5 30
oxyg
0 0 0 20
0 50 100 150 200 150 200 300 400 500 600 700 750
% excess combustion air % excess combustion air
Small quantities of un burnt constituents nearly always occur in the flue products from gas and oi I
fired heating plant. Smoke is a good indication of the incompleteness of combustion on oil fired
plant, whereas on gas fired plant it is indicated by the presence of carbon monoxide and any other
un burnt constituents in the flue gases.
18 Chapter One
If it is suspected that significant quantities of un burnt constituents may be present in the flue gases,
it is necessary to determine their concentrations in order to obtain a reliable assessment of the flue
losses. At low excess air levels, un burnt constituent concentrations of less than 100 ppm may be
ignored as their effect on flue losses and excess air/oxygen/carbon dioxide relationship is small
compared with normal errors in measuring flue gas concentrations. For unburnt constituent concen-
trations in excess of this value or for very high dilution with excess air they must be considered.
In order to determine the true excess air level it is necessary to correct the oxygen and/or carbon
dioxide concentrations to allow for burn out of the flammable gases. The 'corrected' oxygen (0 2 )
and carbon dioxide (C0 2 ) concentrations are determined as follows:
These approximate relationships apply to the combustion of natural gas in air for un burnt consti-
tuent concentrations of 1% (10 000 ppm) or less. The corrected O 2 or CO 2 level can be used with
the relevant graph from Figure 1.5 to determine the excess air level for the plant under consideration.
It is also necessary to make allowance for the calorific value of unburnt constituents leaving the
plant. This additional loss Ll can be determined from the following relationship:
[(Volumes dry products per unit vol. fuel gas) x ~(Vihill /hfuel 1.11
Figure 1.6 shows the flue losses represented by various levels of CO up to 10 000 ppm for excess air
levels represented by oxygen concentrations up to 7%. The calorific values of carbon monoxide and
hydrogen can be considered to be equal. Therefore, where the hydrogen and carbon monoxide
concentrations are known, the curves on Figure 1.6 should be used after adding the dry hydrogen
and carbon monoxide concentrations to produce a 'total unburnt' figure which is used to estimate
the additional gross loss. Where the hydrogen concentration is not known the assumption can be
made that in an appliance flue the ratio of carbon monoxide to hydrogen concentrations is approxI-
mately 4:1. The additional gross flue loss can then be determined. The error introduced by assuming
the CO/H2 ratio in the above manner is small.
Products of combustion must always be analysed for two components and preferably for three.
The variation in combustion products with air factor in Figure 1.1 shows that for a given CO 2
concentration combustion can be occurring with either excess air or with high levels of CO. Ideally
flue gases should be tested for carbon dioxide, carbon monoxide and oxygen. The presence of
oxygen in the flue gases would seem to eliminate the chance of carbon monoxide being found but
this is not necessarily true. There are circumstances in which carbon monoxide and oxygen can exist
together. If the reactant gases, which constitute a flame, impinge directly upon a very cold surface
such as the water cooled surface of a boiler before mixing and hence combustion is complete,
the gases may be cooled to below ignition temperature and so combustion will effectively be
extinguished. Under such circumstances, partly burned fuel gas and unreacted oxygen will pass out
of the system and both carbon monoxide and carbon dioxide will be in the products of combustion,
Combustion 19
5.5 35 40
30
% excess ai r 25
20 10000
5.0 15
10
5
9000
4.5 0
8000
4.0
7000
3.5 ~
~
3.0 6000 '"'"E
]
~
'"
... 5000 E
a.
~
'"'" 2.5 a.
0
<J!. u
2.0 4000
1.5 3000
2000
1.0
1000
0.5
500
0
0 1.0 2.0 3.0 4.0 5.0 6.0 7.0
% O 2 in flue gas
even though excess air may have been provided at the burner. No amount of excess air will resolve
this condition.
In condensing recuperators as used for heat recovery, or in direct contact water heaters, a spray of
water passes through the flue products. If the water is at a temperature below the dewpoint of the
flue products and there is sufficient flow rate, then the flue gases are cooled to around the dewpoint
provided that there is a sufficient degree of contact between water and flue gas in order to obtain
the necessary mass and heat transfer. Dewpoint for natural gas combustion products will be about
50°C and will vary with excess air, as shown in Figure 1.7. Sensible heat in the flue gases and some
latent heat from water formed by combustion are recovered. As some water in the combustion
products has been condensed, heat has been recovered and therefore normal flue gas heat loss
charts based on gross enthalpy (Figure 1.5) are inapplicable.
In order to calculate flue losses for a condensing system the procedure outlined below should be
followed.
Condensing mechanism
The theory of a condensing spray recuperator can be explained with the use of a standard hygro-
metric chart. Most charts are produced only for temperatures close to ambient and a high temperature
version is necessary, as in Figure 1.8. The hygrometric chart is for water in air, not water in flue
products, but the error involved is small. Note that enthalpies are quoted above base 0 °c, and must
be corrected to base 15°C or base ambient air temperature, whichever is the reference standard.
20 Chapter One
55
U
...c:
a
a
c.
:l:
III
"C
50
45
o 50 100
excess air %
Fig. 1.7 Dew point vs excess air for natural gas/air systems
0.3
en thalpy kJ/kg dry 800
700
6oo~
...>
"C
0.2 CI
-..
~
CI
500 .. ...c:
~
...c:
III
400 a
u
...::J
III
300
~1\
0.1 '0
...
III
C E
200
100
.... o
o 50 100 150 200 250
dry bulb temperature °c
Fig. 1.8 Hygrometric chart for flue gases
The flue gases leaving the heat exchanger can be represented by point A. As water is added this
water evaporates and the moisture content increases together with the temperature decreasing. There
is no heat loss from the system and therefore the heat taken out of the flue gas appears as heat in
water vapour carried in the flue gas. The line A-8 is followed.
Combustion 21
At point B the flue gases become saturated. As more cold water is added condensation commences
and the flue gases are cooled to their final temperature, point C. The flue gases are still saturated at
the exit from the spray recuperator.
It should be noted that if insufficient cold water is added the line A-B is followed but saturation is
not reached and there is no heat recovery, only a loss of spray water. If the spray water is at too
high a temperature no condensation will take place.
Example 1.1 A boiler operates at a gross thermal efficiency of 80%, consuming 340 m 3 /h
of natural gas. Excess air level is 20% and flue gas stack temperature is 220°C. If a spray
recuperator is fitted, with water entering at 45°C and leaving at 55 °c, and flue gas temperature
is thereby reduced to 50°C, what gas saving will be made and what will boiler efficiency be
raised to? Assume fuel properties as for mean North Sea Gas (Tables 1.1 and 1.3) and combus-
tion air 15 °c and 50% R.H.
kg/kg fuel
Combustion air - stoich 16.5
- 20% excess 3.3
- total 19.8
Fuel 1.0
Total flue gases 20.8 (by virtue of conservation of mass)
Combustion H2 0 = 9 X H2 % mass 2.1
Dry flue gases by difference 18.7
From standard psychrometric charts, combustion air has 0.0054 kg H 2 0/kg dry air, therefore
total H2 0 from air is 0.0054 x 19.8 = 0.1 kg/kg fuel. Total H 2 0 in flue gases = 2.1 + 0.1 = 2.2,
i.e. 0.118 kg/kg dry gases.
Savings:
Savings and the new boiler efficiency can be calculated by equating before and after heat
balances, but this is laborious. The gain in efficiency equals the drop in flue gas loss (ignoring
what should be minor losses within the heat recovery system):
It has been demonstrated that the heat content of the gases can be derived with a good degree of
accuracy from charts showing flue losses plotted against excess air for different flue temperatures.
These graphical methods can only be as accurate as the data that is used with them. This means that
both temperature and excess air must be measured as accurately as possible.
The sampling position used must give a representative sample. This will normally mean carrying out
a traverse of the flue to determine the extent of stratification. It is necessary to avoid dead-spaces to
ensure that the point of sampling is being continuously swept by the gas to be analysed. Points in
the flue or duct system where reasonable turbulence exists are to be preferred for sampling because
the possibility of stratification exisiting is thereby reduced. Reduction in flue cross-sectional area,
operation of fans and bends all assist in breaking up stratification patterns. Sampling on bends
themselves should be avoided because of the probable existence of dead-space. It should be noted
that in ducts of square or rectangular cross-section, the flow of fluid takes up a roughly circular
profi Ie so that the corners of these cross-sections wi II not be subject to much flow and errors of
dead-space could arise if sampling is carelessly performed.
The flue gas sample should be rapidly cooled, usually by means of a cooling coil, after withdrawal
from the flue. This quenching prevents further reaction between the constituents which would give
erroneous results.
Accurate temperature measurement is important. The use of a simple sheathed thermocouple does
not give accurate readings of gas temperature. Radiation from the thermocouple to the surrounding
flue walls, which are cooler than the flue gases, causes an erroneously low temperature to be
indicated. With low temperature plant, such as boilers, this is not too important but a suction
pyrometer should be used for measurement at higher temperatures. Determination of the percentage
of excess air present is usually done by measuring CO 2 or O 2 , preferably together with CO. For
maximum accuracy all three should be measured. The excess air based on each parameter should be
the same, within experimental error: if it is not; an error in reading must be assumed and both CO 2
and O 2 measurements repeated until acceptable correlation is achieved.
Further details of instrumentation and techniques for flue gas sampling will be found in Chapter 14.
An important aspect of the heat treatment of materials is the effect of the furnace atmosphere
on the stock being heated. In most cases the need is to minimise or eliminate an undesirable
effect of the furnace gases, such as oxidation or decarburisation, but there are processes such as
the carburising of steels in which the interaction between the gases and the stock is the purpose of
the process.
In directly fired processes only a limited control of furnace atmosphere is possible and although this
is adequate for some low-temperature heat-treatment processes, for example stress-relieving of low-
carbon steels, an increasing number of heat-treatment processes are carried out in indirectly fired
furnaces. In these the stock is separated from the heating gases by a muffle or radiant tubes and the
working chamber is fed with a separately generated prepared atmosphere.
Combustion 23
Protective atmospheres find their widest application in metallurgical heat-treatment processes but
are also used for protective and fire-prevention purposes in, for example, the paint, pharmaceuticals
and rubber industries and for purging. Coverage in this section will be limited to their application in
the metallurgical field.
The terms 'oxidizing', 'neutral' and 'reducing' are often applied to furnace atmospheres. A common
misconception is that if a furnace atmosphere contains no free oxygen it is necessarily neutral in its
effects. However carbon dioxide and water vapour are both oxidizing agents and furthermore a
furnace atmosphere which may be neutral to one material may be highly oxidizing to another.
Additionally, the temperature of the process exerts an important influence on the neutrality or
otherwise of the furnace atmosphere. Consequently the terms oxidizing, neutral and reducing
when applied to furnace gases produced from combustion with excess air, stoichiometric air and
substoiciometric air respectively, should not be taken as implying that these effects will occur.
The consideration of metal: gas reactions occurring in the furnace chamber is essentially one of the
chemical equilibria existing at the temperature and pressure of the process.
The main reaction to be considered is oxidation and this may be caused by CO 2 and H 2 0 as well as
free oxygen, i.e.
(1) R + CO2 ~ RO+ CO
(2) R + H2O ~ RO + H2
(3) R+ O ~ RO
The vapour pressures of the metal and its oxides are negligibly small and Equation 1.13 may be
rewritten:
Kl = Pco/PCO:l 1.14
(4) Fe + CO 2 ~ FeO + CO
Figure 1.9 shows the equilibrium ratios for this oxidation-reduction reaction.
1100
Furnace gases and most prepared atmospheres consist of multi-component gas mixtures and so in
practice it is necessary to consider not only the reactions which may take place between the gases
and the solids but between the gases themselves. In addition, the metals being treated, in particular
nickel and chromium and to a lesser extent iron, may act as catalysts on the reactions between the
constituents of the furnace atmosphere. The inter-gas reactions are of course temperature-dependent
and it is necessary to consider the atmosphere conditions at all stages of a heat-treatment process,
including cooling.
Industrial protective atmospheres are derived from the combustion or reforming of fuel gases and
contain some or all of the following constituents: nitrogen, carbon dioxide, carbon monoxide,
hydrogen, water vapour and hydrocarbons. The atmospheres of direct gas-fired furnaces will contain
these constituents and in addition sulphur dioxide will be present in the case of oil or coal firing.
However, an increasing number of metallurgical processes utilise inert gas atmospheres maintained
from e.g. liquid nitrogen delivered by industrial gas suppliers.
I ndustrially the most significant reactions are those betVfleen these gases and the constituents of
steel, i.e. ferrite (a-Fe) and cementite (Fe3C) and these are considered below.
The following gas-metal and inter-gas reactions may be involved in the heat processing of ferrous
metals:
Oxidation-reduction
(4) Fe + CO 2 ~ FeO + CO
(5) Fe + H2O ..--
--'"
FeO + H2
Decarburizing-carburizing
(6) Fe3C + CO 2 --'"..-- 3Fe + 2CO
(7) ......
Fe3C + H 2O ..-- 3Fe + CO + H2
(8) 3Fe + CH 4 ...... Fe3C + 2H2
..--
Producer-gas reaction
(9) C + CO 2 ......
..-- 2CO
Water-gas reaction
(10) CO + H 2O ......
..-- CO 2 + H2
Thermal decomposition
(11 ) CH 4 ......
..-- C + 2H2
Others
(12) H 2O + C
......
..-- CO + H2
(13) H 2O + CH 4
......
..-- CO + 3H 2
(14) CO 2 + CH 4
......
..-- 2CO + 2H2
(15) C + O2
......
..-- CO 2
The gas: metal reactions are oxidation-reduction and decarburization-carburization. The combined
effects of the four gases carbon dioxide, carbon monoxide, water vapour and hydrogen on a-ferrite
can be illustrated by constructing equilibrium diagrams for each reaction, based on Figure 1.9. This
is shown in Figure 1.10.
Reactions (4) to (7) involving reactions with CO 2 and H 20 cause oxidation and decarburisation in
steel. In general it is necessary to keep water vapour and carbon dioxide levels to a minimum in
heat treatment processes. Whether the process is one of oxidation or decarburisation CO and H2
are reaction products and thus an increase in their concentration wi II inhibit those reactions. It is
Combustion 25
1 lOa
o oxidation
CO
2 R reduction
1000 CO
o decarburization
CH4
H2 C carburization
H2O
900 H2
reduction range
D C
oxidizing range
Fe
800
Fe3
3C
u CO
Fe3
C
~ Fe 2 3Fe
C
3c 2CO
2H
~ CO
2
2H
~ 700
2
OJ 3fE 2CO
C D
c.
2
E
3F
0
R
OJ
e
3F
I- Fe
O
e
C
H
600 CO2 Fe CO
4
CH
CO
O
CO
4
Fe
Fe
C
O
O 2
O
500
H2
2
range of simultaneous
H2
C
O
oxidation and reduction O
Fe
Fe
R
CO2
H2
400
H2
CO
FeO
FeO
300
0 0.2 0.4 0.6 0.8 1.0 1.2 1.4
C0 2 /CO, H2 O/H 2
possible, as Figure 1.10 indicates, to produce an atmosphere which is non-scaling but decarburising,
but non-decarburising atmospheres must be non-scaling at all practical temperatures. For high
carbon steels and high temperatures the CO/C0 2 ratio will need to be high, primarily to prevent
decarburisation, as can be seen from Figure 1.10.
Industrial controlled atmospheres are generated from two main sources, single relatively pure gases
in cylinders and fuel gases. As an alternative to the use of a prepared atmosphere a vacuum is some-
times employed, particularly for copper annealing. Of the gases drawn from cylinders pure hydrogen
is used where extreme reducing conditions are required and argon for processes in which nitrogen is
not sufficiently inert.
Atmospheres derived from fuel gases may be termed endothermic or exothermic. The terms refer to
the production processes, the former being produced by catalytic reforming of the gas with a small
amount of air (the mixture being well beyond the rich limit of flammability) in an externally heated
retort and the latter by combustion of the fuel with a controlled amount of air ranging from above-
stoichiometric down to just sufficient to maintain combustion. These atmospheres have a wide
range of composition which can be further extended and controlled, if required, by selective purifi-
cation after generation. The processes are flexible and controllable and the gases obtained range
from active atmospheres containing free oxygen and highly reducing and carburising atmospheres
to inert atmospheres comprising almost pure nitrogen. The typical air/gas ratio requirements for the
type of atmosphere produced by a range of common feedstock gases are shown in Table 1.6.
26 Chapter One
Table 1.6 Typical Air/Gas Ratios Required for Controlled Atmosphere Production
Endothermic Exothermic
Rich Lean
Endothermic atmospheres are produced by the catalysed reforming of fuel gases with a small
amount of air. Air and gas from a machine-premix system are fed into an externally heated nickel-
chrome steel or refractory retort containing nickel-impregnated refractory as a catalyst. The
generator temperature is normally about 1 000 °C. The retort is generally heated by tangential-firing
air-blast burners or in some instances by natural-draught burners.
Normal air-gas ratios on various fuel gases are: natural gas 2.4:1, propane 7.5:1, butane 9.8:1. The
aim is to produce exact conversion to H 2 , CO and N 2 , which can be written approximately for
methane as:
A typical composition using natural gas (reactor temperature == 1050 °C) is: 0.2% CO 2 , 19.8% CO,
39.4% H 2 , 0.2% CH 4 , balance N 2 , dew point 2°C.
Carbon is deposited on the catalyst and this needs to be burned off periodically by reducing the
retort temperature below 700°C and passing air through the retort. The interval between burning-
off runs depends largely on the dew point at which the plant has been running, e.g. for _1°C to -5 °c
the interval is about three weeks. After burning off, the nickel catalyst is partially oxidized and
needs to be reduced to metallic nickel by 'conditioning', i.e. passing a mixture with the correct
air-gas ratio at the rated throughput for a few hours.
For gas-carburizing purposes it is usually necessary to enrich endothermic gas by the addition of
a few per cent of propane or natural gas to increase its carbon potential. This is defined as the
percentage carbon content of the steel that would be in equilibrium with the atmosphere at the
temperature of the heat-treatment process. It is discussed in further detail in Chapter 6.
Combustion 27
Exothermic atmosphere generators consist essentially of a burner, combustion chamber and atmos-
phere cooler. Pre-mix burners are used almost exclusively and, although air-blast burners have been
used, the most common arrangement is a machine-premix system which combines relatively accurate
gas-air ratio control with a high mixture pressure to overcome the appreciable system resistance.
Exothermic atmospheres may be classified as rich or lean, the limit of the former being dictated by
the upper flammability limit. Lean exothermic atmospheres are produced by combustion around
stoichiometric proportions, generally slightly below, to produce a gas consisting of CO 2 , N2 and
H2 a
with a small excess of combustibles, but for a few specialised processes combustion with a
very slight excess of air is used. The air-gas ratios on various fuels are typically:
The method adopted for cooling the combustion products in the generator has a significant effect
on the composition of the atmosphere generated and is of particular importance for lean exothermic
atmospheres close to stoichiometric composition. In addition to containing N 2 , CO 2 and H2 0,
equilibrium considerations dictate that at flame temperature stoichiometric combustion products
also contain small but significant amounts of CO, H2 and O 2 together with a variety of atomic
species. If the combustion products are cooled slowly, equilibrium composition is maintained
throughout the cooling process and the cooled atmosphere consists of N 2 , CO 2 and H2 only. a
If on the other hand the combustion products are rapidly cooled, for example by a direct water
quench, the prepared atmosphere retains the same proportions of the trace constituents as at flame
temperature.
For rich exothermic atmospheres, the product analysis varies but for natural gas is typically 15%
H 2, 10% CO, 5% CO 2 and 1% CH 4 on a dry basis.
Rich exothermic gas may be used 'raw' or alternatively 'stripped', i.e. after CO 2 and water vapour
have been removed. CO 2 and water vapour may be removed by three alternative methods: com-
pression over water, the use of molecular sieves or by the use of organic solvents, e.g. ethanolamines.
Stripped rich exothermic gas is extremely reducing and may be used as an alternative to endothermic
gases. Near-pure nitrogen may be produced by CO 2 and water removal from lean exothermic
atmospheres.
The advent of tonnage oxygen plants and large bulk storage has greatly increased the use of oxygen
for combustion in recent years.
Oxygen enrichment of combustion air to increase flame temperatures by reducing the proportion
of nitrogen used in the combustion process is of particular value to operators of high temperature
furnaces, both for improving efficiency but, in particular, as a means of increasing furnace produc-
tivity.
As the proportion of oxygen used in combustion is increased, the characteristics of the flame
change. Increased levels of oxygen lead to an increase in flame temperature, an increase in burning
28 Chapter One
velocity and a decrease in ignition temperature. Since there is a decrease in the proportion of
nitrogen accompanying the combustion process there will also be a reduction in the waste gas
volume.
The use of a properly designed burner is essential as rapid deterioration results if parts of a burner
become overheated. In general this is not significant for levels of oxygen enrichment approaching
10% in the combustion air. Most oxy/fuel burners are water-cooled and in this respect oil firing
has distinct advantages as the burner size and hence the water cooling requirement are considerably
less than for oxy/gas burners of equivalent heating capacity.
As the proportion of oxygen in the combustion air is increased the flame temperatures are pro-
gressively increased as shown in Figure 1.11.
2800
theoretical flame temperature C
2600 60
50
30
2200 20
10
2000 0
20 40 60 80 100
% oxygen in combustion air
At temperatures above 1 500°C dissociation of the products of combustion occurs. Thus as oxygen
addition levels increase, combustion reactions are increasingly opposed by dissociation reactions
which absorb heat. The flame temperature at the equilibrium state will therefore be considerably
lower than the theoretical flame temperature.
By adopting the correct technique of adding oxygen to a flame, the heat lost by dissociation can be
regained. When the dissociated species contact the relatively lower temperature walls of the furnace
and the charge, the radicals recombine and thereby release the heat of dissociation. This percentage
latent heat of dissociation against oxygen content of the combustion air is shown in Figure 1.11.
The higher temperatures resulting from oxygen addition will give rise to higher heat transfer rates
by radiation, conduction and convection. Clearly the effect on heat transfer rate will depend on
factors such as flame type, the degree of oxygen addition and the furnace type. Similarly, accelerated
Combustion 29
combustion may also markedly increase flame intensity and non-luminous emissivity, thereby
leading to a change in the mode of heat transfer within a furnace.
80
CIl
::c
% net heat of combustion transferable
E 70
....en CIl furnace charge
at charge
C
...cE 60
0 furnace charge
.;; 50
en
:l at 1550°C
.c
E 40
0
....u0
...ra 30
CIl
...
.I:.
CIl 20
c
<f<
10
20 30 40 50 60 70 80 90 100
% oxygen in combustion air
Fig. 1.12 The proportion of net heat of combustion of methane transferable with
increasing levels of oxygen
The reduced waste gas volume resulting from oxy/fuel firing may be significant where some heat
transfer is effected by convection. Thus the heat loss in a furnace will be lowered due to the reduc-
tion in waste gas volume and thus the potential for heat transfer will be increased. The effect
of oxygen addition on the potential heat transfer from the combustion of methane is shown in
Fig. 1.12.
The addition of oxygen gives rise to a higher burning velocity. A comparison of burning velocity of
various fuels with air and oxygen is shown in Table 1.7.
Oxygen addition results in a reduction in flame dimensions and thereby an increase in heat release
per unit volume. Burner design and the selection of the currect technique of oxygen addition is
thus important in order to avoid light back. The techniques can be broadly categorised into enrich-
ment, under-flame enrichment and oxy/fuel combustion and they are discussed in greater detail in
Chapter 2.
30 Chapter One
Since oxygen enrichment improves the mixing of fuel and oxidant the temperature at which com-
bustion becomes self-sustaining will tend to reduce. A comparison of ignition temperatures in air
and oxygen for various gaseous fuels is shown in Table 1.8.
Crude petroleum as first recovered from the earth's crust is a liquid, from light brown to black
in colour and varying very considerably in other properties, according to its origin. It occurs in
sedimentary rock strata and is tapped by drilling, often to a depth of 3000 m or more. In the
initial stages it may emerge from the well under hydrostatic pressure or under gas pressure. The
gas exerting this pressure, which also occurs in large accumulations contiguous to crude oil, is
natural gas and natural gas should therefore be regarded as part of the same series of compounds
as the constituents of crude petroleum. Petroleum is a mixture of many compounds and is refined
into a wide range of saleable products including fuel oils.
Fuel oils include a variety of products, from distillates such as gas oil to residual oils and blends of
distillates, residuals and bitumens. They have a wide range of physical properties but by careful
manipulation of the blending processes the products may be made to conform in properties to
specification with comparatively close tolerances. A range of fuel oils may therefore be marketed
with each specific product falling into a closely defined category. Table 1.9 lists the salient physical
properties of fuel oils sold in the U.K.
With gaseous fuels good mixing of combustion air and gas is possible and combustion can take place
to completion with very little excess air.
Oil requires atomisation to achieve the intimacy of mixing required for good combustion and this is
therefore a stringent design necessity in an oil burner. It entails the supply of oil droplets at a specific
particle size range into a region in which they can be vapourised on ignition, and in which they will
continue to vaporise once the flame is established. The vapours resulting from this distillation can
then behave as gases and can mix with air and produce flames. At the same time, since solid carbon
is a residual from thermal cracking, the combustion process requires adequate air and sufficient time
for combustion of these carbon particles to be completed.
The air-oil system is mainly operated with oil droplets and air meeting at their respective discharge
ports Pre-mix in the sense in which it was used to describe an air-gas system, is possible only when
Combustion 31
Class D E F G
Typical as Supplied
Water %vol nil 0.2 0.4 0.5
Ash % mass nil 0.1 0.1 0.15
Sulphur % mass 0.3 2.2 2.6 3.0
Density at 15°C kg/I 0.840 0.955 0.976 1.000
Mean specific heat capacity between
o°c & 100°C kJ/kg K 2.05 1.93 1.89 1.89
Min storage temp °c - 10 25 40
Min handling temp °c - 10 30 50
Vol correction factor per K 0.00083 0.00072 0.00070 0.00068
a stream of oil particles can be vapourised before emerging from a port. Oil vapourising burners are
restricted to using oils which have a limited distillation temperature range, with a minimum residue
content especially kerosene.
In the majority of cases air and oil meet at the discharge from the oil atomiser, whether this be
operated on oil pressure alone, or with a second atomising fluid (such as steam or compressed air or
another compressed gas), or by mechanical means.
The combustion of fuel oil is considered in detail in Chapter 3. The flame produced, as in the case
of gas combustion, depends on the mixing properties of the system. This factor predominantly will
determine the pattern of heat release, flame dimension, and combustion performance.
Calculations of the air requirement and the quantity and composition of flue gases follows the
chemical equation requirements of the elements of which the fuel oil is composed. The main
properties of fuel oils related to combustion are noted in Table 1.10.
32 Chapter One
D E F G
The combustion gas analysis is calculated from the oxygen requirement of each constituent on a
mass basis. The volumetric composition of the combustion gas is then derived via density from its
weight composition. The volumetric analysis appropriate to each of the main combustible consti-
tuents of fuel oil is summarised in Table 1.11.
29.
Und der Wald schweigt wie von Andacht gepackt;
der erste Schnee liegt tief und schwer.
Aus Höfen und Scheunen vom Talgrund her
tönt gedämpft der Dreschertakt.
Fern, groß, im weißen Sonnenglast,
steht eine Bäurin und worfelt Korn;
zuweilen blitzt ihr Sieb auf wie voll Zorn,
dann flattern Spatzen. Der Mann macht Rast:
30.
Und die Sonne küßt den Schnee vom Dach,
und leise summt die Glut in den Kaminen.
Lächelnd tritt das Weib ins Turmgemach;
breit vom Morgenglanz beschienen
sinnt der Mann auf seine Arbeit nieder.
Er blickt nicht auf. Sie lächelt wieder.
Leise naht sie ihm in heller Freude,
weich umwogt vom Mutterhoffnungskleide:
31.
Und es tanzt der Schnee; kalt flimmern die Flocken
wie Sterne im schwachen Sonnenschein.
Immer stiller starrt das Weib landein.
Aber wärmer immer, als will er sie feien,
streicht der Mann ihre schwarzen Locken:
32.
Doch eine Nacht kommt, da drohn die Weiten;
da hat der Mond Macht. Grausig rein
erleuchtet sein erlauchtes Licht den Hain.
Und das Weib schluchzt auf, wild auf, wie vor Zeiten:
33.
Und es sprießen wohl Sterne aus der Erde,
so strahlt der Schnee im Mittagsglanz,
so sind die Berge Ein Silberkranz.
Aber strahlender noch als all der Glanz
wird nun des Mannes Blick und Geberde:
34.
Doch die Stunde des Scheidens naht und naht,
wie wenn die Zukunft eilender rollte.
Und sie gehn noch einmal den steinigen Pfad,
wo das Werk ihres Geistes wachsen sollte.
Und inmitten der kahlen, vereisten Flächen
muß das Weib einen alten Zweifel aussprechen:
35.
Und Seel in Seele neu begnadet
umschreiten sie die alte Ahnengruft.
In den verschneiten Wäldern badet
ein goldenblauer Morgenduft.
Und Hand in Hand vorbei an Baum und Baum
erzählt der Mann dem Weib einen Traum:
36.
Und Tal und Berge ruhn in bleicher Pracht;
groß blühn die Sterne durch die Bäume,
und lautlos über Raum und Räume
erdehnt ins Leere sich die blaue Nacht.
Und nun ist bald das Schwere vollbracht;
schon rührt sich fern durchs Land, als schlüge
ein Herz im Schnee mit dumpfer Macht,
eisern das Bahngeräusch der Züge.
Und heiß, mit einem Lächeln heiliger Lüge,
haucht das Weib: Nun magst du gehn —
Ausgang
Muttersprache
Vatergruß
ebookgate.com