0 ratings0% found this document useful (0 votes) 276 views12 pagesFlue Gas Analysis
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content,
claim it here.
Available Formats
Download as PDF or read online on Scribd
Gas App
ince Engineers Handbook
ION 3—FLUE GAS ANALYSIS
FLUE GAS ANALYSIS
by Andrew A. Mitchell — Central Chapter
NATIONAL GAE:
FLUE GAs
tis important to determine what is meant by
“Flue Gases” and probably the best deseription is
that flue gases are the final product of a combus-
tion process. The common fueis consist primarily,
of carbon and hydrogea with their compounds and
the flue gases resulting from combustion contain
oxides of these elements and those of the impuri-
together with any excess air and, under some
conditions, other substances evolved from the
heated material,
‘An analysis of the flue gases gives evidence of
efficiency of combustion and is @ prime factor in
controlling the operation for maximum results and
in arriving at improvernents in design. Flue gas
analysis indicates the air-fuel ratio. Good or poor
mixing of the air and fuel determines the signifi-
cance of the flue gas analysis to e considerable
extent. This is shown by the curves given in
Exhibit 1. If the mixing is poor, an excess of air
must be supplied to make sure that every particle
of fuel will contact some air and burn
Before dealing with the analysis of the flue
‘gases, consideration should be given to the prod:
ucts of combustion. It is often desirable to predict
the quantity and analysis of the products of com-
bustion to determine flue sizes and furnace pres-
sure, and to predict the magnitude of stack or flue
gas losses, Exhibit 2 lists the products resulting
from perfect combustion and Exhibit 3 gives equa-
tions to permit calculation of the quantities of
gases in the combustion products of gaseous fuels
either by volume or by weight, The percentages
to be submitted in all of these equations are per-
centages by volume. For cases in which an excess
of air is supplied, it is necessary only to add the
weight or volume of the excess nitrogen and excess
‘oxygen to the other combustion products. When
there is a deficiency of air, it is difficult to predict
the volume or the analysis of the combustion
products, but the total weight of the products is
equal to the weight of the fuel plus the weight of
the air supplied,
31
AWARD — 1965
SIGNIFICANCE OF PERCENT CARBON
DIOXIDE,
With good gas/air mixing, perfect combustion
is obtained when the flue gas analysis shows no
CO, Fl, or O: and when the percentage CO- is at
4 maximum. ‘The theoretical maximum percent
‘0G, in the dry flue gases is termed the ultimate
pereent CO,. If the air-fuel rétio is adjusted until
‘the maximum percent COy is obtained, then the
furnace is operating very ¢lose to the point of
greatest thermal efficiency. This permits the ap-
proximate setting of the air-fuel ratio without
metering the fuel and the air flow,
With poor mixing, the manner in which the
flue gas analysis varies with changes in the alr-
fuel ratio is a function of the physical arrange-
‘ments of the burner and combustion chamber. The
following is the percent CO; readings which
‘should be obtained with perfect mixing of various
amounts of air with natural gas, propane and
butane.
in Combustion Products
ENect of Excess Air on C
ges Teg
Natural (1,000 Bu per 0100 3a
eubie foot, specine gravity 10110 13
96) 2 120 ie
40140 2
60 160 150
80 180 183
100 200, 138
iso 250 235.
200 300 282
Propane (2,629 Blu par 0 100, 29
cuble foct, pecife gravity 10110, 263
158) —Pure measured st 20 120 a7
SOF, 30 Inches mercury 40 140 aa
preisure saturated. Com- 80 180 332
ioercial propane has a &@ 180, a0
heating Value feom 2250 190 200 478
to 2490 Biu'per cubic foot 150 250 501
and’a gravity about 136. 200 300 na
Butane (3,293 Bis per 0 100 at
Sgubic foot, specie gravity 30110 542
2.00)—Pure measured at 20120 ans
Q0F, 30 inches mercury 0140 435
prefsure saturated, Com- 80180, 47
‘mercial grades have heat- 80 180 55.9
Sng values of 9.50 to 190 200 ot
3,250 Btu per cuble foot. 150 250 18
200300 $92Gas Appliance Engineers Handbook
EXHIBIT 1
Flue as analyses are used to indicate the air
fuei ratio and to indicate the degree of complete
ness of combustion. If the mixing is poor, an
excess of air must be supplied so that every par-
ticle of fuel wil! contact some air and burn. Exhibit
1 shows the effects of good or poor mixing.
24
20
Se é
= H,0 2) /%%
Fi a
2
28
3 -—
eo 1,0
é + 6,
— 2
cH, cop
olay WV 1 1 L 1 . 1
60 40S 0 20. 40 60 +80 + +100 +500 + ——*1000
% Deficiency of Air Yo Excess Air
40 6080 100 = 1201401608020 6001100
Effect of FuelAir Ratio on Flue Gas Analysis for 1108
Btu/eu ft natura! gas (0.63 gas gravity) containing 83%
CHy anf 16% Celiy. Values tor rich mixturss dee
Somewhat on contbustion chamber design. ‘The ave
values shown are within ¥% of eoxrect for Hz, CO, an
‘COy, but may be as much ws 29 low for other constitents.
‘Sorhe external heat is Usually required for rmxturos with
jest than 70% aeration {dotted Lines). Dashed lines show
the trends with poor mixing. Gas constituent percenisecs
‘The following table indicates the Carbon Di-
oxide produced in the produets of combustion of
commonly-used domestic appliances, on the dry
basis —
from 2 to § per cent
6 to 10 per cent
3 to 10 per cent
4 to 7 per cent
Ranges
Water Heaters
Room Heaters
Refrigerators
Central Heating furnaces and
‘boilers 6 to 11 per cont
TRACE PRODUCTS
‘Trace products of combustion of fuel gas in-
clude oxides of nitrogen, aldehydes, organic acids
and, if sulphur compounds exist in the fuel, oxides
of sulphur. These trace produets are in such small
32
tke on & Grp basis 10 Allow comparison with of gos an-
alyrers, whieh measure the as volumes after water vapor
Fras condensed out of the sgmple, With correct ratio, this
ss produces 2.11 cu ft HeG, LISP cut Oop, Bab cu ft
= 80%, COs = 100% 1182 eu ft Oy + (1352-4 838)
cul tory Rue gas = 12.08%,
Information came {rom
‘The North American Mig. Co’s Combustion Handbook
amounts that a delicate chemieal analysis is neces-
sary to determine their presence. It is impossible
to detect them by normal flue gas analysis. How-
ever, the presence of oxides of nitrogen and sulfur
has been called to the attention of the gas industry
in the investigation relating to smog control in
the city of Los Angeles. The sharp, penetrating
‘odor which usually accompanies and serves to
identify partially-burned fuel gases is commonly
attributed to the products of incomplete combus-
tion, chiefly aldehydes. The presence of aldehydes
is recognized as an indication of incomplete com-
bustion, When adjusting appliances, no tests for
traces of Carbon Monoxide are made until such
odors have been eliminated. ‘The supply of ade-
quate air for combustion will not always correctGas Appliance Engineers Handbook
EXHIBIT 2
To oT
Specie ‘Theoretical |
imi | Srcigc | rvequied —(Gioss Hul Usimite| Produces of pet combustion!
rut [itive | Sone Isecut| pen | ;
tale | eatvlb coed] wtih susie | '€0s | GO, HO | Ne | SO ] Tow
i rd |
Karbon (© [ass | oo | 29) 30! 0 | ans] oo firs
a | or los fp 0") ao) tm | 0) has
Rydoeen Hd | aero | wR7 239 5 | Bs0 o | soo | ta | 0 | 285
saan | sar | tos | |p | “S| ae | 32
_ ceo tsa | ay 5 cg ffm |e tae | 70" | bam
: sar) jo | um | 36) 100 | ta
tie [ow [we | oe] me fm OE ah Ea
fe oe 2 56) 2265 |i | 2@ Jr 8
wemmiens [eos | na | aoe any Jaws | wae | 28 | 240] 136 |)
; ; yy | {200 | 300 | tsar | 0
Fite (Gti) | x04 | 1230 | rer | ara0 | aos2 | nas |/2@ | 300 | 1322 | @
sonane un | ae }300) 400 tao | 0
roraetety | is | a0 x63 +) 1065 | yo0 [3 | 4 m8
crane (Catton | 2002 | 6 : }300 | $00} ae | 0
suave (atts) | 2082 | 629 wer pro oo (ES) ERS) 8
Octet wo rss | wa! bo | re | ie |e
67 | reer | rae | 1200 | 09 | 94s 0
semen cet] om | vane | nos ve fone | na OST OR | at
| I {
Boldface figures ae in ca f/eu feof foo, lightface Rgutes ate in Ib/lb of fee
Information came trom
‘The North American Atfg, Co.'s Combustion Handbook
this condition, because unbalanced and uneven
distribution of this air and impingement of the
burner flames may be occasioned by burner or
appliance design, permitting incomplete combus-
fon to occur in some part of the flame,
FLUE GAS ANALYSIS —
GENERAL METHODS
Methods for determining flue gas composition,
can be divided into three general categories,
namely, Chemicai, Instrumental and Trace An-
alysis. ‘The latter involves special techniques
which may be either chemical or instrumental and
which usually reguire the services of a trained
chemist or chemical technologist.
In all cases methods for taking the sample are
‘most important and will be discussed briefly. No
attempt is made here to describe specifle apparatus
for sampling or analysis. A bibliography is ap-
pended from which detailed information can be
derived,
SAMPLING
Samples are taken by means of a tube and an
aspirator bulb being drawn either directly into
the analyzer or stored in glass sampling tubes for
analysis at a later date, Sample tubes must be
made of 2 gas-tight material and the kind of tube
3.3Gas Appliance Engineers Handbook
EXHIBIT 8
Equations for Determiniag Products of Combustion of Gaseous
Fuels (all percentages are percents by volume)
1h comb prot
“ST = CO x 0.00256 + WH x 0.001873 4 CH, X 0.00772 +
cafe foek KCiHy X 0.01336 + RCSHy x OID = BC AHia X 00261 +
COs X 0.001164 + BN2 X 0.00074 + $H;O X 0.600476 —
0. x 0.0028
1b.€O,
= = NCO X 0.001165 + #CHy X 0.001165 + SCH, X 0.00230 +
eu (fed RCyHe X 0.00819 + WEAPo X 00465 + HCO: X 00164
1b HO
= tHe x 0.000176 4P#CH4 x 0.000951 + #CiHy x 2.00143 +
cof tua HGHs % 0.001905 + MCyHe x 0.00298 BLO x
‘o.000476
Tb Ne
§CO x 0.001398 + AH: X 0.001398 + KCHe x 0.00361 +
cafe fuel WCaHs X 0.00980 4 CH X.001395 + aCyHio X 0.01821 +
Ns % 0.00074 — Oy X 00028
ce fecom prod
RETR ACO X 00289 +-4Hs X 0.0289 4. KCH, 0.1036 + ACHE
cu fe Fact 0.1824 + 8 Hs 02390 + HCH ie 03360 4 Rimes x
0.01 ~ $05 x 0.0878
eu fe dry comb prod
wr SOO PI = RCO X 0.0289 + Ha X 0.0189 + HCH, X 0.0856 +
cafe fuel GH, x 01524 + KCHE x 02190 4} GH X
0.2860" 4 inenss X DOL ~ WOs x OOS7R
ca fe COs
Sa = 80 x 000k + 8CH X 001 + $GHe X 0002 + SHE x
cu ft fuel 003 + WCiHie X DO + ¥COs X 0.01
2 COs _ cu fe dey com prod
COs in dry due gases = 100
caktud ~— euttfue
cafe 0
= Hs X O01 | HCH, x 0.02 | KCHHy X 008 + CHa x
cw fe fuck 004 + #CyHo X 008 + 6HLO X0O1
eu
= aco X 0.0199 + Hy x 0.0189 + RCH, X 00736 +
cu fe ford $CaHe 0.1824 } MAH X 0.1890 + XCuHe X 0.2460 +
HN? X 0.01 ~ Oz x 00578
Information eame fror
‘The North American Mfg, Co.'s Combustion Nandbook
84Gas Appliance Engineers Handbook
used depends upon the temperature of the flue
being sampied. Here is a general guide—
‘Temperature
up to 250 C (482 F)
up to 500 C (932 F)
abave 500 C
Uncooled Metal Tubes
Combustion Glass
Water Cooled Metal Tubes
Uneooled metal tubes (iron for instance) at high
temperatures not only remove oxygen through the
formation of iron oxide but also may reduce CO.
to CO and produce H, by reaction with water
vapour.
Stainless steel tubes are satisfactory for sam-
pling domestic gas appliances,
Great care must be taken in choosing a loca~
tion that will give a truly representative sample
of the flue produets being tested. Sampling at a
position where the flue produets being tested have
been diluted with air, drawn in through faulty
brick work or through openings in metal ducts,
can give very misleading information. Gases
should not be kept in sample tubes for longer than
necessary particularly where moisture is present
since differential absorption of the various consti-
tuent gases may occur.
A, CHEMICAL METHODS
Chemical methods ere the original and per-
haps best known of alll methods of flue gas analysis
and are applicable to most gases with the excep-
tion of nitrogen, argon, ete. which are chemically
inert,
Flue gases resulting from incomplete combus-
ion also contain methane, hydrogen, carbon mon-
oxide and traces of aldehydes, ete. Determination
of such gases other than CO requires additions to
the ordinary Orsat apparatus. They are not nor-
mally determined in field testing except as com-
bustible components (see instrumental methods)
‘The standard apparatus for analyzing flue
gases is known as the Orset and many types of
Orsat are available, For field work very simplified
forms are available for doing a single gas such
as CO; or oxygen. For example, the Fyrite Tester
is one such device used for testing for earbon diox-
ide. Caustic potash solution is used to absorb
carbon dioxide contained in the flue gas sample,
and this results in a slight vacuum which ean be
measured with @ gage. The temperature of the
flue gas sample may affect the result, and the
apparatus must be completely tight to secure
reasonably accurate results.
a5
Basically an Orsat type apparatus consists of
an accurately calibrated measuring tube (burette)
attached by means of a glass manifold to a suc-
cession of absorbing tubes or serubbers known as
pipettes (See Exhibit 4). There are many designs
of pipettes on the market but all have the same
purpose, namely to bring the gas sample into ine
timate contact with a chemical reagent so that a
specifle component of the flue yas is absorbed by
the chemical as rapidly and completely as possible.
EXHIBIT 4
‘Technical Orsat Assesbly
(Courtesy Fisher Sctentife Cop
In doing an analysis the sample volume (us-
ually 100 ml) is measured in the burette before
and after absorption, The difference in the two
readings is taken as the volume of component
absorbed.
Carbon dioxide and oxygen are very soluble
in water. A 1 per cent solution of sulfuric acid in
distilled water decreases the solubilities of these
gases in water so that it may be used as the con-
fining fiuld in all but the most accurate Orsat
analysis. A permanent acid, one with a very low
vapor pressure, is preferred. A nearly saturated
acid solution of sodium chloride (salt) may be
used but this may lead to difficulties in manipula
tion under conditions of use where a slight lower~
ing of temperature occurs. It is convenient to color
these solutions red with a very small amount ofGas App!
powdered methyl orange for ease of reading and
the methyl orange also aets as an indication that
the solution is always acie.
The burette is enclosed in a water jacket to
help maintain a constant temperature during
analysis, due to temperature change and
variations in barometrie pressure are normally
minimal during the time needed to complete an
analysis, However in flue gas analysis hot samptes
are normally obtained. These samples must be
cooled to the temperature of the water jacket of
the burette or room temperature before it enters
the Orsat. If the sample ix measured and analyzed
above this temperature, each 5F the sample is
above jacket temperature results in an apparent
1 per cent increase in carbon dioxide percentage
due to the contraction of the sample during the
analysis
‘The accuracy of the burette readings are in
the range of 0.2% of the scale reading, It should
be noted that the pipettes are arranged on the
manifold in such @ way that no reagent will re-~
move more than one constituent gas Alkaline
pyrogallol, for instance, will absorb both CO, and
0, and henee must not be used before the C
has been removed by the potassium hydroxide
pipette, The following reagents are used in Orsat
type apparatus for determining the components
in fue gases:
For absorbing CO, —60 per cent potassium
hydroxide,
For absorbing O,—alkaline pyrogallol or
chromous chloride,
For absorbing CO—when the concentration
is over 1% per cent acid cuprous chloride
‘or cuprous sulfate-beta naphol solution
may be used for absorbing carbon monox-
ide, In the concentrations that carbon
monoxide occur in flue products (usually
under .04 per cent) the Orsat apparatus
cannot be used
COMBUSTION METHODS.
Orsat apparatus used for laboratory and more
detailed testing can in addition to the above de-
termine the amount of hydrogen and unburnt
hydrocarbons, These tests are not often made in a
routine flue gas analysis but when required they
can be done chemicaily as follows:
Methane — determined by explosion or slow com-
bustion with a measured volume of O2 and
36
ince Engineers Handbook
determination of the CO, and volume con-
traction due to water vapour. Ignition takes
place either by means of a spark in the
explosion method or over a kot platinum
wire in the combustion method. Combustion
of the hydrocarbon over a catalyst is also
used
CH, } 20, CO, 42H, O
‘Volumes
12 1 2
Hydrogen — determined dy oxidation over hot
platinum catalyst, and de-
contraction due to
copper oxitie
termination of volum
water formed.
2 Oo, °
Hy
ETHANE, PROPANE AND HIGHER
HYDROCARBONS
Only two hydrocarbons such as methane can
be determined in a single mixture by chemical
combustion. Total amount of each is calculated
from the CO, and water formed by using equations
with two unknowns. Instrumental methods have
largely displaced chemical application in this
(See Table in Exhibit 6)
B. INSTRUMENTAL METHODS
Instrumental methods of flue gas analysis
range from portable instruments designed to de-
termine CO, and/or ©, to complicated electronic
assemblies for completely analyzing a flue gas.
‘Two standard methods of analysis are being used
to an ever-greater degree for the latter purpose.
(1) GAS — LIQUID CHROMATOGRAPHY
Gas Chromatography is a comparatively new
technique with a wide feld of applications. Briefly
it consists of the separation of the constituents of a
flue gas by absorption on a packed coluran from a
moving stream of a carrier gas (usually helium)
‘The length of time each constituent gas is retained
on a specific column is constant under given con
ditions. It is known as the retention time and it
depends upon the temperature, flow rate of the
carrier gas, size of column and molecular structure
of the gas. Not all gases can be separated on the
same column and the following are typical columns
and instrument parameters for separating the
constituents of flue gases.Gas Appliance Engineers Handbook
EXHIBIT 5
FORMULAS FOR COMBUSTION ANALYSIS CALCULATIONS
Mixtave | ” x v
Garbon monoxide)
Methane
Ethane J
Garton sree)
cabo Wwsis| cle
firaronen
Ehone
Carbon monoxide
feeeen
Mathene
Methane!
Ethane °
Carbon monoxide! ab
Methane i 3
= 0
Carbon monoxide? z aw
Hydrower i
o
= imuls value contraction; b= mols carbon dioxide produced; «— mols oxygen consumed; w, 2 y and
original sample of carvan motoxide, hydrogen, methane, and ethane, reageetively.
Tatle from Gas Anuiysis— Pestino of ‘Gaseous Moteriol, Published ty American Gas Association.
FOR SEPARATING OXYGEN, NITROGEN For determining very small percentages of
AND METHAN}
Column Material Linde Molecular
Sieve Type 5A.
compounds other types of cells are used. Ioniza-
tion detectors such as the flame and Beta say are
among the more common,
In the ges chromatograph the electronic im-
Column Length 2 Meters pulse emitted by the detector cell is transmitted to
Operating Temperature 40C a moving chart by means of a standard electronic
Flow 50 misma recorder, ‘The position, height and area under the
Soon ge plinin ‘curve drawn by the recorder bears a definite rela-
: tionship to the amount and type of cach gas
separated by the column,
AIR, METHANE, STHANE, CO. AND In doing quantitative anaiysis of a flue ges,
PROPANE the succession of eurves drawn by a given volume
Column Materia! Silica Gel of the sample put into the chromatograph are
: compared to a similar set of curves drawn by an
Column Length 2 Meters equivalent sample of a standard gas, ‘The consti-
Operating Temperature 40C tuents of the standard gas are accurately known,
Helium Flow 50 ml./min, Ifa standard gas is not available air may be used
Sample Stse Imi asa standard when doing nitrogen and oxygen in a
flue gas, Pure COs and air may also be used pro-
vided the amount used as a standard is within
After separation on the column the individual
25%% of the mass of CO, in the sample being tested.
components are moved by the carrier gas into a
detector cell connected to an electronic recorder, The volume of any pure gas varies directly as the
‘There are several forms of detector cells. The one __height of the curve for thet gas,
most commonly used for fiue gas analysis is the ‘Once set up gas chromatographs are easy to
thermal conductivity cell in which the resistance operate but the standardization, and interpretation
of a thermistor bead or platinum wire varies with _of chromatograms should be under the supervision
temperature. This cell operates in much the same _of a well trained technologist.
manner as the indicator cireuit of a combustible ‘This method is satisfactory for separating and
gas indicator. determining different fuel gas components, but is
aTGas Appliance Engineers Handbook
not accurate enough for the small amounts of
carbon monoxide usually found in appliance fue
products.
(2) MASS SPECTROSCOPY
Mass spectroscopy is primarily = laboratory
technique requiring rather more expensive equip-
ment than does gas chromatography
In this procedure the molecular species mak-
ing up a gaseous mixture are toniged and disso
ciated by electron bombardment, the resulting
positive ions of different masses are then acceler
ated in an electcic field and separated magnetically,
A spectrogram is obtained showing the mass of
each constituent and this is compared to a similar
spectrogram for a known mixture.
Instrumentation for mass spectroscopy is ¢on-
siderably more expensive than that required for
gas chromatography. It also requires the attention
of a highly skilled operator.
PORTABLE INSTRUMENTS FOR FIELD
TESTING OF FLUE GASES
There are a number of equally good Combus-
tion Analysers on the market for determining the
CO, and O, content of flue gases. They all operate
on the same general principles and determine the
amount of O, and combustibles. Using the percent
oxygen and curves supplied with the instruments
the percent CO, and the amount of excess air in
the ftue gas can be determined.
‘The principle of operation is catalytic combus-
tion. ‘The combustibles are detected by burning
them with air on a detecting filament (again sim-
ilar method to the combustible gas indicator)
‘The Ox is determined in a similar manner by burn-
ing it in a stream of a combustible gas, usually
propane.
‘As in all instrumentation an operator is re-
quired who is fully conversant with the analyzer
he is using and in the interpretation of results
for example: The O, is limited to @ maximum of
219. Tae Combustibles indication will not be
usable above a fixed concentration of combustibles.
Dilution with air is the limiting factor. When the
Combustibles concentration goes above the value
where enough O; is present the instrument is
liable to give 9 low reading.
C. TRACE COMPONENTS
Of the trace components in flue gases carbon
monoxide, nitric oxide and sulphur trioxide are
the most important,
CARBON MONOXIDE
This is the most important of the minor con-
stituents and certainly the most dangerous. It is
not normally detected in regular Orsat analysis.
‘The following methods are used for detecting ear
bon monoxide in trace amounts:
Hoolamite lodine Pentoxide Indicator — for
0. per cent oc more in contaminated
aimospheres.
Iodine Pentoxide Apparatus—for trace
amounts ‘with precision and accuracy, but
this method is long and arduous,
Pyrotannic Acid Method—for determining
CO in the blood, Fresh blood samples are
needed when used on flue gas samples.
Range of method is 0.01 to 0.2 per cent,
with accuracy of 0.01 per cent.
Hopealite Cell — Fairly accurate method, Dif-
ficult to msiztain and manipulate. Of
historical interest at present.
Infra-Red Absorption — Specific absorption of
infra-red radiation by CO is a distinetive
physical property, which provides a rapid
accurate method when a standard refer~
ence sample is available for comparison.
Used at AG.A. and C.G.A. Laboratories,
its accuracy is 0.001 per cent.
Colorimetric Tubes — A 60 ml, rubber aspirat-
ing bulb is used to draw a sample through
a 15 mm. Jong yellow indicator gel be-
tween white guard gel sections 15 and 49
mm, long. The yellow gel section is com-
posed of @ sulfuric acid solution of am-
monium molybdate and palladous sulfate.
‘Phe yellow gel changes shade uniformly
from yellow to shades of green. The test
color intensity ts @ funetion of the product
of carbon monoxide concentration and
time, The range of concentration for
which this bulb is useful is from 0,001 to
Ol per cent, Other reducing gases also
cause this change in color. Better results
are obtained with these tubes when a
filter made up of potassium dichromate
crystals is used between the sample and
the indicating tube
Length of Stain Indicators — A new tube has
recently been placed on the market in
which the carbon monoxide percentage is,
indicated by the length of stain instead of
the change in color, Preliminary experi-
38Gas Appliance Engineers Handbook
ence with these tubes have indicated good
accuracy.
Nitric oxide and sulphur trioxide in trace
amounts are only determined in special cases
When needed their accurate determination re-
quires rather elaborate equipment and is a job for
the professional chemist.
Water vapour might also be considered as a
minor component. Here once again sampling is
most important and the flue gas must reach the
instrument at a temperature above the dew point
of the gas being tested.
Dew point determinations are only made for
specific pitrposes since a great deal of work has
been done and curves exist from which the dew
point can be obtained when excess air, gas tem-
perature, etc, are known,
One of the simplest and most effective instru-
ments for determining flue gas dew points is
known as the Alnor Dew Pointer. [ts operation is
based on the rapid expansion of compressed gas,
into a sealed chamber containing a very low emis-
sion radioactive source. The rapid expansion of
the sample into the chamber causes a temperature
drop. When this drop is sufficient to lower the gas
temperature below the dew point a fog is visible in
the chamber. A calculator is supplied from which
the pressure differential causing a minimum
amount of fog can be converted into dew point,
The dew point may also be determined by wet
and dry bulb thermometers and cooled cup hy-
grometers. In the latter device a sample of flue
gas is aspirated across the outer surface of a
highly polished metal cup in which is placed a
cooling mixture such as ice and aleohol.
WATER VAPOUR IN FLUE GASES
Water vapour plays a significant role in gas
burning problems. Important points to consider
are
(1) Heat loss due to water vapour can be
calculated at 50.3 Btu per cu. ft. of water
vapour produced (on 60 F., 30 in. Hg.
basis).
Dew point of the fue gases. This is most
marked in the case of domestic appliances
where the moisture may be undesirable
or may be the carrier for corrosive sulphur
acids. It is also important in certain indus-
trial furnace atmospheres.
Radiation from water vapour. Water va~
pour and carbon possess considerable ra-
diating power in the infra-red end of the
spectrum. Thus, they contribute heavily
@
8)
30)
to the radiant heat transmitted from non-
luminous flames.
‘The dew point is of particular importance,
It can be defined as the temperature at which the
vapour phase is in equilibrium with a minute
quantity of liquid phase in any system, More sim-
ply it is the temperature at which condensation
begins to occur: T> minimize eorrosion from gases
coritaining sulphur oxides and water vapour, it is
customary to maintain the temperature higher
than the dew point of the gas. A chart to de-
termine the dew point of the combustion products
of natural gas is given in Exhibit 6, The set of
curves gives the expected dew point of the com-
bustion products when the amount of exeess ait
and the dew point of the air used for combustion
are both known, An example is inserted on the
graph—a burner with 300% theoretical air and
with the dew point of the combustion air at 60 F.
gives the dew point of the combustion product as
107 F. With natural gas, most furnace dew points
tun between 125 F and 135 F.
APPLICATIONS OF FLUE GAS
ANALYSIS
Since the flue gases consist of the end products
of a combustion heating process, their composition
is of interest and importance, chiefly from the
viewpoint of (1) safety; (2) efficiency; and (3)
‘computation of specific values pertaining to either
oF both.
Safety operation requires the elimination,
within reason, of all hazards which might affect,
the continuity of the process. In combustion pro-
cesses the common personnel hazard is carbon
monoxide which can result from incomplete com-
bustion. An operational hazard can be the forma-
tion of an explosive mixture. This hazard can be
present wherever combustible gases or vapours
are in contact with a substance which will support
combustion and the properties are in the explosive
range. Flue gas analysis ean be used to detect,
and/or to confirm and locate suspected conditions
of this nature.
The uses of gas analysis to promote efficiency
are many and varied. Each installation gives rise
to considerations peculiar to itself. Efficiency and
economy do not necessarily imply the most eom-
plete oxidation of the fuel with the least amount
of oxygen or air. In metal heating, scaling loss is
frequently more important economically than fuel
conservation.
In each installation there are spectfic consider
ations which are represented by values reflectingFLUE GAS DEW POINT— DEGREES F
140
130
120
110
107
100
70
Gas Appliance Engineers Handbook
EXHIBIT 6
DEW POINTS of COMBUSTION PRODUCTS
RAL GAS and AIR
8
ATMOSPHERIC DEW POINT — DEGREES F
A
H
100 200 300 400 500 600 700 800 900
% THEORETICAL AIR
Developed by J. C, Clark, Consumers’ Gas Co., Toronto
3-40
1000 1100Gas Appliance Engineers Handbook
either safety, operating efficiency, or both. ‘These
fact as criteria by which the engineer judges and
controls his operations. In most plants, safety and
efficiency are not necessarily separate eonsidera-
tions, Flue gas analysis is the means by whieh
the goals for good utilization can be reached —
(1) for gas appliances to operate safely with-
out liberating injurious quantities of toxie
ases
(2) for gas appliances to operate near op-
timum efficiency.
Exhibit 7 gives an easy formula, sufficiently
accurate, to caleulate excess air in a combustion
‘ocess, also an alignment chart to calculate per-
centage flue loss so that combustion efficiency may
be known,
EXHIBIT 7
1. The following simple formula is useful in des
termining excess air
E= (U-C) xQx 100
AxC
where E = percentage of excess air.
Acs theoretical air required for com
plete combustion of I cu. ft. of gas.
Q= volume of theoretical flue prod-
usts formed by combustion of 1
cu. ft. of gas.
C= % COy is flue gas sample,
U=ultimate 9 CO, in flue gas.
SRE
Example: Suppose « natural yas sample is burned
which requires 10 cu. ft. of air for com
plete combustion and this will develop
11 cu, ft. of flue products per cubic foot
of gas having an ultimate CO, of 12.5%
‘The flue gas sample contains 10.5% CO,
What is the percentage of excess ait?
Exooss air = (125-105) x11 x100 _
Toxis ot”
When using an Orsat-type apparatus in the
field, alignment charts are used to caleulate
percentage flue loss so that combustion efft-
ciency may be known. The following chart
is typical of the type used.
ORGS
:
4
2
won,
(90, 8 FN GASES. SATNAE AND CORE GPE
[OMPERATERE Dirrnce, FLU Gass — ROO
ALIONMENT CHART FOR CALCULATION OF FLUE Lcssey FOR AUTANE,
THOPANE, CORE OVEN AND NATUTAL ales
SatGas Appliance Engineers Handbook
BIBLIOGRAPHY
“Products of Combustion of Gaseous Fuels’
by Edwin L. Hall
Director. A.G.A. Laboratories.
‘orth American Combustion Handbook"
by The North American Manufacturing Co.
Cleveland, Ohio,
“Gaseous Fuels”
An AGA. Publication
edited by Louis Schidman,
“Gas Liguid Chromatography”
by Stephen Dal Wogare, Richard S. Juvety, Jr
Interscience Publishers 1962
“Manual on Hydrocarbon Analysis.”
First Edition — 1983
AATM, Special Technical Publications
No. 932.
American Society for Testing and Materials,
1916 Race Street, Philadelphia 3
“Modern Gas Analysis
by P. W. Mullen
“Gas Analysis — ‘Testing of Gaseous Materials"
hy V. J. Altieri, BS. First Edition
Published by American Gas Association.
312