ap-3-12.dvi


Acta Polytechnica Vol. 52 No. 3/2012

Measurement of O2 in the Combustion Chamber of

Apulverized Coal Boiler

Břetislav Janeba1, Michal Kolovratńık1, Ondřej Bartoš1

1
CTU in Prague, Faculty of Mechanical Engineering, Department of Energy Engineering,
Technická 4, 166 07 Prague 6, Czech Republic

Correspondence to: bretislav.janeba@fs.cvut.cz

Abstract

Operational measurements of the O2 concentration in the combustion chamber of a pulverized coal boiler are not yet

common practice. Operators are generally satisfied with measuring the O2 concentration in the second pass of the boiler,

usually behind the economizer, where a flue gas sample is extracted for analysis in a classical analyzer. A disadvantage

of this approach is that there is a very weak relation between the measured value and the condition in specific locations

in the fireplace, e.g. the function of the individual burners and the combustion process as a whole. A new extraction

line was developed for measuring the O2 concentration in the combustion chamber. A planar lambda probe is used in

this approach. The extraction line is designed to get outputs that can be used directly for diagnosis or management of

the combustion in the boiler.

Keywords: O2 measurement, Lambda probe, Pulverized coal boiler.

1 Introduction

The aim of this study was to obtain more detailed
information about the combustion process in pulver-
ized coal boilers. The volumetric concentration of
O2 in the combustion chamber provides important
information about the burn-out ratio, the combus-
tion quality, and the space distribution of the flame.
Simultaneous application of several classical O2 an-
alyzers for making measurements in multiple places
in the combustion chamber is too expensive for ordi-
nary industrial operation. Another disadvantage of
classical types of O2 analyzers is that a flue gas sam-
ple cooler needs to be used, which makes the entire
measurement system complicated and delays the de-
livery of the acquired information. A new analyzer
device has been developed for making simple and af-
fordable measurements in the combustion chamber
of a pulverized coal boiler [1]. It consists of an ex-
traction line, an oxygen sensor, suction pump, and a
periodic cleaning system. A wide range λ-sensor is
used as the O2 measurement element. This is a type
of sensor that is widely used in the automotive indus-
try. Mass production reduces the price of the whole
analyzer. The flue gas is drawn by the pump directly
through the extraction line of the analyzer, without
using the cooler, and then the flue gas passes by the
λ-sensor. The sensor measures the O2 concentration
directly, i.e. in the wet flue gas. Under suggestion
that the combustion process is finished, the air ex-
cess ratio can be determined by Eq. 1

α =
21 + ωwetO2 · ν

(
VF lueGasW et min

VAirW et min
− 1

)
21 − ωwetO2 · ν

, (1)

where ω denotes concentration and V denotes vol-
ume. Ratio ν is defined as

υ =
VAirWet min
VAirDry min

. (2)

The ratio
VF lueGasWet min
VAirWet min

varies from 1.1 to 1.25,

and has to be taken into consideration.

2 Description of the analyzer

The measurement sensor is a BOSCH LSU 4.9 wide-
range (planar) sensor. This sensor [2, 3] is based on
the principle of two ZrO2 electrochemical cells. The
first cell measures the O2 concentration in the test-
ing volume. The second cell is the inverse electro-
chemical cell, and it removes (or adds) oxygen ions,
to held the air excess ratio in this volume equal to
one. The oxygen concentration in the flue gasses is
determined from the electric current flowing through
the second electrochemical cell. An advantage of this
sensor is that it can determine the air excess ratio in
a wide range of values. This sensor can also identify
sub-stoichiometric combustionas well. The extrac-
tion line of the analyzer is equipped with periodic
blow-through periodical cleaning by compressed air.

35



Acta Polytechnica Vol. 52 No. 3/2012

Figure 1: The extraction line and the head of the λ-sensor

The sensor itself is protected by a high-capacity stain-
less steel filter. Without this cleaning, ash and coal
powder can block the extraction pipe within a few
hours. The temperature along the entire extraction
line is controlled to avoid unfavorable condensation.
The sample transport within the analyzer is executed
by a chemically resistant air pump.

The present design of the analyzer with the ex-
traction line (Figure 1) is the result of development
work based on test measurements performed at the
Mělńık 1 power plant. The analyzer itself is installed
in an isolated, temperature-controlled box. The aim
was to develop a compact structure for the analyzer.
Several different structural solutions were abandoned
because they did not provide the anticipated proper-
ties. A former manner of the sample transport was
provided by the air driven ejector. The tests in the
power plant revealed the problem of the ejector get-
ting blocked by a mixture of ash and condensed liq-
uids from the flue gas. Another disadvantage was
the high consumption of compressed air. The ex-
traction pipe inserted in the combustion chamber is
made from high-temperature resistant kanthal mate-
rial. The velocity of the flue gas in this pipe should
be slow, in order to take advantage of the gravitation
setting of the prevailing part of the ash. This set-
tled powder is blown away by the periodic cleaning.
The compressed air for cleaning is stored in a 5-litre
pressure vessel. The cleaning period of 10 sec ev-
ery hour is activated by opening an electromagnetic
valve. The valve is governed by a programmable logic
controller (PLC).

3 Calibration and test
measurements

Several tests were performed in the laboratory with
the extraction pipe and the LSU 4.9 sensor. The
sensor has its own calibration function between the
measured electric current and the volumetric O2 con-
centration. The purpose of the laboratory tests was
to find the properties of our extraction line with the

sensor. We tested the absolute value of the O2 con-
centration, and the time response to the change in
concentration. For the reference measurement, we
used a paramagnetic analyzer PMA12 (M&C). Both
analyzers measured the same gas sample (a mixture
of N2 and air) at the same time. The uncertainty
detected during the tests was lower than 2 % of the
measurement range for the expected operation mea-
surement. This uncertainty is acceptable for this an-
alyzer.

The time delay is in the order of seconds, depend
on the pump performance and the optimum flue gas
velocity in the extraction pipe. For O2 concentra-
tions higher than 15 %, the time response is longer,
but this condition is, in general, outside the desired
operational condition in the combustion chamber.

More attention was paid to the tests in the power
plant. In a laboratory, it is difficult to simulate the
same conditions as in a combustion chamber, espe-
cially the high temperature of the dewpoint and the
high concentration of solid particles in the flue gas.
The measurements were mostly carried out at the
Mělńık 1 power plant. These measurements focused
on the reliability of the entire analyzer and tests of
the periodic cleaning of the extraction pipe. The
heating system of the analyzer was also tested.

Figure 2: Scheme of the extraction line and the an-
alyzer: 1 – air pump, 2 – connection port, 3 – elec-
tromagnetic valve, 4 – air vessel, 5 – connection to
the compressed air line, 6 – PID-controlled heater,
7 –λ-sensor, 8 – isolated box, 9 – filter

36



Acta Polytechnica Vol. 52 No. 3/2012

Figure 3: The inlet to the extraction pipe after a
one-day test

Figure 3 shows the inlet part of the extraction
pipe after a one-day test. This test showed the good
efficiency of periodic cleaning. Without cleaning, the
tip of the extraction line became blocked by ash. Af-

ter this measurement, no trace of unfavorable con-
densation inside the analyzer was detected.

Figure 4 presents an example of the measured
data from a pulverized coal boiler at the Mělńık 1
power plant. After 3 minutes, the analyzer was
placed in the boiler. In the 16th minute, the clean-
ing event by the compressed air blow through can be
seen.

4 Results

When the final design of the analyzer was settled, a
second analyzer was manufactured. For the verifica-
tion of analyzers were used opportunityto measure
the off-design operation of boiler K6 at Mělńık 1,
in collaboration with I&CEnergoCompany. During
these tests, the data from the analyzers was com-
pared with the operational measurement of the O2
concentration in the second pass of the boiler. This
verification measurement was considered successful,
because the new analyzers follow all changes to the
boiler setting. Further verification tests will be made
and operational experience with the analyzer will be
gained.

Figure 4: Acquired data example of O2 concentration

Figure 5: Acquired dataexample from the left and right side of the boiler measured by the boiler control system
and by the tested analyzers

37



Acta Polytechnica Vol. 52 No. 3/2012

An example of the measurements is presented in
Figure 5. The label CVUT denotes the new analyzer,
and EME denotes operational measurements of the
boiler. L and R refer to the left and right side of the
boiler. In the 18th minute of this test, the amount
of air in the right side of the boiler increased rapidly,
while the left side remained stable. The operational
measurement is placed behind the combustion cham-
ber and the information is blurred by the mixing of
the flue gas behind the combustion chamber.

5 Conclusion

The new O2 analyzer was developed, tested and used
for practical measurements. The main advantage of
the analyzer is that it provides immediate informa-
tion about the O2 concentration at the place of mea-
surement. This information can be used in the boiler
control system to optimize some imbalances in the
combustion chamber. The new analyzer was devel-
oped for pulverized coal boilers, but it can also be
used for other types of boilers. T analyzer costs less
than classical analyzers, which raises the question of
the possibility of using the same analyzer or a similar
analyzer to control boilers with lower performance.
Also biomass combustions systems offer a broad field
of potential applications.

If the combustion process is not finished, mea-
sured value provides cumulative information about
the amount of as yet unburned fuel and the excess
air ratio. These two parameters can be distinguished

only with additional information about the combus-
tion. We expect that the analyzer will be able to
provide accurate results only in the region when com-
bustion is finished.

The present development focuses on reliability en-
hancement, especially on the external side of the ex-
traction pipe. The outer surface can be covered by
melted ash; this layer can increase until the weight
of the stuck ash bends the extraction pipe. Periodic
mechanical cleaning of the outer surface is currently
under development.

Acknowledgement

This work has been supported by Grant TIP FR-
TI1/539 of the Ministry of Industry and Trade of the
Czech Republic and by I&C Energocompany.

References

[1] Janeba, B., Kolovratńık, M., Bartoš, O.: Mě-
řićı technika pro výkonněǰśı ř́ızeńı proces̊u spa-
lováńı, část 2. Zpráva FS ČVUT v Praze,
Ústavenergetiky, Z-575/2011, 2011.

[2] Janeba, B., Kolovratńık, M., Bartoš, O.: Měřićı
technika pro výkonněǰśı ř́ızeńı proces̊u spalováńı,
1. Zpráva FS ČVUT v Praze, Ústav energetiky,
Z-571/2010, 2010.

[3] Shuk, P., Bailey, E., Guth, U.: Zirkonia oxygen
sensor for process application: State-of-the-art,
Sensors and transducers journal, 2008.

38