Note: Descriptions are shown in the official language in which they were submitted.
S
BACKGROUND OF THE INVENTION
1. Field of the Invention
This invention relates to a combustion state deternlin-
ation method and more particularly to a method of quantitatively
determining the combustion state inside a furnace on the basis
of a pulsation pattern of minute pressure overlapping the inner
pressure of the furnace, to an apparatus for said method and
also to a method and an apparatus which control the combustion
inside the furnace on the basis of the results of determination.
2. Description of the Prior Art
In burning various kinds of fuels in a variety of
industrial furnaces using a burner or the like, it is a matter
of the utmost importance how to judge properly the combustion
state for the purpose of preventing the environmental pollution
by nitrogen oxides (NOx) or the smoke involved in the combustion,
for the improvement of the combustion efficiency or for the heat
management such as the effective use of the heat~
It has been possible in a combustion test furnace, etc.
to qualitatively evaluate the combustion state to a certain
extent by inspecting the temperature distribution or the gas
distribution of the flame through sampling holes formed on the
furnace body. However, since a great deal of time and labor
are required for their measurement, the determination is generally
made by observing the flame with the naked eye.
On the other hand, commercial furnaces are not equipped
with the sampling holes because of the structural strength or
for the purpose of energy saving and not a few are not equipped
even with an observation hole. Hence, judgement of the combustion
state has been extremely difficult.
In any case, the fact is that extremely abstract
9~5
1 evaluation of the combustion state has so far been made on the
basis of the visual observation to such effect that flammability
is good or bad or the combustion is rapid or slow.
In order to make effective heat management or pollution
preventiOn management inside a combustion furnace, it is
necessary to suitably control the combustion state. For this
purpose is necessary an effective method of rapidly and
quantitatively determining the combustion state inside the
furnace. If such determination method is possible, full
automatic control of the combustion could also be actualized.
Accordingly, the inventors of the present invention
have made intensive observation and examination in detail of
the combustion state under various combustion conditions using a
combustion test furnace and have found a fact that minute
pressure pulsation below about 20 mmAq overlapping the pressure
inside the furnace exhibits a predetermined change in accordance
with the change in the combustion state. As a result of
further studies, the inventors have also found that the combus-
tion states such as the combustion speed inside the furnace, the
flame length, the property of the flame (transparent flame,
luminous flame, etc.), the smoke amount and so forth reflect
clearly on the minute pressure pulsation pattern inside the
furnace and that the power spectral density distribution
obtained by converting the minute pressure pulsation pattern
signal can be used effectively for the practical application
as an index for the determination of the combustion state. The
present invention is completed on the basis of these novel
findings.
SUMMARY OF THE INVENTION
It is an object of the present invention to provide a
9~5
1 determination method which enables to accurately estimate the
combustion state and also an apparatus for said method.
It is another object of th~ present invention to provide
a method and an apparatus which determine the combustion state
and controls the combustion on the basis of the results of the
determination.
It is still another object of the present invention to
provide a combustion control method which enables to ensure high
combustion effeciency and to minimize discharge of nitrogen
oxides and also an apparatus for said method.
Embodiments of the present invention for aacomplishing
these and other objects thereof will become more apparent from
the following detailed description in conjunction with the
accompanying drawings.
BRIEF DESCRIPTION OF THE DRAWINGS
Figure 1 is a side view in section lI] and a front
view in section [II] each showing the combustion test furnace; ;
Figure 2 is a sectional view showing various burner
tips;
Figure 3 is a schematic view useful for explaining the
principal portion of the burning installation position;
Figure 4 is a block diagram for measuring the minute
pressure pulsation;
Figures 5 through 21 are charts each showing the
minute pressure pulsation pattern wherein (A) shows the wave-
form and (B) does the power spectrum;
Figure 22 is a measuring apparatus for measuring the
minute pressure pulsation in accordance with the present
invention;
! 30 Figure 23 is a combustion state determination apparatus
in accordance with the present invention;
--3--
- ' .
.
9~5
1 Figure 24 is a combustion control apparatus in accord-
ance with the present inven~ion;
Figure 25 shows a co~ustion air quantity (excess air
ratio) controlling apparatus in accordance with the present
invention;
Figure 26 shows an atomizing quantity controlling
apparatus in accordance with the present invention;
Figure 27 shows a burner position controlling
apparatus in accordance with the present invention;
Figure 28 shows a combustion air controlling apparatus
in accordance with the present invention;
Figure 29 shows an atomizing quantity controlling
apparatus in accordance with the present invention; and
Figure 30 shows a burner position controlling apparatus
in accordance with the present invention.
DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS
In performing the combustion inside various industrial
combustion furnaces using various kinds of gas or liquid fuels,
the present invention enables to accurately make quantitative
determination of the combustion state by detecting the minute
pressure pulsation overlapping the inner pressure of the
furnace by means of an inner pressure detection probe disposed
in front of a burner inside the furnace and converting the
minute pressure pulsation signal so obtained into a pulsation
pattern represented such as by a power spectral density
distribution in order to use it as an index for the determination.
Combustion is a phenomenon of the oxidation reaction
of molecules of an inflammable matter involving light and heat,
and flame is their aggregate. Inside the furnace, there occurs
random minute pressure pulsation (generally up to about 20 ml~q)
--4--
1 overlapping the inner pressure of the furnace even under the
normal combustion state due to the "flicker phenomenon" or the
"intermittent phenomenon" of the flame or to the local density
fluctuation arising from the combustion of the turbulent
diffusion flow.
The present invention has clarified that the combustion
state can be expressed in terms of the pulsation pattern of
this minute pressure pulsation inside the furnace and has
established the technique enabling to measure the combustion
state rapidly and quantitatively through simple measuring
operations of the inner pressure of the furnace without measuring
the flame temperature distribution and the gas distribution as
required in the conventional method.
The combustion state inside the furnace varies
depending on the combustion conditions determined by varying such
factors as the shape of burner tips, the positions of the
burner tips, the excess air ratio and so forth. Irrespective
of the kinds of factors that vary, however, a certain combustion
state and a minute pressure pulsation pattern inside the
furnace at that time show a predetermined correspondence. In
other wor~s, a specific minute pressure pulsation pattern
under a given combustion condition always represents a specific
combustion state expressed in terms of such pattern.
The explanation in detail will now be given on the
relationship between the combustion state and the minute
pressure pulsation pattern when the factors such as the shape
of the burner tips, the excess air ratio or the like are varied.
The following combustion test is carried out using a combustion
test furnace ~inner diameter lm x length 4m) shown in Figure
1 [I] and [II] so that the fuel and the air are respectively
s
1 jetted for combustion from a burner 3 and a fuel assistant feed
port 4 inside a burner tile 2 of the main frame 1 of the furnace
and at the same time, the minute pressure pulsation is detected
by a minute pressure pulsation detector 5 ln front of the
burner 3 while the flame is inspected from an inspection hole 6
at the tail of the furnace. The exhaust gas is sampled from an
exhaust sampling hole 7 of the flue.
The shapes of the burner tips, the tip position of the
burner and the excess air ratio are employed as the factors
varying the combustion state. As the shapes of the burner tips,
three kinds of shapes such as shown in Figures 2 [I], ~II] and
[III] are used. Figure 2 [I] shows the so-called ordinary
type which has plural jet holes (a) arranged in the progressively
broadening form and has the highest miscibility between the fuel
and the air for combustion, thus providing good combustibility.
Figure 2 [II] shows the so-called straight type which has one
jet hole (b) at its axis in parallel to the burner shaft and
has the miscibility between the fuel and the air for combustion
ranked between the abovementioned ordinary type and the
eccentric type to be next described. Figure 2 [III] shows the
so-called eccentric type which has a jet hole (c) at a predeter-
mined angle of inclination with respect to the burner axis and
has mild miscibility, thereby allowing the combustion to
proceed gradually.
As shown in Figure 3, the tip o the burner is placed
at positions (a) - (g) from the furnace inner surface (F) of
t~ (i~;)
the burner tile 2 towards an air resister ~ at the back o the
burner tile. Numeric values in the drawing represents the
distance from the furnace inner surface (F) of the burner tile.
The position in the normal combustion is near *he point c
1 (370mm) for the gas type fuel and near the point d (470mm) for
the liquid fuel.
The excess air ratio is changed over four stages in
the range of 0.5 - 9.5% in terms of the 2 concentration of ~ ;
the exhaust gas.
As the other combustion conditions, the combustion
quantity is 40 x 104 Kcal/Hr, the temperature of the air for
combustion is 320C and the open angle of the burner tile is
30 degrees whereby each of these values is kept constant and
each of the abovementioned factors is varied one by one.
On the other hand, as a measuring instrument for the
minute pressure pulsation inside the furnace, there is used an -
apparatus capable of amplifying the pressure detected by a
pressure detector applying a wire strain gauge, directly
reading out the pressure on a meter and enabling to pick up the
output voltage from its record output terminals. Figure 4
shows a block diagram of the apparatus wherein reference numeral
8 represents a pressure pick-up probe; 9 is a detector; 10 is
an amplifier; 11 is a stylus oscillograph and 12 is a data
recorder. The pressure pick-up probe is fitted at a right angle
with respect to the axial direction of the flame at a position
by 400 mm ahead of the furnace inner surface (F) of the burner
tile (see Figure 3) and its tip protrudes by 200 mm from the
side wall of the furnace. The detector 9 is a pressure trans-
ducer by applying a wire strain gauge and its strain quantity
is 430 x 10 6 at 20 mmAq. The amplifier 10 is a dynamic strain
gauge and its output is 2V per 20 mmAq. The oscillograph 11
records the waveform for observation and the data recorder 12
stores the waveform analysis data on a magnetic recording tape.
Figures 5 through 16 show the inter-furnace minute
s
1 pressure pulsation patterns when the shapes of the burner tips,
the tip positions of the burner, the air ratio and so forth are
varied wherein (A) represents the waveform recorded in the
stylus oscillograph and (B) represents the density distribution
of the pulsation energy obtained by subjecting the pulsation
signal stored in the data recorder to the power spectral
density analysis. The waveform shown in (A) is a part of the
record obtained by recording only the pulsation component by
removing the absolute pressure inside the furnace wherein the
ordinate represents the pulsation pressure (mmH20) and the
abscissa doé's the time (second). The power spectrum shown in
(~) represents the pulsation energy per unit number of pulsation~,
of the irregular pulsation shown in (A) wherein the ordinate
represents the energy density and the abscissa does the
frequency (Hz) within the range of from 0 to 45 Hz.
In the abovementioned analysis, the data of the
measuring time of about 2 minutes are used and 100 Hz is
divided approximately equally into 1000 and then digitalized.
Incidentally, the power spectral density of the
random fluctuation can be obtained by the following equation;
Self-correlation coefficient R(~)
(T
- Qim y(t)y(t + ~)dt
2T
T-~ 'r
wherein 2T is an analysis time (zone); y is a pulsation
pressure; and
t is a time.
The power spectral density can be obtained by the
Fourier transformation of the abovementioned equation;
; -8-
s
1 Power spectral density P(f)
= ~ exp (-i2~f~)R(7)d
where ~ is a time interval for reading the ~aveform;
f is a frequency and i is a complex number.
Figure 5 [I] - [VII] shows the minute pressure
pulsation patterns when the combustion is made using the butane
gas as the fuel and the ordinary type shown in Figure 2 [I] as
the burner tip, and adjusting the excess air ratio so that the
exhaust gas 2 concentration becomes 3.0 + 0.2 % with the other
combustion conditions being at the respective predetermined
values as mentioned previously and when the tip positions of
the burner are variously changed. The tip position of the
burner is at the point (a) (70 mm, see Figure 3) for [I], at
the point (b) (270 mm) for [II], at the point (c) (370 mm) for -
[III], at the point (d) (470 mm) for [IV], at the point (e)
(570 mm) for [V], at the point (f) (670 mm) for [VI] and at the
point (g) (770 mm) for [VII].
As to the combustion state initially, the closer the
tip position of the burner to the inside of the furnace, the
slower the combustion whereby the shape of the flame is great
and exhibits the long flame. As the burner position is
separated away from the inside of the furnace, the flame
becomes a short flame until the flame is found to become a
transparent flame and the combustion is found shifting into the
rapid combustion state approximate to the so-called premix
combustion flame.
When the pulsation waveforms shown in Figure 5(A) are
examined along with the shift of the combustion state, it is
found that the pulsation describes a large and gradual waveform
.
' :
9~5
1 (Figure 5 [I}, [II]) when the combustion is slow (when the
burner position is near the inside of the furnace), the wave-
form changes from a large waveform to small one as the combustion
changes to the rapid combustion state, and the ~requency
becomes extremely high under the state of the transparent flame
(Figure 5 [VI], (VII]).
When the change in the abovementioned waveforms is
examined with reference to the results of the power spectral
density analysis shown in Figure 5(B), it is found that whereas
the frequency components of about 2 - 4Hz are the principal
components of the minute pressure pulsation in the slow
combustion state (Figure 5 ~I], [II], etc.) the principal
components shift to about 12 - 13Hz as the combustion becomes
quicker (Figure 5 [III¦) and further to about 20 - 25Hz
(Figùre 5 [VI], [VII]).
As the combustion shifts from the slow state to the
rapid state near the premix flame in this manner, the frequency
components of the minute pressure pulsation inside the furnace
exhibits a tendency of increase and hence, it is recognized
that the combustion state corresponds to the pulsation pattern.
Figure 6 shows the pulsation pattern when the combus-
tion is made under the same conditions as the abovementioned
Figure 5 except that the straight type is used as the burner-
tip, and when the tip position of the burner is variously
changed. Figures 6 lI] - lVII] show respectively the cases
where the tip positions of the burner are at the points ~a) -
lg) (see Figure 3).
Since the straight type burner tip in this embodiment
has inferior miscibility in comparison with the ordinary type
burner tip, the luminous flame is yet present even at the
--10--
.
s
1 point (d) in the case of the straight type whereas the com-
bustion becomes considerably rapid and the flame is almost
transparent at the point (d) in the case of the ordinary type.
Though there is such a difference of the combustion state, it
is confirmed that the same predetermined relationship is present
between the combustion state and the pulsation pattern as in the
case of the ordinary type burner tip shown in Figure 5. In
other words, when the combustion is slow and the size of the
flame is great, the pulsation waveform is large and gradual
and its power spectrum consists principally of the frequency
components of about 2 - 5Hz (Figure 6 [I] - [III]). As the
combustion becomes rapid, the power spectrum exceeds about lOHz
and shifts to 20 - 25Hz (Figure 6 lV] - [VII]).
When the difference in the combustion state is observed
from the pulsation pattern by placing respectively the ordinary
type burner tip and the ~traight type burner tip at the same
position, e.g., at the point (d), the power spectrum in the
former no longer contains the frequency components o 2 - 4Hz
but consists principally of 20 - 25Hz (Figure 5 [IV]), whereas
the frequency components of 2 - 4Hz still remain in the latter
~Figure 6 [IV]). It is thus assumed that the combustion
state is slower in the straight type than in the ordinary type.
Figure 7 shows the pulsation pattern when the position
of the burner tip is varied under the same conditions as in
Figure 5 except that the eccentric type is used as the burner
tip. Figures 7 [I] [IV] show respectively the burner tip
positions at the points (a), (b), ~c) and (d). The combustion
state is relatively slow over the entire points (a) - ~d) and
a considerable quantity of the luminous flame is yet to be
observed even at the burner position of the point (d). As to
--11--
P5
1 the pulsation pattern, on the other hand, the waveform is loose
in each case and the frequency components are primarily about
2 - 4Hz, thus indicating that the combustion state is slow
(Figure 7 [I] - [IV]) and that the miscibility is lower than in
the case of the ordinary type or straight type burner tip.
In eachof the aforementioned experiments, the
combustion tests are carried out by use of the burner tip
position as ~he variable factor. Hence, the relationship
between the combustion state and the minute pressure pulsation
pattern is next measured under the same condition as in the
aforementioned experiments except that the burner tip position is
fixed at the point (c) and the air ratio is used as the variable
factor (whereby the adjustment of the air ratio is expressed
in terms of the percentage of the exhaust gas 2 as an index).
The results are shown in Figures 8 - 10.
Figure 8 shows the results when the ordinary type is
used as the burner tip wherein [I] shows the pulsation pattern
when the exhaust gas 2 is 0.75%, lII] does the pulsation
pattern when the exhaust gas 2 is 3.1%, [III] does the pulsation
pat*ern when the exhaust gas 2 is 6.4~ and ~IV] does the
pulsation pattern when the exhaust gas 2 is 9.3~.
In this case, the cornbustion is slow and the luminous
flame can be observed on the low 2 side, but the flame
becomes transparent and the combustion shifts to the rapid
combustion on the high 2 side. Under the condition where the
exhaust gas 2 is at the lowest value of 0.75%, the combustion
also is the slowest and its frequency componenets are
principally of about 2 - 4Hz (Figure 8 [I]). As 2 is
gradually increased, however, the principal frequency components
shift to about 10 - 15Hz ([II], [III]). When 2 is further
~9~5
1 increased to attain the rapid combustion state, the frequency
components as high as about 20 - 27Hz are dominant ([IV]).
Figure 9 shows the case where the straight type tip
is used wherein [I] shows the case where the exhaust gas 2 is
o.6%, [II] is the case where 2 is 3.1~, [III] is the case
where 2 is 6.2~ and 1VI] is the case where 2 is 9.35%. In
the same way as in the cases shown in Figure 8, the increase
in oxygen improves the combustibility. S.ince the miscibility is
lower in this case than in the case of the ordinary type tip,
the combustion state is slower even at the excessive 2 f about
6% and the fre~uency components corresponding thereto are
principally of about 2 - 4Hz up to the 2 concentration of
about 6%. When 2 is about 9%, the combustion shifts to the
rapid combustion and its frequency also shifts to about 20 - 25Hz.
. Figure lO shows the case where the eccentric type tip
is used wherein lI] is the case where the exhaust gas 2 is
1.0%, [II] is the case where 2 is 3.0%, [III] is the case
where 2 is 6.4% and [IV] is the case where 2 is 9.3~. In
comparison with the ordinary type tip and with the straight
type tip, the combustion state is slower at the same percentage
f 2 and its frequency components are principally of about
2 - 4Hz. Even at the time of combustion of higher percentage
f 2~ the frequency components of high frequency are extremely
small and this corresponds to the fact that the combustion state
is slow.
As described in the foregoing paragraph, it is obvious
that the combustion state and the minute pressure pulsation
pattern inside the furnace exhibit a specific correlation
between them when the combustion is carried out by using the
butane gas as the fuel and each of the ordinary type, the
1 straight type and the eccentric type as the burner tip and by
varying the burner tip position or the air ratio (excessive air
amount) as the variable factor for the combustion condition.
Where the combustion is slow and the luminous flame is present,
the power spectrum consists principally of low frequency
components of about 2 - 4Hz, and where the combustion is rapid
and the transparent flame is present, the frequency components
shift to higher components of about 20 - 25Hz. Thus, this
predetermined tendency has been confirmed. Clear correspondence
is also observed to the difference in the miscibility arising
from the shape of the burner tip. Namely, it is found that the
combustibility becomes rapid even at the exhaust 2 f 3~ in
the ordinary type tip and the frequency components of about
20 - 25Hz are dominant whereas the combustion state is yet slow
even at the exhaust 2 of 9.8% and the frequency is still in the
low ranye of about 2 - 4~z in the straight type and the
eccentric type.
Next, the correlationship between the burner tip
position and the combustibility and its minute pressure pulsation
pattern is examined when the combustion is made using kerosene
instead of the butane gas used in the aforementioned experiments
and also using each of the ordinary-, straight- and eccentric
types of burner tips. The air ratio is set to the exhaust gas
2 f 3.0 + 0.2%. The burner is a commercially available
steam system intermixing type gas burner and high pressure air
is used as an atomizing fluid. The results are shown in
Figures 11 - 13, respectively.
Figure 11 shows the results of the experiments using
the ordinary type burner tip wherein Figures [I] through [III]
represent the cases where the burner tip positions are at the
-14-
s
1 point (a) ~70 mm), at the point (b) (270 mm) and at the point
(d) (470 mm), respectively.
It is found that in the same way as in the afore-
mentioned experiments using the butane gas as the fuel, the same
relationship is present also in the case of the experiments
using kerosene as the fuel between the influence of the burner
position over the combustion state and the pulsation pattern
in that combustion state. Namely, when the burner tip is near
the inside of the furnace, the combustion is slow, but when
it is located at the point (d), the combustion becomes
gradually the transparent flame and shifts to a rapid combustion.
On the other hand, the waveform changes from a large and loose
waveform to a waveform of a high frequency correspondingly. The
frequency components are principally of about 2 - 4Hz during the
slow combustion (Figure 11 [I],[II]), but also as the combustion
shifts gradually to the rapid combustion, the components of
about 2 - 4Hz decrease and those of about 10 - 15Hz and of about
20 - 25Hz start appearing (Figure 11 [III]). The correlation
between the combustion state and the pulsation pattern is
exactly the same as in the aforementioned experiments using
the butane gas as the fuel.
Figure 12 shows the results of the experiments using
the straight type tip wherein [I] - [III] show the cases where
the burner tip positions are located at the point (a), at the
point (b) and at the point (d), respectively.
At each of the points (a) - (d), the degree of the
luminous flame is stronger in comparison with the case when the
ordinary type burner tip is used and the frequency comes to
possess the components of a higher range of about 30Hz,
correspondingly (Figure 12 [III]).
-15-
9~5
1 Figure 13 shows the results of the experiments using
the eccentric type burner tip wherein [I] - [III] show the
cases where the burner tip positions are located at the point (a),
at the point (b) and at the point (d), respectively. In each
case, the combustion state is slow and the frequency components
- are of those in the low range of about 2 - 4Hz, correspondingly.
Figures 14 and 15 show the combustion state and the
pulsation pattern when the combustion is made using kerosene as
the fuel in the same way as the abovementioned experiments,
fixing the tip of the ordinary or eccentric type burner at the
point (d) (470 mm), respectively and varying the air ratio.
Figure 14 shows the results of the experiments using
the ordirary type tip wherein [I], [II] and [III] represent the
cases where the exhaust gas 2 is 2.95%, 6.55% and 9.4~,
respectively. Though the luminous flame is observed on the
low 2 side, the transparent flame is observed on the high 2
side. The frequency distributes over the wide range of about
2 - 4~z, about 10 - 15Hz and about 20 - 30Hz on the low 2
side ~[I~), but consists principally of high frequency
components of about 20 ~ 26Hz on the high 2 side ([II], [III]).
Thus, the frequency exhibits the tendency of shifting to the
waveform having higher frequency components with an increasing
speed of combustibility.
Figure 15 shows the results of experiments using the
eccentric type tip wherein lI], [II] and [III] show respetively
the cases where the exhaust gas 2 is 0.3 - 0.6~, 3.05~, 6.5%,
and 9.6%, respectively. The combustion is slow when the exhaust
gas 2 is up to 6.5%, but it approaches to the transparent
flame when the exhaust gas 2 is 9.6~. On the other hand, the
frequency components are principally of about 2 - ~Hz on the
.
-16-
~ 99~5
low 2 side (lI] ~ [III]) but exhibit the pulsation pattern
having greater portions of components of about 10 - 15HZ and
about 20 - 25Hz at the exhaust 2 of 9.6% ([IV]).
As explained above, the combustion state and the
pulsation pattern exhibit a predetermined correlation between them
irrespective of the kind of the variable factors for the
combustion condition also when kerosene is used as the fuel.
Where the combustion is slow and a large soft flame is formed,
the frequency components are principally of those of about
2 - 4Hz. As the combustion becomes quicker, the frequency
components come to contain those of about 10 - 15Hz and further
those of about 20 - 26Hz. It is thus obvious that detection of
the pulsation pattern enables to estimate the combustion state.
Figures 16 - 19 show the pulsation pattern and the
combustion state when the combustion is carried out using a
heavy oil as the fuel and when either the burner tip position
or the air ratio is varied as the variable factors far the
combustion condition. Incidentally, when the burner position is
used as the variable factor, the air ratio is fixed at the
20 exhaust ga5 2 of 3.0% + 0.2%, while the burner tip position is
fixed at the point (d) (470 mm) when the air ratio is used as
the variable factor. The burner and the atomizing fluid are
the same as those used in the aforementioned experments using
the kerosene fuel.
Figure 16 shows the experiments wherein the ordinary
type is used as the burner tip and the burner tip position is
changed. Figures 16 [I], [II] and [III] represent the pulsation
patterns when the burner tip is located at the point (a), at
the point (b) and at the point (d), respectively.
Figure 17 shows the experiments wherein the eccentric
-17-
:
.. , j ,
s
1 type is used as the burner tip and the burner tip position is
changed. Figures 17 [I], [II] and 1III] represent the
pulsation patterns when the burner tip is located at the point
(a), at the point (b) and at the point (d), respectively.
Figure 18 shows the experiments wherein the ordinary
type is used as the burner tip and the exhaust gas 2 (~) is
changed. Figures 18 [I], [II], [III] and [IV] represent the
pulsation patterns when the exhaust gas 2 is 0.6~, 3.1%, 6.6
and 9.5%, respectively.
Figure 19 shows the experiments wherein the eccentric
type is used as the burner tip and the exhaust gas 2 (%) is
changed. Figures 19 [I], [II], [III] and [IV] represent the
pulsation patterns when the exhaust gas 2 is 3.1~, 6.3% and
9.15%, respectively.
The following may be summarized from the pulsation
pattern and the combustion state shown in Figures 16 - 19. The
pulsation pattern consisting principally of the frequency
components of about 2 - 4Hz is detected when the comhustion
state is judged as being slow from the flame photograph in the
same way as in the aforementioned experiments and on the other
hand, when the flame turns to be a sharp short flame and the
combustion is judged as being rapid, the frequency components
of about 10 - 15Hz and about 20 - 27Hz are detected~
In order to further confirm the correlation between
the combustibility due to the burner tip and the pulsation
pattern, the correlation between them is examined by carrying
out the combustion using a straight type burner tip, which is
made especially and has a varying fuel injection speed and a
butane gas as the fuel while changing either the burner tip
position or the air ratio as the variable factor for the
-18-
,: .
1 combustion condition. The other combustion conditions are the
same as those in the aformentioned experiments, i.e., the
combustion quantity of 40 x 104 Kcal/hr, the air temperature
of 320C and the open angle of the burner tile of 30 degrees.
When the burner position is used as the variable factor, the
air ratio is fixed at the exhaust gas 2 of 3.0% + 0.2% while
the burner tip position is fixed at the point (c) when the air
ratio is used as the variable factor. The results are shown
in Figures 20 and 21.
Figure 20 shows the experiments wherein the burner tip
position is varied. Figures 20 [I] through [IV] represent the
cases where the burner tip is located at the point (a), at the
point (b), at the point (c), at the point (d), at the point (e)
and at the point (f), respectively. According to this burner
tip, the combustion is slow when the burner tip position is up
to the point (b) (270 mm), starts shifting to the rapid combus-
tion already at the point (c) (370 mm) and becomes the trans-
parent flame at the point (d) onwards. In response to the
shift of the combustion state, the frequency of about 10 - 25Hz
starts appearing in the pulsation pattern already at the point
(c) and the low ~requency of about 2 - 4Hz disappears at this
point ([III]). Thereafter, the pulsation pattern having the
frequency of about 20 - 25Hz as the principal components can be
seen at the subsequent points (d) - (f). If has been confirmed
that this b~rner tip has flammability as good as the afore-
mentioned ordinary type tip.
Figure 21 shows the pulsation pattern when the exhaust
gas 2 is changed wherein [I] through [IV] represent the cases
where the exhaust gas 2 is 0.8%, 3.0%, 6.3% and 9.45%,
respectively. As can be seen from these charts, the combustion
--19--
~9~5
1 is good even when the exhaust gas 2 is below 1% and its
frequency contains the components of about 10 - 25Hz (lI]).
As the exhaust gas 2 exceeds this level, the combustion
becomes rapid and becomes the transparent flame at 2 of 9 45~
and the frequency corresponding thereto shifts to the pulsation
pattern consisting principally of about 20 - 25Hz. Hence, the
combustibility of the tip is found as good as the ordinary type
burner tip.
; As can be appreciated from the foregoing explanation,
when the variable factors for combustion are changed, the
- combustion state reflects on the minute pressure pulsation
pattern overlapping the inner pressure of the furnace,
irrespective of the kinds of gas or liquid type fuels and .
irrespective of the shape of the burner tip. Accordingly, it
is obvious that the combustion state inside the furnace can be
properly judged from the pulsation pattern. Namely, when the
combustion proceeds slowing and a large long flame is formed,
there is detected the pulsation pattern having the fr.equency
components o about 2 - 4Hz and as the combustion changes to
the rapid state, it shifts further towards the pulsation pattern
having about 20 - 25Hz. The pulsation pattern and the combustion
characteristics in the combustion test furnace may be
summarized as shown in Table 1.
.
-20-
.~ '.
s
1 Table 1: P~elation between the pulsation pattern
and combustion characteristics
Frequency Combustion
characteristics characteristics
_ _ . . _ . _ . . . _
1 Only components of Combustion state is Fig. 6 [I]
about 2 - 4Hz extremely slow.
Large and soft flame Fig. 9 [II]
is formed.
Some is apt to occur. Fig. 11 [I]etc.
_ . . . _ _ _ _ _ . _
2 Consists principally Combustion is yet slow. Fig. 5 [III]
of about 2 - 4Hz and Though flame is large, F 8 [I]
contains components less column of flame. lg.
of about 10 - 15Hz Fig. 15 1IV]
and about 20 - 25Hz.
O Fig. 18 [I]e~.
. . .
3 Components of about Ordinary combustion Fig. 5 [IV]
2 - 4Hz become less state.
and those of about Flame becomes sharp. Fig. 6 [IV]
10 - lSHz and about
20 - 25Hz are Fig. 14 (III]
dominant.
.
4 Components of about Combustion becomes Fig. 5 [VI]
2 - 4Hz disappear. quicker and sharp
Components of about short of flame is Fig. 20 [IV]
10 - 15Hz become formed.
smaller and those Combustion becomes Fig. 21 tIII],
of about 20 - 25Hz near the transparent etc.
are dominant. flame in the gas
fuel.
.. . . .
5 Only components of Rapid combustion state. Fig. 6 [VII]
about 20 - 25Hz Transparent flame in
the case of the gas Fig. 14 [IV]
fuel. Very few
luminous flame even Fig. 20 [VI],
in the liquid fuel. etc.
. _ . , .
The frequency range corresponding to the abovementioned
specific combustion state is not necessarily sationary, but
exhibits peculiar frequency components in accordance with the
type and the capacity of various kinds of furnaces. It
contains a range of low frequency componentC such as about
2 - 4Hz when the size of the flame is large and the combustion
is slow and comes to contain a range of higher frequency
-21-
~ 9~)S
1 components as the combustion becomes quicker. Furthermore,
under the combustion state wherein the flame approaches to the
transparent flame or near thereto, only the high frequency
components of about 20 - 25Hz appear. In this manner, a pre-
determined frequency peculiar to a given furnace appears under a
specific combustion state of the furnace. If the pulsation
pattern peculiar to the various combustion states of a given
oombustion furnace is graphed in advance, therefore, it becomes
possible to accurately judge the combustion state of that
furnace only by detecting the pulsation pattern.
Furthermore, if determination is made in advance how
much the combustion characteristics change and how much the
pulsation pattern (frequency characteristics) changes
correspondingly when the burner tip position or the air ratio as
the combus*ion state-limiting factor is changed in a predeter-
mined quantity, it becomes possible to optionally control the
combustion state on the basis of the pulsation pattern. Based
on the predetermined correlation betwen the frequency
characteristics and the combustion characteristics inside a
given furnace to be used, such as shown in Table 1, it ls
pos~ible to quantitatively estimate the various states inside
the furnace such as the rapid or slow combustion, the combustion
at the long or short flame or the combustion directed primarily
to the low N0x or the low smoke, for example, in terms of the
objective index, i.e., the frequency. By feeding back the
pulsation pattern to the control system of the combustion
state-limiting factor such as the burner tip position or the
a~r ratio, it is po8sible to correct the deviation of the
combustion from the desired combustion cha~acteristics and
to carry out the combustion in a stable and desirable manner.
-22-
1 Hence, it becomes possible to easily embody the perfectly
automatic control of the combustion state through the
quantitative judgement of the furnace condition according to
the minute pressure pulsation pattern.
Incidentally, an ordinary manometer may be used in
principle as the instrument for the detection of the minute
pressure pulsation pattern inside the furnace. It is practically
preferred, however, to use a pressure transducer applying a
wire strain gauge as the detector in combination with an
amplifier to be connected thereto. The instrument used is such
that it would provide a strain quantity corresponding to the
pressure of about 20 mmAq since the minute pressure pulsation
pattern inside the furnace as the object for the measurement
is at about 20 mm~q.
The pulsation pattern may be read out directly by
reading out the pressure using a stylus oscillograph. It is
however more convenient and more practical to store the pulsa-
tion signal to a data recorder connected to the amplifier
(dynamic strain gause), subjecting the signal to the power
spectral density analysis and reading out the pulsation pattern
on the basis of the resulting frequency characteristics.
Next, the detailed explanation will be given how to
reflect the results of determination of the combustion state on
the combustion control.
For the purpose of the energy saving and the preven-
tion of the environmental pollution, there may generally be used
a number of factors limiting the combustion state inside the
combustion furnace besides the air ratio, the atomizing quantity
of the oil burner and the burner position. However, the
explanation will hereby be given on a method and an apparatus
for controlling these three factors.
-23-
- - \
1 As to the control of the air ratio, it has been
confirmed that when the energy level of a set frequency band
exceeds a set value, a signal for decreasing the air ratio is to
be produced in accordance with the deviation and when the former
becomes lower than the latter, the air ratio is to be increased.
As to the position of the burner, it has been con~
firmed by various experiments that when the energy level of a
set frequency band becomes higher than a set value, a signal is
to be produced so that the burner position is moved in the
rightward direction in Figure 3 or in the direction of the
insi~e of the furnace in accordance with the deviation and
when the former becomes lower than the latter, on the contrary,
a signal is to be produced so that the burner position is
moved in the leftward direction in Figure 3, that is, towards
the throat of the furnace.
As to the atomizing quantity of the oil burner,
urther, it becomes possible to make the stable combustion
control by producing a signal for reducing the atomizing
quantity of the oil burner in accordance with the deviation when
the energy level of a set frequency band becomes higher than a
set value, and by producing a signal for increasing the
atomizing quantity when the former is lower than the latter,
on the contrary.
The detail of the definite example of the apparatus
for controlling these combustion-limiting factors is as follows.
Figure 22 isa block diagram of the apparatus for meaSuring the
minute pressure pulsation wherein the minute pressure inside
the furnace is detected by a detector 9 through a minute pressure
pick-up probe 8 and its signal is amplified by an amplifier 10.
The minute pressure pulsation pattern is determined by a
-24-
.
S
1 frequency analyzer 18 and used as a control signal source. For
example, the power spectrum can be obtained in ~his manner.
Symbols (a) and ~b) in Figure 23 represent respectively a
combustion state determination apparatus wherein a filter 18' -
for picking up only a specific band signal from broad band
frequency components of the minute pressure pulsation is disposed
at the pre-stage of the frequency analyzer 18 while an operator
18" for using the minute pressure pulsation signal which has
been converted into an electric signal by the frequency analyzer,
as the control signal source is disposed at the post-stage of
the frequency analyzer 18. Though only a specific frequency
may be picked up by the filter 18' and then fed to the frequency
analyzer 18, the minute pressure pulsation signal may be applied
as input directly to the frequency analyzer 18 and/or to the
operàtor 18" over the whole frequency band. Alternatively,
control may be made using both whole frequency band signal and
specific frequency band signal.
Figure 24 shows a combustion control apparatus in
which the signal produced as output from the frequency analyzer
18, that is to say, the signal as a resul~ of the analysis,
enters a pulsation energy controller 19 and is compared with the
value of an energy level of a predetermined frequency or of
a predetermined frequency band so that each of the combustion-
limiting factors is suitably controlled via a controller 20 and
a relay 21 shown respectively in Figures 25, 26, 27, 28, 29 and
30 in accordance with the deviation signal. In the combustion
air quantity (air ratio) controller shown in Figure 25, a
combustion air quantity valve 14 is disposed while in the
cascade type controller shown in Fiyure 28, there are disposed
an automatic variable air ratio setter 38, an air flow controller
-25-
s
1 39 and an air flow meter 15, in order to adjust and control
the air quantity.
Next, Figure 26 shows an atomizing quantity controller
wherein the atomizing quantity is controlled by means of a
relay 21 and an atomizing flow control valve 23 of the oil
burner. Figure 29 shows an apparatus equipped with an automatic
variable type atomizing ratio setter 26, an atomizing flow
controller 22 and a flow meter 24.
Figure 27 shows a burner position controller which
enables to move the position of the burner 3 to a suitable
position by means of a burner driving device 30. Figure 30
shows a burner position controller which is equipped with a
guide roller 31 and a burner position control meter 32 in addi-
tion to the burner driving device 30.
Though the present invention has been explained with
reference ~ the abovementioned embodiments, the present
invention is not particularly limited thereto. For example, the
same effect would be obtained by measuring acoustic pressure
using a microphone. Alternatively, combustion state-limiting
factors other than the abovementioned factors may also be
controlled. For example, the fuel injection speed from the
burner, the air velocity or the pressure inside the furnace
may be controlled.
As described in the foregoing paragraph, the present
invention has established in the judgement of the aombustion
state an entirely novel technique capable o~ accurately
estimating the combustion state through a simple operation of
detecting the minute pressure pulsation overlapping the inner
pressure of the furnace, and has thus succeeded in replacing
the conventional method which wholely depends on the observation
-26-
'' ' ~ ' '
9~5
1 with the naked eye through the inspection hole of the furnace
and hence, has been of a qualitative nature and has extremely
low reliability. (Moreover, there are not a few commerical
furnaces not having the inspection hole itself due to the
limitation imposed on the construction of the furnaces.)
Through this quantitative judgement of the state of the furnace
the present invention has made it possible to properly control
the combustion state, to fully automate the combustion control
and to eliminate the troublesome procedure of the observation
with the naked eye and also the disposition of the inspection
hole for that purpose. Thus, the present invention has also
solved the problem of the furnace construction.
-27-
