2. In this DOE, I am trying to attain the best efficiency of a coal boiler. I
wish to improve this process to maximize the efficiency of how the
coal is burned.The current process is as such:
Coal is deposited on to the chain driven stoker which feeds it into
the boiler.
The fire arch composed of radiant fire brick continually ignites coal
via heat radiation from initial light off.
The boiler operator then strives to meet the continually changing
steam load from the boiler by adjusting the FD fan, which effects
the changes the O2% of the combustion gases, the gate level,
which effects the amount of bed on the stoker in inches, and the
stoker speed, which effects how fast the coal move through the
boiler to be burned.
3. The current method used by operators is to simply maintain
the desired out of the coal boiler in thousands of pounds per
hour of steam.The coal boilers are usually base-loaded to
maintain a particular output over their shift.The fluctuating
load may force the operator to either decrease or increase
the output, within specifications, in order to keep the gas
boiler(s) output at a minimum to save natural gas fuel costs.
In analyzing the process I have determined that while output
is maintained to satisfactory specifications, the efficiency of
this output ranges from 60% to 85%.The higher the
efficiency, the less coal we have to use to achieve our desired
output results.The efficiency results are often over looked or
neglected to maintain output specifications.
6. SPC data input We have a stable process.
0.00
2.00
4.00
6.00
8.00
10.00
12.00
14.00
16.00
18.00
20.00
Frequency
Midpoints
Histogram of SPC Project Data
SPC Data
SPC Prob.
7. Since we have determined our process is stable, we are now
ready to run a DOE full factorial analysis.
I have chosen the three main factors used in controlling the
output of these boilers to be used in the experiment.They
are the O2 %, the depth of the coal bed, and the stoker
speed. The set up matrix will be as follows:
8. Design Matrix for
Test
Using our low and high setting
for all three variables, I will try
to determine which set up
produces the highest efficiency
of burn form the boiler.
The stoker is a controller
setting where as each digit
represents a .05 inch per
minute movement.
The settings of 27 and 32
represent a 1.35 and 1.60
inches/minute speed.
DOE with 3 factors (23)
K (run) 8 X1 O2%
n (rep) 3 4% (low) -1 8% (high) +1
Fact 7 X3 Stoker Speed
d.o.f 16 27 -1 32 1 27 -1 32 1
X2gatedepth
5 inches
-1
68.20 70.50 79.80 75.30
72.30 77.00 74.00 70.20
71.30 72.20 81.20 78.20
--- run 1 --+ run 5 +-- run 2 +-+ run 6
8 inches
+1
74.00 77.20 83.20 80.50
70.00 72.20 81.20 82.20
68.20 78.50 77.20 85.60
-+- run 3 -++ run 7 ++- run 4 +++ run 8
Efficiency in percentages
10. Pareto Chart
Here we have the results
from the DOE matrix.
0.000
1.000
2.000
3.000
4.000
5.000
6.000
7.000
X1 X2 X1X3 X2X3 X1X2 X3 X1X2X3
Effects
Factors
Pareto Chart
Factors
11. Here we see that factors X1 and X2 are significant:
The resulting Math Model would be:
75.842 + 3.208 *X1 + 1.658 *X2 + 0 *X3 + 0 *X1X2 + 0 * X1X3 + 0 *X2X3 + 0
*X1X2X3
12. Based on the DOE matrix the maximum efficiency that can be achieved is
80.71 percent.This is achieved by factors X1 and X2 which would be the (+)
effect of 8% oxygen with a gate depth of 8 inches. Stoker speed is not a
factor in this instance.The F ratio andVariance analysis indicate that the
interaction of X1 and X2 are important to consider in the optimum settings,
which has already be decided above. I will recommend these optimum
settings to the Chief Engineer for use in the Plant.