STORM TECHNOLOGIES, INC.

Specialists in Combustion and Power

Storm Service Report 16-127-01

Engie: Tractebel Energia Jorge Lacerda
Unit 4: Boiler Tuning & Pulverizer Performance Testing

Onsite: June 19th 30th, 2017

Submitted: September 8, 2017

Storm Technologies, Inc.

P.O. Box 429 411 N Depot St. Albemarle, NC 28002
Phone (704) 983-2040 Fax (704) 982-96
 STORM TECHNOLOGIES, INC
411 Depot Street, PO Box 429 Albemarle, NC 28001
Phone: (704) 983-2040Fax: (704) 982-9657

September 8, 2017

Mr. Fbio Mendes

Maintenance Engineer
Tractebel Energia
Phone: + 55(48) 3621-4190
Email: fabio.mendes@engie.com
Subject: Storm Service Report 16-127-01 Boiler Tuning and Pulverizer Performance Testing

Mr. Mendes,
Storm Technologies was onsite June 19th-30th to complete testing and tuning on the pulverizers
and boiler on Unit 4 of Tractebel Energias Jorge Lacerda station. Storm personnel conducted
clean airflow balancing, venturi airflow calibrations, pulverizer testing, and overall combustion
tuning. Storm personnel conducted clean air balancing on each mill at minimum airflow and
normal operating temperatures. Storm utilized the newly installed orifice housings to make orifice
adjustments to bring each mills clean air pipe-to-pipe balance to within 2%.
After the completion of clean air balancing, Storm personnel tested each of the installed venturis
located at the total primary air, tempering air, and secondary air ducts. After the initial baseline
tests, the tempering and secondary air differential pressure measurement was calibrated and then
a new K-factor was installed to bring the deviation to within 3% indicated vs. measured. The
primary air differential pressure scale was adjusted on mills A, C, and D to bring the deviation to
within Storms recommended limits. Mill B tested within 3% and required no adjustments.
During the pulverizer testing, Storm personnel conducted tests on each mill to find the classifier
position that produced the best fineness. Mill As classifier position was adjusted multiple times
but yielded no increase in mill motor amps or fuel fineness. Mill C displayed the best fineness with
76% passing 200 mesh sieve.
Storm conducted boiler testing at the furnace exit to analyze the combustion in the furnace.
Measurements of the flue gas and temperature were recorded using the plants water cooled HVT
probe. The testing was completed with the bottom three mills; D, B, and C in operation. Test 2
resulted in the best results with an average of 3.4% O2, 3 ppm CO, and 843 ppm NOx. The collected
flyash LOI was found to be 4.85% (based on the weighted average from sieve analysis). The
following report discusses the testing that was completed on Unit 4 at Jorge Lacerda as well as our
findings during testing. It also details our recommendations moving forward to optimize
performance and improve combustion. It was a pleasure working with you and the entire Tractebel
team and we look forward to working with you again in the future.
Respectfully Submitted,

Matthew Rutherford
Project Engineer
Storm Technologies, Inc.
Executive Summary

Storm Technologies was contracted to conduct boiler tuning and pulverizer performance testing
on Unit 4 at Tractebel Energias Jorge Lacerda station. A Storm designed classifier reject system
was installed on all mills to improve the recirculation of large coal particles in the pulverizer. In
addition Storm classifier blades with coarse particle guides and an extended and flared outlet
cylinder were installed to assist with coal classification. Storm designed orifice housings were
installed in each fuel line to improve fuel balance and distribution. Storm designed venturis were
also installed in the tempering air, total primary air, and secondary air ducts to provide accurate
airflow measurement through each system. Note that Mill C was the only mill that had Storm
burner modifications. Modifications were also completed on the windbox to reduce the tendency
of ash to build up and reduce airflow to the bottom level of burners.

Storm began by conducting clean air balancing on all the mills on Unit 4. Storm recommends that
clean air be balanced to within 2% pipe-to-pipe deviation. Achieving this clean air balance is the
first step to attaining optimum air and fuel distribution. From the baseline tests with no orifices
installed, some of the deviations were as high as ~11%. Storm conducted multiple tests while
making adjustments to orifice sizes to bring all the mills deviations to within the recommended
limits.

Storm designed venturis were also installed for measurement of the tempering air, total primary
air, and secondary airflow. Storm recommends that these airflows be monitored to within 3%
indicated vs. measured airflow. From the initial test, the tempering and secondary airflow were out
of Storms recommended deviation. Only mill Bs primary air indication was within Storms
recommended limits. It was found that the differential pressure (P) in the control logic on the
tempering and secondary air venturis was not matching the local P on the transmitters. When this
was corrected, Storm implemented new K-factors to bring the deviation within 3%. The primary
air transmitter differential pressure range was adjusted until the indication was within Storms
recommended limits.

Storm personnel completed dirty airflow testing and coal sampling utilizing a calibrated dirty air
probe and isokinetic coal sampler. Storm Technologies recommends 5% pipe-to-pipe deviation
from the average dirty airflow and 10% pipe-to-pipe deviation in fuel balance. Storm
Technologies also recommends 75% passing a 200 mesh sieve for optimum combustion. Only
mill A was below Storms recommended 75% passing 200 mesh. Even though all the mills were
clean air balanced before completing the dirty air and coal sampling, only Mill A met Storms
recommended dirty air and fuel balance limits. The recommended splatter plate at the exit of the
main fuel take off of each mill was not installed while Storm personnel were onsite. It is believed
that the dirty air and fuel imbalances measured were impacted from the roping of coal and locations
of the three way splitter in comparison with the exit of the bend.

Storm personnel also completed boiler testing at the furnace exit utilizing two pre-installed
observation ports. Each port was traversed 14 feet (~4.3 meters) with a water cooled HVT probe
and measurements of temperature, O2, CO, and NOx were recorded. Flyash samples were collected
simultaneously during each test to be analyzed for LOI or loss on ignition. Test 1 and 2 were
completed with the top three mills in service; A, D, and B. The baseline test (Test 1) resulted in an

Storm Technologies, Inc.

Tractebel Energia Jorge Lacerda
Storm Service Report 16-127-01
Proprietary and Confidential Page |3
overall average of 1,029C, 3.40% O2, 20 ppm CO, and 798 ppm NOx. Test 2 was conducted after
adjustments to the burner registers. Test 2 resulted in an overall average of 1,046C, 3.16% O2, 3
ppm CO, and 843 ppm NOx.

The following report details our findings from the testing that was performed on Unit 4 at Jorge
Lacerda Station and also details our recommendations moving forward to optimize performance
and combustion.

Clean Air Testing

Clean air flow balancing was completed on each of the four mills, in an attempt to improve the air
flow balance to each fuel line per mill to within 2% of the mean air flow. Baseline clean air
testing revealed that none of the mills met Storms recommend deviation. After the completion of
several tests and orifice plate changes, Storm personnel were able to achieve clean air balances of
2% on all mills. The results of the As Found and As Left clean air testing can be found in
Figure 1 and Figure 2.

Figure 1: "As Found" Clean Air Balance

Figure 2: "As Left" Clean Air Balance

Storm Technologies, Inc.

Tractebel Energia Jorge Lacerda
Storm Service Report 16-127-01
Proprietary and Confidential Page |4
Listed in Table 1 below are the final orifice plate sizes that correspond with the As Left clean
air flow balances. Please note that pipes marked open does not contain an orifice plate.

Table 1: Final Orifice Sizes

Orifice Sizes Mill A T3 Mill B T4 Mill C T2 Mill D T6
Pipe 1 mm (in.) 356 (13.75) 349 (13.75) 343 (13.5) 316 (12.50)
Pipe 2 mm (in.) 343 (13.5) 337 (13.25) OPEN 325 (12.75)
Pipe 3 mm (in.) OPEN OPEN 349 (13.75) OPEN

Secondary Airflow Testing

Storms secondary air venturis were installed in each secondary air duct. Storm manually traversed
each duct and compared the measured airflow to the indicated airflow. The deviation between
indicated and measured airflows was outside of Storms recommended deviation limits of 3% on
the A side venturi. Storm investigated the installed equation in the DCS and found that the
differential pressure value input value did not match the local transmitter display differential
pressure. Storm adjusted the differential pressure input to match the transmitter value as well as
installing new K-factors, shown in Table 3. The results of the As Found and As Left are shown
in the tables below.

Table 2: "As Found" Secondary Airflow Summary

SA - A SA - B
K-Factor "As Found" 1.451 1.451
Indicated Flow (tons/hr) 109.6 103.6
Measured Flow (tons/hr) 99.5 104.2
Deviation (%) 10.10 -0.6

Table 3: "As Left" Secondary Airflow Summary

SA - A SA - B
K-Factor "As Tested" 1.662 1.703
Indicated Flow (tons/hr) 99.8 97.8
Measured Flow (tons/hr) 97.1 98.4
Deviation (%) 2.8 -0.6

Tempering Airflow Testing

Storms tempering air venturis were installed in each tempering air duct. Storm manually traversed
each duct and compared the measured airflow to the indicated airflow. The deviation of indicated
vs. measured was outside of Storms recommended deviation limits of 3%. A similar situation
was found with the differential pressure measured locally in the field not matching the DCS
indicated differential. Therefore, Storm worked with the plant to adjust the differential pressure
input to match the transmitter value as well as installing new K-factors. The results of the As
Found and As Left are shown in the Table 4 and Table 5.

Storm Technologies, Inc.

Tractebel Energia Jorge Lacerda
Storm Service Report 16-127-01
Proprietary and Confidential Page |5
 Table 4: "As Found" Tempering Airflow Summary
TA - A TA - B
K-Factor "As Found" 14,512 14,512
Indicated Flow (tons/hr) 37.5 45.3
Measured Flow (tons/hr) 12.1 17.1
Deviation (%) 208.4 165.5

Table 5: "As Left" Tempering Airflow Summary

TA - A TA - B
K-Factor "As Tested" 3482.2 3810.8
Indicated Flow (tons/hr) 11.9 14.1
Measured Flow (tons/hr) 12.1 14.4
Deviation (%) -2.0 -2.7

Primary Airflow Testing

Storm designed primary air venturis were installed in each mills total primary air duct. While
onsite Storm manually traversed each duct and compared the measured airflow to the indicated
airflow at two mill load points. The deviation between the indicated airflow and the measured
airflow was outside of Storms recommended deviation limits of 3% on mills A, C, and D. Storm
investigated the installed equation in the DCS but found that there is no primary airflow control
logic installed in the DCS. Therefore, adjustments were made to each transmitters differential
pressure range to bring the deviation to within Storms recommended limits. The results of the As
Found and As Left are shown in the tables below.

Table 6: "As Found" Primary Airflow Summary

"As Found" Primary Air Results Mill A Mill B Mill C Mill D
Indicated Flow (tons/hr) 23.8 23.2 22.6 23.5
High Flow
Measured Flow (tons/hr) 26.8 23.9 27.1 25.3
Test
Deviation (%) -11.5 -2.9 -16.5 -7.0
Indicated Flow (tons/hr) 23.7 22.6 22.3 23.1
Mid Flow
Measured Flow (tons/hr) 26.5 23.1 25.5 25.1
Test
Deviation (%) -10.8 -2.2 -12.7 -8.0
Average Deviation (%) -11.2 -2.5 -14.6 -7.5

Table 7: "As Left" Primary Airflow Summary

"As Left" Primary Air Results Mill A Mill C Mill D
Indicated Flow (tons/hr) 24.7 23.3 24.1
High Flow
Measured Flow (tons/hr) 24.4 23.9 24.7
Test
Deviation (%) 0.9 -2.7 -2.7

Storm Technologies, Inc.

Tractebel Energia Jorge Lacerda
Storm Service Report 16-127-01
Proprietary and Confidential Page |6
Pulverizer Performance Testing

Storm personnel conducted dirty airflow and isokinetic coal sampling on each of the fuel lines.
During each test, the mill was placed in manual operation at approximately 60% feeder speed.
Storm personnel traversed each fuel line utilizing multiple ports to evaluate the dirty air velocities
and collect an isokinetic coal sample. Each coal sample was then dried and sieved by Tractebel
personnel. The following figures below show the measured dirty airflow and fuel balance
deviations for each test completed. Storm recommends 5% deviation for dirty airflow balance
and 10% in fuel balance deviation. Although the dirty air balance was not within these limits, the
deviations did improve compared to the testing completed in 2014. Deviations of -25% were
balanced closer to 10% deviation on Mill B.

Figure 3: Dirty Air Balance

Figure 4: Fuel Balance

Storm Technologies, Inc.

Tractebel Energia Jorge Lacerda
Storm Service Report 16-127-01
Proprietary and Confidential Page |7
Figure 4 illustrates the weighted fineness average for each test completed. It can be seen that the
fineness results between Test 1 through Test 3 on Mill A showed no change from the classifier
positions 50%, 0%, and 30% during the three tests respectfully. During the classifier adjustments
on Mill A, no amp change was seen in the DCS even with the classifier blades at the full closed
position (0). The classifiers on Mill A were left at 30% as indicated on the position indicator. Mill
C and Mill D classifier positions were set at 40% during all tests were as Mill B was at 30%.

Storm Technologies, Inc.

Tractebel Energia - Jorge Lacerda
Unit 4: Weighted Fineness Averages
100.0%
95.0%
90.0%
85.0%
% Passing

80.0%
75.0%
70.0%
65.0%
60.0%
55.0%
50.0%
45.0%
50 100 140 200
Recommended 99.9% 95.0% 88.0% 75.0%
Mill A T1 (Class. 50%) 99.3% 90.3% 78.8% 65.5%
Mill A T2 (Class. 0%) 99.2% 90.1% 78.4% 65.1%
Mill A T3 (Class. 30%) 99.3% 91.5% 80.3% 66.3%
Mill B T1 (Class. 30%) 99.4% 93.0% 82.9% 69.7%
Mill B T2 (Class. 30%) 99.4% 93.2% 83.4% 70.5%
Mill C T1 (Class. 40%) 99.5% 95.8% 88.4% 76.0%
Mill D T1 (Class. 40%) 99.8% 96.5% 87.8% 73.6%
Mill D T2 (Class. 40%) 99.8% 96.5% 87.8% 73.4%

Figure 5: Weighted Fineness Average

Boiler Testing

Following the mill and airflow testing, Storm personnel completed boiler testing at the furnace
exit utilizing two observation ports and a water cooled HVT probe. Each port was manually
traversed a total of fourteen feet while measuring flue gas constituents every two feet. Test 1 and
Test 2 were completed with the top three mills in service: A, D, and B. Test 3 was completed with
the bottom three mills in service: D, B, and C. Table 8 below shows the testing results from the
boiler testing completed on Unit 4.
Table 8: Boiler Testing Summary
Overall Side A Side B
O
Test Temp. (C) 2 (%) CO (ppm) NO x (ppm) O
Temp. (C) 2 (%) CO (ppm) NOx (ppm) Temp. (C) O2 (%) CO (ppm) NOx (ppm)
1 1,029 3.40 20 798 1,098 2.80 30 850 959 3.98 10 746
2 1,046 3.16 3 843 1,072 2.61 5 843 1,020 3.72 0 842
3 1,105 3.13 873 834 1,149 1.44 1,746 758 1,060 4.81 0 909

Storm Technologies, Inc.

Tractebel Energia Jorge Lacerda
Storm Service Report 16-127-01
Proprietary and Confidential Page |8
From the testing results, the following figures were generated to show the measure flue gas
constituents and their quantities at the approximate measured location.

Figure 6: Unit 4 HVT Test 1 and 2 - Temperature (F)

Figure 7: Unit 4 HVT Test 1 and 2 - Oxygen (%)

Figure 8: Unit 4 HVT Test 1 and 2 - CO (ppm)

Storm Technologies, Inc.

Tractebel Energia Jorge Lacerda
Storm Service Report 16-127-01
Proprietary and Confidential Page |9
 Figure 9: Unit 4 HVT Test 1 and 2 - NOx (ppm)

Storm personnel also collected flyash samples simultaneously at the air heater outlet test locations
during each boiler test. The samples collected were sieved through a 200 mesh sieve and then
analyzed by Tractebel personnel for LOI quantities. The results from the flyash sieve and LOI
analysis are shown below. Note that the sieve results for test 1 are believed to been influence by
debris collected from inside the test port. The initial test results from Storms 2014 trip have been
provided in Table 10 to show the improvement in LOI results from the modifications implemented
on the boiler.

Table 9: Flyash Sieve Results

Table 10: Flyash LOI Results

From the measurements take while onsite, an abbreviated boiler efficiency was calculated. The
results as well as the losses measured onsite are shown in Table 11.

Storm Technologies, Inc.

Tractebel Energia Jorge Lacerda
Storm Service Report 16-127-01
Proprietary and Confidential P a g e | 10
 Table 11: Unit 4 Approximate Boiler Efficiency
Engie - Jorge Larcenda - Unit 4
Heat Loss Components Units Value
Losses due to unburned carbon in total dry refuse % 1.97
Losses due to heat in dry flue gas % 4.25
Losses due to moisture in the "as-fired" fuel % 0.76
Losses due to moisture from burning hydrogen % 1.97
Losses due to moisture in air % 0.14
Losses due to air infiltration % 0.00
Unmeasured Losses to be assumed % 1.50
Heat Credit Components
Credits due to heat from entering air % 0.27
Boiler Efficiency % 89.68

Air In-Leakage Testing

Storm personnel conducted a leakage test utilizing the air heater inlet test ports and the air heater
outlet test ports in each duct. Storm measured the oxygen at the inlet and outlet of each air heater
to calculate an approximate oxygen rise between the two test locations. Note that the test locations
were not directly at the inlet and outlet of each air heater so possible leakage measured could be
influenced by duct leakage between the two test locations. Table 12 shows the average measured
oxygen at each test location along with the calculated leakage between the two test locations. Table
12 shows the measured oxygen at each testing location.

Table 12: Air Heater Leakage Results

Air Heater Leakage
Air Heater O2 Entering/Exiting Leakage Corrected Gas Temperature (C) X-ratio Efficiency (%)
2.78
A 21.47 163.24 0.71 59.47
6.27
4.19
B 19.95 169.58 0.72 57.84
7.22

Storm Technologies, Inc.

Tractebel Energia Jorge Lacerda
Storm Service Report 16-127-01
Proprietary and Confidential P a g e | 11
 Figure 10: Measured Oxygen Rise

Recommendations
Storm was originally contacted by the plant in 2013 to review current operations and help provide
a plan of action to improve the heat rate and efficiency of the boilers. Based off of Storms initial
visit and baseline testing Storms approach to improving performance at the plant was presented
to the plant and has been working towards installing Storm designed engineered solutions to work
towards improved heat rate and efficiency. With that said, the main focus has been on the inputs
to the boiler for the combustion process. Pulverizer performance and accurately controlled airflow
is essential to good combustion. Achieving good fuel distribution to each burner is the ultimate
goal of the pulverizer improvements. From the LOI sampling, a significant reduction in LOI was
measured between Storms initial and most recent visit. The weighted LOI was reduced from
15.0% during the 2014 visit to 4.85% LOI. This reduction of LOI results in an approximate 4.4%
increase in boiler efficiency. Storm and Tractebel personnel have been working towards that with
the fabrication of new primary air flow venturis, fuel line orifice housings and a new classifier
system in the mills. The testing after the classifier systems were installed showed significant
improvement in the fuel fineness levels on mills B, C, and D, which met Storms initial
performance guarantees. However, the dirty air and fuel balance were still outside of Storms
recommended levels. With that said, the recommendations in the following pages are provided
based off of the test results gathered while onsite coupled with Storms past experiences with
similar mill and boiler designs that will further help maintain and improve upon the current heat
rate.

Ensuring that there is equal distribution of primary air flow to each of the fuel lines is the first step
to achieving even fuel distribution to each of the three burners on a given pulverizer. Although
each mill was clean air balanced to 2% pipe-to-pipe deviation, the dirty air and fuel balance were
still outside of Storms recommended limits on three of the four mills. During the initial site visit
and following previous testing on Unit 3 and 4s pulverizers it was found that there appeared to be
significant roping prior to and in the splitter box forcing more coal into certain fuel lines. Upon
the initial review of the test data it was suggested and expected that some sort of rope breaker
would be needed to optimize fuel distribution. While onsite during this visit it was discovered that
Storm Technologies, Inc.
Tractebel Energia Jorge Lacerda
Storm Service Report 16-127-01
Proprietary and Confidential P a g e | 12
the proposed splatter plates were not installed on any of the mills. The splatter plate, as shown
below, should be installed at the exit of the bend in the main fuel supply line exiting each mill.

Figure 11: Storm Technologies' Splatter Plate

The purpose of the splatter plate is break the rope that occurs downstream of the 90 bend in the
fuel line. The splatter plate is designed as a D shaped orifice with the blocked chord opposite of
the elbow to break the coal rope which hugs the fuel line ID. This will help disperse the fuel
providing a more homogenous mixture before the fuel is separated into each individual fuel line.
With coal being roughly 1,000 time denser than air the fuel will rope at the elbow and has little
time to recover before entering the splitter. This has been seen in virtually all testing that has been
completed in the past at the plant. The venturi upstream of the splitter was removed when the
orifice housings were installed. The venturi was used to assist with distribution however a splatter
plate at the elbow along with clean air balancing will assist with fuel distribution to the burners.
The result of breaking up the coal rope along with the clean air being balanced should help improve
the dirty air and fuel balance which will greatly improve combustion. Balancing of the airflow in
the fuel pipes can also allow for adjustments to optimum air/fuel ratios while ensuring none of the
fuel pipe velocities drop below 16.8 meter per second.

Even though the fineness levels had greatly improved, most of the mills met the 70% that was
originally targeted. However, Mill C was able to achieve 75% passing 200 mesh. Therefore, it is
recommended that additional tuning be completed to try and achieve fineness levels of 99.9%
passing 50 mesh and 75% passing 200 mesh on each of the mills. During testing on Mill A, the
classifier blades were stroked to the fully close position. Typically this will cause a significant
increase in mill amps as well as have an effect on mill outlet temperatures. Neither was seen when
adjusting the classifier position. After inspection of the pulverizer, the bottom of the classifier cone
was off center at its installed location causing one inch gap between the top and bottom of the
classifier cone as shown

Figure 12: Inner Classifier Cone

Storm Technologies, Inc.

Tractebel Energia Jorge Lacerda
Storm Service Report 16-127-01
Proprietary and Confidential P a g e | 13
This can cause air and fuel flow to bypass the classifier blades and exit the mill through this gap
as well as disrupting the centrifugal spin inside the classifier cone. Storm recommends the bottom
half of the classifier cone be readjusted and sealed to remove the gaps. The reject doors should
also be inspected routinely to assure they open freely and close fully. If the doors stick open or
closed mill performance suffers. Furthermore, upon inspection it was noticed that the springs were
at different heights during the internal inspection. Although the mill was not fully clean, Storm
recommends that each spring setting be checked and set to the same setting for each mill and be
set to a loading of 1,600 lbf or 7,117 N.

The current classifier position gauge is in percentage. Storm recommends that the distance between
the classifier blades tip and the next blade be measured and recorded at each 10% intervals during
the next outage. This will allow for better tuning of the classifier as the internal position of the
classifier will be known. During internal inspections, Storm technologies noticed the close
proximity of the coarse particle return line to the springs due to the large raw coal feed pipe
diameter. Storm learned that the coal feed pipe diameter was increased to reduce coal pluggage on
the inlet of the mill before Storm was involved onsite. This however resulted in a larger diameter
at the reject doors in order to provide adequate area for the coarse particles to be freely recirculated.
This causes the reject of the coal to be located in close proximity of the spring. In Storms
experience, the ratio of coal recirculation to coal output in the mill is approximately 4:1 to achieve
the Storms recommended fineness levels. This means four times as much coal will be recirculating
through the reject system and released near the spring. Storm recommends that the springs be
check for wear during each outage. The large coal feed pipe diameter could also cause the rejected
coal to be released into the airflow vectors towards the classifier. This would restrict the coal from
reaching the grinding surface for regrind. A long term recommendation would be to reduce the
coal feeder pipe diameter to increase the recirculation area increasing recirculation capacity and
ensuring smooth coarse particle return to the grinding surface. This would also require review of
the existing reject system to assure proper sizing.

From the leakage test, an approximate ~20% leakage was measured on both air heaters. The
measured leakage from the furnace exit to the air heater inlet was approximately ~5%. Based off
design a Ljungstrom air heaters can typically achieve 5-7% leakage. Storm has found that with
proper inspection and maintenance that these air heaters can easily achieve less than 10% air in
leakage between outages. Storm recommends an inspection of the air heater as well as the duct
and expansion joints to inspect for areas of leakage. By reducing the air-in leakage through the air
heater, a reduction of dry gas loss as well as an increase in ID fan and FD fan capacity can be
obtained.
While Storm was onsite, burner C1 air register was connected in the opposite direction. Indication
of 100% open was actually 0% fully closed position. Storm recommends that during the next
outage all registers be fully stroked with the correct indication of the damper arm. For optimum
combustion, it is essential to ensure good mixing of the fuel and air. Storm recommends that the
burner modifications that were installed on Mill C as be implemented on each of the other burner
rows. With the spinner spreaders installed, this should increase the mixing of the fuel and air at
the burner front further improving combustion. Furthermore, with all the burners with the same
design and essentially the same dimensions, the balance of secondary air flow in the windbox
should improve. While onsite the new ash hopper system was not operational. From past
experience, the ash carryover into the windbox restricting the bottom burners from necessary

Storm Technologies, Inc.

Tractebel Energia Jorge Lacerda
Storm Service Report 16-127-01
Proprietary and Confidential P a g e | 14
secondary airflow for combustion. Storm recommends that this system be repaired to remove any
ash build up from restricting the bottom row of burners.

Closing

It has been a sincere pleasure working with you and the Tractebel team and we look forward to the
continued relationship. If you have any questions concerning the information presented in this
report or any future testing please feel free to contact us.

Respectfully,

Matthew Rutherford
Project Engineer
Storm Technologies, Inc.

Storm Technologies, Inc.

Tractebel Energia Jorge Lacerda
Storm Service Report 16-127-01
Proprietary and Confidential P a g e | 15