Improving Gas Turbine – HRSG output using Inlet Air Chilling and Converted Evaporator

International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056
Volume: 04 Issue: 11 | Nov -2017 www.irjet.net p-ISSN: 2395-0072
© 2017, IRJET | Impact Factor value: 6.171 | ISO 9001:2008 Certified Journal | Page 428
Improving Gas Turbine – HRSG output using Inlet Air Chilling and
Converted Evaporator
Utkarsh Adhyaru
---------------------------------------------------------------------***---------------------------------------------------------------------
Abstract - Every machine ever created has some output
limitations with respect to the inputsgiventothem. Inorder to
change the output rate some design and condition
modifications have to be done in order to get effectiveness
beyond the designed parameters and at thesametimealsonot
hamper the metallurgical and operational stability of the
machine. In order to get more out of a GT-HRSG, sincetheyare
temperature controlled machines, some design modifications
proved a considerable increaseinoutputkeepingthe efficiency
at bay.
1. INTRODUCTION
With a continuous improvement and expansion in
process plants going on, there is a vital need of utilities to
have more efficiency to put in more effort to give the intend
output and keep production at par. Gas turbineswork on the
principle of hot gas expanding. They have a fixed loading
capacity. In order to increase the loading, the ambient
temperature plays an important role as it eventually
determines the output temperature. In case of peak load, if
the exhaust temperature reaches the critical value, the
machine will not allow any more power loading and the
machine will go into temperature control.
Another problem related to over temperature is related to
HRSG. In order to increase the steam loading of HRSG there
is a provision of supplementary firing. Incaseofelevated gas
turbine exhausted temperature, added with supplementary
firing the metal temperature of superheater-II tends to
increase as the supplementary firing is followed by the
superheater-II. The idea of improvingtheconditionsofthese
machines at peak load was to improve the reference
temperatures of turbine and at the same time improve the
metallurgy in order to keep the HRSG healthy.
The modifications thus done was done in a GE Frame 6B Gas
Turbine with a power capacity of 34.5MW. Along with this
another design change was done in the 120 MT HRSG which
followed.
2. INLET AIR CHILLING
The Frame 6B has a capacity of goinguptoa loadingcapacity
of around 34 MW which was observed on cold winter days
where the ambient temperature was less and the reference
exhaust temperature (TTXM) allowed the machine to go as
much as 34-34.5 MW on preselect mode, before it can go to
temperature control. On summer days this can come down
to as low as 29 MW and the machine is almost on the verge
of accepting no more MW input.
The turbines used in this case had very high demanding
complex loads and importing power was proving very
expensive. To avoid this the idea of Inlet Air Chilling (IAC)
was tried to see if this can prove out to be beneficial. The
idea was to cool the incoming air so as to get more window
for MW input and run the machine on maximum loads.
To put the idea in use, a new set of filter module was
installed. It had air inlet from three sides. The fourth side
was attached to a bellmouth which then ended in the inlet
plenum. The filter module had chilled water tubes running
throughout. The filters were placed in an order so has to
have the best possible air in inlet. It had one mesh to avoid
any big particle enter. Then it had a primary filter followed
by chilled water tubes, and then a denser secondary filter.
A 1000 TR vacuum absorption cycle based chiller was
commissioned and it could sufficeforanyclimateconditions.
Fig-1: Solid model of Inlet air bellmouth
This filter assembly not only providedanoptiontohaveinlet
air chilling but also provide uniform air flow throughout the
three filter modules. The CFD analysis iterations provided a
final design to achieve optimum airflow with minimum
backpressure.
The following figures clearly shows uniform entry of
streamlines to the aero-acoustic diffusers at 3D modeling
and top view of CFD results respectively, which reduce the
vortex and turbulence with in the Gas Turbine inlet duct.
This would also reduce flow induced vibration in the duct.

Fig-2: Velocity drop
Fig-3: Velocity distribution
The analysis showed total pressure drop values,whichwere
quite low compared to previous model. In the proposed
model the pressure drop was expected to be around 1.97”
WC From rain hood to the Gas Turbine inlet.
There also was significant reductionintheductflowinduced
vibration due to effective diffusion of stratified high velocity
air entering compressor inlet plenum. It was confirmed by
having a condition monitoring team check the vibrations.
Flow Rate 130 Kg/S @ 20oC
Pre filter 0.3”WC
Water coil 0.50”WC
Final filter 0.35”WC
Pressure drop:
Ambient – Filters 1.22”WC
Filters – Inlet plenum 0.75”WC
Total pressure drop 1.97”WC
2.1 Importance of compressor bellmouth
differential pressure.
The static pressure difference between the inside of the
bellmouth and static ambient pressure will give an
indication of mass flow as total pressure will remain the
same, the static drop converts into dynamic component
(Velocity) along with flow area will give mass flow of less
dense inlet air.
Mass flow = Speed * Density * Area
Pitot tube is significant in determining the inlet air flow
which was vital to see the changeinflowcharacteristicsafter
inlet air chilling. When the Pitot tube is connected to a
manometer, the resultant displacementoffluidismeasureof
the “Velocity Pressure”.
Velocity Pressure= P1-P2 = ½*ρ*v2,
ρ=density of air
V = sqrt 2(P1-P2) / ρ
But P1-P2 = ρw * g * H,
ρw = Density of water in manometer is 1000kg/m3
H= Height of water column in meters
g=9.81 m/s2
V = sqrt 2* ρw * H * g / ρ
= sqrt 2* 1000 * H * 9.81 / ρ
H =height of liquid column to be expressedinmmofwater hv
H = hv / 1000
V = sqrt 2* 1000 *hv /1000*9.81/ρ m/sec
So, Velocity air flow = 4.43 sqrt hv/ρ m/sec
In practice the density of air is affected by temperature and
by static pressure and so it is necessary to know the air
density at some reference temperature and pressure, and
then by applying corrections the value of test conditions are
obtained. The density of air at 1.013 bar 0deg is 1.29 kg/m3
and it varies inversely to the absolute temperature and
directly to absolute pressure.
The density ρfor any temperature and pressure condition is
thus given by
ρ = 1.29*[273/T]*(10363+hs/10363)*B/760
Where, 1.29 is density of air at 0degC and at 10.13 bar in
kg/m3
T is air temperature, K
hs is static pressure in duct in mmWC
B is barometric pressure in mmHg
10363 is standard pressure in mmWC

So using this the condition monitoring was done with and
without chilling and mass flow was thus calculated.
Gas turbine inlet mass flow without chilling:
Temperature of air inside the duct = 34 degC
Static pressure of air as per Pitot tube measurement = 18.2
(-) mmWC
Barometric pressure = 760 mmHg
Velocity at Elbow exit (from velocitydistributionasperCFD)
= 22.5 m/Sec
Cross sectional area of duct at Elbow = 4.61 m2
Volume flow = 103.7 m3/sec
Air density = 1.1451 kg/m3 at duct conditions
Thus, Mass flow = 118.74 Kg/Sec
Gas turbine inlet mass flow with chilling:
Temperature of air inside the duct = 21 degC
Static pressure of air as per Pitot tube measurement = 32.92
(-) mmWC
Barometric pressure = 760 mmHg
Velocity at Elbow exit (from velocitydistributionasperCFD)
= 22.5 m/Sec
Cross sectional area of duct at Elbow = 4.61 m2
Volume flow = 103.7 m3/sec
Air density = 1.1941 kg/m3 at duct conditions
Thus, Mass flow = 123.82 Kg/Sec
The above calculations were done against the base load
design mass flow of 135 Kg/Sec. The calculations clearly
indicate a higher density of air flowing after using the inlet
air chilling. In case of any sweating, a condensate collection
pan is provided leaving air side of each dehumidifyingpan.It
also had a condensate trap of at least 2.5 times the expected
negative static pressure of the unit.
Fig-4: Coil sweating recovery U-tube
Expected negative pressure is
0.1+.05+.09=1.5” WC
d= 1.5” WC * 2.5 = 3.75”
Trap used was of 4” (100mm).
The following graph shows how the gas turbine behaves
without any inlet air chilling in normal conditions.
Table-1: Gas turbine data without chilling
GT DATA WITHOUT INLET AIR CHILLING
Parameter Unit 1200 1300 1400 1600
Speed RPM 5020 5062 5057 5030
Load MW 26.5 26.8 26.2 25.5
GT speed % 98.55 99.37 99.29 98.75
Flow divider speed % 68.42 68.59 67.7 66.2
FSR % 69.16 69.04 68.38 67.54
Liq fuel flow kg/sec 2.6 2.67 2.65 2.59
Mass fuel flow TPH 8.13 8.19 8.03 8.02
Compressor inlet temp. degC 30 32 34 33
Inlet air filter D.P INWC 1.37 1.39 1.97 1.3
Compressor outlet pressure Bar 8.7 8.79 8.67 8.52
I.G.V. Angle DGA 83.9 83.9 83.9 83.8
Exhaust Temperature degC 573 574 575 576
Spread Limit degC 69.4 69.4 69.4 69.4
HRSG flue gas inlet pressure MMWC 274 276.7 273 262
Compressor discharge temperature degC 366 371 373 372
Heat rate (open cycle) 3169 3169 3169
Steam/fuel ratio 19.67 20.01 18.23
Time
The following graph shows how the gas turbine behaves
with inlet air chilling in normal conditions.
Table-2: Gas turbine data with chilling
GT DATA WITH INLET AIR CHILLING
Parameter Unit 1200 1300 1400 1600
Speed RPM 5018 5047 5030 5010
Load MW 26 26 26 26
GT speed % 98.5 99.09 98.75 98.35
Flow divider speed % 65.12 65.53 65.2 65.53
FSR % 66.45 65.79 66.6 67.11
Liq fuel flow kg/sec 2.56 2.56 2.57 2.57
Mass fuel flow TPH 7.93 7.89 7.9 7.91
Compressor inlet temp. degC 19 20 19 19
Inlet air filter D.P INWC 1.52 1.54 1.51 1.47
Compressor outlet pressure Bar 8.97 9.04 9 8.93
I.G.V. Angle DGA 83.7 83.7 83.7 83.7
Exhaust Temperature degC 526 522 524 526
Spread Limit degC 172.7 176.6 176.7 177
HRSG flue gas inlet pressure MMWC 297 298.6 299.6 296
Compressor discharge temperaturedegC 351 353 351 350
Heat rate (open cycle) 3169 3169 3169
Steam/fuel ratio 17.7 19.18 18.22
Time
Based on the observations taken at different time of the day,
it was observed that the inlet air chilling does a considerable
amount of change in the machine. InitiallyitreducestheFSR,

which in turns also improves the fuel going inside the GT.
According to the observations both noted with the flow
meter and mass flow with micro motion,it wasobservedthat
less fuel goes inside the machine thus reducing the flow
divider speeds. With low inlet temperature the compressor
outlet pressure is high with low exhaust temperatures, low
compressor discharge temperatures and high HRSG flue gas
pressures. Thus allowing more temperature window to
increase the capacity of the machine without going into
temperature control.
The only downside observed with IAC is the increase in
suction air filter D.P. which has to be monitored during
operations. Also at the end of HRSG NOx and Sox have to be
monitored to avoid self-condensing that might result in
damaging of HRSG tubes.
On longer runs and more report gathering it was observed
that the heat rate of the turbine also improved. Also
additionally the steam generation of HRSGimproved.Alsoat
the end of HRSG the temperature was high so an additional
LP evaporator was installed which couldgivea steamoutput
of 6barg which was used to run the 1000 TR chiller so the
entire energy was conserved quite efficiently.
Table-3: Data comparison
WithChilling
Day 1
Time
Fuel
Totalizer
Power
totalizer
Totalpower
generater
Fuel
consume
Open
cycle MW
12:15 131987 402703 80 26 3013 29
15:00 132013 402783 120 39 3013 30
19:00 132052 402903 98 32 3027 28
22:00 132084 403001 46 14 2821 31
0:00 132098 403047 197 58 3004 30
6:00 132156 403226 523 169 2995 30
Day 2
132372 403878 227 74 3022 30
Day 3
132467 404171 293 95 3006 29
Withoutchilling
Day 4 685.6 223.8 3026 29
Day 5 682.6 218.9 2973 28
Day 6 670.5 220.8 3053 28
Total 2038.7 663.5 3017 29
3. CONVERTED EVAPORATOR
With higher gas turbine loads and more output required out
of HRSG, supplementary firing is required tocompensatefor
minor heat loss and add up to the heat input to further
increase the steam flow and also compensated for any
process abnormalities to keep the header pressureconstant.
With all the modifications thus made, there was a constant
problem of tube failures in HRSG at certain point of time. It
was observed that the super heater-II which was most
exposed to the supplementary heat, was exposed to the
maximum damage and failed eventually. To avoid this some
modification was done in the evaporator.
The HRSG used in line with the mentioned gas turbine had a
design capacity of 100 TPH capacity of steam generation at
100 barg 495oC. It consist of twoeconomizers,an evaporator
module and three super heater module. TheentireHRSG ran
on gas turbines outlet flue gas. To compensate or any minor
heat loss or process steam requirement change, it had six
burners in between super heater II and III. At peak plant
loads all the burners were used with full GT load to give
maximum steam output. The problem faced had a lot of
effect on machine life and also demanded big shutdowns
creating a problem for process requirements.
To avoid this problem, the initial two modules of
superheater II were modified. They were cut and replaced
with the same tubes as used in evaporator. It also had coper
fins attached to it for maximum heat transferandabsorption
in case of supplementary firing and the tubes ran water
unlike other superheater tubes which ran steam. This
created a jacket for the superheater II section and absorbed
the conductive heat coming from the burner flame so as to
not damage the superheater tubes.
Fig-5: Converted Evaporator in a Heat recovery steam
generator
Before the change, at peak loads the temperature of
superheater II used to go over 515 degC. After the project,

the temperature remained below 478 degC. Another benefit
was that the flow output of HRSG improved ant increased to
125 TPH from 105 TPH. Additionally there was quite a bit of
hunting seen in HRSG when put on flow control which also
improved when kept in three element control.
The gas turbine used for the particular HRSG is a black start
machine and so it’s total time to reach FSNL is about 11
minutes. So the heating of HRSG can be done parallel to the
startup or after it reached certain temperature. Both the
cases referred to as cold start and hot start. The new tube
bundle integrity was to be checked with both startups so as
to confirm if it can withstand thermal shocks in case of hot
startups.
It was observed that a temperature difference persisted
between old evaporator and the converted evaporator. But
it did not affect the metal temperature of the main
superheater II tubes and served the purpose of jacketing the
tube module from flame heat.
Chart-6: Fatigue effect on modification
The study showed that the maximum fatigue effect
transpired in the piping at the bottom header of the
evaporator, in specific, the feeders that linked the manifold
to the lower headers. It was also observed that the tube
bundle metal temperature was on the higher side in the
transient condition during critical time period of hotstartup
phase compared to cold startup and operation.
The fatigue analysis results approve that the system is
adequate to support the thermal growth without yielding
excessive fatigue stresses.
Chart-7: Effect on tube life
As the graph suggests there is a higher life consumptionrate
in the life of tubes, especially that of the converted
evaporator as the temperature in that zone is highest for
water tubes of 135 barg pressure and so it was to beavoided
to take the HRSG in line with hot start procedure and was to
be done only in extreme cases as required.
4. CONCLUSION
The entire project was to developanincreaseinproductivity
in case of a gas turbine – HRSG closed cycle operation and
not only did it help to increase the throughput of the
machines without affecting the reliability but also made
failure intensity less.
Table-4: Benefits after modification
Gas turbine IAC benefit
Description UOM
Present With IAC
GT output at ISO MW 37.52 37.52
O/C heat rate at ISO kcal/kwh 2764 2764
Ambient temperature degC 35 20
Frequency Hz 50 50
Inlet pressure drop mmWC 100 100
Outlet pressure drop mmWC 225 225
Correction factor for degradation/aging % 5 5
Correction Factors:
Temperature CF 1.0350 1.0080
Frequency CF 1.0050 1.0050
Inlet pressure drop CF 1.0000 1.0000
Outlet pressure drop CF 1.0063 1.0063
Part load CF 1.0470 1.0470
Degradation CF 1.0500 1.0500
GT heat rate for conditions kcal/kwh 3166 3083
Values

The overall heat rate reduction was of about 83 kcal/kwh. It
also improved the loading capacity of GT in hot conditions
and avoided the machine to go into temperature control.
The bellmouth design proved a huge reduction in pressure
drops and uniform velocity throughout the filters.
Fig-8: Pressure loss
Inlet plenum loss was calculated at two points. One in the
filter house and another before the compressor inlet. Atfirst
point, a loss of 85-95 mmWC was observed and at point two,
a loss of 210 mmWC was observed. According to the gas
turbine vendor, a guaranteed inlet plenum loss of 101.6
mmWC was constant.
As per the performance curves, output loss for every 4 inch
of H2O is 1.55% and results in a heat rate elevation of
0.5%.According to it, the GTG had an output loss of above
1.65% leading to a heat rate increase of 0.53%.
Post changing of bellmouth, a reduction of 4.9 inch of H2Oin
pressure drop was observed along with a power gain of2.08
%. This led to an increase of 8000 operation hours per year
and a power gain of 4664.8 MWhr. The payback time was a
little under 5 months.
With a reduction in heat rate at 31kcal/kwh, the machine
saw a fuel savings of 38kg/hr. with an additional steam
generation benefit of 1.5 TPH.
Additional benefits from uniform flow lead to uniform
loading on GT compressor blades and a significantreduction
on flow induced vibrations. Also a uniform and laminar flow
distribution resulted in better heat transfer coefficient
resulting in better supplementary firing efficiency.
The converted evaporator completelyeradicated theneed of
frequent shutdowns required for HRSG at peak loads and
also increased the output rate of HRSG.Alsothecombination
of tube diameter, thickness of the wall of HRSG and the
thermal growth differences was adequate for hot startups.
The fatigue damage will increase in the tubes due to sudden
thermal shock but this will not have much effect on the
planned life of the finned evaporatortubes.Therewasa 22%
increase in the steam loading capacity of the HRSG.
5. Challenges faced
The major challenge faced with the new bellmouth and
filters is the automatic puffing system to auto clean the
filters became redundant. Even through the life of the new
filters is long enough, they have to be replaced after they
retire. The entire changing need the shutdown of the GT and
the vacuum refrigeration cycle machine needs excess steam
to run. There is also seen a coating of dust and oil
accumulated in the chiller coil of the IAC filters and has to be
cleaned to maintain the heat transfer. Running the chiller
creates a little spike on the differential pressure of thefilters
and has to be monitored carefully with proper alarms. The
filter also was additionally provided with blast windows
which open automatically when the differential pressure
reaches beyond a specific limit and was after the filters to
make sure that the compressor never suffocates for inletair.
The HRSG also needed a complete changing of the tube
structure in superheater II module. The control valves were
tuned to feed additional water to compensate for the
increase in evaporator tube bundle. Overall the entire
project proved as a benefit to the efficiency and output and
the limitations are feasible for the resultant output of the
project.
References
[1]. Claire Soares, Gas TurbinesA Handbook ofAir,Land
and Sea Applications (Elsevier, 2008).
[2]. Vernon L. Eriksen, Heat Recovery Steam Generator
Technology (Elsvier, March 2017).
[3]. V. Ganpathy, Industrial Boilers and Heat Recovery
Steam Generators: Design, Application and
Calculations (CRC, 2002).
[4]. Venkateswarlu K.S., WaterChemistryIndustrial and
Power Station Water Treatment, (New Age
International Publishers).
[5]. Prabir Basu, Boilers and Burners (Springer 1999).
[6]. Patil R.D., (1992), Water Management Practices in
Fertilizer Industry, Ibid. G.I – G.20, November.
[7]. Frederick M. Steingress, Harold J. Frost, Daryl R.
Walker, High PressureBoilers (AmericanTechinical
Publishers, 2012)
[8]. Gas Turbine Note No. 7-Performance curve
No.499HA542, GTF6SP, GTF6SH Kadambi, V. and

Manohar Prasad, An Introduction to Energy
Conversion Vol. III: Turbomachinery (Wiley
Eastern, 1997)
BIOGRAPHY
Utkarsh Adhyaru is a Mechanical
Engineer and has been working in
a captive power plant as a
production and maintenance
engineer. He has also been part of
erection and commissioning of an
850 MW CPP and worked
extensively in changing designs of
boilers and burners. He also looks
for the utilities of CPP and their
process optimization.

Improving Gas Turbine – HRSG output using Inlet Air Chilling and Converted Evaporator

More Related Content

What's hot (20)

Similar to Improving Gas Turbine – HRSG output using Inlet Air Chilling and Converted Evaporator (20)

More from IRJET Journal (20)

Recently uploaded (20)

Improving Gas Turbine – HRSG output using Inlet Air Chilling and Converted Evaporator