Applied Thermal Engineering: Ik Hwan Kwon, Do Won Kang, Tong Seop Kim

You might also like

Download as pdf or txt
Download as pdf or txt
You are on page 1of 10

Applied Thermal Engineering 60 (2013) 285e294

Contents lists available at SciVerse ScienceDirect

Applied Thermal Engineering


journal homepage: www.elsevier.com/locate/apthermeng

Using coolant modulation and pre-cooling to avoid turbine blade


overheating in a gas turbine combined cycle power plant fired
with low calorific value gas
Ik Hwan Kwon a, Do Won Kang a, Tong Seop Kim b, *
a
Graduate School, Inha University, Inha-ro 100, Nam-gu, Incheon 402-751, Republic of Korea
b
Department of Mechanical Engineering, Inha University, Inha-ro 100, Nam-gu, Incheon 402-751, Republic of Korea

h i g h l i g h t s

 The blade overheating problem in firing low calorific value gas in gas turbine was examined.
 Several measures to suppress blade overheating were compared.
 Coolant modulation was shown to result in a much lower power penalty than under-firing.
 Pre-cooling of the coolant reduces the power penalty further by reducing the coolant supply.

a r t i c l e i n f o a b s t r a c t

Article history: Overheating of turbine blades is one of the major concerns in using low calorific value fuels in gas
Received 2 April 2013 turbines. In this work, we examined the deviation of operating conditions of a gas turbine fired with a
Accepted 8 July 2013 low calorific value gas fuel, with a focus on the turbine blade temperatures. Several measures to suppress
Available online 16 July 2013
blade overheating were compared in terms of the power output and efficiency of the gas turbine
combined cycle plant. Blade overheating can be prevented by decreasing the firing temperature without
Keywords:
the need for hardware modifications, but the accompanying power reduction is considerable. As a
Gas turbine
remedy to this large reduction in power, modulation of the coolant supply to each blade row was
Combined cycle
Low calorific gas
simulated, and a much lower power penalty was observed. Moreover, pre-cooling of the coolant en-
Turbine blade overheating hances the power output further by reducing the coolant supply. Pre-cooling recovers 80% of the
Under-firing available maximum augmentation of the combined cycle by simply switching the fuel from natural gas to
Coolant modulation low calorific value gas. Pre-cooling also provides higher overall combined cycle efficiency compared to
Pre-cooling under-firing.
Ó 2013 Elsevier Ltd. All rights reserved.

1. Introduction past decade including performance summaries and modeling of


existing plants [1,2], and examinations of the effects of major
Fuel diversity is the major advantage of gas turbines over other design parameters such as the integration between a gas turbine
types of power generators. In addition to natural gases, a wide and auxiliary components [3,4]. Design limitations regarding the
range of low calorific gaseous fuels (such as synthetic gases from operating conditions of major components, especially turbine and
coal and biomass gasification, and various kinds of biogas) can be compressor, have been studied [5,6]. In addition, the influence of
used in gas turbines. The integrated gasification combined cycle syngas composition on the performance and operability of gas
(IGCC) is considered to be the most environmentally friendly turbines has been examined [7]. Attempts to use biomass as a fuel
method of using coal. Several full-size plants are under operation in gas turbines and combined cycle power plants have also been
and a number of projects are ongoing worldwide. Various perfor- initiated recently. Various basic studies on the use of biomass in gas
mance analyses and comparisons have been undertaken during the turbine-based power plants have been published, such as per-
spectives on the use of biomass in combined cycle plants [8],
different strategies for using biomass [9], and the influence of firing
* Corresponding author. Tel.: þ82 32 860 7307; fax: þ82 32 868 1716. biomass on gas turbine components [10]. Also, the possibility of co-
E-mail address: kts@inha.ac.kr (T.S. Kim). firing biomass with natural gas has been investigated [11,12]. The

1359-4311/$ e see front matter Ó 2013 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.applthermaleng.2013.07.008
286 I.H. Kwon et al. / Applied Thermal Engineering 60 (2013) 285e294

Nomenclature T temperature (K)


U blade speed (m/s)
A area (m2) V relative velocity (m/s)
C cooling constant, absolute velocity (m/s) W_ power (MW)
CC combined cycle a absolute flow angle
CDP compressor discharge pressure b relative flow angle
CDT compressor discharge temperature f cooling effectiveness
cp specific heat (kJ/kg K) g specific heat ratio
ECO economizer h efficiency
EVA evaporator k constant
GT gas turbine
HP high pressure Subscripts
IGCC integrated gasification combined cycle 1 nozzle inlet
IP intermediate pressure 2 nozzle outlet, rotor inlet
LCG low calorific value gas 3 rotor outlet
LHV lower heating value (kJ/kg) a axial component
LP low pressure b turbine blade
m_ mass flow rate (kg/s) c coolant
NG natural gas d design point
P pressure (kPa) g gas
PR pressure ratio in inlet
R gas constant (kJ/kg K) rel relative total property
SH superheater N asymptotic
ST steam turbine

use of biogas (digester and landfill gas) in relatively small gas coolant can be pre-cooled by water or steam from the bottoming
turbine-based combined heat and power systems has also been cycle before it is supplied to the turbine. Pre-cooling would mini-
studied [13,14]. mize the coolant supply because with a lower the coolant tem-
Even though low calorific value gas fuels are good resources for perature, less coolant is needed. We demonstrated the relative
use in gas turbine-based power plants, there are concerns advantage of modulating coolant flow rates and pre-cooling the
regarding the effect of these fuels on the operability and lifetimes of coolant compared to simple under-firing. A full gas turbine com-
gas turbine components. Overheating of hot sections such as tur- bined cycle plant was modeled. Changes in plant performance and
bine blades is one of the major concerns, as pointed out in recent the operating condition (especially the turbine blade temperatures)
publications [5,6,10,15]. The common understanding is as follows. when using a low calorific value gas were analyzed. In addition, the
The calorific values of syngases are much lower than that of natural effects of different methods to restore the blade temperatures to
gas, which gas turbines are designed for. Therefore, when the fuel is the reference values were simulated and compared.
switched from natural gas to a low calorific value gas, more fuel In the detailed turbine design and analysis stage, blade tem-
must be supplied to the combustor, resulting in greater mass flow perature distribution and lifetime analysis using numerical
in the turbine. The larger turbine mass flow results in a rise in the methods, especially the conjugate heat transfer analysis [16e19],
compressor pressure ratio if there is no hardware modification in can be used to optimize the cooling system accounting for the
the turbine. This causes an increase in the cooling air temperature, operating condition change. However, performing a numerical
which increases the blade metal temperature over the design analysis is beyond the scope of this study, and thus we have focused
temperature in coal and biomass integrated gasification power only on thermodynamic system level analysis in this paper. The
plants [6,10]. It has been reported that firing biogas might cause a result of this study may provide useful basic data for the detailed
similar overheating problem of hot sections [14]. Thus, hot section numerical analysis.
overheating is a common phenomenon when firing low calorific
value gas in a gas turbine. Therefore, fuel switching should be 2. System modeling
accompanied by proper measures to suppress the overheating.
We investigated the degree of turbine blade overheating in a 2.1. Gas turbine
modern state-of-the-art gas turbine for combined cycle power
plants, and comparatively analyzed several measures to suppress Fig. 1 shows the gas turbine combined cycle system considered
overheating. Decreasing the firing temperature would be the in this study. The performance of the entire system was simulated
simplest way to avoid turbine overheating [6,7,15], but this reduces using GateCycle [20]. A state-of-the-art F-class gas turbine that is
the performance of the gas turbine noticeably, especially with widely used for combined cycle plants was adopted. Design spec-
respect to power output [6,7]. Alternative methods are associated ifications were taken from a manufacturer’s report [21e23] and the
with the turbine coolant supply. The flow rate of the coolant can be open literature [24]. The engine consists of an eighteen-stage
modulated to supply an appropriate amount of coolant to keep the compressor with a pressure ratio of 16, and a three-stage turbine.
blade temperature below a target value. The feasibility of coolant The design fuel is a natural gas consisting of 90.1% CH4 by mole and
flow rate control was examined in a previous study [6], wherein the other miscellaneous hydrocarbons, and its lower heating value
authors used a simplified approach that focused on the first stage (LHV) is 49,244 kJ/kg. The turbine blade cooling was modeled as
nozzle blade. The present study adopted a more detailed analysis close to the actual design as possible using the reference data, as
based on a calculation for each cooled blade row. Another distinct depicted by the coolant lines shown in Fig. 1. Of course, the coolant
feature of this study is the adoption of coolant pre-cooling. Turbine pre-cooling lines were not adopted in the reference engine, but
I.H. Kwon et al. / Applied Thermal Engineering 60 (2013) 285e294 287

Fig. 1. Schematic of system configuration (dotted lines are only for the coolant pre-cooled case).

were used only in the pre-cooled case simulated in this study. For modeled using the performance map shown in Fig. 2. We used a
each turbine stage, the cooling of nozzle and rotor blades was multi-stage axial compressor map with a similar design pressure
separately modeled to predict the temperature variation of each ratio embedded in GateCycle [20], with proper scaling, taking into
blade row. Five rows (2.5 stages) were cooled by air from different account the design point (pressure ratio and mass flow) of the gas
sources, as shown in Fig. 1. The second stage nozzle/rotor and the turbine used in this study. The off-design operation of the turbine
third stage nozzle were cooled by air bled from the compressor was modeled by the following constant swallowing capacity
middle stages. Since the exact locations were not described in the (choking condition), which is very reasonable for heavy-duty in-
manufacturer’s references, we selected appropriate stages that dustrial gas turbines [20]:
have sufficient pressure to be injected into the corresponding tur- sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
bine sections. The simulated design performance of the gas turbine pffiffiffiffiffiffi  gþ1
_ in Tin
m g 2 g1
is shown in Table 1. All of the three major performance parameters ¼ constant; where k ¼ (1)
kAin Pin R gþ1
(power output, thermal efficiency, and exhaust temperature) were
in good agreement with the reference data, which demonstrates
the feasibility of the reference gas turbine modeling. The coolant flow rate for each off-design operation was calculated
When a low calorific value gas is supplied to the combustor as as follows [20]:
fuel in an existing gas turbine designed for natural gas, the oper-
ating conditions of both the compressor and the turbine deviate
2.5 1.2
from their design conditions. Therefore, a full off-design analysis is
required to perform a realistic simulation. The compressor was
efficiency
2.0 1.0

Table 1
Gas turbine specifications. 110
1.5 105 0.8
relative speed
d
PR/PR

Parameter Reference Modeling 100


d

Ambient condition 273.2 K, 1013 kPa, 60% RH


1.0 95 0.6
Pressure ratio 16 16
Compressor isentropic efficiency (%) NA 83.7 90
Total coolant flow relative to inlet air (%) NA 17.4 pressure ratio
Turbine inlet temperature (K) NA 1670.2 0.5 0.4
Turbine rotor inlet temperaturea (K) 1600.2 1600.2
Number of turbine stages 3 3
Turbine stage efficiency (%) NA 88.5
0.0 0.2
Exhaust gas flow (kg/s) 445.0 444.9 0.6 0.7 0.8 0.9 1.0 1.1 1.2
Net power (MW) 171.7 171.5
LHV efficiency (%) 36.5 36.7 Relative corrected mass flow
a
Temperature at the first stage rotor inlet. Fig. 2. Compressor performance map.
288 I.H. Kwon et al. / Applied Thermal Engineering 60 (2013) 285e294

! 0:5 Table 2
Pc Tc;d
m _ c;d
_c ¼ m (2) Fuel compositions and heating values.
Pc;d Tc
Component Mole fractions (%)
Once the turbine inlet temperature is given, thermodynamic NG LCG
matching between the compressor map and the turbine charac- CO 35.1
teristic equation determines the operating condition of the gas CO2 13.1
turbine. Fig. 3 exemplifies the feasibility of the off-design calcula- H2 31.4
tion. It shows the variation of the full load (fully fired) performance H2O 16.4
N2 þ Ar 0.19 3.9
of the gas turbine versus the ambient temperature. The simulated
CH4 90.09 0.07
variations of the power output and the efficiency are in very good C2H6 6.04
agreement with the manufacturer’s reference data [23], which C3H8 2.54
proves the validity of the off-design calculation based on the soft- C4H10 1.12
Others 0.02
ware used in this study.
LHV (kJ/kg) 49244.2 8624.7

2.2. Fuel

We selected coal syngas as a typical low calorific value gas fuel. from the bottoming cycle were used for pre-cooling, as indicated by
Table 2 shows the compositions and calorific value of the fuel [1] the dotted lines in Fig. 1. Thus, no heat loss outside the entire
used in this study. The table lists the properties of the low calo- combined cycle system was allowed. The recovery of the thermal
rific value gas, and those of natural gas. Hydrogen and carbon energy released from the cooling air by the water/steam of the
monoxide are the major components, and the calorific value bottoming cycle was beneficial in terms of the overall plant per-
(shown as the lower heating value in the table) is about one-sixth formance in the conventional natural gas-fired gas turbine and
that of natural gas. combined cycle plants [25]. We adopted this observation to the low
calorific gas-fired system as a way to minimize the performance
penalty while maintaining the target blade temperature. We
2.3. Bottoming cycle
assumed that the bottoming steam cycle was optimally designed
with respect to the cycle parameters (steam pressures and tem-
A triple pressure bottoming steam turbine cycle was used, as
peratures, condenser pressure, temperature difference, pressure
shown in Fig. 1. The major design parameters of the bottoming cycle
drop, etc.) given in Table 3 for each gas turbine condition.
and the predicted the combined cycle design performance using
natural gas are listed in Table 3. Due to the gas-side pressure drop at
the heat recovery steam generator, the gas turbine power in the 2.4. Turbine blade cooling
combined cycle plant is slightly less than the reference power
shown in Table 1. The simulated combined cycle efficiency was very The variation in the temperature of each turbine blade row was
close to that reported in the literature (56.5%) [24]. investigated using a cooling model [26]. The model describes a
In the simulated case in which the turbine coolant is pre-cooled, relationship between the cooling effectiveness and the ratio of
some of the intermediate pressure and low pressure water streams thermal capacities (the mass flow multiplied by the specific heat)
between the coolant and the mainstream gas. The cooling effec-
tiveness is defined by
120
Tg  Tb
f¼ (3)
Tg  Tc
115 Reference
Power output Simulation Once the temperatures of the mainstream gas, cooling air, and
blade metal are given at the design point, the cooling effectiveness
110 can be specified. Table 4 shows the coolant properties for each
blade row. The gas and coolant temperatures are total tempera-
Relative variation (%)

tures. In case of the rotor blades, the total temperature relative to


105
the rotating frame (i.e., the relative total temperature) should be
used. The blade temperature in Eq. (3) represents an average
100 temperature. In real engines, the blade surface temperature must
Efficiency have a distribution affected by the non-uniformity of the gas and

95
Table 3
Bottoming cycle specifications and combined cycle plant performance.
90 HP pressure (bar) 180
IP pressure (bar) 40
LP pressure (bar) 30
85 Condenser pressure (bar) 0.07
Steam temperaturea (K) 838.9
Pinch temperature difference (K) 11.1
80 Gas-side pressure drop (bar) 0.042
-10 0 10 20 30 40 GT power (MW) 168.2
o ST power (MW) 94.3
Ambient Temperature( C) Total CC plant power (MW) 262.5
CC Plant LHV efficiency (%) 56.1
Fig. 3. Example of off-design calculation: performance variation versus ambient
a
temperature. Both at HP and IP inlets.
I.H. Kwon et al. / Applied Thermal Engineering 60 (2013) 285e294 289

Table 4 relative speed V2 at the off-design condition was calculated. Finally,


Coolant properties for each cooled blade row. the relative total temperature of the rotor blade (T02,rel) was pre-
Stage Blade Coolant to gas Coolant Gas dicted and used for Tg in Eq. (3).
mass flow ratio temperature (K) temperature (K) The cooling performance was predicted using the following
1st Nozzle 0.088 691.4 1670.2 equation [26], which describes a relationship between the cooling
Rotor 0.050 691.4 1422.1 effectiveness and the ratio of thermal capacities between the
2nd Nozzle 0.032 600.4 1312.0 coolant and the mainstream gas:
Rotor 0.020 600.4 1158.7
_ c $cp;c
m f
3rd Nozzle 0.007 520.1 1065.1 ¼ C (5)
_ g $cp;g
m fN  f

where fN represents the asymptotic cooling effectiveness corre-


coolant temperatures, and the temperature of local hot spot must sponding to a very high thermal capacity ratio, and C represents the
be higher than the average temperature. However, we did not technology level of the cooling scheme. The cooling performance is
consider a detailed temperature distribution because it can only be usually presented as a curve showing the functional relations be-
predicted through a computational analysis, which is beyond the tween the cooling effectiveness and the thermal capacity ratio, as
scope of the present study. Examples of conjugate heat transfer exemplified in Fig. 5. The trend of the cooling curve is similar to
analyses can be found in literature [16e19]. Accordingly, we those of real engines [27]. Thus, Eq. (5) can be used to simulate the
adopted the average temperature approach and compared the behavior sufficiently well as illustrated in Refs. [6,7]. The value of
average temperatures among different operating strategies. fN was set to 0.92 for the first and second stages, and to 0.83 for the
To estimate the rotor-relative total temperature, a velocity tri- last stage, with a reference to the literature [28].
angle between the rotor inlet and outlet was considered at every At the design point, C of each blade row was determined using
turbine stage, as shown in Fig. 4. The absolute total temperature at Eq. (5), using all of the other parameters given by the cycle calcu-
the rotor inlet (T02) is the mixed-out temperature after nozzle blade lation. The calculated C values are 0.062 and 0.061 for the first stage
cooling. Based on reasonable design values for the blade speed (U), nozzle and rotor; 0.079 and 0.077 for the second stage nozzle and
flow coefficient (Ca/U), loading coefficient (stage work/U2), and rotor; and 0.058 for the third stage nozzle, respectively. The design
reaction, all of the absolute and relative velocities (C, V) and angles temperatures of the first and second stage blade rows were set to
(a, b) were determined at the design point. The relative total 870  C (1143.2 K), and that of the third stage nozzle blade was set to
temperature was obtained using the relation 750  C (1023.2 K). Then, the complete turbine blade cooling model
for each blade row was established. For an off-design condition, all
V22 of the parameters in Eqs. (3) and (5) (except for the blade metal
T02:rel ¼ T2 þ (4)
2cp temperature (Tb)) were known from the cycle calculation. Then, the
metal temperature was predicted using Eq. (3).
At each off-design condition, the absolute nozzle exit flow angle
The estimations of the variations in turbine blade temperatures
(a2) was assumed to remain constant at the design value (constant
described in this section were performed by using the macro-
flow deviation). Then, the absolute flow speed (C2) was calculated
function in GateCycle [20]. For low calorific value gas-fired opera-
using the mass flow equation. U is fixed because the rotational
tion, four cases were simulated. Except for the under-firing case, the
speed of the shaft should be kept constant in gas turbines used for
turbine inlet temperature remains at the design value. The first is
electric power generation (3600 rpm for a 60 Hz machine that
the baseline case in which the fuel was switched from natural gas to
directly drives the generator without a gear box). As a result, the
syngas. In the second case, we reduced the turbine inlet tempera-
ture to keep all of the blade temperatures below the design values.
In the third case, all of the cooled blade rows were kept at the
design values by modulating (actually increasing) the cooling air
flows. Finally, the effect of coolant pre-cooling in the third case was
investigated. Table 5 summarizes the five cases.

Fig. 4. Velocity triangle of a turbine stage. Fig. 5. Example of cooling effectiveness curve.
290 I.H. Kwon et al. / Applied Thermal Engineering 60 (2013) 285e294

Table 5 This result is positive in terms of gas turbine performance


Descriptions of different operating cases. because the power output increases substantially, but efficiency
Case Fuel Description decreases only slightly. However, the result causes a critical prob-
Design NG Design operation
lem with respect to the lifetime of the hot section. The increased
Baseline LCG NG is simply switched to LCG turbine inlet gas flow causes the turbine inlet pressure to rise ac-
Under-firing LCG Modulating firing temperature to keep cording to the mass flow increase (see the turbine characteristic
blade temperatures below the design values described by Eq. (1)). Thus, the compressor pressure ratio increases
Coolant modulation LCG Modulating coolant flow rates to keep blade
proportionally, which causes the turbine coolant temperature to
temperatures below the design values
Coolant modulation LCG Simultaneous use of pre-cooling and rise. Fig. 6 shows the shift of the operating point on the compressor
with pre-cooling modulation of flow rate of coolant to keep blade map. The compressor discharge pressure increases from 1620.8 to
temperatures below the design values 1782.9 kPa, and the discharge temperature increases from 691 to
726 K. This is negative in terms of turbine blade temperature.
Another factor affecting the blade temperature variation is the
The net plant power output and efficiency are defined as thermal capacity ratio between the coolant and gas flow. Fig. 7
follows: shows this variation in the first stage nozzle. Although the main-
stream gas flow increased significantly by firing low calorific value
_
_ net ¼ W _ _
W GT þ W ST  W Aux (6) gas, the thermal capacity ratio did not change appreciably. This is
because the coolant flow also increased, which was affected by the
W_ net increased source pressure (see Eq. (2) e the total coolant fraction
hnet ¼ (7) increased from 17.3% to 19.0%). A slight decrease in the cooling
_
ðm$LHVÞfuel
effectiveness was predicted, as shown in Fig. 7. The coolant tem-
The auxiliary power is the sum of all of the power consumptions perature was predicted to rise from 691 to 726 K as previously
of the additional components (mostly pumps). mentioned, and the cooling effectiveness was predicted to decrease
from 53.8% to 52.6%. Both of these factors (the rise in the coolant
temperature and the decrease in effectiveness) contributed to the
3. Results and discussion 30 K increase in the first stage nozzle temperature, but the former
factor is dominant. Similar patterns occur in all of the other turbine
The predicted performance of the gas turbine fired with the low blade rows, resulting in 24e30 K increases in blade temperature.
calorific value gas is compared with its design performance using The increase in the compressor pressure ratio could be an issue
natural gas in Table 6. To achieve the same turbine inlet tempera- from the viewpoint of safe engine operation. Active hardware
ture, a much larger fuel (6.9 times) should be supplied to the modifications such as an increase in the turbine annulus area [9,30]
combustor in the low calorific value gas case compared to the could be an ultimate solution to the surge issue. In this study, we
natural gas case, which would cause a considerable increase in the focused only on the issue of turbine blade overheating because we
turbine gas flow (more than a 10% increase in terms of the exhaust still have a 10% surge margin in the baseline operation and want to
gas flow). This results in a considerable improvement in the net gas present a remedy to the overheating problem. Other three opera-
turbine power output. The predicted power output was 204.5 MW, tions have larger surge margins than the baseline operation.
which is 19.2% greater than the design power output of 171.5 MW. Under-firing (reduction of the turbine inlet temperature) is the
This power augmentation is realizable in the view point of me- simplest way to suppress blade overheating without engine hard-
chanical design of gas turbines. Gas turbines are usually designed to ware modification. Fig. 8 shows the effect of reducing the turbine
accommodate the thermodynamically available power generation inlet temperature on the temperatures of the cooled blade rows. To
capacity in cold ambient conditions (see Fig. 3) which is much reduce the temperatures of all of the blade rows below the design
larger than the power output at standard ambient temperature temperatures, the turbine inlet temperature must be reduced to
(15  C). The gas turbine dealt with in this study is known to have a 1610 K, which is 60 K lower than the design value. Fig. 9 shows the
maximum power capacity around 200 MW [29]. Some re-design
and reinforcement of the shaft mechanical system may be
required depending on the actual shaft design features. The gas
turbine efficiency was predicted to decrease slightly. This is due to
the fact that the thermal energy input required to achieve the same
combustor outlet temperature of the product gas in the low calo-
rific value gas-fired case is slightly larger because low calorific value
gas includes non-reacting (inert) components such as water vapor,
carbon dioxide, and nitrogen.

Table 6
Parameter comparison between the natural gas fired case and the baseline low
calorific value gas fired case.

Parameter NG LCG

GT Power (MW) 171.5 204.5


GT Efficiency (%) 36.7 36.3
Turbine inlet temperature (K) 1670.2 1670.2
Pressure ratio 16 17.6
Exhaust mass flow (kg/s) 444.9 491.1
Fuel mass flow (kg/s) 9.5 65.3
1st stage nozzle blade temp. (K) 1143.2 1173.6
Fig. 6. Deviation of compressor operating condition caused by firing low calorific value
1st stage rotor blade temp. (K) 1143.2 1172.4
gas.
I.H. Kwon et al. / Applied Thermal Engineering 60 (2013) 285e294 291

0.65 205.0 36.4

Power
Efficiency

0.60 200.0 36.3

Efficiency [%]
Power [MW]
0.55
195.0 36.2
φ

0.50
190.0 36.1

NG design
baseline
0.45 LCG under-firing
coolant modulation 185.0 36.0
coolant modulation 1610 1620 1630 1640 1650 1660 1670
with pre-cooling Turbine inlet temperature [K]
0.40
Fig. 9. Variations in gas turbine power output and efficiency versus a reduction in
0.06 0.065 0.07 0.075 0.08
turbine inlet temperature.

Fig. 7. Cooling effectiveness of the first stage nozzle. turbine performance corresponding to such an operation was
evaluated. The cooling effectiveness of the first stage nozzle in this
coolant-modulated operation is shown in Fig. 7 together with other
variations in the gas turbine power output and efficiency. With a cases. The required total coolant fraction was estimated to be 23.3%.
1610 K turbine inlet temperature, the power output was reduced by Modulation of the coolant flow rate requires some modifications of
7.4% from 204.5 MW, reaching 189.3 MW. The efficiency also the coolant passage between the compressor and the turbine (e.g.,
decreased slightly (from 36.3 to 36.0%). The decrease in the turbine control of the valve in the coolant flow loop). The relative advantage
inlet temperature caused the pressure ratio, and thus the coolant of the coolant flow rate modulation compared to under-firing is
temperature, to decrease slightly, which provides a small advantage that it causes a lower power penalty from the baseline case (a
in terms of the blade temperature. simple fuel switch from natural gas to low calorific value gas). The
Increasing the coolant flows also decreases the blade tempera- gas turbine power output of the coolant-modulated case was pre-
tures by increasing the cooling effectiveness. For each blade row, dicted to be 196.7 MW, which is larger than that of the under-firing
we calculated an appropriate amount of coolant flow to maintain case by 7 MW. The efficiency was predicted to decrease slightly to
the blade metal temperature at the design value, and the gas 35.9%.
Based on the coolant-modulated operation, pre-cooling of the
coolant was simulated. If the coolant temperature is reduced, the
1200

200.0 36.0
Power

1150 199.5 Efficiency


Blade temperature [K]

35.9
1st, 2nd stage nozzle/rotor
design temperature 199.0
Efficiency [%]

35.8
Power [MW]

1100 1st Nozzle 198.5


1st Rotor
2nd Nozzle
2nd Rotor 198.0
35.7
3rd Nozzle
1050
197.5

35.6
3rd stage nozzle design temperature 197.0

1000
1610 1620 1630 1640 1650 1660 1670 196.5 35.5
Turbine inlet temperature [K] 0 20 40 60 80 100
Degree of pre-cooling [K]
Fig. 8. Variations in turbine blade temperatures versus a reduction in turbine inlet
temperature. Fig. 10. Effect of coolant pre-cooling on gas turbine performance.
292 I.H. Kwon et al. / Applied Thermal Engineering 60 (2013) 285e294

210.0 37.5 120 30


NG LCG NG LCG
Power Coolant flow rate
Efficiency 110 25
200.0 37.0 Coolant fraction

Total coolant flow rate[ kg/s]


Gas turbine efficiency [%]
Gas turbine power [MW]

Total coolant fraction [%]


100 20
190.0 36.5

90 15
180.0 36.0

80 10

170.0 35.5

70 5

160.0 35.0
Design Base- Under- Coolant Coolant
60 0
line firing modulation modulation Design Base- Under- Coolant Coolant
with 50K line firing modulation modulation
pre-ccoling
with 50K
pre-cooling
Fig. 11. Gas turbine performance comparison for different operations.
Fig. 13. Total coolant flow rate and fraction relative to compressor inlet air flow rate.

coolant supply could be decreased while maintaining the same


blade temperatures. The reduction in the coolant flows, especially and the total coolant fraction relative to compressor inlet air flow in
those of the blade rows after the first stage nozzle, results in the five cases. The advantage of coolant modulation versus under-
increased turbine inlet gas flow. This enhances the gas turbine firing is clear. Despite a marginal gas turbine efficiency penalty,
power output. The impact of the degree of pre-cooling on gas tur- coolant modulation ensures a sensible advantage in terms of power
bine performance is shown in Fig. 10. A reduction of the coolant generation. For example, with coolant modulation and 50 K pre-
temperatures of all blade rows by the same amount was simulated. cooling, we obtained a 19% power augmentation and a 1% point
A 100 K reduction would enhance the power output by 2.3 MW efficiency penalty compared to the gas turbine design performance.
(1.2%), while reducing the efficiency only slightly (0.34% points). Fig. 14 shows the flow rate and temperature at the gas turbine exit.
Fig. 11 summarizes the performance comparison among four Increases of exhaust gas flow in cases of firing the low calorific
different low calorific value gas-fired operations and the reference value gas is due to the increased fuel flow as we have already dis-
natural gas-fired operation. Fig. 12 compares the compressor cussed. The baseline case has a higher exhaust temperature than
discharge pressures and the corresponding air temperatures (i.e., the design case even with a higher turbine expansion pressure ratio
the first turbine stage coolant temperatures except for the case with because it exhibits a slightly higher specific heat of the turbine gas
pre-cooling). Fig. 13 compares the total turbine coolant flow rates compared to the design case due to an increase in the water content

2000 500 640 540


NG LCG NG LCG
CDP 630 Temperature
Compressor discharge temperature [ C]

CDT
Compressor discharge pressure [kPa]

Flow 520
1800 480
Turbine exhaust temperature [ C]
o

620
Turbine exhaust flow [kg/s]

500
610
1600 460

600 480

1400 440
590
460

580
1200 420
o

440
570

1000 400 560 420


Design Base- Under- Coolant Coolant Design Base- Under- Coolant Coolant
line firing modulation modulation line firing modulation modulation
with 50K with 50K
pre-cooling pre-cooling

Fig. 12. Compressor discharge air pressure and temperature. Fig. 14. Turbine exhaust temperature and flow rate.
I.H. Kwon et al. / Applied Thermal Engineering 60 (2013) 285e294 293

58.0 (3) Modulation of the coolant flow rates of all cooled turbine blade
NG LCG rows provides a greater gas turbine power output compared to
320 Power under-firing. Moreover, coolant pre-cooling enhances the po-
Efficiency 57.0 wer output further by reducing the coolant supply.
(4) Coolant modulation is slightly disadvantageous in terms of gas
300 turbine efficiency. However, the relatively higher gas turbine
56.0 exhaust temperature provides much higher combined cycle

Efficiency [%]
Power [MW]

efficiency compared to the under-firing, and provides 80% of


the maximum available power augmentation. The recovery of
280 55.0
the rejected heat from the coolant by the steam cycle in the
case of coolant pre-cooling is advantageous from the viewpoint
54.0 of combined cycle performance. Therefore, coolant modula-
260 tion, especially when assisted by coolant pre-cooling, is a better
solution than under-firing.
53.0
240 Acknowledgements

52.0 This work was supported by the New & Renewable Energy
D esign Base- U nder- Coolan t C oolant
line firin g m odulation m odu lation Center/Korea Energy Management Corporation through the
w ith 50K
pre-cco lin g
“Design and Construction of 300 MW IGCC Demonstration Plant in
Korea” project funded by the Korean Ministry of Knowledge
Fig. 15. Combined cycle performance comparison among different operations. Economy.

References
of the combustion gas. Under-firing, of course, causes a drop in the
exhaust gas temperature. The exhaust temperatures of the two [1] J. Hoffmann, J. Tennant, G.J. Stiegel, Comparison of Pratt and Whitney Rock-
etdyne IGCC and Commercial IGCC Performance, 2006. DOE/NETL-401/
coolant modulated cases are lower than that of the baseline case 062006.
because of the increased coolant supply to the turbine side but they [2] R.A. Dennis, W.W. Shelton, P. Le, Development of Baseline Performance Values
are sufficiently higher than that of the under-firing case. for Turbines in Existing IGCC Applications, 2007. ASME paper GT2007-28096.
[3] J.J. Lee, Y.S. Kim, K.S. Cha, T.S. Kim, J.L. Sohn, Y.J. Joo, Influence of system
Fig. 15 shows the combined cycle performance. The trend in the integration options on the performance of an integrated gasification com-
power output comparison is the same as in the gas turbine per- bined cycle power plant, Applied Energy 86 (2009) 1788e1796.
formance. However, the efficiency trend is different. In particular, [4] M. Rieger, R. Pardemann, H. Rauchfub, B. Meyer, Effects of ASU integration on
IGCC performance and gas turbine operation, VGB Power Tech 5 (2008) 102e107.
coolant-modulated operation exhibits a slightly higher combined
[5] E.O. Oluyede, J.N. Phillips, Fundamental Impact of Firing Syngas in Gas Tur-
cycle efficiency than the under-firing operation. This is because bines, 2007. ASME paper 2007-27385.
coolant-modulated operation has a relatively higher gas turbine [6] Y.S. Kim, J.J. Lee, T.S. Kim, J.L. Sohn, Y.J. Joo, Performance analysis of a syngas-
exhaust temperature as shown in Fig. 14, which is definitely posi- fed gas turbine considering the operating limitations of its components,
Applied Energy 87 (2010) 1902e1911.
tive in terms of the bottoming cycle performance (i.e., it generates [7] Y.S. Kim, J.J. Lee, T.S. Kim, J.L. Sohn, Effects of syngas type on the operation and
more steam turbine power). The efficiency gap between normal performance of a gas turbine in integrated gasification combined cycle, Energy
coolant modulation and coolant modulation with pre-cooling is Conversion and Management 52 (2011) 2262e2271.
[8] A. Franco, N. Giannini, Perspectives for the use of biomass as fuel in combined
nearly insensible because the heat rejected from the coolant is cycle power plants, International Journal of Thermal Sciences 44 (2005) 193e197.
transported into the bottoming cycle. Thus, more power is gener- [9] M. Rodrigues, A. Walter, A. Faaij, Performance evaluation of atmospheric
ated, which minimizes the efficiency penalty due to pre-cooling. biomass integrated gasifier combined cycle systems under different strategies
for the use of low calorific gases, Energy Conversion & Management 48 (2007)
The maximum available power augmentation by firing low calo- 1289e1301.
rific value gas (i.e., by simply switching fuel from natural gas to low [10] S. Colantoni, S.D. Gatta, R. De Prosperis, A. Russo, F. Fantozzi, U. Desideri, Gas
calorific value gas) is 47 MW (309 MW of baseline operation versus turbines fired with biomass pyrolysis syngas: analysis of the overheating of
hot gas path components, Journal of Engineering for Gas Turbines and Power
262 MW of design operation). However, this operation significantly 132 (2010) 061401-1e061401-8.
affects hot section life. Coolant modulation using 50 K pre-cooling [11] A. Walter, J. Llagostera, Feasibility analysis of co-fired combined-cycles using
produces 300 MW of net power output, which guarantees 80% of biomass-derived gas and natural gas, Energy Conversion & Management 48
(2007) 2888e2896.
the maximum available power augmentation while maintaining
[12] A. Bhattacharya, D. Manna, B. Paul, A. Datta, Biomass integrated gasification
blade metal temperatures at the design levels. Therefore, coolant combined cycle power generation with supplementary biomass firing: energy
modulation, especially when assisted by coolant pre-cooling, is a and exergy based performance analysis, Energy 36 (2011) 2599e2610.
better solution compared to under-firing. [13] K.K. Gupta, A. Rehman, R.M. Sarviya, Bio-fuels for the gas turbine: a review,
Renewable and Sustainable Energy Reviews 14 (2010) 2946e2955.
[14] D.W. Kang, T.S. Kim, K.B. Hur, J.K. Park, The effect of firing biogas on the
performance and operating characteristics of simple and recuperative cycle
4. Conclusion gas turbine combined heat and power systems, Applied Energy 93 (2012)
215e228.
[15] F. He, A. Li, P. Liu, L. Ma, E.N. Pistikopoulos, Operation window and part-load
The results of this study are summarized as follows: performance study of a syngas fired gas turbine, Applied Energy 89 (2012)
133e141.
(1) Firing low calorific value gas in the gas turbine increases power [16] S. Amaral, T. Verstraete, R.V. Braembussche, T. Arts, Design and optimization
of the internal cooling channels of a high pressure turbine blade e part I:
output due to the increase in fuel flow. The increased turbine
methodology, ASME Journal of Turbomachinery 132 (2010) 021013.
gas flow causes the pressure ratio to increase, which is the [17] T. Verstraete, S. Amaral, R.V. Braembussche, T. Arts, Design and optimization
major cause of turbine blade overheating. of the internal cooling channels of a high pressure turbine blade e part II:
(2) Under-firing could be an easiest solution to keep the blade optimization, ASME Journal of Turbomachinery 132 (2010) 021014.
[18] G. Nowak, W. Wróblewski, Optimization of blade cooling system with use of
temperature below the design values. However, it causes a conjugate heat transfer approach, International Journal of Thermal Sciences
considerable reduction in power output. 50 (2011) 1770e1781.
294 I.H. Kwon et al. / Applied Thermal Engineering 60 (2013) 285e294

[19] G. Nowak, I. Nowak, Shape design of internal cooling passages within a tur- [26] T.S. Kim, S.T. Ro, Comparative Evaluation of the Effect of Turbine Configu-
bine blade, Engineering Optimization 44 (2012) 449e466. ration on the Performance of Heavy-duty Gas Turbines, 1995. ASME paper
[20] GE Power-Enter Software, GateCycle Version 6.0, 2006. 95-GT-334.
[21] T. Ginter, T. Bouvay, Uprate Options for the MS7001 Heavy Duty Gas Turbine, [27] K. Kawaike, N. Kobayashi, T. Ikeguchi, Effect of new blade cooling system with
GE Power Systems, 2006. GER3808C. minimized gas temperature dilution on gas turbine performance, Journal of
[22] E. Gebhardt, The F Technology Experience Story, GE Power Systems, 2000. Engineering for Gas Turbines and Power 106 (1984) 756e764.
GER3950C. [28] A. Cohn, M. Waters, The Effect of Alternative Turbine Cooling Schemes on
[23] F.J. Brooks, Ge Gas Turbine Performance Characteristics, GE Power Systems, the Performance of Utility Gas Turbine Power Plants, 1982. ASME 82-JPGC-
2000. GER3567H. GT-19.
[24] R. Farmer, Gas Turbine World, Handbook, vol. 27, Pequot Publishing Inc., [29] R. Martin, J. Donohue, The 7FA Gas Turbine: A Class Reimagined, in: GE Gas
2009, pp. 75e91. Turbine Technology Symposium, Greenville, SC, USA, 2009.
[25] I.H. Kwon, D.W. Kang, S.Y. Kang, T.S. Kim, Influence of precooling cooling air [30] R. Chacartegui, D. Sanchez, J.M. Munoz de Escalona, A. Munoz, T. Sanchez, Gas
on the performance of a gas turbine combined cycle, Transactions of the and steam combined cycles for low calorific syngas fuels utilization, Applied
Korean Society of Mechanical Engineering B 36 (2012) 171e179 (in Korean). Energy 101 (2013) 81e92.

You might also like