STROJN ICKY CASOPIS, 59, 2008, .

4 c

189

A computer-based model for gas-turbine power augmentation by inlet-air cooling and water/steam injection
MOHAMED M. EL-AWAD

The paper presents a computer-based thermodynamic model that estimates the eects of power augmentation methods on the gas turbine power and heat rate. The model takes into consideration the eect of evaporative or refrigerative inlet-air cooling and allows for water or steam injection into the combustion chamber. The formulation of the model improves upon the conventional standard-air analysis by including the eects of water on the working-uid properties, by accounting for the mass of fuel, and by adopting the exact variable specic-heat approach. The model also improves the accuracy and extends the scope of the analysis by solving an energy equation for the combustion process, which enables either the combustion temperature, or the required rate of fuel consumption, to be specied and the other to be calculated. The paper describes the thermodynamic model and the computer program based on it. Verication of the computer-based model against relevant published data is also presented and discussed. K e y w o r d s: power generation, gas turbine, inlet-air cooling, power aumentation, computer model

1. Introduction Gas turbines have become the favoured power generation machines in many parts of the world. Standing alone in open-cycle, gas turbines are used to meet the peak-loads and combined with steam turbines in combined cycles they are used as ecient base-load machines. The easy-to-start capability, fuel-exibility, low-fuel cost, and low environmental pollution oer the gas-turbine clear advantages over other power-generation machines. Unfortunately, the performance of these machines is greatly degraded by adverse ambient conditions due to the high air temperatures and dusty environment. Being a constant volume-ow machine,
Mechanical Engineering Department, Faculty of Engineering & Architecture, University of Khartoum, P. O. Box 321, Khartoum, Sudan e-mail address: [email protected]

190

STROJN ICKY CASOPIS, 59, 2008, . 4 c

the power of the gas turbine is directly proportional to the mass ow rate of air through the turbine. A high ambient temperature (or humidity) reduces the air density and, accordingly, the mass-ow rate through the gas turbine. Therefore, the turbines loose signicant portions of their design generation capacity at standard ISO condition (15 C) when operating in hot and humid climate. The gas turbins output typically decreases by 10 % to 18 % for every 10 K of increase in e inlet air temperature [1, 2]. A high ambient temperature also increases their heat rate and, according to McCracken [2], gas turbines produce 2535 % less power in summer than in winter at 510 % higher heat rate, which means an average increase of 6 % in fuel consumption. Power producers apply various power augmentation methods in order to compensate for the eect of high temperatures on the gas-turbine output [27]. These methods include refrigerative or evaporative inlet-air cooling. The refrigerative cooling can be of the mechanical or the absorption type and can be applied with or without thermal energy storage [35]. The evaporative cooling can be achieved by a conventional media-type method or by fogging [6, 7]. By applying inlet-air cooling, additional megawatts can be obtained from existing gas turbines at a fraction of the cost of installing new generation plants. According to Boyce [8], a new power plant costs 500 $ per kilowatt while inlet air-cooling costs 110120 $ per kilowatt of augmented power. Besides inlet-air cooling, power-augmentation can also be achieved by injecting water or steam into the combustion chamber [911]. Unfortunately, water injection, which is usually used for NOx control, increases the heat rate of the plant and, therefore, the generation cost [11]. The problem can be solved by injecting steam, rather than liquid water. Since the ue gases are usually exhausted at a temperature exceeding 500 C, the generation of steam in a heat-recovery steam generator enables some of this lost energy to be recovered and the plant eciency to be improved. The required initial investments and running costs of the dierent power augmentation systems vary considerably [1, 8]. The suitability and economical feasibility of power augmentation methods depend strongly on the existing power-plant equipment and layout, the ambient atmospheric conditions, the eect of the method on the turbine power and heat rate, and the electricity-consumption pattern. Therefore, a given power augmentation method may have certain advantages or disadvantages, compared to other methods, when a particular gas-turbine power plant is considered. The present work is part of an on-going eort to develop a computer-based model that estimates the eect of the dierent power augmentation methods on the turbine output. The model, which takes into consideration the eect of inlet air cooling on the turbine power and heat rate and allows for water or steam injection into the combustion chamber, can assist in the selection of the suitable power augmentation method. Sections 2 and 3 of this paper describe the analytical thermodynamic model and the computer program based on it while Section 4 deals with the model verication against published data.

 STROJN ICKY CASOPIS, 59, 2008, . 4 c

191

2. The thermodynamic model Figure 1a,b shows the schematic diagram of a two-shaft gas-turbine plant together with the cycle T -s diagram. The system consists of an axial-ow air compressor followed by the combustion chamber and the two-shaft gas turbine. In the present model, it is assumed that the gas exhausted by the gas-generator (i.e. the high-pressure) turbine is passed directly to the power (i.e. the low-pressure) turbine without reheating. An air-cooler cools the air inlet to the compressor and, therefore, increases its density and mass ow rate through the plant. The air-cooler can be of the evaporative or refrigerative type. Before discharged to the atmosphere, the ue gases pass through a heat-recovery steam generator (HRSG), which uses the heat of the ue gases leaving the turbine to produce steam. Some of this steam can be injected into the combustion chamber for power augmentation and NOx reduction. A pump is needed to raise the water pressure before the HRSG to a suitable value. Standard textbooks of thermodynamics usually analyse the gas-turbine cycle by adopting the standard-air method, according to which the working uid is

Fig. 1. The steam-injected gas-turbine plant (a) and the cycle (b).

192

STROJN ICKY CASOPIS, 59, 2008, . 4 c

treated as pure air whose specic heats are evaluated at some average temperature or even kept constant at their room-temperature values [12, 13]. Moreover, the eect of the fuel mass on the turbine work is completely ignored. These assumptions allow an approximate method of analysis to be applied, which greatly simplies the analysis. However, the results of the approximate method are known to deviate from the exact values by signicant margins [13]. The present thermodynamic model includes the eects of water on the working-uid properties, adopts the exact variable specic-heat approach [13] and accounts for the fuel mass. By solving an energy equation for the combustion process, the model extends the analysis scope since this enables either the combustion temperature or the required rate of fuel consumption to be specied and the other calculated. In order to simplify the model, pressure and heat losses are ignored anywhere in the cycle. 2.1 T h e a i r c o m p r e s s o r Given the inlet-air temperature (T1 ) and humidity (1 ), the enthalpy of humid air (per kg of dry air) at the compressor inlet (h1 ) is determined from: h1 = h1da + 1 h1w , (1a)

where h1da and h1w are the specic enthalpies of dry air and water-vapour, respectively. The enthalpy of humid air after an ideal isentropic compression (h2s ) is determined using the compressor pressure ratio (Prc ) and the relative pressure (Pr ) as follows [12]: T1 Pr1 ; Pr2 = Pr1 Prc ; Pr2 T2s h2s . (1b)

Note that an iterative method is needed to nd T2s from Pr2 of the humid air. The actual enthalpy after compression (h2 ) is calculated from the isentropic head as follows: h2s h1 , c

h2 = h1 +

(1c)

where c is the compressor isentropic eciency. The input of the compressor (Wc ) is calculated from the mass ow rate and air enthalpy increase within the compressor as follows:

 STROJN ICKY CASOPIS, 59, 2008, . 4 c

193 (1d) (1e)

mha = (1 + )mda ; Wc = mha (h2 h1 ),

where (mha ) and (mda ) are the mass ow rates of humid air and dry air, respec tively. The compressor discharge temperature (T2 ) is found from the discharge enthalpy of humid air (h2 ) by iteration. 2.2 T h e c o m b u s t i o n p r o c e s s The presence of excess air into the combustion chamber, which is usually the case with gas turbines, assures complete combustion of the hydrocarbon fuel. Neglecting kinetic energy and potential energy changes between inlet and exit of the combustion chamber, the energy balance over an adiabatic chamber is given by: 1 Hf0 + nh2 = n h3
da

+ n h2

+ n h8

wi

+ n h3

wi

+ n h3

H2 O

+ n h3

(2a)
CO2

+ n h3

excess da

nh3

,
O2

where Hf0 is the molar heat (enthalpy) of formation of the hydrocarbon gas-fuel, n is the number of moles of the reactant or product of combustion per mole of the fuel, and h is the molar enthalpy. The suces H2 O and CO2 refer to the two products of combustion (water and carbon dioxide) and O2 refers to oxygen. The suces da, w and wi refer to the dry air, air-borne moisture, and injected water/steam, respectively. Suces 2 and 3 refer to the conditions (i.e. temperatures) before and after the combustion chamber. The number of oxygen-moles appearing in the product side is obtained from: 1 (n)O2 = (n)CO2 + (n)H2 O . (2b) 2 Equation (2) is used to nd T3 when the fuel-air ratio is specied, or to nd the fuel-air ratio (mass of fuel per unit mass of dry air) when the combustion temperature (T3 ) is specied. 2.3 T h e g a s - g e n e r a t o r t u r b i n e The enthalpy after the gas-generator turbine (h4 ) is obtained by equating the gas-generator turbine work with the compressor work. The discharge temperature (T4 ) is then obtained from h4 by iteration:

194 h4 = h3 Wc ; mha

STROJN ICKY CASOPIS, 59, 2008, . 4 c

h4 T 4 .

(3a)

The enthalpy (h4s ) and temperature (T4s ) after an ideal, isentropic expansion are then obtained using the generator turbine eciency (gt ): h4s = h3 h3 h4 ; gt h4s T4s . (3b)

The pressure at the outlet of the gas-generator turbine is determined from the known temperatures T3 and T4s using the isentropic expansion relationship: T3 Pr3 ; T4s Pr4 ; P4 = P3 Pr4 . Pr3 (3c)

2.4 P o w e r t u r b i n e a n d t h e r m a l e f f i c i e n c y Assuming the discharge pressure from the power turbine (P5 ) to be the same as the inlet pressure P1 , the enthalpy after an ideal, isentropic expansion (h5s ) is determined from the known temperature (T4 ) and pressure ratio (P5 /P4 ) as follows: T4 Pr4 ; Pr5 = Pr4 P5 ; P4 Pr5 T5s h5s . (4a)

The actual enthalpy at the outlet of the power turbine (h5 ) is then calculated from the isentropic eciency of the turbine (pt ) and enthalpy dierence (h4 h5s ): h5 = h4 (h4 h5s ) pt . (4b)

The net work of the gas turbine per unit mass ow rate of dry air is equal to that of the power turbine (Wpt ), which is calculated as follows: Wpt = (mha + mfuel + mwi )(h4 h5 ), (4c)

where mfuel is the fuel consumption and mwi is the rate of water/steam injection per unit mass of dry air. The heat energy of the fuel (Qin ) and thermal eciency (th ) are then obtained from: Qin = mfuel LHV ; (4d)

 STROJN ICKY CASOPIS, 59, 2008, . 4 c

195 Wpt , Qin (4e)

where LHV is the fuel lower heating value. Finally, the exhaust-gas temperature (T5 ) is obtained from the calculated enthalpy (h5 ) of the gas mixture by iteration and the energy balance over the HRSG is given by: (mha + mfuel + mwi )(h5 h6 ) = mwi (h8 h7 ). 3. The computer program A computer program has been developed based on the above thermodynamic model to analyse the gas-turbine cycle with power augmentation by inlet-air cooling and/or water/steam injection. Developed in FORTRAN, the program initially reads the basic data of the gas turbine (total pressure ratio, compressor and turbine eciencies, combustion temperature, and rate of fuel injection) and requires the user to specify a reference atmospheric condition as described by the inlet values of the pressure, temperature and humidity. It then asks the user about the actual inlet-air condition and the method of power augmentation to be considered. If inlet-air cooling is to be used, the program determines the dew-point and wet-pulp temperatures and informs the user with their values since these two temperatures determine the limiting values for refrigerative and evaporative cooling, respectively. If water/steam is to be injected, the program asks about the temperature of water/steam to be injected in order to calculate the maximum rate of injection that saturates the air at the compressor discharge. It then asks the user to give the required rate of water/steam injection, which should be less than the maximum value, before proceeding to calculate, from the relationships described in the preceding section, the compressor work, combustion temperature (or fuel rate), gas-generator turbine work and exit pressure, power turbine work and exit temperature, and thermal eciency of the plant. The important details of the computer program are given below. 3.1 T h e r e f e r e n c e a t m o s p h e r i c c o n d i t i o n and reference volume flow rate The mass ow rate through the gas turbine at a particular site depends on the air density, which depends on the atmospheric pressure, ambient temperature, and humidity. To be able to compare the gas-turbine performance at dierent atmospheric or inlet-air conditions, the present computer model analyses the cycle on the basis of volume ow rate rather than mass ow rate. A reference volume ow rate (VR ) is dened as the volume that contains a unit mass ow rate (1 kg s1 ) of dry air at a chosen reference inlet-air condition. The reference inlet-air condition can be the design condition or the ISO standard condition of 101.325 kPa, 15 C, (5)

196

STROJN ICKY CASOPIS, 59, 2008, . 4 c

and 60 % humidity. But it can also be a certain atmospheric conditions at the gasturbine site, say in a winter day. At atmospheric or inlet-air conditions which are dierent from the reference condition, the reference volume will contain dierent mass ow rates of dry air and moisture. Applying thermodynamic relations for a mixture of two ideal gases (dry air and water-vapour), the reference volume ow rate is given by [12]: Ra T R Ra TR VR = = , (6) R Psat PR,da 100 where Ra is the gas constant for dry air. TR , PR , R , and PR,da are, respectively, values of the pressure, temperature, relative humidity (%), and partial pressure of dry air at the reference atmospheric condition. Psat is the saturation pressure for water at the reference temperature. The total mass ow rate of humid air (dry air + moisture) entering the compressor through the reference volume at the reference condition is equal to (1 + R ) kg s1 , where R is the absolute humidity (or humidity ratio) of air at the reference condition. 3.2 A c t u a l m a s s f l o w r a t e o f a i r t h r o u g h t h e c o m p r e s s o r At atmospheric condition other than the reference condition, the total mass ow rate of dry air through the reference volume is calculated from the given values of atmospheric pressure (Patm ) and dry-bulb temperature (Tatm ) as follows: mda = Using Eq. (6), this becomes: mda = Pda TR . PR,da Tatm (7b) Pda VR . Ra Tatm (7a)

The total mass ow rate of humid air (mha ) is then given by: mha = (1 + atm )mda . (7c)

The specic humidity at actual atmospheric condition (atm ) is either given or determined from the relative humidity (atm ) according to the known thermodynamic relations for a mixture of air-water vapour mixture [12]. 3.3 E f f e c t o f w a t e r / s t e a m i n j e c t i o n on the combustion temperature The injection of liquid water or steam into the combustion chamber causes the combustion temperature to drop unless more fuel is added. Two options have been considered and included in the program:

 STROJN ICKY CASOPIS, 59, 2008, . 4 c

197

1. to maintain the combustion temperature at its value prior to water/steam injection, 2. to keep the fuel consumption constant and allow the combustion temperature to drop. The amount of additional fuel required to maintain the combustion temperature equal to that prior to water/steam injection depends on the temperature and injection rate of water/steam. More fuel is consumed when the injection rate of water/steam increases and/or when its temperature decreases. The required fuel ow rate is obtained from the energy equation (Eq. 2). If no fuel is supplied in addition to that prior to water/steam injection and the combustion temperature is allowed to drop, Eq. (2) is solved to give the new combustion temperature. In this case, the work from air and the combustion products will be reduced as a result of the reduced temperature and enthalpy at the turbine inlet. However, additional work is provided by the injected water/steam, which increases the mass ow rate in the turbine. Therefore, both the turbine power output and thermal eciency depend on the additional work of the injected steam or water. 3.4 M o d e l l i n g t h e w o r k i n g f l u i d The thermodynamic model described above requires values of the enthalpy (h) and relative pressure (Pr ) for the working uid, which consists of two gases or more. The computer program evaluates the required thermodynamic properties by treating the pre-combustion working uid as a mixture of two ideal gases, viz. dry air and water vapour. After combustion the working uid consists of these two components plus the products of combustion (CO2 , N2 and H2 O). If the temperature is known, the thermodynamic properties of the pro-combustion working uid can also be found easily by treating it as a mixture of ideal gases. However, if an inverse solution is required (e.g. nding T from a known value of h) it will be dicult to perform an iterative solution for more than two components (see section 3.6 below). Therefore, the pro-combustion working uid is also treated as a mixture of dry air and water vapour, but the program oers the user two options for modelling the water vapour part: 1. Model 1: all products of combustion including the combustion water are treated as dry air. Thus, the water vapour part includes the original air-borne moisture plus the water vapour that results from the injected water/steam. 2. Model 2: the combustion water is also included into the water vapour component. Thus, the dry air component includes only CO2 and N2 . For the purpose of code verication and comparison with available data, the program oers a third option that treats all the working uid as dry air. This option, called Model 0, is similar to the air-equivalent method of Bathie [13], who also used variable specic heats but did not take into consideration the eect of inlet-air humidity.

198

STROJN ICKY CASOPIS, 59, 2008, . 4 c

3.5 T h e r m o d y n a m i c p r o p e r t i e s o f t h e w o r k i n g f l u i d The computer program obtains the thermodynamic properties of the working-uid components using various relations. Most of these relations were adopted from the published literature [12, 13], but some relationships have been derived directly from thermodynamic property tables. Properties of the dry air, water vapour, products of combustion, and liquid water are determined from the following relations: Air : Treated as an ideal gas, the enthalpy of air depends on the temperature alone and is obtained from the following relationship [12]: h = Ra a T + + c d b T2 + T3 + T4 2 3 4 [kJ kg1 ], (8a)

e T 5 302.67, 5

where a, b, c, d, and e are constants and T is the absolute temperature in Kelvin. The constant 302.67 is added to the original equation given by Moran and Shapiro [12] in order to unify their reference temperature (originally taken as 0.0 K) with that of Bathie [13], taken as 298.15 K or 25 C. The relative pressure is also obtained from a polynomial in temperature. For the temperature range 5001000 K, this is given by the following 4th order polynomial: Pr = a + b T + c T 2 + d T 3 + e T 4 , [kJ kg1 ], (8b)

where the constants a, b, c, d, and e have the values of 8.2488968, 0.060512094, 0.00016927346, 1.8744272 107 and 1.8445728 1010 , respectively. For temperatures higher than 1000 K, the 4th order polynomial requires a dierent set of coecients. For temperatures lower than 500 K, a 10th order polynomial is used due to the rapid change with temperature. Water vapour and products of combustion: The enthalpies of the water vapour and products of combustion (CO2 and O2 ), which also depend on the temperature alone, are obtained from the following relation [12]: h = Ru T a + b c d e f T + T2 + T3 + T4 + 2 3 4 5 T , [kJ kmol1 ], (9)

 STROJN ICKY CASOPIS, 59, 2008, . 4 c

199

where Ru is the universal ideal gas constant and T is the absolute temperature in Kelvin. The values of the constants a, b, c, d, e, and f are obtained by tting polynomials to the tabulated values of h for each gas/vapour. For a better accuracy, two sets of these constants are used depending on the temperature level, the levels being 3001000 K and 10005000 K. The model also requires the relative pressure for the air-water vapour mixture. This is obtained by tting 4th order polynomials to the required data of the air and water vapour with dierent coecients for three temperature levels: less than 500 K, 5001000 K and 10001500 K. Liquid water : Two thermodynamic properties of liquid water are required in the model, which are the enthalpy (h) and saturation pressure at a given temperature (Psat ). The enthalpy of water at the injection temperature is approximated by that of saturated liquid at the given temperature (hf ), which is obtained from the following relationship: h hf = hsteam hfg , [kJ kg1 ], (10a)

where hsteam is the value of enthalpy for water vapour as obtained from Eq. (9) and hfg is the latent heat of vaporisation [kJ kg1 ] at the given temperature obtained from the following third-order polynomial: hfg = a + b T + c T 2 + d T 3 , [kJ kg1 ], (10b)

in which T is in C and values of the coecients a, b, c, and d are obtained by tting a polynomial to the tabulated values of hfg for water [12]. The respective values of these constants are 2501.3687, 2.3680605, 0.00056653491 and 1.3212214 105 . The polynomial is applicable in the range of temperatures 5100 C. The saturation pressure of water (Psat ) at a given temperature is obtained from the following relation [12]: ln(Psat ) = 70.4346943 7362.6981 + 0.006952085T 9.0ln(T ), T (10c)

where T is in Kelvin and Psat is in atmospheres. The relationship is applicable for temperatures in the range 0200 C. 3.6 T h e i t e r a t i v e s o l u t i o n p r o c e d u r e In all the thermodynamic property relations given above, the properties are functions of the temperature. When the value of a given property (say h) is known and the corresponding temperature is to be determined, an inverse solution of these

200

STROJN ICKY CASOPIS, 59, 2008, . 4 c

relations is required which the present computer program obtains by an iterative procedure. This iterative solution is particularly needed since the working uid in the present model is a mixture of gases and not simply dry air. First, a value of the temperature (Tguess ) is assumed, which must be lower than the expected temperature (if a better guess cannot be made, this initial guess is made equal to 273 K, or 0 C). The calculated value of the required property at the guessed temperature (say hguess ) is then compared with the known value of the property (hknown ). Since the thermodynamic properties increase with temperature, the calculated property value will be lower than the required value. The temperature is then increased by a reasonably large increment T (typically 10 K) and hguess is recalculated and compared with hknown . If hguess is still less than hknown , the temperature is increased again by T . This is repeated until hguess exceeds hknown . Once this happens, the increment T is reduced (typically by a factor of 10) and the last step of the iterative solution is repeated using the smaller T . The increment T is reduced until a prescribed criterion = (hguess hknown )/hknown is met at which the value of Tguess is taken as the required solution for T . A similar iterative solution is applied when the rate of fuel combustion is xed and the combustion temperature is to be determined from the energy balance equation (Eq. 2).

4. Validation of the model In order to verify the computer program with its property relationships, it was initially tested against the solutions given by Bathie [13] for the following two-shaft gas turbine: Inlet air pressure = 101.35 kPa Inlet air temperature = 15 C Inlet air humidity = 0.0 % (dry air) Compressor pressure ratio = 12 : 1 Fuel = n-octane (C8 H18 ) Fuel mass ow rate at rated load = 0.0214773 kg kg1 dry air Flue gas temperature at generator-turbine inlet (at rated load) = 1400 C Flue gas temperature at power-turbine exit (at rated load) = 497.63 C Compressor eciency = 87 % Gas-generator and power turbines eciency = 89 % Bathie [13] analysed the cycle of the above gas turbine with and without steam injection. Table 1 shows seven primary quantities in the cycle without steam injection as computed by the present computer model compared to their respective values obtained from Bathie [13]. The table shows three solutions for the present model (M0, M1 and M2), which represent the dierent options given by the computer program for modelling the working uid as discussed in section 3.4. Bathie

 STROJN ICKY CASOPIS, 59, 2008, . 4 c

201

T a b l e 1. Comparison of the main cycle parameters as computed by the present computer program with those given by Bathie [13] Key Parameters Compressor work [kJ kg1 ] T2 [K] T4 [K] P4 [kPa] T5 [K] Power turbine work [kJ kg1 ] Thermal eciency [%] SA 343.0 623.0 1109.0 408.2 815.0 333.0 37.7 Bathie [13] AE AF 343.0 343.0 623.0 623.0 1115.8 1122.2 418.5 422.8 815.0 831.1 347.5 355.6 36.4 37.2 Present model M0 M1 M2 344.78 344.79 344.79 622.92 622.90 622.90 1114.50 1114.50 1123.70 419.50 419.49 432.85 814.31 814.30 818.67 346.14 346.15 362.69 36.30 36.30 38.03

[13] also gave three solutions using standard-air (SA), air-equivalent (AE) and actual uid (AF) assumptions. As the table shows, the values of the two parameters prior to combustion (the compressor work and T2 ) as obtained by the present model are close to one another and agree well with the three solutions given by Bathie [13] who also used variable specic heats but did not take into consideration the eect of air humidity. Therefore, the present agreement between all six solutions for the rst two parameters is anticipated since the inlet air humidity is zero and the inlet air is assumed to be dry. The dierent solution methods gave dierent solutions for the ve parameters after combustion. The gures on the table show that both Model 0 (air-equivalent option) and Model 1 of the present model yield values for the ve parameters, which are close to the air-equivalent (AE) solution given by Bathie [13]. The agreement between Model 0 and the AE solution of Bathie [13] is anticipated but the agreement with Model 1 is again due to the assumption used in the above case with zero inlet air humidity. It should be expected that, for the general case of humid inlet air, the results obtained by Model 0 and Model 1 will be dierent. The values obtained by Model 2 are signicantly dierent from those of Models 0 and 1, giving higher values for T4 , P4 , T5 , the power turbine work and the thermal eciency. This option is closest to the actual fuel (AF) solution of Bathie [13]. Also, note the dierence between the standard air analysis (SA) of Bathie [13] and all other solutions. Bathie [13] analysed the gas turbine cycle with steam injection at a temperature of 380 C and a rate of 2.5 % to that of the dry air ow. For a dry air mass-ow rate of 1 kg s1 , his values for the power and thermal eciency were 375.4 kW and 37.5 %, respectively. The corresponding values obtained by Model 1 of the present model, which is similar to his AE method, are 377.2 kW and 37.7 %. The values obtained by the present Model 2 for the power and thermal eciency are 394.6 kW and 39.4 %, respectively. Figure 2a,b compares the model estimation for the power and heat rate of the gas turbine at dierent inlet-air temperatures with other estimates obtained from

202

STROJN ICKY CASOPIS, 59, 2008, . 4 c

Fig. 2. Eect of ambient temperature on the performance of gas turbines: (a) power, (b) heat rate.

the literature [3, 7, 14]. Punwani [3] provided estimates for both industrial and aeroderivative gas turbines. As can be seen from the gure, the model estimate of the turbine power agrees well with those given by all three sources for industrial gas turbines. Although the model estimate for the heat rate is also comparable to the data of Punwani [3], it slightly underestimates that of the General Electric PG6581B model [14] at high inlet air temperatures. Figure 2 clearly shows that the model is suitable for industrial gas turbines rather than aeroderivative ones. Since the eect of inlet-air humidity on the turbine power and heat rate is less signicant than that of air temperature, the dry-air solutions shown on Fig. 2 are practically useful for humid air as well as dry air conditions. Also, since the eect of inlet-air temperature on the turbine power is more signicant than that on the

 STROJN ICKY CASOPIS, 59, 2008, . 4 c

203

heat rate, an inlet-air cooling system will pay for itself by the increased megawatts more than by the reduced fuel consumption of the gas turbine. 5. Conclusions Verication of the present computer-based thermodynamic model against relevant published data shows that the model estimates for the selected key cycle parameters compare well with these data. The eect of inlet-air cooling and steam injection on the power and heat rate of the gas turbine is also predicted accurately by the model. Therefore, the present model can be used to study the eect of evaporative or refrigerative inlet air cooling and water/steam injection on the gas turbine cycle and to analyse the eect of dierent combinations of these power augmentation methods on the gas turbine performance. The model can also be extended to calculate the expected annual revenues that result from the increased megawatts and reduced fuel consumption from inlet air cooling and/or water/steam injection.
Acknowledgements The author acknowledges the contribution of Phoon H. Y., a former student of Universiti Tenaga Nasional (UNITEN), Malaysia, to the initial development and verication of the present computer model.

REFERENCES [1] CORTES, C. R.WILLEMS, D. F.: In: Proc. of POWER-GEN International Conference. Tulsa, PennWell Corporation 2003. [2] MCCRACKEN, C. D.: ASHRAE J., 33, 1991, p. 12. [3] PUNWANI, D.: Energy-Tech Magazine, Dec 2003, p. 20 (see also internet: http://www.energy-tech.com/) [4] HASNAIN, S. M.ALAWAJI, S. H. AL-IBRAHIM, A.SMIAI, M. S.: International Journal of Energy Research, 23, 1999, p. 117. [5] YONG, V. Y.: In: Proc. of 7th annual POWER-GEN Asia 1999. Volume 1. Ed.: Gose, R. Tulsa, PennWell Corporation 1999, p. 187. [6] FARMER, R.: Gas Turbine World. March-April issue, 1999, p. 35. [7] Premier Industries, Inc., 2006, http://www.ctiac.com/turbine1.htm. [8] BOYCE, M. P.: A seminar held at Tenaga National Research & Development Centre (TNRD), Malaysia 1999. [9] RICE, I. G.: ASME Transactions, Journal of Engineering for Gas Turbines and Power, 117, 1995, p. 347. [10] TUZSON, J.: ASME Transactions, Journal of Engineering for Gas Turbines and Power, 114, 1992, p. 682. [11] BROOKS, F. J.: GE Gas Turbine Performance Characteristics. GE Power Systems, Schenectady, NY, GER-3567H, 2008. http://www.gepower.com/prod serv/products/tech docs/en/downloads/ger3567h.pdf

204

STROJN ICKY CASOPIS, 59, 2008, . 4 c

[12] MORAN, M. J.SHAPIRO, H. N.: Fundamentals of Engineering Thermodynamics. 5th edition. New York, John Wiley & Sons 2004. [13] BATHIE, W. W.: Fundamentals of Gas Turbines. 2th edition. New York, John Wiley & Sons 1996. [14] Garri Power Station Manuals. Private communication. Received: 12.3.2007 Revised: 23.9.2008