**2.2 Gas turbine configurations with internal heat recovery**

Compared to simple cycle gas turbines, the higher costs of constructing a combined cycle plant are not always offset by higher efficiency, especially for small and medium size plants. The need to combine the high efficiency of combined cycles with the low cost of simple cycles has raised the interest in new technologies that enable internal waste heat recovery from the gas turbine.

Internal heat can be recovered through the working fluid (fuel, air) or an auxiliary fluid (usually water). In the first case internal heat recovery is defined as "direct", in the second as "indirect"[5].

Thermodynamic regeneration is a direct internal recovery technique, since thermal energy is transferred directly from exhaust gas to air at the compressor exit. This produces an efficiency gain due to the reduction in primary thermal energy requirements without changing, as a first approximation, the mechanical power output.

Steam injection on the other hand is an indirect internal recovery technique. In this case the recovered thermal energy is transferred to an auxiliary fluid (water), which is then injected into the combustor. This increases the primary thermal energy required to keep the temperature at the turbine inlet constant, but results in a power increase and, consequently, an efficiency gain.

Direct and indirect recovery can also be combined, as for instance in HAT and CRGT cycles. In humid air plants (HAT) , the saturation of air at compressor exit extends the regeneration margins, thanks to the greater temperature difference between the exhaust gas at turbine exit and the compressed air at regenerator inlet. In chemically recuperated plants (CRGT), exhaust heat is recovered through an endothermic steam-reforming process of the primary fuel. More specifically, a portion of the recovered heat is transferred directly to the fuel, while the remainder is used to produce the required steam.

### **2.2.1 Heat recovery without auxiliary fluid**

In thermodynamic regeneration, the exhaust heat at the turbine exit is used to preheat the air entering the combustion chamber. The heat exchange between the two gas streams is achieved by means of a countercurrent heat exchanger, known as a regenerator or recuperator. Figure 3 shows a schematic diagram of the regenerative cycle.

The thermal efficiency of the Brayton cycle is enhanced since regeneration decreases the heat input required to produce the same net work output. Heat recovery through a gas-to-gas heat exchanger is limited by a characteristic value of the compression ratio, beyond which the temperature of the exhaust gas falls below that of the air at the compressor outlet, thereby deteriorating efficiency.

The efficiency gain achieved through regeneration strongly depends on the heat exchanger effectiveness, defined as the ratio of the actual heat transfer rate to the air and the maximum possible heat transfer rate, that would exist were the heat exchanger to have infinite heat transfer surface area. More specifically, gas turbine efficiency increases with heat exchanger effectiveness, as the air at the combustion chamber inlet is preheated at higher temperatures, resulting in greater fuel savings.

The Recovery of Exhaust Heat from Gas Turbines 171

The steam pressure should be sufficient to enable injection into the combustor. Typical values of this parameter range from 1.25 to 1.4 times the maximum pressure of the cycle. In addition, the water used for steam production must be demineralized to minimise salts and oxides content so as to prevent fouling the turbomachinery or chemical attack at high

A practical concern with steam injection is water consumption, that typically ranges from 1.1 to 1.6 kg of high purity water per kWh of electrical output. The water purification system required for large scale plant would represent about 5% of total capital costs, whereas

As the temperature of injected steam must be raised to combustion chamber temperature, a small temperature difference at the approach point (i.e. at boiler outlet) can be advantageous. However, since the evaporation process takes place at constant temperature,

Since the high steam temperature requirements conflict with large heat recoveries, there is a

Efficiency can be improved using multi-pressure systems whereby the water vapour temperature profile matches that of the exhaust gases more closely, resulting in a better approximation to a reversible process [1]. However, this complicates the design, while plant

As shown in Figure 5, humid air turbines (HAT) combine regeneration and steam injection. Compared to a traditional regenerative gas turbine, HAT cycle requires the addition of a surface heat exchanger and a saturator. The air at the compressor outlet passes through the heat exchanger, where liquid water is preheated; then air passes through the saturator, where it mixes with the steam, undergoing at the same time a reduction in temperature, as water evaporation absorbs latent heat from the gas stream. The saturation of air at compressor exit extends the regeneration margins, thanks to the greater temperature difference between the exhaust gas exiting the turbine and the compressed air at the regenerator inlet. Moreover humidification reduces the heat capacity difference between air

and exhaust gas, resulting in increased efficiency of the regenerator heat recovery [2].

high steam temperatures are necessarily accompanied by low heat recovery.

limit to the maximum attainable efficiency with STIG cycles.

simplicity is one of the strengths of this technology.

temperatures.

running costs add about 5% to the fuel cost [12].

Fig. 4. Schematic diagram of STIG cycle

However, to achieve greater effectiveness requires a larger heat transfer area. This translates into higher capital costs and larger pressure drops on both air and gas sides of the heat exchanger, which reduce the turbine pressure ratio and therefore the turbine work.

Generally, the air pressure drop on the high-pressure side should be kept below 2% of the total compressor discharge pressure. The effectiveness of most regenerators used in practice is below 0.85.

Fig. 3. Schematic diagram of the regenerative cycle

The regenerator takes up a very large area because of the low heat transfer coefficients of air and exhaust gases. This results in high costs and loss of compactness, affecting the main features of gas turbines. In addition, the exhaust gas flow in the regenerator leads to the formation of carbon deposits, resulting in a reduction of heat transfer coefficient, which becomes more pronounced over time.

The regenerator has met with little success in the industrial gas turbines sector, because of the low efficiency gain accompanied by the significantly higher capital costs.

Only in recent years this option has started to make a comeback, as it can be advantageously integrated with other technologies, such as steam injection.

Moreover regeneration is successfully applied in micro - gas turbines, where the low compression ratios are related to the simplicity of turbomachines.

## **2.2.2 Heat recovery through steam generation**

In the steam injection cycle (STIG), proposed by Cheng in 1978 [10], the gases exhausting the turbine are used to produce steam in a heat recovery steam generator, that is then injected into the combustion chamber (Figure 4).

In addition to drastically reducing the formation of nitrogen oxides, steam injection increases both efficiency and power output [11]. The efficiency gain is about 10%, lower than that obtained with a conventional combined cycle, because the steam expands in a less efficient manner in the gas turbine than in the steam turbine. On the other hand, the power increase varies between 50 and 70% [12].

However, to achieve greater effectiveness requires a larger heat transfer area. This translates into higher capital costs and larger pressure drops on both air and gas sides of the heat

Generally, the air pressure drop on the high-pressure side should be kept below 2% of the total compressor discharge pressure. The effectiveness of most regenerators used in practice

The regenerator takes up a very large area because of the low heat transfer coefficients of air and exhaust gases. This results in high costs and loss of compactness, affecting the main features of gas turbines. In addition, the exhaust gas flow in the regenerator leads to the formation of carbon deposits, resulting in a reduction of heat transfer coefficient, which

The regenerator has met with little success in the industrial gas turbines sector, because of

Only in recent years this option has started to make a comeback, as it can be advantageously

Moreover regeneration is successfully applied in micro - gas turbines, where the low

In the steam injection cycle (STIG), proposed by Cheng in 1978 [10], the gases exhausting the turbine are used to produce steam in a heat recovery steam generator, that is then injected

In addition to drastically reducing the formation of nitrogen oxides, steam injection increases both efficiency and power output [11]. The efficiency gain is about 10%, lower than that obtained with a conventional combined cycle, because the steam expands in a less efficient manner in the gas turbine than in the steam turbine. On the other hand, the power

the low efficiency gain accompanied by the significantly higher capital costs.

exchanger, which reduce the turbine pressure ratio and therefore the turbine work.

Fig. 3. Schematic diagram of the regenerative cycle

integrated with other technologies, such as steam injection.

**2.2.2 Heat recovery through steam generation** 

into the combustion chamber (Figure 4).

increase varies between 50 and 70% [12].

compression ratios are related to the simplicity of turbomachines.

becomes more pronounced over time.

is below 0.85.

The steam pressure should be sufficient to enable injection into the combustor. Typical values of this parameter range from 1.25 to 1.4 times the maximum pressure of the cycle. In addition, the water used for steam production must be demineralized to minimise salts and oxides content so as to prevent fouling the turbomachinery or chemical attack at high temperatures.

A practical concern with steam injection is water consumption, that typically ranges from 1.1 to 1.6 kg of high purity water per kWh of electrical output. The water purification system required for large scale plant would represent about 5% of total capital costs, whereas running costs add about 5% to the fuel cost [12].

Fig. 4. Schematic diagram of STIG cycle

As the temperature of injected steam must be raised to combustion chamber temperature, a small temperature difference at the approach point (i.e. at boiler outlet) can be advantageous. However, since the evaporation process takes place at constant temperature, high steam temperatures are necessarily accompanied by low heat recovery.

Since the high steam temperature requirements conflict with large heat recoveries, there is a limit to the maximum attainable efficiency with STIG cycles.

Efficiency can be improved using multi-pressure systems whereby the water vapour temperature profile matches that of the exhaust gases more closely, resulting in a better approximation to a reversible process [1]. However, this complicates the design, while plant simplicity is one of the strengths of this technology.

As shown in Figure 5, humid air turbines (HAT) combine regeneration and steam injection. Compared to a traditional regenerative gas turbine, HAT cycle requires the addition of a surface heat exchanger and a saturator. The air at the compressor outlet passes through the heat exchanger, where liquid water is preheated; then air passes through the saturator, where it mixes with the steam, undergoing at the same time a reduction in temperature, as water evaporation absorbs latent heat from the gas stream. The saturation of air at compressor exit extends the regeneration margins, thanks to the greater temperature difference between the exhaust gas exiting the turbine and the compressed air at the regenerator inlet. Moreover humidification reduces the heat capacity difference between air and exhaust gas, resulting in increased efficiency of the regenerator heat recovery [2].

The Recovery of Exhaust Heat from Gas Turbines 173

energy recovered at a relatively low temperature is made available at a higher temperature,

Moreover CRGT cycles have ultra low NOx emissions, due to the large amount of steam contained in the reformed fuel, lowering the temperature in the primary zone of the

Recently there has been a revived interest in this technology, in view of the potential use of

The different waste heat recovery techniques can be analysed using a unified thermodynamic approach. Heat recovery influences gas turbine performance by means of two concurrent effects. The first is related to the variation of primary thermal energy (*ΔQ1*) supplied from the outside, the second to the increase in power (*ΔP*) produced by the

> \* \* 1 1

+ Δ <sup>=</sup> + Δ (1)

Δ =Δ −Δ *QQ Q* 1 1,*AF DR* 1, (2)

*P P Q Q*

1 / \* 1 /

Neglecting water pumping power and pressure losses of heat recovery devices, the term *ΔP* is only related to the auxiliary fluid expanding through the gas turbine. The term *ΔQ1*

The first term *ΔQ1*,AF is the additional primary thermal energy required to attain the maximum cycle temperature when an auxiliary fluid is introduced. The second term *ΔQ1,DR* refers to the heat reintroduced into the cycle, on the air or fuel side, through direct recovery. Equation (1) can be taken as a basic relation for characterizing the capabilities of internal and

For direct external heat recovery, i.e. in the case of combined gas-steam cycles, the term *ΔP* refers to the steam turbine power output and *ΔQ1* is positive only when supplementary

For direct internal heat recovery, such as thermodynamic regeneration or dry chemical recovery, the power remains practically unchanged (*ΔP*=0) and the term *ΔQ1,AF* is nil. Therefore the efficiency increase is only due to the reduction in the primary thermal energy supplied (*ΔQ1= -ΔQ1,DR*<0). For thermodynamic regeneration *ΔQ1,DR* can be immediately interpreted; for dry chemical recovery it represents the energy required by the endothermic decomposition process, recovered from the exhaust gases and transferred to the reformed

η

η

during oxidation of the reformed fuel.

strategic fuels, such as methanol.

auxiliary fluid, where present.

**3. A unified thermodynamic approach** 

**3.1 Equations for exhaust heat recovery** 

generally comprises two contributions

external heat recovery of gas turbines.

fuel, by increasing the heating value.

firing is performed.

Denoting with *η* overall plant efficiency, we can write

where *η\**, *P\** and *Q1\** refer to the non-recovery baseline plant.

combustor.

HAT cycles overcome the intrinsic limits of steam injection, further enhancing efficiency.

The main issue of humid air turbines is the difficulty in containing the pressure drops related to the compressed moist air flow through the saturator and the regenerator. The saturator can operate with any clean and filtered water source, as long as the dissolved substances at the water outlet remain below their precipitation concentration under operating conditions [13]. Water consumption is a problem as for steam injection cycles but the consumption rate is only about one third [1].

Fig. 5. Schematic diagram of HAT cycle

The chemically recuperated gas turbine (CRGT) is an extension of the steam-injected gas turbine concept [14]. As shown in Fig. 6, the thermal energy available in the exhaust gases is used to promote an endothermic reaction in the primary fuel, that can occur with or without water addition. In the first case the process is called steam reforming, whereas in the second simple decomposition. The process requires the presence of a nickel based catalyst and results in the production of a reformed fuel, composed of CO, CO2, H2, excess steam and unconverted fuel, which is fed directly to the combustion chamber.

Fig. 6. Schematic diagram of CRGT cycle

The reformed fuel absorbs heat thermally and chemically, resulting in a potentially greater recovery of waste heat than conventional recovery techniques [14]. In fact the thermal

The main issue of humid air turbines is the difficulty in containing the pressure drops related to the compressed moist air flow through the saturator and the regenerator. The saturator can operate with any clean and filtered water source, as long as the dissolved substances at the water outlet remain below their precipitation concentration under operating conditions [13]. Water consumption is a problem as for steam injection cycles but

The chemically recuperated gas turbine (CRGT) is an extension of the steam-injected gas turbine concept [14]. As shown in Fig. 6, the thermal energy available in the exhaust gases is used to promote an endothermic reaction in the primary fuel, that can occur with or without water addition. In the first case the process is called steam reforming, whereas in the second simple decomposition. The process requires the presence of a nickel based catalyst and results in the production of a reformed fuel, composed of CO, CO2, H2, excess steam and

The reformed fuel absorbs heat thermally and chemically, resulting in a potentially greater recovery of waste heat than conventional recovery techniques [14]. In fact the thermal

unconverted fuel, which is fed directly to the combustion chamber.

HAT cycles overcome the intrinsic limits of steam injection, further enhancing efficiency.

the consumption rate is only about one third [1].

Fig. 5. Schematic diagram of HAT cycle

Fig. 6. Schematic diagram of CRGT cycle

energy recovered at a relatively low temperature is made available at a higher temperature, during oxidation of the reformed fuel.

Moreover CRGT cycles have ultra low NOx emissions, due to the large amount of steam contained in the reformed fuel, lowering the temperature in the primary zone of the combustor.

Recently there has been a revived interest in this technology, in view of the potential use of strategic fuels, such as methanol.

### **3. A unified thermodynamic approach**

The different waste heat recovery techniques can be analysed using a unified thermodynamic approach. Heat recovery influences gas turbine performance by means of two concurrent effects. The first is related to the variation of primary thermal energy (*ΔQ1*) supplied from the outside, the second to the increase in power (*ΔP*) produced by the auxiliary fluid, where present.

#### **3.1 Equations for exhaust heat recovery**

Denoting with *η* overall plant efficiency, we can write

$$\frac{\eta}{\eta \, ^\ast} = \frac{1 + \Delta P \, ^\ast P \, ^\ast}{1 + \Delta Q\_1 \, ^\ast Q\_1 ^\ast} \tag{1}$$

where *η\**, *P\** and *Q1\** refer to the non-recovery baseline plant.

Neglecting water pumping power and pressure losses of heat recovery devices, the term *ΔP* is only related to the auxiliary fluid expanding through the gas turbine. The term *ΔQ1* generally comprises two contributions

$$
\Delta Q\_1 = \Delta Q\_{1,AF} - \Delta Q\_{1,DR} \tag{2}
$$

The first term *ΔQ1*,AF is the additional primary thermal energy required to attain the maximum cycle temperature when an auxiliary fluid is introduced. The second term *ΔQ1,DR* refers to the heat reintroduced into the cycle, on the air or fuel side, through direct recovery.

Equation (1) can be taken as a basic relation for characterizing the capabilities of internal and external heat recovery of gas turbines.

For direct external heat recovery, i.e. in the case of combined gas-steam cycles, the term *ΔP* refers to the steam turbine power output and *ΔQ1* is positive only when supplementary firing is performed.

For direct internal heat recovery, such as thermodynamic regeneration or dry chemical recovery, the power remains practically unchanged (*ΔP*=0) and the term *ΔQ1,AF* is nil. Therefore the efficiency increase is only due to the reduction in the primary thermal energy supplied (*ΔQ1= -ΔQ1,DR*<0). For thermodynamic regeneration *ΔQ1,DR* can be immediately interpreted; for dry chemical recovery it represents the energy required by the endothermic decomposition process, recovered from the exhaust gases and transferred to the reformed fuel, by increasing the heating value.

The Recovery of Exhaust Heat from Gas Turbines 175

Equation (5) defines the slope of curves with constant *ξ*, at each value of *ηAF* related to the

<sup>t</sup> =TFG/TFG <sup>x</sup> =0,4 \*

p =0,6

c =DQ1/Q1 \*

p =DP/P\* x =DQ1,DR/Q1

p =0,8

**A**

t

\*

The curve with the maximum slope (i.e. maximum *ηAF* value) is obtained at the maximum auxiliary fluid temperature, that occurs, in the case of steam injection, inside the combustor at the maximum degree of superheat permitted by the steam generator, as well as for steam reforming at the maximum temperature allowable by the exhaust gas at the turbine exit.

x =0,1

p =0,4

x =0,2


**O**

x =0,0

Decreasing *ηAF*, that is for a lower enthalpy of the auxiliary fluid introduced into the combustor, the slope of curves at constant *ξ* diminishes eventually becoming negative in the

Figure 8(a), for instance, shows two *ηAF* curves for steam injected at the maximum degree of superheat (*ηAF,1*) and under saturated conditions (*ηAF,2*). For a given plant with no heat recovery (*P\**, *η\**), a generic point Q on the characteristic plane, as shown in Figure 8(b), may represent different internal heat recovery techniques characterized by various combinations of direct and indirect recovery (*ξ1*< *ξ2* and *ηAF,1*> *ηAF,2*), based on plant configurations defined

The constant *τ* curves indicate the extent of the recovery. For a fixed value of *τ*, the maximum efficiency increase is obtained for *π*=0. Instead, when an auxiliary fluid is introduced, at constant *τ*, the efficiency increase is lower due to the unrecoverable latent

Assuming a limit value for the flue gas temperature, the corresponding curve, together with the curves at *π*=0 and *ξ*=0 define a characteristic region (OAC in Fig. 7) which represents the possible recovery conditions. Each point inside this region does not represent a specific

thermodynamic conditions of auxiliary fluid at the combustor inlet.

p =0,2

p =0,0

**C**

x =0,3

Fig. 7. A performance plane for exhaust heat recovery

0.8

1.0

1.2

1.4

η/η\*

1.6

1.8

2.0

case of water injection.

by different value of *ξ* and *ηAF*.

heat of steam at the turbine exit.

For indirect internal heat recovery, such as steam injection, the introduction of an auxiliary fluid increases power output (*ΔP* >0) and primary thermal energy (*ΔQ1= ΔQ1,AF*>0). Efficiency increases only if

$$
\eta\_{\rm FA} = \frac{\Delta P}{\Delta Q\_{1, \rm AF}} > \eta^\* \tag{3}
$$

where the term on the left hand side can be interpreted as the marginal efficiency of the auxiliary fluid. State of the art gas turbine technology satisfies this condition for steam injection, even for low degrees of superheat, but not for water injection, that produces a power increase with an efficiency penalty.

In the case of combined (direct-indirect) internal heat recovery, such as humid air regeneration or steam reforming, efficiency increases as a result of two effects. The first refers to the power output increase (*ΔP),* the second to the primary energy variation (*ΔQ1*), that can be negative or positive in accordance with Eq. (2). Since *ΔP* and *ΔQ1,AF* are proportional to the mass flow rate of the auxiliary fluid, for a given value thereof, the efficiency gains are greater the more *ΔQ1,DR* increases.

### **3.2 A performance plane for exhaust heat recovery**

Using Eq. (1) it is possible to define a characteristic plane, that allows to compare different techniques for recovering exhaust heat from gas turbines, highlighting their application limits.

As shown in Figure 7, the performance plane of waste heat recovery indicates the trend of the ratio *η/η\** as a function of *χ=ΔQ1/Q1\** and *π=ΔP/P\**. The relation between *η/η\**, *χ* and *π* does not depend on gas turbine characteristics, which are instead introduced by two other families of curves. These define the conditions for constant values of the direct recovery parameter, defined as *ξ=ΔQ1,DR/Q1\**, and for those of the non-dimensional flue gas temperature, defined as *τ=TFG/TFG\** [5]. From Eqs. (2) and (3), we can derive a relationship among different non-dimensional parameters of internal heat recovery

$$
\mathcal{X} = \pi \left( \eta^\* / \eta\_{\text{AF}} \right) - \xi \tag{4}
$$

This establishes, for a given gas turbine (*η\**) and recovery technique (*ξ* and *ηAF*), the relationship between *χ* and *π*.

For *π*=0 – simple direct recovery – from Eq. (4) we get *χ=-ξ*; therefore, each point on *π*=0 curve of Fig. 11 is characterized by a different *ξ* value.

Each point P on this curve defines an envelope of curves at constant *ξ* , but characterized by different *ηAF* values. Combining Eqs. (1) and (4), we get

$$\left. \frac{\partial \left( \eta \,/ \, \eta^\* \right)}{\partial \mathcal{X}} \right|\_{\xi = \text{cost}} = \frac{1}{\left( 1 + \mathcal{X} \right)^2} \left[ \frac{\eta\_{\,\_{AF}} \left( 1 - \xi \right)}{\eta^\*} - 1 \right] \tag{5}$$

For indirect internal heat recovery, such as steam injection, the introduction of an auxiliary fluid increases power output (*ΔP* >0) and primary thermal energy (*ΔQ1= ΔQ1,AF*>0).

> *AF P Q*

1,

where the term on the left hand side can be interpreted as the marginal efficiency of the auxiliary fluid. State of the art gas turbine technology satisfies this condition for steam injection, even for low degrees of superheat, but not for water injection, that produces a

In the case of combined (direct-indirect) internal heat recovery, such as humid air regeneration or steam reforming, efficiency increases as a result of two effects. The first refers to the power output increase (*ΔP),* the second to the primary energy variation (*ΔQ1*), that can be negative or positive in accordance with Eq. (2). Since *ΔP* and *ΔQ1,AF* are proportional to the mass flow rate of the auxiliary fluid, for a given value thereof, the

Using Eq. (1) it is possible to define a characteristic plane, that allows to compare different techniques for recovering exhaust heat from gas turbines, highlighting their application

As shown in Figure 7, the performance plane of waste heat recovery indicates the trend of the ratio *η/η\** as a function of *χ=ΔQ1/Q1\** and *π=ΔP/P\**. The relation between *η/η\**, *χ* and *π* does not depend on gas turbine characteristics, which are instead introduced by two other families of curves. These define the conditions for constant values of the direct recovery parameter, defined as *ξ=ΔQ1,DR/Q1\**, and for those of the non-dimensional flue gas temperature, defined as *τ=TFG/TFG\** [5]. From Eqs. (2) and (3), we can derive a relationship

( ) \*

This establishes, for a given gas turbine (*η\**) and recovery technique (*ξ* and *ηAF*), the

For *π*=0 – simple direct recovery – from Eq. (4) we get *χ=-ξ*; therefore, each point on *π*=0

Each point P on this curve defines an envelope of curves at constant *ξ* , but characterized by

( ) \*

χ 2 \*

η

*AF*

η 1

 ξ

( )

<sup>∂</sup> <sup>−</sup> = − <sup>∂</sup> <sup>+</sup>

/ 1 1

1

ξ

(4)

(5)

πη η*AF* = −

among different non-dimensional parameters of internal heat recovery

χ

cos

ξ*t*

=

Δ = >

\*

<sup>Δ</sup> (3)

η

*FA*

η

Efficiency increases only if

limits.

relationship between *χ* and *π*.

power increase with an efficiency penalty.

efficiency gains are greater the more *ΔQ1,DR* increases.

**3.2 A performance plane for exhaust heat recovery** 

curve of Fig. 11 is characterized by a different *ξ* value.

different *ηAF* values. Combining Eqs. (1) and (4), we get

η η

( )

χ Equation (5) defines the slope of curves with constant *ξ*, at each value of *ηAF* related to the thermodynamic conditions of auxiliary fluid at the combustor inlet.

Fig. 7. A performance plane for exhaust heat recovery

The curve with the maximum slope (i.e. maximum *ηAF* value) is obtained at the maximum auxiliary fluid temperature, that occurs, in the case of steam injection, inside the combustor at the maximum degree of superheat permitted by the steam generator, as well as for steam reforming at the maximum temperature allowable by the exhaust gas at the turbine exit.

Decreasing *ηAF*, that is for a lower enthalpy of the auxiliary fluid introduced into the combustor, the slope of curves at constant *ξ* diminishes eventually becoming negative in the case of water injection.

Figure 8(a), for instance, shows two *ηAF* curves for steam injected at the maximum degree of superheat (*ηAF,1*) and under saturated conditions (*ηAF,2*). For a given plant with no heat recovery (*P\**, *η\**), a generic point Q on the characteristic plane, as shown in Figure 8(b), may represent different internal heat recovery techniques characterized by various combinations of direct and indirect recovery (*ξ1*< *ξ2* and *ηAF,1*> *ηAF,2*), based on plant configurations defined by different value of *ξ* and *ηAF*.

The constant *τ* curves indicate the extent of the recovery. For a fixed value of *τ*, the maximum efficiency increase is obtained for *π*=0. Instead, when an auxiliary fluid is introduced, at constant *τ*, the efficiency increase is lower due to the unrecoverable latent heat of steam at the turbine exit.

Assuming a limit value for the flue gas temperature, the corresponding curve, together with the curves at *π*=0 and *ξ*=0 define a characteristic region (OAC in Fig. 7) which represents the possible recovery conditions. Each point inside this region does not represent a specific

The Recovery of Exhaust Heat from Gas Turbines 177

Having defined the non-recovery gas turbine, the construction of this plane requires the evaluation of performance increases achieved by different heat recovery configurations compared to the baseline gas turbine. For this purpose the General Electric software GateCycle has been used [15]. Using this modelling tool specific plant configurations have been developed to simulate the non-recovery baseline simple cycle (SC), the regenerative cycle (RG), the steam injected cycle (SI), the regenerative steam injected cycle (RG+SI), the humid air regenerative cycle (HAT) and the chemically recuperated cycle (CRGT). In all cases with exhaust heat recovery, combustor and turbomachinery design data are taken to be the same as for the corresponding simple non-recovery cycle, while operating data for

In order to evaluate the influence of pressure ratio and turbine inlet temperature on capabilities of different recovery techniques, the characteristic plane of heat recovery has been defined with reference to four non-recovery gas turbines, that differ in terms of

Referring to these characteristic planes, represented in Figure 9, the different internal and external heat recovery techniques will be discussed in more detail in the following subsections, highlighting the influence of pressure ratio and temperature inlet temperature

<sup>π</sup> = 0,6 <sup>π</sup> = 0,9

<sup>π</sup> = 0,6 <sup>π</sup> = 0,9

**B**

**A**


χ

π = 0,3

**A**

**I**

**K**

π = 0,3


χ

**D**

**G E**

**D**

**I**

**G E <sup>J</sup> <sup>K</sup> <sup>H</sup>**

**O**

**J**

**H**

**O C**

**F**

**F**

**C**

π = 1,2

*TIT* =1200°C β=20 η\*=39,6%

π = 1,2

π = 1,5

0.9 1.0 1.1 1.2 1.3 1.4 1.5

*η/η\**

π = 1,5

0.9 1.0 1.1 1.2 1.3 1.4 1.5

*η/η\**

**B**

*TIT* =1300°C β=20 η\*=40,2%

Fig. 9. Characteristic plane of heat recovery for different non-recovery gas turbines *(*Internal heat recovery *- OC : Thermodynamic regeneration, OAB: Steam injection, CD: Regeneration with steam injection, CE: Humid air regeneration (HAT), FG: Steam methane reforming.* External heat recovery *- HI: Combined cycle with one pressure level HRSG, HI:* 

π = 1,5

*Combined cycle with three pressure levels HRSG and reheat.)* 

<sup>π</sup> = 0,6 <sup>π</sup> = 0,9

<sup>π</sup> = 0,6 <sup>π</sup> = 0,9

**A**


χ

π = 0,3

**A**

**I**

π = 0,3


χ

π = 1,2

*TIT* =1200°C β=15 η\*=37,6%

π = 1,2

**B**

**B**

*TIT* =1300°C β=15 η\*=37,9%

π = 1,5

each heat recovery device have been examined over significant ranges.

pressure ratios and turbine inlet temperatures.

on efficiency increase.

**C**

0.9 1.0 1.1 1.2 1.3 1.4 1.5

0.9 1.0 1.1 1.2 1.3 1.4 1.5

*η/η\**

*η/η\**

**O**

**D**

**E**

**G**

**<sup>J</sup> <sup>K</sup> H**

**O**

**F**

**C**

**F**

**D**

**E**

**G**

**<sup>J</sup> <sup>K</sup> <sup>H</sup>** I

plant configuration, since the same performance can be obtained with different heat recovery techniques.

Fig. 8. Influence of marginal auxiliary fluid efficiency on combined recovery capabilities

Points on the curve *π*=0 (curve OC in Fig. 7) therefore indicate plant configurations with direct recovery (*χ=-ξ*) alone, those on the curve OA (*ξ*=0) on the contrary, solutions with indirect recovery (*χ=π η\*/ηAF*) alone and are characterized by positive values of *χ* denoting an increase of primary energy supplied to the cycle.

The plane region above curves OC and OA represents conditions for combined direct and indirect recovery. Negative values of *χ*, denoting a net reduction of primary energy to the cycle, are possible if direct recovery effects predominate over those associated with steam injection. One limitation to the extension of the OAC region is the minimum flue gas temperature attainable inside the stack, also taking into account the characteristics of the single internal heat recovery techniques (regenerator effectiveness, saturation conditions, pinch point at HRSG, steam-to-methane ratio at reformer). A further restriction may also arise from operational problems with existing combustor and turbomachinery, especially if high steam flow rates are injected [5].

## **4. Capabilities of exhaust heat recovery techniques**

The characteristic plane of heat recovery can be used to determine the capabilities of different internal and external heat recovery techniques for a variety of gas turbines.

plant configuration, since the same performance can be obtained with different heat

c <sup>P</sup> 0

(a)

p =const

x1= - c <sup>1</sup>

p =const x=const

h AF,1 h AF,2

**Q**

h AF,2

h AF,1

p =0

p =0

x2= - c <sup>2</sup>

h AF,1>h AF,2

**P**

h AF,1>h AF,2

1

1

c

*c*

Fig. 8. Influence of marginal auxiliary fluid efficiency on combined recovery capabilities

c 2

an increase of primary energy supplied to the cycle.

**4. Capabilities of exhaust heat recovery techniques** 

high steam flow rates are injected [5].

Points on the curve *π*=0 (curve OC in Fig. 7) therefore indicate plant configurations with direct recovery (*χ=-ξ*) alone, those on the curve OA (*ξ*=0) on the contrary, solutions with indirect recovery (*χ=π η\*/ηAF*) alone and are characterized by positive values of *χ* denoting

(b)

c <sup>1</sup> 0

The plane region above curves OC and OA represents conditions for combined direct and indirect recovery. Negative values of *χ*, denoting a net reduction of primary energy to the cycle, are possible if direct recovery effects predominate over those associated with steam injection. One limitation to the extension of the OAC region is the minimum flue gas temperature attainable inside the stack, also taking into account the characteristics of the single internal heat recovery techniques (regenerator effectiveness, saturation conditions, pinch point at HRSG, steam-to-methane ratio at reformer). A further restriction may also arise from operational problems with existing combustor and turbomachinery, especially if

The characteristic plane of heat recovery can be used to determine the capabilities of

different internal and external heat recovery techniques for a variety of gas turbines.

recovery techniques.

Having defined the non-recovery gas turbine, the construction of this plane requires the evaluation of performance increases achieved by different heat recovery configurations compared to the baseline gas turbine. For this purpose the General Electric software GateCycle has been used [15]. Using this modelling tool specific plant configurations have been developed to simulate the non-recovery baseline simple cycle (SC), the regenerative cycle (RG), the steam injected cycle (SI), the regenerative steam injected cycle (RG+SI), the humid air regenerative cycle (HAT) and the chemically recuperated cycle (CRGT). In all cases with exhaust heat recovery, combustor and turbomachinery design data are taken to be the same as for the corresponding simple non-recovery cycle, while operating data for each heat recovery device have been examined over significant ranges.

In order to evaluate the influence of pressure ratio and turbine inlet temperature on capabilities of different recovery techniques, the characteristic plane of heat recovery has been defined with reference to four non-recovery gas turbines, that differ in terms of pressure ratios and turbine inlet temperatures.

Referring to these characteristic planes, represented in Figure 9, the different internal and external heat recovery techniques will be discussed in more detail in the following subsections, highlighting the influence of pressure ratio and temperature inlet temperature on efficiency increase.

Fig. 9. Characteristic plane of heat recovery for different non-recovery gas turbines *(*Internal heat recovery *- OC : Thermodynamic regeneration, OAB: Steam injection, CD: Regeneration with steam injection, CE: Humid air regeneration (HAT), FG: Steam methane reforming.* External heat recovery *- HI: Combined cycle with one pressure level HRSG, HI: Combined cycle with three pressure levels HRSG and reheat.)* 

The Recovery of Exhaust Heat from Gas Turbines 179

is bounded by two curves at *ξ*=0 (OA and OB) and by the curve AB corresponding to the minimum temperature difference at pinch point of the heat recovery steam generator (*ΔTPP*=10°C). As mentioned above, the slope of the curve with *ξ*=0 depends on the gas

In particular the curve OA refers to the case of maximum marginal efficiency *ηAF* obtained by injecting steam superheated to the same temperature as the turbine exhaust, while curve OB refers to the case of minimum *ηAF*, corresponding to injection of saturated steam. The entire region can be covered varying the hot side temperature difference in the superheater

Efficiency gains due to steam injection diminish with steam temperature, while both steam mass flow rate and power produced increase. In practice the power increase is limited by problems associated with the large water requirements and compressor-turbine

To evaluate the influence of *β* and *TTI* on steam injection capabilities, points A and B are examined. In both cases, steam mass flow rate as well as power increase with temperature,

In particular, at point A (superheated steam) the mass flow rate increase passes from a minimum of 20% at low maximum gas temperature (*TTI*=1200°C) and high pressure ratio (*β*=20) to 30% at *TTI*=1300°C and *β*=15. On the other hand, at point B (saturated steam) the mass flow rate increase ranges from 34 to 55%, under the same pressure and turbine inlet

Performances associated with points inside the region OCDA (Fig. 9) can only be obtained considering recovery techniques that combine direct (thermodynamic regeneration) and indirect recovery (steam injection). The effects associated with the auxiliary fluid occur in different ways, with regard to marginal efficiency value (*ηAF*), which depends on the thermodynamic conditions of the auxiliary fluid upstream from the combustion chamber. The region OCDA can be covered varying the hot side temperature difference in the

Heat can also be indirectly recovered using unconventional techniques, such as humid air

The performance of HAT plants is represented by the points on curve CE. Keeping the hot side temperature difference in the regenerator at 40°C, curve CE has been obtained increasing the mass of water introduced into the saturator from zero (point C) to the maximum permissible value for saturation of the compressed air upstream from the regenerator (point E). On this curve the value of *ξ*, defined at point C, remains constant, since this parameter is established by the capabilities of the plant with no heat recovery with

Gas turbine plants with chemical recovery are represented on curve FG, where methane is used as primary fuel. The methane reforming process is described by the following two

turbine plant with no heat recovery, *η\** and on marginal efficiency *ηAF*.

while the opposite trend is observed when pressure ratio is increased.

regenerator and the minimum temperature difference in the evaporator.

and the minimum temperature difference in the evaporator.

matching.

temperature conditions.

**4.4 Combined internal heat recovery** 

regeneration and steam reforming of the fuel.

respect to thermodynamic regeneration.

reactions:
