Energy Conversion and Management 50 (2009) 2083–2097

Contents lists available at ScienceDirect

Energy Conversion and Management

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

Joining semi-closed gas turbine cycle and tri-reforming: SCGT-TRIREF

as a proposal for low CO2 emissions powerplants
Daniele Fiaschi *, Andrea Baldini
Dipartimento di Energetica ‘‘Sergio Stecco”, University of Florence, Via C. Lombroso 6/17, 50134 Firenze, Italy

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

Article history: Methane conversion to a rich H2 fuel by reforming reactions is a largely applied industrial process.
Received 19 August 2008 Recently, it has been considered for applications combined to gas turbine powerplants, as a mean for
Accepted 21 March 2009 (I) chemical recuperation (i.e. chemical looping CRGT) and (II) decarbonising the primary fuel and make
Available online 8 May 2009
the related power cycle a low CO2 releaser.
The possibility of enhancing methane conversion by the addition of CO2 to the steam reactant flow (i.e.
Keywords: tri-reforming) has been assessed and showed interesting results. When dealing with gas turbines, the
Gas turbine
possibility of applying tri-reforming is related to the availability of some CO2 into the fluegas going to
Tri-reforming
CO2 capture
the reformer. This happens in semi-closed gas turbine cycles (SCGT), where the fluegas has a typical
Semi-closed gas turbine 14–15% CO2 mass content. The possibility of joining CRGT and SCGT technologies to improve methane
Chemically recuperated cycle reforming and propose an innovative, low CO2 emissions gas turbine cycle was assessed here. One of
the key issues of this joining is also the possibility of greatly reduce the external water consumption
due to the reforming, as the SCGT is a water producer cycle.
The SCGT-TRIREF cycle is an SCGT cycle where fuel tri-reforming is applied. The steam due to the refor-
mer is generated by the vaporization of the condensed water coming out from the fluegas condensing
heat exchanger, upstream the main compressor, where the exhausts are cooled down and partially recir-
culated. The heat due to the steam generation is recuperated from the turbine exhausts cooling. The
reforming process is partially sustained by the heat recovered from the turbine exhausts (which gener-
ates superheated steam) and partially by the auto thermal reactions of methane with fresh air, coming
from the compressor (i.e. partial combustion).
The effect of CO2 on methane reforming (tri-reforming effect) increases with decreasing steam/meth-
ane ratio: at very low values, around 30% of methane is converted by reactions with CO2. At high values of
steam/methane ratio, the steam reforming reactions are dominant and only a marginal fraction of meth-
ane is interested to tri-reforming.
Under optimised conditions, which can be reached at relatively high pressure ratios (25–30), the power
cycle showed a potential efficiency around 46% and specific work at 550 kJ/kg level. When the amine CO2
capture is applied, the specific CO2 emissions range between 45 and 55 gCO2 =kW h.
Ó 2009 Elsevier Ltd. All rights reserved.

1. Introduction
There is an ever increasing consensus, worldwide, that one of
the main reasons of earth’s climate change is the CO2 emissions
from anthropogenic activities involving the combustion of hydro-
carbons (i.e. industry, power generation and utilities) [1]. For this
reason, many researchers addressed their efforts to the develop-
ment of new power cycle concepts, including those with integrated
CO2 capture and subsequent sequestration [2–9]. It seems to be the
most promising option to tackle greenhouse issue in the short to
midterm [1]. About one third of the anthropogenic carbon dioxide
emissions comes from fossil fuelled powerplants [10], thus they
* Corresponding author. Tel.: +39 055 4796776; fax: +39 055 4224137.
E-mail address: danif@de.unifi.it (D. Fiaschi).
deserve special attention when dealing with CO2 capture and
sequestration. One of the main challenges to succeed in the task
of reducing CO2 emissions from powerplants is closely related to
the chance of achieving cost effective ways to capture CO2 from
fluegas [11]. Typically, it is largely diluted with nitrogen when, as
in most cases, combustion is done with air. CO2 concentration in
the exhausts ranges from 4% to 6% of gas turbines up to 14–15%
of coal fired plants. For this reason, its capture often represents
the most energy and costs expensive operation in the overall CO2
sequestration process and must be carefully integrated and opti-
mised into the whole basic power cycle layout, in order to mini-
mise its impact on performance and costs. Currently, adding CO2
capture and sequestration to standard GTCCs leads to efficiency
drops ranging from 4% to 10% and to an increase in cost of energy
from 30% to 60% [11,12].

0196-8904/$ - see front matter Ó 2009 Elsevier Ltd. All rights reserved.
doi:10.1016/j.enconman.2009.03.034
2084 D. Fiaschi, A. Baldini / Energy Conversion and Management 50 (2009) 2083–2097
Three main ways of CO2 capture from powerplants were widely
investigated [1,10]:
– Post-combustion capture, by removing CO2 from exhausts. The
main advantage is the possibility of applying capture devices
as retrofit to existing powerplants, with only modest changes
to the cycle layout and to the components design. On the other
hand, the main disadvantage is the very low concentration of
CO2 into the atmospheric pressure fluegas, which makes separa-
tion and subsequent recompression for storage difficult and high
energy demanding [12–14].
– Pre-combustion capture, by fuel decarbonisation, which is a pro-
cess often integrated into the powerplant layout and shifts the
fuel to high hydrogen content while separating and removing
a fraction or the whole of carbon content. The main advantage
of this technique is the high pressure of fuel, which simplifies
CO2 separation, and the low mass flowrate to be treated. On
the other hand, it requires supplemental reactors (for example
reforming and shifting [2–5,8]) and relevant modification to
the existing cycles layouts.
2. Basic idea and principles of tri-reforming process
Reuse of carbon dioxide released by human activities (energy
production, transports, industry, etc.) could assume a very impor-
tant role in tackling the worldwide CO2 emissions issue [15]. Re-
cent studies [15,16] proposed a relatively innovative process for
the production of hydrogen rich syngas (with fixed H2/CO ratio)
from methane reforming. This process, named tri-reforming (TRI-
REF) joins the reactions of CO2 reforming, steam reforming, partial
oxidation and complete oxidation of methane into the same reac-
tor. The following reactions were considered:
CH4 þ 2O2 $ CO2 þ 2H2 O ½DH ¼ 880:3 kJ=kmol
CH4 þ 1=2O2 $ CO þ 2H2 ½DH ¼ 35:6 kJ=kmol
CH4 þ H2 O $ CO þ 3H2 ½DH ¼ þ206:3 kJ=kmol
CH4 þ CO2 $ 2CO þ 2H2 ½DH ¼ þ247:3 kJ=kmol
When compared to some more conventional reforming pro-
cesses (i.e. standard steam reforming, auto thermal reforming, par-
tial oxidation, etc.), the CO2-reforming reaction (D) is added by CO2
enrichment of reactants mixture. The ‘‘stand alone” process is well
known but, recently, some advantages when combined with stan-
dard reforming processes were shown [15,17]. Coupling CO2
reforming and steam reforming allows the achievement of syngas
with the required H2/CO ratio, while avoiding the formation of so-
lid carbon coming from the reactions CH4 $ C þ 2H2 O and
2CO $ C þ CO2 . Some experimental campaigns of Song and Pan
showed that the introduction of CO2-reforming reaction may also
improve the life of eventual catalysers and significantly reduce
the temperature range of solid carbon formation [17]. In the same
studies, they showed that it is possible to reach up to 95% of pri-
mary methane fuel conversion at equilibrium temperature in-
cluded into the 1073–1123 K range when (I) CO2–H2O ratio is
close to 1, (II) O2/CH4 ratio is 0.1 and (III) (CO2 + H2O + O2)/CH4 ra-
tio is 1.05.
A further important chemical reaction taking place in the TRI-
REF process is the water gas shift:
CO þ H2 O $ CO2 þ H2 ½DH ¼ 41:2 kJ=kmol
which may have an important role in the whole process.
Some interesting studies were carried out when the process is
fed with methane and exhausts coming from combustion cham-
bers of powerplants [15,17]. The endothermic reforming reactions
are thermally sustained by the heat content of the exhausts and
the partial combustion of primary methane fuel, which is done
with the excess oxygen of the exhausts. In this way, the TRIREF be-
comes specially attractive for gas turbines, where large excess air is
present into the exhausts and high methane conversions may be
sustained. Often, nickel and magnesium oxide catalysers play an
important role in CO2-reforming reactions [17].
The interest toward tri-reforming process is connected to its
possible integration in gas turbine based power cycles with low
CO2 emissions. The objective is to produce a relatively highly H2
and CO2 concentrated fuel, to be decarbonised upstream the com-
bustion chamber.
The methane conversion levels are closely related to the condi-
tions of gas at reformer inlet (specially its CO2 content, composi-
tion and temperature), thus its application to semi-closed gas
turbine cycles seems to be attractive [18,19]. Here, the relatively
high CO2 concentration into the fluegas is useful to provide the re-
quired amount of carbon dioxide to the TRIREF reactions chain (A)–
(E). Moreover, SCGT is a water producer cycle [18,19], which means
that the amount of steam required for methane reforming is inter-
nally produced, thus the resulting cycle is not water consuming.
3. Modelling and parametric analysis of tri-reforming process
The calculations related to tri-reforming reactions were carried
out by a dedicated homemade software, developed at the Energy
Engineering Dept. of University of Florence by the means of thermo-
dynamic EES software [20]. Once the inlet conditions of reactants
into the tri-reforming reactor are known, the Gibbs free energy of
each compound in products and reactants is evaluated and the over-
all difference between Gibbs free energy of products and reactants is
minimised through a general procedure providing, as a final result,
the equilibrium composition of the syngas fuel. The reactor was con-
sidered adiabatic and the kinetics of reactions were neglected.
ðAÞ The thermodynamic model of gas turbine cycle, again devel-
ðBÞ oped with EES software at the Energy Engineering Dept. of Univer-
ðCÞ sity of Florence, includes a rather classic and well tested blade
ðDÞ cooling procedure, which is able to provide reliable results on cur-
rent production machines [21–23]. The thermodynamic code of
SCGT was developed again within EES environment, starting from
the existing one, developed some years ago at the Energy Engineer-
ing Dept. of University of Florence in Fortran environment [18,19].
The sensitivity analysis of tri-reforming process to the following
four main parameters was carried out:
 Steam flowrate at reformer inlet per unit methane flowrate
(H2Oratio).
 Oxidizer flowrate at reformer inlet per unit methane flowrate
(mref ratio).
 Gas turbine pressure ratio (b).
 Inlet turbine temperature (here assumed as the maximum GT
cycle temperature Tmax).
This analysis was done to assess (I) the methane conversion po-
tential at different cycle parameters and (II) the contribution of
CO2-reforming reaction to the overall process. To do this, we re-
ferred to a simplified scheme of SCGT-TRIREF (Fig. 1), where no
CO2 sequestration was applied. It allowed a special focus on the
TRIREF reactions and the way they are affected by the reformer in-
let conditions and main gas turbine parameters. Subsequently, the
ðEÞ
results and analysis were extended to the complete SCGT-TRIREF
with CO2 removal.
Compared to the standard SCGT configurations (combined cy-
cles SCGT/CC and recuperative–evaporative cycles SCGT/RE [18]),
the following main differences in the cycle scheme can be
remarked:
 D. Fiaschi, A. Baldini / Energy Conversion and Management 50 (2009) 2083–2097
CH4
REF
IC
STACK
MIX
AIR
Fig. 1. Simplified scheme of SCGT-TRIREF with no CO2 capture.
 The split of exhausts stack flowrate is downstream the condens-
ing heat exchanger (SEP in Fig. 1), to increase the mass flow of
condensed water needed to feed the reformer without external
addition.
 Subsequently, the exhausts going to the stack must be reheated
above 100 °C to allow their expulsion. It is done by the heat
released at the intercooled compression.
 The water condensed at SEP is vaporised and superheated (at the
required pressure) into the HRSG by the heat content of the
exhausts at turbine outlet, thus it is sent to the reformer.
The combined effects of mref ratio and H2Oratio on the uncon-
verted fraction of methane is shown in contour plot of Fig. 2. Rel-
atively high conversion ratios were achieved at high oxidiser
flowrates, mainly due to its low O2 content (about 10% mass)
which is needed to the partial combustion reactions that sustain
the endothermic reforming process. Consequently, a syngas with
relatively low heating value is produced (Fig. 3).
Fig. 2 also shows that high values of H2Oratio favour the methane
conversion, mainly due to the increasing development of the
slightly exothermic water gas shift reaction (E), which makes a
higher amount of energy available to sustain the endothermic
reforming reactions at fixed oxidiser mass flow (i.e. fixed mref ratio).
Actually, the steam does not have a primary role in this reaction,
but only gives a contribution to drive it rightward, which implies
an increase of H2/CO ratio (Fig. 4). An inversion of methane conver-
sion trend was found at H2Oratio higher than 7.5, due to the insuf-
ficient amount of exhausts heat content, which leads to lower
steam temperatures at reformer inlet (when H2Oratio > 7.5 steam
superheating is no longer possible).
The conversion to H2 is maximised in the range 6–7.5 of H2Oratio
and 8–11 of mref ratio (see Fig. 5). Generally speaking, the conversion
to H2 increases with H2Oratio up to 7.5, corresponding to the highest
possible amount of steam produced into the HRSG. The behaviour
of conversion to H2 is the result of some processes, that are favoured
by the increase of the amount of steam into the reactants:
 Water gas shift reaction, which converts CO to H2.
 Methane reforming reaction (C), having H2/CO ratio = 3, higher
than all the other considered reactions.
2085
SYNGAS
CC
COOLING
Dehumidified exhausts
HRSG
SEP
When a low amount of H2O is present among the reactants, the
reaction (E) is leftward shifted, which penalises the reaction (C)
compared to the CO2-reforming reaction (D), having a lower H2/
CO ratio = 1. Thus, the CH4 conversion is shifted toward CO. On
the contrary, at high H2Oratio, reactions (C) and (E) are favoured
and H2 content of syngas is maximised. Moreover, the slightly exo-
thermic water gas shift reaction helps partial methane oxidation
reaction (favoured by the low amount of oxygen into the reaction
environment [24]) in sustaining the endothermic reforming
reactions.
It is also important to notice that, at fixed H2Oratio, a value of
mref ratio optimising the conversion to H2 was found. All these val-
ues are approximately distributed along a line on the mref ratio–
H2Oratio plane (see straight line in Fig. 5). Working on the right side
of this line means that a too high amount of O2 is sent to the reac-
tor, which implies that reaction (A) (combustion) is shifted right-
ward, thus a larger amount of CH4 is consumed and the ‘‘useful”
reforming reactions (C) and (D) are shifted leftward, with subse-
quent reduced production of hydrogen, on account of the increased
outlet reformer temperature [17]. When the line of maximum con-
version to H2 of Fig. 5 is reported on contour plot of Fig. 3, it
roughly corresponds to the 5% unconverted CH4 line, up to H2Oratio
below 6 kg=kgCH4 . From Fig. 5 it is possible to determine the lower
limit of mref ratio allowing the maximum possible conversion to H2,
corresponding to 9 kg=kgCH4 : it is the cross point between the line
of maximum conversion at the highest possible H2Oratio (straight
line) and the line representing the maximum possible superheated
steam produced into the HRSG (dashed line).
From the above discussion, we concluded that 95% methane
conversion is the upper bound for the fixed b = 20 and
Tmax = 1538 K gas turbine conditions.
The syngas volumetric composition at reformer outlet vs.
H2Oratio, at three different values of mref ratio is shown in Fig. 6
((a) = combustible species, (b) = non combustible species).
At relatively high values of H2Oratio and mref ratio, the fraction of
carbon initially contained into methane and converted to CO2 is
relatively high, which is an important feature for the subsequent
decarbonisation of the produced syngas fuel. On the contrary, at
low values of H2Oratio, carbon is almost equally shared between
CO and CO2. In this case, in order to favour the decarbonisation
2086 D. Fiaschi, A. Baldini / Energy Conversion and Management 50 (2009) 2083–2097

8
0,05
7

H2Oratio [kg/kg CH4]

6

5
0,10
4
0,20
0,30
3
0,40
0,50
2

0
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18
m ref ratio [kg/kg CH4]

Fig. 2. Unconverted fraction of methane [kg=kgCH4 ] vs. mref ratio and H2Oratio.

9
Tmax = 1538 K
8 β = 20

7 2
H2Oratio [kg/kg CH4]

4 4

3
6

2 10 8

1 14
12
16
0
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18
m ref ratio [kg/kg CH4]

Fig. 3. Syngas heating value [kJ/kg] vs. methane [kg=kgCH4 ] vs. mref ratio and H2Oratio.

process, a supplemental downstream water gas shift reactor
(WGSR) may be useful. Under favourable conditions, the steam
contained into the fluegas could usefully take part to the water
gas shift reactions, without any external water supplies.
The use of H2 rich fuels into the current low NOx GT combus-
tion chambers might result into troubleshoot due to self ignition.
Generally, the upper bound here found for volumetric fraction of
H2 into the syngas is less than 20% in the whole investigated field
of mref ratio and H2Oratio (Fig. 6a). It falls within the recently up-
dated limits experienced, for example, by the 10 years operation
of GE energy F-class gas turbines with hydrogen content up to
45% [25]. The CO2 fraction remains almost unchanged with vary-
ing H2Oratio and mref ratio. The combustible species (CO, H2 and
CH4) are consistently diluted into N2 and vapour. The dilution
into vapour increases with increasing steam flowrate at the refor-
mer inlet. The H2O mass balance of the reactor at relatively high
steam inlet flowrates shows that no appreciable amounts of
steam are involved in reforming reactions and a large fraction
flows unreacted through the reformer. In fact, H2O is both reac-
tant and product in the reactions chain (A), (C), (E) and the latter
gives a limited contribution to the H2O consumption, in spite of
the large amount needed as reactant specie. From Fig. 7, it is
clear that the highest consumption of steam corresponds to
H2Oratio and mref ratio optimising the conversion to H2. Under these
conditions, the reactions (C) and (E) are favoured, while (A) is
limited.
 D. Fiaschi, A. Baldini / Energy Conversion and Management 50 (2009) 2083–2097 2087

9
Tmax =1538 K
8 β=20

Molar ratio H2/CO

6

1
0 1 2 3 4 5 6
H2Oratio [kg/kgCH4]

Fig. 4. H2/CO molar ratio vs. H2Oratio.

7 0.27
H2Oratio [kg/kg CH4]

5 0.24

0.21
4
0.18

3 0.15
0.12
2 0.09
0.06

1 0.03

0
0
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18
m ref ratio [kg/kg CH4]

Fig. 5. H2 conversion ratio [kg=kgCH4 ] vs. mref ratio and H2Oratio.

4. Interactions of TRIREF process with the integrated SCGT-
TRIREF power cycle: effects of the main reforming parameters
on the cycle working conditions
As above remarked, when mref ratio increases, the reactivity at
reformer is increased, as well as the amount of oxygen demand
(i.e. fresh air). If the compressor inlet flowrate is kept constant
(like in the present case), the amount of recirculated fluegas at
the mixer upstream the compressor is reduced. It leads to an in-
crease in O2 mass fraction and, at the same time, to a reduction in
CO2 mass fraction, leading to higher stack flowrate. Thus, the
power cycle shifts toward a ‘‘more opened” configuration, with
an average 10% increase of the ratio between fresh air and overall
compressor inlet flowrate. A similar behaviour was found vs.
H2Oratio, but the shift toward a more opened configuration is even
more remarkable. It is confirmed by the increase of the ratio be-
tween external fresh air and overall compressor inlet flowrate,
which rises up to 15–20% levels. It leads to higher H2O concentra-
tion in fluegas and, subsequently, higher condensed flowrate at
SEP.
When H2Oratio grows above 2, the turbine mass flowrate in-
creases at values 10% higher than the compressor flowrate, up to
20% at the optimising H2Oratio = 20, whereas the sensitivity to
mref ratio is more limited (2%), as shown in Fig. 8. It suggests an
almost complete redesign of the expander.
The lower amount of recirculated exhausts leads to a reduction
of CO2 concentration at the stack (Fig. 8), which is disadvantageous
for post-combustion capture.
2088 D. Fiaschi, A. Baldini / Energy Conversion and Management 50 (2009) 2083–2097

0.7

Volumetric fraction of syngas non combustible species

0.18
Volumetric fraction of syngas combustible species

0.17 H2
0.65
0.16
0.6
0.15 H2
0.14 0.55
0.13 0.5
N2
0.12
0.45
0.11 H2
0.1 0.4
0.09 0.35
0.08 CH4
0.3
0.07
0.25 H 2O
0.06
mref ratio = 8
0.05 mref ratio = 8 0.2
CO mref ratio = 12 mref ratio = 12
0.04 0.15
CH4 mref ratio = 16
0.03 mref ratio = 16
0.1
0.02
0.01 CH4 0.05
CO2
0 0
0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9
H2Oratio [kg/kgCH4] H2Oratio [kg/kgCH4]

Fig. 6. Volumetric composition of syngas vs. H2Oratio at three different mref ratio.

9
Tmax = 1538° K
8 β = 20

7
H2Oratio [kg/kgCH4]

1
-0,4
5

0,8
4

-0,2
3

2
0
0,2
1 0,4
0,6

0
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18
m ref ratio [kg/kgCH4]

Fig. 7. H2O balance at reformer (difference between outlet and inlet mass).

The outlet turbine and HRSG temperatures vs. H2Oratio at differ- achieved at about 1057 K, slightly decreasing with increasing
ent values of mref ratio are shown in Fig. 9. It is clear that the recu- H2Oratio. The noticeable increases in equilibrium temperature at
perative potential of the power cycle increases with H2Oratio and values of H2Oratio and mref ratio higher than those of maximum con-
mref ratio. Obviously, at higher H2Oratio, the improved heat recovery version to H2, confirm the trend of the reaction chain toward the
is due to superheating of the higher amount of required steam. It is combustion of methane fuel.
confirmed by the reduction of the HRSG outlet temperature with The outlet temperature of the second compressor is another
increasing H2Oratio (Fig. 9). important variable, consistently affected by the two main reform-
An important parameter of the analysis is the reformer outlet ing parameters H2Oratio and mref ratio. Their influence reflects di-
temperature, also representing the equilibrium value of the related rectly on the conditions at reformer inlet, combustion chamber
chemical reaction chain. Its behaviour vs. H2Oratio, at different and turbine blade coolant. Specifically, as reported in Fig. 11, this
values of mref ratio, is shown in Fig. 10. It reveals a high sensitivity temperature decreases with increasing H2Oratio and mref ratio,
to mref ratio, as happens for the conversion ratio (see Fig. 2). mainly due to the reduction of the inlet temperature at the second
The dashed line represents the couples of parameters (H2Oratio, compressor. This effect is disadvantageous for the combustion
mref ratio) allowing 95% methane conversion, corresponding to chamber and reforming, whereas blade cooling is favoured. Under
those identified in Fig. 5 with the straight line of maximum conver- conditions of maximum conversion to H2, the reduction of the
sion to hydrogen. With no steam injection (i.e. H2Oratio = 0), it is outlet temperature at the second compressor, compared to the
 D. Fiaschi, A. Baldini / Energy Conversion and Management 50 (2009) 2083–2097 2089

1.45

Turbine/compressor inlet flowrate ratio [kg/kgcomp]

0.104

1.4 0.102

0.100

CO2 molar fraction at stack

1.35
0.098
1.3
0.096
mref ratio = 4
1.25 mref ratio = 8 0.094

mref ratio = 12 0.092

1.2
mref ratio = 16 0.090
1.15 0.088

0.086
1.1
0.084
1.05
0.082

1 0.080
0 1 2 3 4 5 6 7 8 9
H2Oratio [kg/kgCH4]

Fig. 8. Turbine/compressor inlet flowrate ratio and CO2 molar concentration at the stack vs. H2Oratio at different values of mref ratio.

850 840
mref ratio =2
800 mref ratio =4
mref ratio =8 835
750

Outlet turbine temperature [K]

mref ratio =12
Outlet HRSG temperature [K]

mref ratio =16

700 830

650
825
600

550
820
500

450 815

400
810
350

300 805
0 1 2 3 4 5 6 7 8 9
H2 Oratio [kg/kgCH4]

Fig. 9. Outlet turbine and HRSG temperatures vs. H2Oratio at different values of mref ratio.

non intercooled version of the power cycle, ranges from 60 K (at ciency of power cycle increases almost linearly with H2Oratio. As
mref ratio = 16 and H2Oratio = 0) to about 80 K (at mref ratio = 10 and previously discussed, there is a bound over which almost the
H2Oratio = 6). whole amount of steam entering into the reformer flows unre-
acted, thus the beneficial effects on the efficiency are related to
5. Performance analysis of the SCGT-TRIREF the heat recuperation into the HRSG. Actually, the reformer has
mainly the same effect as steam injection upstream the combus-
The behaviour of SCGT-TRIREF efficiency and specific work vs. tion chamber. Anyhow, the chemical recuperation at the reformer
H2Oratio at different values of mref ratio is shown in Fig. 12a and b, gives a valuable contribution, as it reduces the required oxygen
respectively, at the reference pressure ratio b = 20. The efficiency mass flow (i.e. mref ratio), as previously shown in the cases with
of power cycle decreases with increasing mref ratio. It is attributable fixed CH4 and H2 conversion ratios (Figs. 2 and 5). At the different
to the increased methane fuel demand, due to the higher extent of fixed mref ratio, values of H2Oratio ranging between 6.5 and 7.5
oxidation reactions into the reformer. On the contrary, the effi- optimising gGT were found. They roughly correspond to those
2090 D. Fiaschi, A. Baldini / Energy Conversion and Management 50 (2009) 2083–2097

1400
1350
1300

Syngas outlet reformer temperature [K]

1250
1200
1150
1100
1050
1000
950
900
850
800
750
mref ratio =2 mref ratio =8 mref ratio =14
700
mref ratio =4 mref ratio =10 mref ratio =16
650 mref ratio =6 mref ratio =12 mref ratio =18
600
0 1 2 3 4 5 6 7 8 9
H2 Oratio [kg/kg CH4]

Fig. 10. Syngas reformer outlet temperature (chemical equilibrium) vs. H2Oratio at different values of mref ratio.

680
Fluegas temperature at second compressor outlet [K]

670

660

650

640

630
mref ratio =2
620 mref ratio =4
mref ratio =6
610 mref ratio =8
mref ratio =10
600 mref ratio =12
mref ratio =14
mref ratio =16
590
mref ratio =18

580
0 1 2 3 4 5 6 7 8 9
H 2Oratio [kg/kg CH4]

Fig. 11. Outlet fluegas temperature at second compressor outlet.

optimising the conversion to H2. Above these values, the mass flow (2) the temperature of oxidiser fluegas at the compressor outlet,
of water to be superheated is too high and the heat recuperation increasing with b, that favours the methane conversion,
into the HRSG is no longer convenient. leading to lower extent of partial combustion reactions;
The specific work (see Fig. 12b) is only marginally influenced by (3) the outlet turbine temperature, which decreases with
mref ratio, whereas it shows a consistent improvement with increas- increasing b and reduces the superheated steam tempera-
ing H2Oratio. ture at the reformer inlet, penalizing the reforming
One of the main gas turbine parameters, having a dominant ef- reactions.
fect on the methane conversion ratio at the reformer, is the pres-
sure ratio b. It influences three main parameters: The overall effect of b on the powerplant performance and
effectiveness of methane conversion is given by the sum of the
(1) the pressure into the reactor, increasing with b, which penal- above contributions. The behaviour of H2 conversion ratio vs. b at
ises the reforming reactions; fixed H2Oratio = 3, for different values of mref ratio, is shown in
 D. Fiaschi, A. Baldini / Energy Conversion and Management 50 (2009) 2083–2097 2091

a 0.5
0.49
b 900
mref ratio =2 mref ratio =2
850
0.48 mref ratio =4 mref ratio =4

0.47 mref ratio =6 800 mref ratio =6

mref ratio =8 mref ratio =8
0.46

Specific work [kJ/kg comp ]

mref ratio =10 750 mref ratio =10
0.45 mref ratio =12 mref ratio =12
mref ratio =14 700
0.44 mref ratio =14
mref ratio =16 mref ratio =16
0.43 650
mref ratio =18 mref ratio =18
0.42
η GT

600
0.41
0.4 550
0.39
500
0.38
0.37 450
0.36 400
0.35
350
0.34 β = 20 β = 20
0.33 300
0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9
H2 Oratio [kg/kg CH4] H2 Oratio [kg/kg CH4]

Fig. 12. SCGT-TRIREF efficiency and specific work vs. H2Oratio at different fixed values of mref ratio. (a) SCGT-TRIREF efficiency and (b) SCGT-TRIREF specific work.

Fig. 13a. On the whole, it is clear that increasing pressure ratio ing to the value not allowing superheating of the water due to
penalizes the conversion to H2. Nevertheless, it is interesting to no- the reformer, a different behaviour of performance maps was
tice that this behaviour is inverted at very high values of mref ratio found.
(i.e. 16 in Fig. 13a). This couple of H2Oratio and mref ratio values is be-
yond the line of maximum conversion to H2 shown in Fig. 5. It
means that this line shifts toward higher mref ratio with increasing 6. Evaluation of the relative influence of the tri-reforming (A)–
b (Fig. 13b). (E) reactions on the overall chemical equilibrium
The performance maps of the SCGT-TRIREF power cycle are
shown in Fig. 14a (at fixed H2Oratio and different values of mref ratio) The evaluation of the influence of (A)–(E) reactions on the
and Fig. 14b (at fixed mref ratio and different values of H2Oratio). Gen- chemical composition at the reformer outlet is an important issue,
erally, the cycle efficiency takes advantage of high pressure ratios, as it allows to verify the actual development of each one and, spe-
that inhibit the H2 conversion to the reformer, thus allowing a cifically, the effect of the CO2 reforming (D) and its contribution to
more limited methane oxidation. In fact, as previously discussed, the overall reaction process compared to the standard steam
at higher pressure ratios correspond higher reformer inlet temper- reforming. This analysis was carried out under conditions of max-
atures, that allow more limited values of mref ratio to thermally sus- imum methane conversion, corresponding to mref ratio and H2Oratio
tain the reforming reactions. The efficiency also shows the on the 5% contour line of Fig. 2 (roughly the same as maximum
existence of optimising values of b, ranging between 32 and 38 conversion to H2 in Fig. 5). The inlet–outlet molar balance of the
at different mref ratio. Also the specific work shows an optimising five chemical species at the reformer, referred to 1 mole of CH4
b, ranging between 16 and 20 at different mref ratio. It must be no- at different values of H2Oratio, is shown in Table 1. Negative values
ticed that, differently from efficiency, specific work is favoured at mean that the specie is consummated, positive values mean that
higher mref ratio. The performance maps at different H2Oratio are the specie is produced through the reactions. The values of Table
shown in Fig. 14b. A consistent improvement of performance is 1 are the basis to evaluate the relative influence of each reaction
achieved up to H2Oratio = 5–7, at relatively high pressure ratios. on the overall chemical process at equilibrium. This evaluation
Anyway, from H2Oratio = 6, beyond the pressure ratio correspond- was performed by a numerical method. Said a the number of moles

a 0.28 b 9
0.26 β = 10
8 β = 20
H2 conversion ratio [kg/kg CH4]

0.24
β = 30
0.22 7
β = 40
H2Oratio [kg/kg CH4]

0.2
6
0.18
0.16 5
0.14
0.12 4

0.1
3
0.08
0.06 mref ratio = 4 2
mref ratio = 8
0.04 mref ratio = 12 1
0.02 mref ratio = 16 H2O ratio = 3

0 0
10 12 14 16 18 20 22 24 26 28 30 32 34 36 38 40 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18

β m ref ratio [kg/kg CH4]

Fig. 13. H2 conversion ratio vs. b and shift of maximum H2 conversion line. (a) H2 conversion ratio vs. b at different values of mref ratio and (b) shift of maximum H2 conversion
line on the H2Oratio–mref ratio plane with increasing b.
2092 D. Fiaschi, A. Baldini / Energy Conversion and Management 50 (2009) 2083–2097

a 0.44 H2O ratio = 3

b 0.46
0.43
Δβ = 2 0.44
0.42
m ref ratio= 2 0.42
0.41
β = 40
0.4 m ref ratio= 4 0.4
m ref ratio = 3
0.39 m ref ratio= 6
0.38 Δβ = 2
η GT

η GT
0.38 m ref ratio= 8
0.36 β = 40 H2O ratio= 0
0.37 H2O
m ref ratio= 10 ratio= 1

0.36 0.34 H2O ratio= 2

m ref ratio= 12 H2O ratio= 3
0.35 0.32
H2O ratio= 4
m ref ratio= 14 H2O ratio= 5
0.34 H2O ratio= 6
β = 10 0.3 H2O ratio= 7
0.33 m ref ratio= 16 β = 10

0.32 0.28
380 400 420 440 460 480 500 520 200 300 400 500 600 700 800
Wsp [ kJ/kg comp ] Wsp [ kJ/kg comp ]

Fig. 14. Performance maps of the SCGT-TRIREF power cycle. (a) Performance maps at different mref ratio and (b) performance maps at different H2Oratio.

of each specie involved in reaction (A) per 1 mole of CH4, b the
number of moles of each specie involved in reaction (B) per 1 mole
of CH4 and so on, the following equations may be written (see reac-
tions (A)–(E)):
nO2 t ¼ 2a  0:5b; nCH4 t ¼ a  b  c  d;
nH2 Ot ¼ 2a  c  e; nCOt ¼ b þ c þ 2d  e;
nH2 t ¼ 2b þ 3c þ 2d þ enCO2 t ¼ a þ e  d
where nO2 t ; nCH4 t ; nH2 Ot ; nCOt ; nH2 t ; nCO2 t are the moles of each spe-
cie generated (or consummated) by the reactions (A)–(E). These
must differ minimally from the values generated by the equilibrium
model (nO2 ; nCH4 ; nH2 O ; nCO ; nH2 ; nCO2 ) here used for the calculation,
thus the minimisation of the function:
F ¼ ðnO2  nO2 t Þ2 þ ðnCO2  nCO2 t Þ2 þ ðnCH4  nCH4 t Þ2
þ ðnH2 O  nH2 Ot Þ2 þ ðnCO  nCOt Þ2 þ ðnH2  nH2 t Þ2
Allowed the calculation of the a, b, c, d, e parameters, providing
the relative influence of each reaction on the overall tri-reforming
process.
The minimisation of F was carried out using some EES built in
procedures [20], named genetic method and variable metric meth-
od. The first method is often used as the starting point for the sec-
ond one, which is generally much more accurate but
computationally more complex and thus slower when the whole
field of variables has to be scanned. The genetic method was ap-
plied at H2Oratio = 0, thus the variable metric method was applied
for the other values of H2Oratio, using the previous solution as a
starting point for the current one. The results of this analysis, at
different values of H2Oratio and normalized to 1 mole of converted
methane, are shown in Fig. 15 and Table 2.
When no steam is injected to the reformer, almost 30% methane
is involved into the CO2-reforming reaction and only 25% into the
steam reforming. With increasing H2Oratio, the amount of methane
involved in steam reforming increases up to more than 50%,
whereas the CO2-reforming reaction reduces its contribution, down
to 7%. The oxidation (including partial and complete reactions) in-
volves a slightly decreasing amount of methane (45.5–41.7%). The
partial oxidation involves a very small amount of CH4 (2.5%) when
no steam is injected into the reformer, which becomes rather con-
sistent (7.7%) at the highest values of H2Oratio. It can be justified
considering that this analysis was carried out at fixed methane con-
version ratio, which means that at higher H2Oratio correspond lower
mref ratio, thus a reduced oxygen molar fraction is present into the
reactants. In agreement with the previous thermodynamic studies
of Zhu et al. [24], a partial methane oxidation takes place when
CH4/O2 molar ratio is about 0.5, whereas the complete oxidation
is achieved when the above molar ratio is 2. A mix of both processes
is registered at intermediate values. Considering the moles of CH4
involved in oxidation, the ratio CH4/O2 is 1.9 for H2Oratio = 0 and
drops to 1.7 for H2Oratio = 6, justifying the small increase of partial
oxidation. The overall reactions enthalpy drop across the reformer
(calculated as weighted average on the moles of the enthalpies of
reactions: DH = 880.3a  35.6b  41.2e + 206.3c + 247.3d, slightly
decreases of about 4%) with increasing H2Oratio, which is in agree-
ment with the decreasing outlet equilibrium temperature at the
fixed methane conversion ratio.
An important role in the whole reactions process is played by
the CO and the e parameter, representing the number of moles in-
volved in the water shift reaction per mole of converted CH4. Neg-
ative values of e (found at low H2Oratio) mean that the shift reaction
is directed leftward. With increasing H2Oratio, e increases less than
linearly, suggesting that the influence of the steam amount on the
water gas shift reaction is rather attenuated.
7. The SCGT-TRIREF-R power cycle with CO2 removal
The schematic of SCGT-TRIREF-R power cycle is shown in
Fig. 16. It is equipped with upstream CO2 removal. Inlet syngas fuel

Table 1
Molar balance at the reformer.

H2Oratio (kg=kgCH4 ) mref ratio (kg=kgCH4 ) CO (mol=molCH4 ) CO2 (mol=molCH4 ) H2O (mol=molCH4 ) H2 (mol=molCH4 ) O2 (mol=molCH4 )
0 16.36 0.9011 0.09895 1.351 0.6491 0.874
1 15.03 0.7163 0.2837 1.617 0.3827 0.8332
2 13.9 0.5802 0.4198 1.817 0.1835 0.8017
3 12.9 0.4768 0.5232 1.971 0.0295 0.7763
4 12.01 0.3967 0.6033 2.093 0.0929 0.7552
5 11.18 0.3336 0.6664 2.193 0.1926 0.7369
6 10.41 0.2832 0.7168 2.276 0.2758 0.7205
 D. Fiaschi, A. Baldini / Energy Conversion and Management 50 (2009) 2083–2097 2093

Methane reactions
100%
90%
80%
70%
60%
50%
40%
30%
20%
10%
0%
0 1 2 3 4 5 6
H2Oratio [kg/kgCH4]
(A) Oxidation (B) Partial Oxidation
(C) Steam reforming (D) CO2 reforming

Fig. 15. Relative influence of the methane conversion reactions vs. H2Oratio on the line of maximum CH4 to H2 conversion.

Table 2
Behaviour of a, b, c, d and e coefficients and overall process enthalpy variation at different H2Oratio (DH is).

H2Oratio (kg=kgCH4 ) a (mol=molCH4 ) b (mol=molCH4 ) c (mol=molCH4 ) d (mol=molCH4 ) e (mol=molCH4 ) DH (kJ=kmolCH4 )

0 0.4307 0.02507 0.2505 0.2937 0.03813 254.2
1 0.4065 0.04018 0.3262 0.227 0.1042 240.2
2 0.388 0.05153 0.3827 0.1779 0.2098 229.1
3 0.3731 0.06023 0.4262 0.1406 0.2907 219.9
4 0.3608 0.06708 0.4606 0.1116 0.354 212.0
5 0.3503 0.07267 0.4885 0.08867 0.4048 204.9
6 0.3409 0.07727 0.5115 0.07033 0.4462 198.3

CH4
REF

WGSR
RIG 2
CC
RIG 1

Syngas

CO2
IC SEP
CO2
EXHAUSTS SH

STACK

MIX

EVA
SEP
AIR
COOLING WATER

Fig. 16. Schematic of SCGT-TRIREF-R power cycle.

is decarbonised by chemical absorption with an aqueous solution
of ethanolamine [26].
The main layout of the power cycle resembles the simple ver-
sion of SCGT-TRIREF with no CO2 capture shown in Fig. 1. Here,
the syngas treatment section for CO2 removal was introduced.
The reformed fuel coming from REF is cooled down in two different
recuperative heat exchangers, RIG1 and RIG2, thus it is sent into
the WGSR, which increases the CO2 concentration (typically rang-
ing from 10% to 15% at REF outlet) and favours its capture. The heat
recovery into RIG1 and RIG2 is rather consistent, due to the rela-
tively high syngas flowrate and temperature. This configuration
is convenient when values of H2Oratio and mref ratio are located in
2094 D. Fiaschi, A. Baldini / Energy Conversion and Management 50 (2009) 2083–2097

the region below the hydrogen maximum conversion line (see the lower CO2 concentration of the current syngas compared to the
Fig. 5). On the contrary, when the values of these parameters are previous reference case, which should increase the heat consump-
above this line, a condensing heat exchanger for water extraction tion for a fixed capture level. In the SCGT-TRIREF-R power cycle,
is better suited, due to the high vapour concentration into the syn- this heat is supplied by the gas turbine exhausts, upstream the
gas. Anyway, both processes are effective at temperatures below condensing heat exchanger (SEP), that are cooled down to the low-
400 K, thus cooling syngas coming out from the reformer at 800– er temperature limit, here fixed at 120 °C [27].
1000 K is mandatory. Moreover, the reduction of syngas tempera- Another important difference between the SCGT-TRIREF-R and
ture increases the CO2 capture effectiveness [22,23]. The water the simplified scheme of Fig. 1 is the introduction of steam blade
needed to the WGSR is contained into the wet syngas itself, thus cooling, which is a well known way for improving cycle perfor-
no external supply is needed and, once the required thermody- mance when steam is available in some sections of powerplant
namic conditions are reached, the shift reaction takes place. The [28]. For this reason, the recirculation of condensed water coming
heat released from this reaction preheats the water coming out from SEP has undergone substantial changes, as a fraction of it
from SEP, up to conditions close to saturation. The syngas coming must be provided to blade cooling. The whole amount of con-
out from WGSR is then decarbonised into the downstream etha- densed water coming out from SEP is preheated into the WGSR,
nolamine absorption section. The CO2 removal effectiveness and then vaporised and slightly superheated (10 °C) into the vaporiser
heat demand for the regeneration of amine solution were fixed, by heat recovery from the exhausts. Thus, it is splitted in two flows,
in agreement with previous research activities in the field of auto one of which goes to blade cooling while the other one is further
thermal reforming [22,23], at 90% and 2000 kJ=kgCO2 , respectively. superheated (again exploiting the turbine exhausts heat content)
Actually, the latter was a bit overestimated, in order to account for and subsequently sent to the reformer. The amount of steam due

650
630
SCGT -T RIREF-R H2
610
T at amine regeneration outlet [°K]

SCGT -T RIREF-R H3
590
SCGT -T RIREF-R H4
570
550
530
510
490
470
450
430
410
390
370
350
10 12 14 16 18 20 22 24 26 28 30 32 34 36
β
Fig. 17. Exhausts temperature at amine regeneration outlet vs. GT pressure ratio at three different values of H2Oratio.

a 0.48 b 60

0.46 β =30 58
CO2 emissions [g/kWh]

β =26 56
0.44 β =40 SCGT -T RIREF-R H2
β =36 β =16 54 SCGT -T RIREF-R H3

0.42 SCGT -T RIREF-R H4

52
η GT

0.4 50

0.38 48
β =10
β =10
46
0.36
β =10 SCGT -T RIREF H6
SCGT -T RIREF-R H2
44
β =10
0.34 SCGT -T RIREF-R H3
42
SCGT -T RIREF-R H4
0.32 40
460 480 500 520 540 560 580 600 620 10 12 14 16 18 20 22 24 26 28 30 32 34 36 38 40

Wsp [ kJ/kg comp ] β

Fig. 18. Performance maps and specific CO2 emissions of the SCGT-TRIREF-R. (a) Performance maps and (b) specific CO2 emissions.
 D. Fiaschi, A. Baldini / Energy Conversion and Management 50 (2009) 2083–2097 2095

to blade cooling consistently reduces the allowable range of 8. Parametric analysis and performance of the SCGT-TRIREF-R
H2Oratio, which may be set above 3 kg=kgCH4 only at low pressure power cycle
ratios while providing, at the same time, the minimum required
temperature level for amine regeneration [26,27] (see Fig. 17). The performance analysis of the SCGT-TRIREF-R was carried out
at three different values of H2Oratio (2, 3 and 4), under conditions of
95% methane conversion into the reformer. The performance maps
Table 3 are summarised in Fig. 18a, where the comparison with the simple
Volumetric composition of the reformed decarbonised syngas upstream GT SCGT-TRIREF without CO2 removal also appears (SCGT-TRIREF H6).
combustor.
As shown in Fig. 17, the maximum allowable pressure ratio de-
CH4 CO2 CO H2O H2 N2 O2 creases with increasing H2Oratio. With this constraint, the best re-
0.004 0.0088 0.0002 0.199 0.220 0.568 0 sults were achieved at high pressure ratios. Thus, further
increases of H2Oratio lead to a reduction of the maximum allowable

a 0.48 b 400
SCGT -T RIREF-R (conv 95%) SCGT -T RIREF-R (95%)
H2Oratio = 3
0.46 SCGT -T RIREF-R (conv 75%) 350 SCGT -T RIREF-R (75%)
SCGT -T RIREF-R (conv 50%) SCGT -T RIREF-R (50%)
SCGT -T RIREF-R (conv 25%) β =28 β =30 SCGT -T RIREF-R (25%)
0.44 β =26 β = 26
300

CO2 emissions [g/kWh]

0.42 250
η GT

0.4 200

0.38 150

0.36 100
β =10
β =10 β =10
β =10
0.34 50
H2Oratio = 3
0.32 0
480 500 520 540 560 580 600 10 12 14 16 18 20 22 24 26 28 30

Wsp [ kJ/kg comp ] β

Fig. 19. Performance maps and specific CO2 emissions of the SCGT-TRIREF-R at reduced methane conversion ratios. (a) Performance maps and (b) specific CO2 emissions.

SH

AIR
COOLING WATER

Fig. 20. Layout and basic operating data of the SCGT-TRIREF-R with supplemental condensing heat exchanger (SEP1).
2096 D. Fiaschi, A. Baldini / Energy Conversion and Management 50 (2009) 2083–2097
pressure ratio and, consequently, of the related highest possible
efficiency levels. Under these conditions, the most efficient case
is the one with H2Oratio = 3 and b = 26. When compared to the opti-
mised basic SCGT-TRIREF without CO2 capture, only two points
efficiency are lost, which is an interesting result, as in most cases
the introduction of CO2 capture leads to 4–8 points drop
[7,8,11,19]. On the other hand, the CO2 emissions are reduced from
428.8 to 43.5 g/kW h under optimised conditions (see Fig. 18b).
As in the case of simple SCGT-TRIREF, the low syngas heating
value leads to a relatively high increase in turbine inlet flowrate
(12% more than compressor), due to the high demand of fuel. Any-
way, this is lower than SCGT-TRIREF (17%), due to the CO2 captured
and removed from the syngas. This consistent amount of fuel input
should require the redesign of GT combustor and a different oper-
ating point of the turbine–compressor system, that could be com-
pensated through different strategies like variable guide vanes,
increased pressure ratio, re engineering of the machine [29]. The
volumetric composition of decarbonised syngas at combustor inlet
shows an almost complete removal of carbonaceous compounds
and the only combustible specie is H2, highly diluted into nitrogen
and vapour (see Table 3).
The resulting very low syngas heating value, makes it very hard
to join gas turbines. For this reason, the analysis of the SCGT-TRI-
REF-R performance potential was carried out also at lower values
of methane reforming conversion ratios (75%, 50% and 25%). The
resulting performance and CO2 emissions are summarised in
Fig. 19a and b. The maximum possible pressure ratio to match
the allowed minimum exhausts temperature at the amine regener-
ation outlet increases with decreasing methane conversion, thus
rather higher b may be adopted.
At fixed b, the penalization of efficiency is modest compared to
the 95% methane conversion, whereas a larger increase in specific
work was observed (up to 9%). It might be regarded as a possibility
of peakload shaving of the powerplant, which would be achieved
by the simple modulation of mref ratio at compressor outlet (i.e.
the amount of oxidiser at reformer inlet). Under these conditions,
the main drawback is the increase of CO2 emissions (Fig. 19b).
0.48
0.47
0.46
0.45
0.44
0.43
0.42
0.41
0.4 β =34
β =22
ηGT
0.39
0.38
β =14
0.37
0.36
0.35 β =10
0.34 β =10 β =10
0.33
β =10
0.32
0.31
0.3
380 400 420 440 460
Fig. 21. Comparison of performance maps of SCGT-TRIREF-R with and without supplemental condensing heat exchanger (SEP1).
Another investigated way to increase the syngas heating value
was the introduction of a supplemental condensing heat exchanger
(SEP1) on the syngas line, downstream the WGSR, in order to re-
move a consistent fraction of water vapour (there is potentially
about 20% improvement margin, see Table 3). Thus, the power cy-
cle was modified to the scheme of Fig. 20: this condensing heat ex-
changer was placed downstream the WGSR and cools the syngas
down below the dew point by heating an external stream of water
taken from the environment (which could be considered a cogen-
eration potential of the powerplant). The additional condensed
water was remixed with the stream coming from SEP1 and then
partially splitted to the reformer and partially to blade cooling.
The performance values shown in Fig. 20 and maps of Fig. 21
put in evidence a consistent drop compared to the basic configura-
tion without SEP1 (both in efficiency and specific work). This is
mainly due to the reduction of inlet fuel flowrate (about 2 kg/s,
corresponding to the condensed and separated water, see data in
Fig. 20), which implies the reduction of turbine inlet flowrate
and the related power output. Practically, the vapour content of
the syngas fuel plays the role of steam injection into the combus-
tor. Thus, we can conclude that, given the high performance drop,
it is not worthy to add SEP1 to condense water to partially recover
design conditions of the gas turbine, but rather accept higher off
design degree or, either proceed to a dedicated redesign.
9. Conclusions
The application of tri-reforming to semi-closed gas turbines was
proposed and investigated, by the means of a thermodynamic code
developed into EES software.
The sensitivity analysis of tri-reforming process to the four
main parameters (H2Oratio, mref ratio, b, Tmax) showed that high val-
ues of H2Oratio favour the methane conversion. At fixed H2Oratio, a
value of mref ratio optimising the conversion to H2 was found,
corresponding to about 5% unconverted methane. Without steam
injection to the reformer, almost 30% methane is involved into
the CO2-reforming reaction and only 25% into the steam reforming.
β =30
β =26 β =40
β =36
β =16
β =10 β =10
SCGT-TRIREF H2Oratio =6
β =10 SCGT-TRIREF-R H2Oratio=2
SCGT-TRIREF-R H2Oratio =3
SCGT-TRIREF-R H2Oratio =4
SCGT-TRIREF-R without SEP1 H2Oratio =4
SCGT-TRIREF-R without SEP1 H2Oratio =3
SCGT-TRIREF-R without SEP1 H2Oratio =2
480 500 520 540 560 580 600 620 640
Wsp [ kJ/kgcomp ]
 D. Fiaschi, A. Baldini / Energy Conversion and Management 50 (2009) 2083–2097
With increasing H2Oratio, methane involved in steam reforming in-
creases up to more than 50%, whereas the CO2-reforming reaction
is reduced to 7%. The oxidation involves a slightly decreasing
amount of methane (45.5–41.7%).
With increasing mref ratio and H2Oratio, the reactivity at reformer
is enhanced and the operating mode of TRIREF power cycle shifts
toward a ‘‘more opened” configuration, with a consistent reduction
of CO2 concentration at the stack, resulting into a more difficult
post-combustion capture. At relatively high values of H2Oratio (5–
6) and mref ratio (10–13), the fraction of carbon initially contained
into methane converted to CO2 is relatively high, but at low values
of H2Oratio, carbon is almost equally shared between CO and CO2. In
order to favour the decarbonisation process, a supplemental down-
stream water gas shift reactor was applied. Given the high fluegas
steam content, it has the interesting feature of working without
any external water supply.
The syngas fuel species (CO, H2 and CH4) are highly diluted into
N2 and vapour, resulting into a modest heating value. It is a major
drawback for GT combustion chamber, which requires novel and
not currently available design.
Cycle efficiency is optimised at H2Oratio ranging between 6.5 and
7.5, roughly corresponding to the values maximising H2 conver-
sion. Optimising values of b, ranging between 32 and 38 at differ-
ent mref ratio and H2Oratio between 6.5 and 7.5 were found.
When CO2 removal was added to the power cycle (SCGT-TRI-
REF-R), the performance were optimised at H2Oratio = 3 and
b = 26. Compared to the optimised basic SCGT-TRIREF without
CO2 capture, only two efficiency points are lost, while CO2 emis-
sions are reduced from 428.8 to 43.5 g/kW h. Once again, the very
low syngas heating value makes it very hard to join gas turbines,
even though the increase in turbine inlet flow rate (12%) is lower
than SCGT-TRIREF without CO2 removal.
In order to increase the syngas heating value, a consistent
amount of water vapour was taken out from the stream by the
introduction of a supplemental condensing heat exchanger on
the syngas line. The resulting power cycle showed a consistent effi-
ciency drop compared to the basic configuration.
As a final remark, we can say that the analysis showed that tri-
reforming process has an interesting potential when coupled to
semi-closed gas turbines cycles. The proposed SCGT-TRIREF-R is
an interesting partially chemically recuperated power cycle, with
relatively high efficiency (44%) and low CO2 emissions. No external
water supply is needed and there is the possibility of powerplant
peakload shaving by tuning methane to H2 conversion. Unfortu-
nately, a new design for combustion chamber is required.
References
[1] Figueroa José D, Fout Timothy, Sean Plasynski, McIlvried Howard, Srivastava
Rameshwar D. Advances in CO2 capture technology—the US Department of
energy’s carbon sequestration program. Int J Greenhouse Gas Control
2008;2(1):9–20.
[2] Andersen T, Kvamsdal M, Bolland O. Gas turbine combined cycle with CO2-
capture using auto-thermal reforming of natural gas. In: Proceedings of ASME
turbo expo, 8–11 May, Munich, Germany; 2000.
[3] Cau G, Cocco D. Performance assessment of semi-closed chemically
recuperated gas turbine systems. In: Proceedings of ASME turbo expo, 8–11
May, Munich, Germany; 2000.
[4] Lozza G, Chiesa P. Natural gas decarbonization to reduce CO2 emission from
combined cycles. Part a: partial oxidation. In: Paper 2000-GT-0163, ASME IGTI
conference and exhibition, Munich; 2000.
2097
[5] Lozza G, Chiesa P. Natural gas decarbonization to reduce CO2 emission from
combined cycles. Part b: steam-methane reforming. In: Paper 2000-GT-0164,
ASME IGTI conference and exhibition, Munich; 2000.
[6] Mathieu P, Nihart R. Sensitivity analysis of the MATIANT cycle. Energy Convers
Manage 1999;40(15–16):1687–700.
[7] Carcasci C, Cosi L, Fiaschi D, Manfrida G. Semi-closed HAT (SC-HAT) power
cycle. In: 2000 ASME IGTI conference and exhibition, Munich May 2000, 2000-
GT-154.
[8] Fiaschi Daniele, Tapinassi Libero. Exergy analysis of the recuperative auto
thermal reforming (R-ATR) and recuperative reforming (R-REF) power cycles
with CO2 removal. Energy 2000;29(12–15):2003–24.
[9] Fiaschi Daniele, Gamberi Francesco, Bartlett Michael, Griffin Timothy. The air
membrane-ATR integrated gas turbine power cycle: a method for producing
electricity with low CO2 emissions. Energy Convers Manage 2005;46(15–
16):2514–29.
[10] DOE. Carbon sequestration technology roadmap. Office of Fossil Energy, NETL,
US Department of Energy; 2002.
[11] Rubin Edward S, Chen Chao, Rao Anand B. Cost and performance of fossil fuel
power plants with CO2 capture and storage. Energy Policy 2007;35:4444–54.
[12] Riahi Keywan, Rubin Edward S, Schrattenholzer Leo. Prospects for carbon
capture and sequestration technologies assuming their technological learning.
Energy 2004;29:1309–18.
[13] Aroonwilas Adisorn, Veawab Amornvadee. Integration of CO2 capture unit
using single- and blended-amines into supercritical coal-fired power plants:
implications for emission and energy management. Int J Greenhouse Gas
Control 2007;1(2):143–50.
[14] Bounaceur Roda, Lape Nancy, Roizard Denis, Vallieres Cécile, Favre Eric.
Membrane processes for post-combustion carbon dioxide capture: a
parametric study. Energy 2006;31(14):2556–70.
[15] Song Chunshan. Global challenges and strategies for control, conversion and
utilization of CO2 for sustainable development involving energy, catalysis,
adsorption and chemical processing. Catal Today 2006;115:2–32.
www.elsevier.com/locate/cattod.
[16] Pan W. Tri-reforming and combined reforming of methane for producing
syngas with desired H2/CO ratios. Ph.D. Thesis, Pennsylvania State University,
USA; 2002.
[17] Song C, Pan W. Tri-reforming of methane: a novel concept for catalytic
production of industrially useful synthesis gas with desired H2/CO ratios. Catal
Today 2004;98:463–84.
[18] Corti A, Failli L, Fiaschi D, Manfrida G. Exergy analysis of two second-
generation SCGT plant proposals. In: 1998 IGTI gas turbine conference and
exhibition, Stockholm; June 1998.
[19] Facchini B, Fiaschi D, Manfrida G. SCGT/CC: an innovative cycle with advanced
environmental and peakload shaving features. Energy Convers Manage
1996;38(15–17):1647–53. 1997.
[20] Klein SA. EES (engineering equation solver). Academic – professional V 7.934-
3D (07/15/07), F-Chart Software.
[21] Fiaschi Daniele, Lombardi Lidia. Cellular thermodynamic model for gas turbine
blade cooling. In: Proceedings of ECOS’01 conference, 4–6 July, Istanbul,
Turkey; 2001.
[22] Fiaschi D, Tapinassi L, Lombardi L. The recuperative-auto thermal reforming
and the recuperative-reforming gas turbine power cycles with CO2 removal—
Part I: the recuperative-auto thermal reforming cycle. J Eng Gas Turbine Power
2003;125(October):933–9.
[23] Fiaschi D, Tapinassi L, Lombardi L. The recuperative-auto thermal reforming
and the recuperative-reforming gas turbine power cycles with Co2 removal—
Part II: the recuperative reforming cycle. J Eng Gas Turbine Power
2004;126(January):62–8.
[24] Zhu J, Zhang D, King KD. Reforming of CH4 by partial oxidation:
thermodynamic and kinetic analysis. Fuel 2001;80:899–905.
[25] PEI Power Engineering International. Heavy-duty gas turbines’ fuel pick ‘n’
mix”; 2008. <http://pepei.pennnet.com/display_article/330188/17/ARTCL/
none/none/1/Heavy-duty-gas-turbines%E2%80%99-fuel-pick-
%E2%80%98n%E2%80%99-mix/>.
[26] Lombardi Lidia. Life cycle assessment comparison of technical solutions for
CO2 emissions reduction in power generation. Energy Convers Manage
2003;44(1):93–108.
[27] Lombardi Lidia. Life cycle assessment (LCA) and exergetic life cycle assessment
(ELCA) of a semi-closed gas turbine cycle with CO2 chemical absorption.
Energy Convers Manage 2001;42(1):101–14.
[28] Facchini Bruno, Fiaschi Daniele, Manfrida Giampaolo. Exergy analysis of
combined cycles using latest generation gas turbines. J Eng Gas Turbine Power
2000;122(April).
[29] Chiesa Paolo, Lozza Giovanni, Mazzocchi Luigi. Using hydrogen as gas turbine
fuel. J Eng Gas Turbine Power 2005;127(January):73.