1

Hydrogen Combustion Turbine Cycles

A. Miller1, J. Lewandowski1, K. Badyda1, S. Kiryk1, J. Milewski1

Abstract
Thermodynamics of hydrogen combustion turbine cycles is presented. The GRAZ,
TOSHIBA, WESTINGHOUSE and MNRC cycles are analyzed. Research was performed in
co-operation with the Mechanical Engineering Laboratory Tsukuba.

Introduction
Power generation systems based on hydrogen can be, in near future, an important alternative
to conventional power systems based on combustion of fossil fuels. Hence hydrogen as a
clean fuel is a subject of interest of many research institutions all over the world. National and
international projects for the hydrogen utilization have been progressed in several countries.
One of the most important program here is the Japanese: the International Clean Energy
Network Using Hydrogen Conversion WE-Net. The WE-Net Program predicts an

Institute of Heating Engineering, Warsaw University of Technology (www.itc.pw.edu.pl)

2
implementation of the Hydrogen-Fuelled Combustion Turbine Cycle (HFCTC) as a new
energy source for power sector. In this respect, a configuration and performance study of the
HFCTC was conducted. The research was performed in terms of the Polish research grant
Nr 8T10B00918 and in cooperation with the Mechanical Engineering Laboratory Tsukuba,
Japan (at the present time: the Institute of Energy Utilization, AIST). Further on it is presented
the results obtained in the range of hydrogen turbine cycle thermodynamics.

The cycle selection

High costs of hydrogen production cause that the thermal efficiency over 60% HHV (it means
about 71% LHV) of the energy generation cycle is a must in WE-Net system. Nowadays,
thermal efficiency of the most advanced combined cycles (natural gas fuelled) is close to
60% LHV and it is comparable to 50,4% HHV if hydrogen as a fuel is used.
It is obvious that efficiency should be increased of about 10 percent points, which is an
equivalent of about 20 % in comparison to the most efficient contemporary power plant units.
It is a very serious technological challenge to meet such requirement (a qualitative change not
only quantitative). Besides increasing a working medium temperature at the turbines inlet to
1700 oC (at present it is a level of 1500 oC), it is essential to implement a new approach
(different than traditional one) to both conceptual design of the system (its configuration and
working parameters) and detailed construction solutions.
Until now several concepts of the HFCTC have been proposed, out of which the most
important are as follows:
I. Combined Steam Cycle with Steam Recirculation prepared by prof. H. Jericha (Technical
University of Graz) by convention called the GRAZ cycle [3], [4], [];
II. Direct-Fired Rankine Steam Cycle (New Rankine Cycle) which was studied in the
following variants:

proposed by Toshiba Co., by convention called the TOSHIBA cycle [], [];

proposed by Westinghouse Electric Co., by convention called the WESTINGHOUSE

cycle [];

Modified New Rankine Cycle, as authors own concept, by convention called the
MNRC cycle [].

A common feature of the a.m. cycles is that only one working medium (steam) is used for
both the topping as well as bottoming cycle. A replacement of an external firing (as in the
Rankine steam cycle) by a direct firing (similar to gas turbines or piston engines) makes it
possible. Main assumption made here is stochiometric combustion of hydrogen and oxygen
mixture. This combustion takes place inside a stream of cooling steam, which cuts down
combustion temperature to 1700 oC. It is assumed also that hydrogen and oxygen at the
ambient temperature are available at a pressure level that allows providing them to combustor.
It means that hydrogen would be provided as cryogenic liquid, however, cryogenic energy
could be utilized for pure oxygen production in an air-separator unit.
GRAZ, TOSHIBA, WESTINGHOUSE and MNRC cycles were analyzed in comparable
conditions to evaluate their performance. The analysis was undertaken in the same, specific
conditions with the same assumptions and property tables. Research results on a.m. cycles,
which were published so far, do not provide an opportunity for such comparison because of
incomparable and/or not clear conditions and assumptions set taken for the research. Cycles
without cooling system were only taken into account in order to create an opportunity for an
explicit evaluation.

4
Cycle performance in nominal conditions.
Analysis of the HFCTC in nominal conditions was conducted for assumed values of
parameters as follows:

compressor stages group internal efficiency 0.9

turbine stages group internal efficiency 0.9

combustor efficiency 0.99

heat exchanger pressure loss 0.043

combustor pressure loss 0.05

pump efficiency 0.9

electric generator efficiency 0.99

cycles overall mechanical efficiency 0.99

overall power output 500 MW

temperature after combustor 1700 oC

condenser pressure 0.005 Mpa

condensate temperature 33 oC.

Working medium parameters were determined using the ITC-PAR calculation routines [11]
based on the NIST/ASME steam property tables.
For cycle performance calculations appropriate computer program was elaborated. Pressure
splits between turbine parts were optimized according to maximum overall HHV thermal
efficiency. Efficiency of analyzed cycles was determined in relation to the Carnot cycle
(c), the Rankine cycle (R, high heating value of fuel (HHV) and low heating value of fuel
(LHV) as:

C = 1

Td
100%
Tg

(1)

Where Td temperature (K) of heat sink and Tg temperature (K) of heat resource;

LHV =

P
100%
m LHV

(2)

where P denotes the internal power output of the cycle and m hydrogen mass flow rate of
hydrogen supplied to the cycle

HHV =

R =

P
100%
m HHV

(3)

P
100%
P+Q

(4)

where Q denotes the heat amount taken out from the cycle. Efficiencies HHV and R should
be equal (at exact calculations).
Besides the turboset efficiency was determined:
el = LHV * g * m

(5)

where g generator efficiency and m overall mechanical efficiency.

GRAZ cycle

Fig.1. Scheme of GRAZ cycle

The GRAZ cycle proposed by prof. H. Jericha from Graz University of Technology
(Austria) [][] is an original connection of Joule and Rankine cycles (Fig.1). In the zone of
high parameters the Joule cycle is utilized in a semi closed configuration, coupled with the
Rankine cycle, which operates in low parameters zone. The Rankin cycle plays here
simultaneously the role of heat sink for the Joule cycle. The hydrogen combustion chamber is
the high temperature source of heat. The original idea which differs the GRAZ cycle from any
others, is application of extraction of partially cooled working medium from the Joule cycle
and use of it as a working fluid in the Rankine cycle. Efficiency rise is obtained here due to
significant diminishing of compression work of working medium in the Joule cycle.
Heat flow scheme and h-s diagram of the Graz cycle is shown in Fig.1. and Fig.2.

Fig.2. H-s diagram of GRAZ cycle.

Toshiba cycle.

Fig.3. Scheme of TOSHIBA cycle.

The Toshiba cycle belongs to the group of steam cycles with direct combustion direct fired
Rankin steam cycles. Toshiba cycle is also called MORITS cycle Modified Rankin Cycle
Integrated Turbine System. Scheme of the cycle is shown in Fig.3. It consists of four turbine
parts, where two of them: the first and the last one (HHP and LPT) do not have combustion
chamber before. The regenerator (heat recovery boiler) is located before the last turbine part
in which superheated steam is produced. Calculation result and h-s diagram are shown in
Fig.4.

Fig.4. H-s diagram of TOSHIBA cycle.

WESTINGHOUSE cycle.

Fig.5. Scheme of WESTINGHOUSE cycle.

The Westinghouse cycle is also a variant of steam cycle with direct combustion. In
comparison with Toshiba cycle it does not poses the turbine part operated in the highest
pressure region. So it can be classified as a kind of new Rankin cycle with single reheat.
Calculation results and h-s diagram are shown in Fig.5 and 6.

Fig.6. H-s diagram of WESTINGHOUSE cycle.

10

MNRC cycle

Fig.7. Scheme of MNRC cycle.

The MNRC cycle is kind of own concept that has arisen after analysis of earlier mentioned
and published cycles. Contrary to other cycles, heat recovery steam generator (HRSG) is
placed here after the last turbine stage group in low-pressure zone. This concept has been
already mentioned in elaboration []. Arrangement of the cycle allows adding another reheat
stage before the low-pressure turbine stage group. As a result a very high thermal efficiency
of the cycle is achieved. Flow diagram and calculation results of the MNRC cycle are shown
on Fig.7 , and h-s diagram of the process illustrates Fig. 8.

Fig.8. H-s diagram of MNRC cycle.

11

Confrontation of cycle performance.

The main performance parameters of analyzed cycles are shown in Tab.1. Comparison and
cycles evaluation could be done from different points of view. Basic criteria here is maximal
overall thermal efficiency, at least 60 %HHV for the 500 MW class unit.

Tab. 1

Cycle

GRAZ

TOSHIBA

WESTINGHOUSE

MNRC

pmax [bar]

350

380

250

250

tmax [oC]

1700

1700

1700/1600

1700

Inner Power [MW]

513

513

513

513

LHV [%]

70,8

71,2

74,0/72,8

79,0

HHV [%]

59,5

59,8

62,2/61,2

66,4

Specific Power [kJ/kg]

2202

3331

3489

4706

Electricial Power [MW]

500

500

500

500

elLHV [%]

69,0

69,4

72,2/71,0

77,0

elHHV [%]

58,0

58,3

60,6/59,7

64,7

t [oC]

1700

1700

1700/1600

1700

p [bar]

50

73

250

250

p [bar]

350

343

277

277

t [oC]

650

876

517

463

[MW]

315

329

256

165

Parameter

The most
loaded element
temperature
The most
loaded element
pressure
The heat
exchanged/HRSG
heat load

As it is shown, efficiencies of GRAZ and TOSHIBA cycles are practically on a limit of the
WE-Net program requirements. The WESTINGHOUSE cycle fulfils these requirements with
some overlap and the MNRC cycle is far beyond. It is possible to increase the GRAZ cycles
efficiency by adding to the basic configuration a compressors inter-stage cooling and
recuperation (high-temperature regeneration). All cycles discussed achieve very high specific
power (related to a maximal mass-flow in the cycle). Those are values much higher

12
(extremely of a range of magnitude) than ones achieved by contemporary heavy-duty gas
turbines (450500 kJ/kg), combined steam-gas cycles (600700 kJ/kg) or steam turbo-sets
(12001400 kJ/kg). It would be possible then to build extremely compact power units with
minimal usage of construction materials.

Sensitivity study
Examination results of influence of main cycle parameters change on efficiency of
TOSHIBA, WESTINGHOUSE and MNRC cycles are shown further on. The GRAZ cycle
was analyzed in their respect in Japan [][]. All cycles are relatively little sensitive on change
of main parameters. It depends especially the MNRC cycle. It allows for instance to reduce
combustion temperature from 1700 oC to 1300 oC maintaining the required 60% HHV overall
thermal efficiency.
The results of sensitivity study are presented from Fig.9. to Fig.11.

13

Fig.9. Sensitivity study of TOSHIBA cycle.

14

Fig.10. Sensitivity study of WESTINGHOUSE cycle.

15

Fig.11. Sensitivity study of MNRC cycle.

16

Summary
All the HFCTC cycles with direct firing which were studied within the hereby research
(500MW class units) have shown very high thermal efficiency (minimum 60%HHV,
71%LHV) what is far above the performance of the known cycles. It has been fully confirmed
a possibility to achieve the efficiency at least 10 percentage points higher than the efficiency
of the most efficient contemporary power units (which is about 20% more). Similarly all the
HFCTC cycles have very high elementary power (22004700 kJ/kg), which is many times
higher (extremely for the range of magnitude higher) than performance of the contemporary
gas or steam turbines or combined cycles. Extremely compact power units with minimal
usage of construction materials would be possible to be built then. Considering, additionally,
a fact that the HFCTC cycles almost totally eliminate CO2 and NOx emission this solution can
be recognized as an interesting alternative for the future power technology development trend
comparing to conventional power technologies.

Reference
[1] Reseatch on Hydrogen-Fuelled Gas Turbine Polish Research Grant KBN, Warsaw 2001
[2] IKI N., HAMA J. TAKAHASKI S., MILLER A., KIRYK S.: Future Hydrogen
Technologies in Power Engineering and Semi-Closed Gas Turbine System. Politechnika
Warszawska, Prace Naukowe, s. Mechanika z.181 (1999).
[3] JERICHA H.: A new combined gas steam cycle promising up to 60% thermal efficiency.
15th International Congress on Combustion Engines, CIMAC 1984, Paris; 1984
[4] MORITSUKA H., KODA E.: Hydrogen-Oxygen Fired Integrated Turbine Systems Comparison on MORITS and GRAZ. Proceedings of the International Gas Turbine
Congress, Kobe, 14-19 Nov. 1999, IGTC'
99Kobe TS-18, page 401-404, Copyright 1999 by
GTSJ; 1999

17
[5] OKAMURA T., KAWAGISHI H., KOGA A., ITO S.: Development of Hydrogen-Oxygen
Combustion Turbine Cooling Blades in WE-NET Project. Proceedings of the International
Gas Turbine Congress, Kobe, 14-19 Nov. 1999, IGTC'
99Kobe TS-46, page 599-606,
Copyright 1999 by GTSJ; 1999
[6] DESIDERI V., ERCOLANI P., YAN J.: Thermodynamic Analisis of Hydrogen Cmbustion
Turbina Cycles. International Gas Turbine Congerss & Exibition, New Orlean, 2001-GT-95;
2001
[7] MOURI K.; R&D of 1700C Hydrogen Combustion Turbine in WE-NET Program;
Proceedings of the International Gas Turbine Congress, Kobe, 14-19 Nov. 1999,
IGTC'
99Kobe TS-11, page 351-358, Copyright 1999 by GTSJ; 1999
[8] SUGISHITA H., MORI H., UEMATSU K.: A study of thermodynamic cycle and system
configurations of hydrogen combustion turbines. Proceedings of the 11th World Hydrogen
Energy

Conference,

Stuttgart

Germany,

pp

1851-1860,

June

23-26

1996

http://www.enaa.or.jp\WE-NET\ronbun\p_1996_e.html"; 1996
[9] UEMATSU K., MORI H., SUGISHITA H.: Topping Recuperation Cycle for Hydrogen
Combustion Turbine in WE-NET. "RAN Extended Abstracts, NAGOYA Japan , 1998
http://www.enaa.or.jp/WE-NET/ronbun/1998/16/1698.htm"; 1998
[10] BANNISTER R., NEWBY R.A., YANG W.-C.: Development of a Hydrogen-Fueled
Combustion Turbine Cycle for Power Generation. ASME, International Gas Turbine &
Aeroengine Congress & Exhibition, Orlando, Floride - 2-5 June 1997, 97-GT-14 ; 1997
[11] Kiryk S., Miller A.: Obliczanie wa ciwo ci termodynamicznych pary i wody w wysokich
parametracyh - zestaw procedur ITC PAR. Zeszyty Naukowe Politechniki Warszawskiej s.
Mechanika, z. 195 (2001).