Emission level calculation for water injection in gas turbines.
Popescu, Jeni ; Vilag, Valeriu ; Barbu, Ene 等
1. INTRODUCTION
The technical-scientific development the last decades led to
modernization of terrestrial and aero-transportation constituting
sources of noxious emissions raising acute ecological problems and
forcing the research of fast and efficient limitative solutions.
The gas turbine chemical pollution consists in production of
complete combustion resultant (C[O.sub.2], water), incomplete combustion
resultants (smoke, CO, hydrocarbons), nitrogen oxides (N[O.sub.2],
[N.sub.2]O and particularly NO) and ash particles. Some important
achievements have been reached concerning the reduction of the visible
smoke and the total emission level through advanced injection
techniques, therefore the problem of the emissions produced by gas
turbines can be restricted to reducing the N[O.sub.x] level, mainly
caused by the spectacular increase in thermodynamic parameters, such as
compression ratio, admission temperature in the combustion chamber and
power turbine (Carlanescu, 1998).
The data presented in the paper are the result of the researches
made at INCDT COMOTI Bucharest in the framework of the contracts
21-056/2007 and 22-108/2008 ("Partnership in prior fields"
Programme)
2. INFLUENCES AND REDUCTION METHODS
The problem of emission level reduction is particularly complex due
to the fact that the efficient solutions for decreasing some pollutants
prove to be stimulant for producing others. The concentration levels of
numerous pollutant agents coming from gas turbines are directly
connected to temperature distribution, concentration and residence time
in the combustion chamber, distributions modifying form a chamber to
another and from a working regime to another. The incomplete combustion
products having maximum values on low regimes (low pressures,
temperatures, poor mixture). In these conditions, the N[O.sub.x]
production is also low, to increase fast with the regime, caused by the
fact that improving the combustion conditions leads to nitrogen oxides
production, proportionally to the temperature increase (Khartchenko,
1997).
The ratio and atomization of the fuel-air mixture influences
directly the combustion efficiency, therefore the quantity of the
products. The N[O.sub.x] emission is not affected by the atomization,
but tends to increase with the fuel quantity in the mixture. The
compression ratio influences the air pressure at the combustion chamber
inlet, its increase leading to the decrease of incomplete combustion
products and the increase in N[O.sub.x] production. From the point of
view of the velocity, a quick passing of the air, respectively of the
burning gases, through the combustion chamber involves an increase in
the incomplete combustion products level, while reducing the N[O.sub.x]
generation which needs high temperatures and pressures, as well as
relatively long time.
Improving combustion conditions, by using pre-mixing rooms and low
pressure water injection, along with additives based on manganese,
barium and calcium, and depending on the fuel type (Giampaolo, 2006),
pulls in step the decrease of smoke and soot emission. Another solution
for theoretical high efficiency and emission control consists in
variable geometry combustion chambers, characterized by using or closing
a set of additional orifices, based on the working conditions. This
solution for optimizing the ratio and gases velocity constitutes the
main technological conception for future gas turbines, particularly
turbojets. The most utilized solutions for reducing the emission level
in gas turbines, by internal control, are catalytic combustion chambers
and water or steam injection.
The water injection is a method of cooling the combustion chambers
by adding water or water-methanol to the fuel-air mixture. The initial
water injection cools significantly the fuel-air mixture increasing the
density and the mass of the fluid, therefore the thrust for turbojets
and the power for turboprops, with the subsequent effect, during
combustion, of absorbing the energy and reducing the temperature peak
and the N[O.sub.x] formation (Carlanescu, 1998).
The importance of the data in the present paper comes from
establishing the basics in applying a method of N[O.sub.x] reduction by
completely modifying the fuel system of an existing aviation gas
turbine, as part of a complex project dedicated to decrease the emission
level for gas turbines by involving both a water injection system and a
supplementary firing system.
3. THEORETICAL CALCULATION METHOD
The combustion analysis and the theoretical calculation of
emissions were made with the help of the CEA program developed by NASA Lewis Research Centre. The program allows obtaining the chemical
equilibrium for thermodynamic states imposed by the user through
defining the parameters, completed with included data bases for
substances composition and common thermodynamic data (McBride, 1996).
A theoretical analysis was made for kerosene and methane
combustion, with seven different injected water concentrations, in a
known combustion chamber of an aviation turbo-shaft, in the conditions
of pressure and temperature parameters recommended for the nominal
regime and experimental for the idle regime of the gas turbine. The
combustion reactions of kerosene and methane, with water injection, are
given by equations (1) and (2), where the coefficient "a" of
the water was considered as a function of the fuel coefficient.
[C.sub.12][H.sub.23] + 17.75([O.sub.2]+3.76[N.sub.2]) + a[H.sub.2]O
[right arrow] 11.5[H.sub.2]O + C[O.sub.2] + 66.74[N.sub.2] (1)
C[H.sub.4] + 2([O.sub.2]+3.76[N.sub.2]) + a[H.sub.2]O [right arrow]
2[H.sub.2]O + C[O.sub.2] + 7.52[N.sub.2] (2)
The maximum temperature of the original gas turbine, of 1063 [K],
was tracked in order to establish the necessary air/oxidant excess for
the two regimes, along with the concentrations of the combustion
resultants, particularly N[O.sub.x]. Knowing the recommended kerosene
mass flow rate, the theoretical air quantity necessary for the
combustion process and the air/oxidant excess, the air/oxidant mass flow
rate and the fuel mass flow rate can be calculated for each case, in
order to be further used as input data in numerical simulations of the
combustion process.
4. NUMERICAL APPROACH
For the most relevant reduction of emissions, the methane
combustion at nominal working regime of the gas turbine, another method
allowing the calculation of maximum temperatures and N[O.sub.x] level is
applied, consisting in simulating the combustion process in CFD environment.
The combustion chamber of the aviation gas turbine, along with the
injection system, initially designed for kerosene, were geometrically
adjusted, the annular combustion chamber, well known, was next divided
in eight identical slices, corresponding to the injectors number, for
reducing the necessary time and calculating resources. The meshing of
the used geometry was made in three variants, based on fineness, the
evolution being required by the necessity of obtaining most relevant
results.
The most important numerical approaches are characterized by the
definition of the fluid material of the domain, as a reactive mixture of
perfect gases, where the reactions are controlled by the methane-air
combustion, with the tracking of the NO level from the CFD code internal
library (Ansys, 2007). The pressure and temperature conditions, as well
as the mass flow rates and the mass fractions of the reactants are
dictated by the known regimes of the gas turbine.
The results are presented in the following figures as temperature
plots on a radial plan defined in the centre of the domain, Fig. 1 and
3, and on the outlet of the combustion chamber, Fig. 2 and 4, for the
methane combustion at nominal regime without water injection,
respectively with water injection. For the second approach, the purpose
was to restore the maximum temperature of the gases on the outlet of the
combustion chamber. The figures show the shape of the flame and the
maximum temperature inside the combustion chamber. The post-processing
of the results also includes the calculation of the thermodynamic
parameters for regions of interest, particularly the outlet of the
combustion chamber.
Table 1 presents the comparative results of the theoretical and
numerical calculation from the point of view of the obtained maximum
temperature and N[O.sub.x] emission level.
[FIGURE 1 OMITTED]
[FIGURE 2 OMITTED]
[FIGURE 3 OMITTED]
[FIGURE 4 OMITTED]
5. CONCLUSIONS
Comparing the theoretical and numerical results of the methane
combustion, with and without water injection, the temperature is
obtained at nearly the same value, while the N[O.sub.x] emission level
has different values, however on the same order of magnitude. This is
caused by several hypotheses used for the two methods. The theoretical
calculation involves NO, NO2 and NO3 emissions, while the numerical case
only calculates the NO. The injection area in the combustion chamber is
not defined, both methods considering only the mass fractions of the
reactants, slightly different following a recalculation of the air
excess. Another important factor is that, even for stable processes, the
experiments and measurements made on international level proved an
unstable NO concentration.
The paper establishes two methods for calculating the N[O.sub.x]
emission level from gas turbines, a theoretical one by using the CEA
NASA program and a numerical one by using a commercial CFD code, with
comparable results. Future work will consist in defining a water
injection system suitable for the studied gas turbine, testing the
modified gas turbine on experimentation bench and comparing the
experimental results with the already obtained ones.
6. REFERENCES
Carlanescu, C. (1998). Turbomotoare. Fenomenologia Producerii si
Controlul Noxelor (Gas Turbines. Phenomonology of Production and
Emission Control), Ed. Academiei Tehnice Militare, Bucharest, Romania
Khartchenko, N.V. (1997). Advanced Energy Systems, First Edition,
Taylor & Francis, ISBN 978-1560326113, Bristol, UK
Giampaolo, T. (2006). Gas Turbine Handbook: Principles and
Practices, Third Edition, The Fairmont Press, Inc., ISBN 0-88173-515-9,
London, UK
McBride, B.J. & Gordon, S. (1992). NASA RP-1271, Computer
Program for Calculating and Fitting Thermodynamic Functions, NASA Lewis
Research Center, NSN 7540-01280-5500, Cleveland, Ohio, USA
*** (2007) Release 11.0 Documentation for Ansys, SAS, IP
Tab. 1. Comparative results for methane combustion
Without water injection With water injection
Results Theoretic Numeric Theoretic Numeric
Temp [K] 1063 1069 1063 1059
N[O.sub.x] [ppm] 53.38 29.03 25 11.02