1. science
2. ecology
3. pollution

THE AMERICAN SOCIETY OF MECHANICAL ENGINEERS
345 E. 47th St., New York, N.Y. 10017
97-GT-286
The Society shall not be responsible for statements or opinions advanced in papers or cibbussion at meetings of the Society or of its Divisions or
Sections, or printed in its publications. Discussion is printed only if the paper is published in an ASME Journal. Authorization to photocopy
material for Internal or personal use under circumstance not Wang within the fait' use -provisions of the Copyright Act is granted by ASME to
libraries and other users registered with the Copyright Clearance Center (eCC) Transactional Reporting Service provided that the base lee of $0.30
per page Is paid directly to the CCC, 27 Congress Street Salem MA 01970. Requests for special permission or bulk reproduction should be addressed
to the ASME Tedmical Putishing Department
Copyright 0 1997 by ASME
All Rights Reserved
Printed in U.S.A
ENVIRONMENTAL INFLUENCE ON THE THERMOECONOMIC OPTIMIZATION
OF A COMBINED PLANT WITH NO ABATEMENT
11 11 111 111,11E1 111 1111 111
A. Agazzani, A.F. Massardo
University of Genova
Istituto di Macchine e Sislemi Energetici
Genova, Italy
CA. Frangopoulos
National Technical University of Athens
Department of Naval Architecture and Marine Engineering
157 10 Zografou, Greece
ABSTRACT
Methods to analyze, improve and optimir , thermal energy sys_
tans have to take into account not only energy (exergy) consumption
and economic resources but also pollution and degradation of the
environment. The tam "environomics" implies a method which takes
thermodynamic, economic and environmental aspects systematically
into consideration for the analysis and optimization of energy systans For optimization of energy systems, the environmental aspects
are quantified and introduced into the objective function.
In this particular work, the environomic approach is followed for
the analysis and optimal design of a combined-cycle plant In addition to the basic configuration, two alternatives for NO* abatement
are studied: Selective Catalytic Reduction (SCR) and steam injectioa
The optimization problem is solved for each configuration and the
results are compared with each other. The effect of the unit pollution
penalties and of the limits imposed by regulations is studied. Some
general conclusions are drawn.
NOMENCLATURE
c_ unit cost [Ski] or unit pollution penalty [SAcg]
Cs) purchased cost [S]
F objective function
fp pollution penalty factor (harmfulnms factor)
g inequality constraint function
h equality constraint function
k constant related with the specific properties of a selected SCR
catalyst
M molecular weight.
th mass flow rate (kg/s]
0 measure of pollution
p pressure [Pa]
SV SCR space velocity [11 1]
t
primary zone residence time [s]
T temperature [K]
V volume [m3]
volumetric flow rate (m 3/s]
x, independent decision variable
x set of independent variables
y, dependent variable
y set of dependent variables
Z
annualized capital cost (including investment, depreciation,
maintenance, etc.) (S/s]
a intensive property of the pollutant
ac. intensive property of the pollutant in the environment
5 critical limit of the intensive property (harmfulness limit)
pressure ratio
•
degree of abatement of the pollutant
rate of costs [Vs]
-
Subscripts
g
flue gas
w water
INTRODUCTION
The thermal-energy systems consume not only fuel and economic
resources but they have also an harmful effect on the environment
Pollutant emissions from combustion have become of great concern
due to their impact on human health and nature. Restrictions in specific pollution components are introduced in several countries.
Fmission taxes are also introduced to restrict air pollutioa
Therefore methods and techniques utilized for the optimintion of
thermal-energy systems have to deal not only with energy consumption and economics but also with the pollution and degradation of the
environment
Presented at the International Gas Turbine & Aeroengine Congress & Exhibition
Orlando, Florida — June 2–June 5,1997
Downloaded From: http://proceedings.asmedigitalcollection.asme.org/ on 12/29/2014 Terms of Use: http://asme.org/terms
The sensitivity of the particular environment to a certain pollutant is taken into consideration by the pollution penalty factor, as it
will be explained in one of the following sections.
The optimization problem in environomics, considered under
steady-state conditions, is mathematically stated as follows:
Environomics was introduced by Frangopoulos (1991, 1992) as
an extension of thermoeconomics, in order to include environmental
aspects in the thermoeconomic analysis and optimization of energy
systems. In addition to flows of energy, exagy, and costs, flows of
other consumed resources as well as flows of pollutants are considered, degradation of the environment is taken into consideration by
treating the environment as a consumed resource.
The objective of this work is to analyze and prove the potential of
environomics by means of application examples. With the introduction of environmental costs due to pollution, which depend on essential parameters such as unit pollution penalties (i.e. costs imposed by
society or governmental institutions) and limits imposed by regulations, several scenarios can be analyzed for the same plant As application examples, two NO. abatement systems for a combined-cycle
plant have been selected: steam injection and SCR.
The steam injection is a method frequently utilized for the NO
abatement, since it is easily feasible and suitable for traditional existing combustors. Water or steam injection decreases the flame temperature and lowers the NO 1 emissions but involves several remarkable adverse effects: it produces an increase in carbon monoxide
(CO) and unburned hydrocarbon (UHC) emissions, and causes some
losses of thermal efficiency, water consumption, if considerable,
causes supplying problems in certain localities and an increase of the
plant and management costs because water must be treated and
demineralized carefully, the combustor useful life decreases because
of foaling and scaling. Furthermore, in locations where the ambient
concentration of NO1 exceeds the standards, emission regulations
may become too stringent to be met by modification of the combustion process with the steam injection. Therefore, since emission with
steam injection normally reaches values of around 40-60 ppmvd, this
situation could require different treatment of the exhaust gas to remove NO..
Selective Catalytic Reduction (SCR) is now commonly used to
control NO1 emissions from combustion sources in countries with
strict emission regulations, even if its capital cost is greater than a
system with steam injection. For gas turbines, NO\ emission reduction to as low as 10 ppmvd is sometimes required by the authorities,
and is generally easily achieved by the SCR systems.
Several of these aspects have been taken into consideration here
from an environomic point of view comparison among systems with
and without NQ, abatement is carried out and the influence of the
two aforementioned essential parameters is analyzed and discussed
in depth.
+ Ea4
min F =
„antiadi
1.1
i= 1
subjected to Igx,y)=0
g.(x,y)50
where
x={2e}
Y=IY1)
is =zs (x,y)
:Zits.%
(0
j=1,.., number of equality constraints
k=1,.., number of inequality constraints
i=1,.., number of independent variables
i=1,.., number of dependent variables
tS
pallution = rpollition (Ia3r)
item, = rrn,„„ x (x,y)
The first term in the objective function, Eq. (1), represents the capital
costs of the system components (including investment, depreciation,
maintenance, etc.), the second term is the cost of resources bought by
the system, the third term is the cost due to pollution of the environment resulting from the operation of the energy system (Frangopoulos
and von Spakovsky, 1993). If the revenues of all the system products
are included (fourth term), then F is the negative of the system profit
SYSTEM ANALYSIS AND SIMULATION
A modular simulator tool for thamoeconomic analysis of thermal-energy systems has been utilized here; the code, called TEMP ThermoEconomic Modular Program, has been presented in detail by
Agazzani and Massardo (1996). The tool has been aimed at the following targets: thermodynamic and exergy analysis, thennoeconomic
analysis, optimization.
The thamoeconomic technique utilized, is similar to the T.F.A.
(Thermoeconomic Functional Analysis) developed by Fran,gopoulos
(1983, 1991) and the E.F.A. (Engineering Functional Analysis) developed by von Spakovsky (1993). By means of a junctional productive analysis in which each component has several inputs and one
output (product), the code allows the functional exergy flows (i.e. the
productive relationships, in addition to the physical exergy flows,
expressed by the extensive variable exagy) among the components to
be calculated. Besides, marginal and average unit costs of each functional many flow are determined, and therefore the global thermoeconomic performances of the components (internal economy) are
defined. The calculation of the marginal and average unit costs is
carried out only at the end of the optimization procedure, corresponding to the optimum point (Agazzani et al., 1995).
The code has been expanded by means of the introduction of a
new module (the SCR), the modification of the combustor module in
order to insert the injected steam and the pollutant emissions estimation and the introduction of an apt module for the make-up water.
Furthermore, damage costs due to the pollutant emissions have been
enclosed.
The modularity of the code allows the study of several configurations of multi-pressure combined plants (Agazzsni and Massardo,
1996). However, since the aim of this work is to present and clarify
the methodology, a relatively simple plant has been selected it is a
combined-cycle plant of a single-pressure heat recovery steam generator, equipped with an SCR abatement unit inside the HRSG or
BRIEF PRESENTATION OF THE ENVIRONOMIC APPROACH
In the environomic analysis and optimization of energy systems,
the effect of the system construction and operation on the environment is taken into consideration. In order to do so, the effect is quantified and a penalty is imposed. To cope with the penalty, pollution
abatement equipment and techniques can be introduced, the perfomtonce of which is measured by the degree of abatement
8=—
where pi is the initial pollution, before abatement, and p is the pollution after abatement A discussion on appropriate measures of pollution, p , can be found in the literature (Frangopoulos, 1991b, 1992,
Frangopoulos and von Spakovsky, 1993).
2
Downloaded From: http://proceedings.asmedigitalcollection.asme.org/ on 12/29/2014 Terms of Use: http://asme.org/terms
Steam injection model
In the model of steam injection utilized for the NO, emissions
abatement, the steam has been considered a perfect gas and injected
inside the combustor (into the primary zone) in order to decrease the
combustion temperature. In such a case, in accordance with the correlations above mentioned, the NO, emissions decrease but the CO
emissions increase (see the opposite influence of the temperature into
the two correlations). The values of NO, and CO emissions are dotamined by calculating the new value of the stoichiometric combustion temperature (Ta) or the primary zone combustion temperature
(Tpe) and then utilizing the same correlations mentioned above. A
cost of 0.5 Vm3 has been set for the demineraliz.ed make-up water
into the steam plant (its mass flow rate is equal to the steam mass
flow rate injected into the gas turbine).
SCR model
Selective catalytic reduction with ammonia is one of the most effective techniques for removing NO, from combustion gases. The
ammonia, injected into the exhaust gas upstream from a catalytic reactor, is applied as a suitable reduction medium on a technical scale.
The catalyst bed is situated in the heat recovery steam generator
(Fig. 1) at the appropriate location to provide the optimum reaction
temperature. Ammonia reacts with NO and 1402 as follows:
6 NO2 +8 NH3 -+12 1120 + 7 N2
6 NO + 4 NH3 *61-1201-5 N2
When oxygen is present, the corresponding reactions are:
2 NO2 + 4 NH3 +02 —s 1-120 + 3 N2
4 NO +4 NH3 +02 —s 6 H20 + 4 NI
Since oxygen is usually present in combustion gas, and practically all
of the NO, is present as NO, the Nib/NO, stoichiometric ratio is 1.0,
i.e. one mole of NH3 is required for removing one mole of NO,. The
homogeneous reactions occur in a narrow temperature range at about
800 °C, and are accompanied by undesirable side reactions; besides,
it is difficult to achieve a high level of NO, removal with a process
based on homogeneous reduction. Utilizing a catalyst, the operating
temperature can be reduced to the range of 220-420 °C (the temperature depends upon the type of catalyst), but with an increase in capital cost
The NO, removal is called selective because the reduction
chemicals should react exclusively with the pollutant
In the presence of oxygen, the NO reduction by ammonia over ceramic catalyst is in many cases in accordance with an Eley-Rideal
mechanism (Weber etal. 1991). Due to this reaction mechanism, the
NO conversion is independent of the initial NO concentration. Considering the catalytic reactor as an ideal flow tube, the NO reduction
for a given type of catalyst must be written as:
In( 1-8220,,)=-IrJSV
(4)
where No„ is the NO, degree of abatement and SV is the space velocity [WI. SV is given by the ratio of the volume flow rate to the
catalyst volume and is thus proportional to the reciprocal of the mean
residence time of the gas in the catalyst The rate constant k is related with the specific properties of a selected catalyst and uerally
shows an exponential dependence according to the Arrhenius equation. On the other hand, it has to be taken into account that with increasing temperatures the oxidation of NH3 into NO may be enhanced as well. Taking into consideration these two opposite aspects,
the degree of NO conversion reaches maximum values near temperatures of 250-300 °C. This allows the SCR to be inserted between the
evaporator and the economizer inside the FIRSG.
The constant k of industrially applied catalyst is about 7500 W I at
optimum reaction temperature (above 300 *C).
TO lrfACIC
Fig. 1. Combined plant with an SCR abatement unit inside the
HRSG or steam injection for removing NO,; the greyfilled units represent components with a capital cost.
with steam injection inside the combustor chamber of the gas turbine
for removing NO, (Fig. 1).
g missions estimation
In this work, a gas turbine with a typical diffusion flame combustor and therefore with high NO,, emissions has been considered. The
formation of NO„ and CO emissions has been established employing
correlations. Since the emissions from a combustor depend on the
type and the geometry, in order to not complicate a lot the thermoeconomic study, some parameters have been fixed for all the calculations in particular, the primary zone residence time (which depends on the geometry) and the excess air in the same zone (which
has influence on the primary zone combustion temperature) have
been fixed.
The following correlations have been employed for the NO, and
CO emissions estimation (Rizk and Mongia, 1993):
NO
71100 -us (apru
•—
(2)
5.10- 10 14 (t — 03. t ,)" expEH p
CO r- 018 .103
ezarE7 1100y
Ap
— p2 (t — 0.4 t ) • (— Ts]
(3)
where the emissions are in g/kg fuel, te is the fuel evaporation time
(for a gaseous fuel this is fixed equal to zero), Tu is the stoichiometric flame temperature in degrees Kelvin, p the combustion pressure
in Pa, Ap/p the nondimensional linear pressure drop; t is the primaryzone residence time in seconds and has been fixed equal to 2-10 -3 s;
Tp: is the primary zone combustion temperature in degrees Kelvin (in
order to calculate the temperature Tpz , an air aCeSS of about 3% has
been assumed in the primary zone).
The unit of the emissions is 'g/kg fuel' and therefore the mass
flow rate of fuel which is burnt, must be known. The combustor
model considered here is a diffusion-flame model and almost all of
the fuel is burnt into the primary zone. Therefore, the mass flow rate
of NO, and CO at the exit of the combustor has been roughly calculated by means of the product between the value obtained with the
correlation and the mass flow rate of fuel utilized by the gas turbine.
3
Downloaded From: http://proceedings.asmedigitalcollection.asme.org/ on 12/29/2014 Terms of Use: http://asme.org/terms
The amount of ammonia to be injected into the flue gas stream is
usually controlled by the flow rate of the flue gases and the inlet gas
NO concentration. In order to avoid significant clean gas concentration, the molar NH../NO1 ratio is voiatly selected lower than 1 (about
0.9-0.95). Therefore the ammonia mass flow rate is given by
151 NH, =
MNH
8 NOx .1i1NOchict at/
"NO
r
4.5E02
40E-CQ
15E02
3.0E02
25602
(5)
1.0E02
5.0E43
ammonia molecular weight
Ma0=30.0061 NO molecular weight
molar Nib/NO. ratio
NO mass flow rate at the SCR inlet [kg/s]
IhNO:cs,
The pressure drop of an industrial ceramic catalyst which has
been designed to remove more than 90% of NO is in the range of
about 500 to 900 Pa On the basis of this data, the pressure loss of an
SCR can be so estimated:
k-700
Ap —
(6)
— 0.9)
The catalyst cost is estimated by means of its volume. This last is
given by
In(l —8)
Vc,„ = —3600 v —
(7)
k
• where ts is the flue gas volumetric flow rate. The catalyst cost Ow
depends on its composition (base metal or noble metal) since the
catalyst cost is about 70% of the SCR total cost (Eskinazi et al.,
1989), the SCR cost can be so estimated (for a noble metal catalyst):
M=17.030
0.2
Fig.
cr-V.N/0.7
+ +it,
0.4
0.13
KO degree V abate mon!
2. Influence of 8Nox
0.8
on the SCR capital cost
pollution of the environment resulting from the NO and CO emissions ( FsNo ,
The selected independent decision variables are (see Table 1):
pressure ratio and maximum temperature of the gas turbine, pressure
of the produced steam, NO degree of abatement (for the plant with
the SCR), mass flow rate ratio between steam injected and fuel (for
the plant with steam injection). It would be useful to introduce a degree of abatement also for the steam injection system; however, the
mass flow rate ratio between steam injected and fuel is a variable
often utilized in literature.
Table 2 reports the quantities considered known and fixed once
and for all at the beginning of calculation; other hypotheses done for
the calculation are reported in Agazzani and Massardo (1996). The
net electrical power has been fixed equal to 60 MWe. The maximum
allowable steam temperature has been set equal to 565 °C. Two inequality constraints have been taken into consideration during the optimintion: outlet turbine steam quality (>0.86) and stack gas temperature (>100 °C).
The equality constraints, imposed by the physical and economic
model of the system, consist of mathematical relations describing the
thermodynamic performances as well as the costs of components and
other economic parameters as functions of the independent variable
Optimization
The goal of the optimization for both analyzed plants, is to
minimize the rate of the total cost (including construction and operation) of the system:
+
ammonia cost
0.0E100
(8)
where c 1 16245 Sim' (Frey et al., 1991).
The ammonia unit cost cwa, is 0.230 S/kg (Rosenberg et al.,
1992).
As an example of a SCR system cost, Fig. 2 shows the influence
of the NO degree of abatement for a unit with a flue gas mass flow
rate of 100 kg/sand 150 ppmv (15% 02, dry) of NO at the inlet
mint!, (x,yEtsi
flue goo mats flow rats SOO kg/s
Wm at Inlet = 150 pummel
20E02
1.5E02
where
C:at = C:11 /0.7 =
Vs
Table 1. Decision variables.
(9)
Decision variables
Optimization range
max gm tants= MI
rinsaure ratio
stem went= IMPal
11004403
10.30
Nat deg= of abatement (for the plant with the SCR)
man flow nite ntio hetweeo steam injected and fuel
(for the Mimi with steam injection)
where
tles = Crod lil gia LHVAid
6-14
0.0.95
0-2
Table 2. Nominal values of main parameters.
flat = c,S Of itan.„ = Cmi,f1INE,
10
min. AT between gas and ewporat-
the set of decision variables and y is the set of all functional exergy flows. Equation (9) is derived from the Eq. (1), in which the
revenue term has not been considered since the only system product,
the electrical power, has been fixed in this work (there is no need to
include it since addition of a constant does not change the optimum
point).
The total cost includes the annualized capital costs it of the
plant units, including investment, depreciation, maintenance, etc.
(see the Appendix), the fuel cost, the costs of auxiliary resources
Fs
(i.e. ammonia cost for the plant with the SCR or make-up
water cost for the plant with steam injection), and the costs due to the
is
67 Sweat inlet copcsoor 10
ins tteam (pinch point) rC1
ST between gas and supobtued
stem cc water rel
pastabine agenda isotopic eff
tacrator pressure f MP')
camktroa man= [MPal
race isentropic efficiency
alternator efficimcy
maseaum steam lane:rem ("C]
0.88
0.14
0.005
0.86
0.985
s565
fixl UN (methane) Nag]
50030
pmm elearicalhnechanical eft
annual fixed charm tee
numb= of spathe hours pa year
teest snag( filel cost Car [Ski]
WOO
4.0.10'
cpua IS/kg NOJ
cos Isma CO)
outlet nabine steam quality
>016
stack gas temperance MI
25
4
Downloaded From: http://proceedings.asmedigitalcollection.asme.org/ on 12/29/2014 Terms of Use: http://asme.org/terms
and outlet economizer ft]
AT between ecodoniog slam
and cooling weer 1°C1
=bine mechanical efficiency
pump isotopic efficiency
net *end power [MW1
steam Patine isentrapic eff
aretronment temperature r 2C1
5
0.995
0.83
60
0.85
15
0.90
18.7%
7.5
1.014
>103
set 1. A non-linear algorithm of optimization described in depth by
Massardo et al. (1990), is directly applied to the objective function.
Therefore the physical and cost models are used to describe the objective function of the system, while the mathematical formulation of
the T.F.A.-EFA is not requested here to find the optimum design
conditions. In this way, the introduction of many dependent variables
related to the internal economy of the system is not necessary to locate the minimum, and therefore it can be avoided with a strong reduction in computation time (Agazumi and Massardo, 1996). However, the calculation of the internal economy (marginal and average
unit costs) is pained at the optimum conditions.
The last two terms of Eq. (9) are the pollution costs due to the
NO and CO emissions from the gas turbine exhaust to the environment These are damage cost terms and can be written as (von Spakovsky and Frangopoulos, 1994):
mcci=cso.c....datio, tsco=ccoccothe.
What does it happen if C is greater than one? A plant should not
operate at all if its emissions are higher than limits set by regulations. Consequently, there would be no condition for imposing the
penalty, except if the meaning of the regulations is changed: instead
of setting limits that can not be violated, limits which can be violated
are set but, since such a change would have a negative effect on the
environment, a new greater penalty is imposed.
In conclusion, a penalty (i.e. c licohNo„ or e co firm ) is imposed
for every kilogram of a pollutant emitted by a plant no matter
whether the concentration is higher or lower than a limit set by
• regulations. Then, the introduction of the factor fp increases or decreases this penalty according to the relationship between the real
concentration and the one set by regulations.
Several scenarios can be analyzed in order to establish how 4
must be considered when it reaches values between 0 and 1; three
decision cases have been set here when fel :
• case (a): fp retains its value, as it results from the calculations,
• cue (b): fp is set equal to 1;
• case (c): fp is set equal to O.
In case (a), a pollution penalty is still considered but less than the
one of case (b) In case (c), if the concentration of the pollutant is
lower than a limit imposed by regulations, then there is no pollution
penalty.
The introduction of these pollution costs, corrected by the factor
fp, is different from the direct utilization of inequality constraints
which may be imposed by safety considerations for each pollutant
with this procedure, the optimization keeps into account also those
configurations which have emissions greater than the limits imposed
by regulations. However, only the case (c) could be considered very
similar to an inequality constraint since a penalty takes put in the
objective function only if emission is greater than the limits imposed
by regulations. It will be shown how the case (c) forces the decision
variables towards values which allow the analyzed plant to have
emissions just equal to the limits imposed by regulations
(harmfulness limits).
( 10)
in which the mass flow rates of NO and CO to the environment have
been selected as the pollution measure. clue( and cco (Table 2) are
the unit pollution penalties which are taken equal to the damage
caused to the environment per unit mass of a pollutant (Goswami,
1993). These represent costs imposed by society, governmental institutions, international organizations, etc., on the systems as a result of
the damage caused by the emissions in the environment
Several suggestions have been made for the pollution penalty
factors fp which appear in Eqs. (10) (Frangopoulos and von Spakovsky, 1993); the Izarmfuhiess factor has been used here:
ao
(11)
where
car intensive ptu .1y of the ith pollutant
cco, intensive property of the ith pollutant in the environment
Et; harmfulness limit of the intensive property of the ith pollutant
(if it is exceeded, the pollution may become particularly harmful).
The concentration has been selected as the intensive property for
the pollutants in this example. In such a case, if the value of the
standard air quality limit concentration is selected as the intensive
property of the pollutant in the environment (tree) and emissions limit
concentration from a power plant is selected as the harmfulness limit
( ) of the intensive property, then the value of ax can be neglected
In comparison to the values of . As an example, according to the
values imposed by Italian decrees on pollutant emissions, for the NO
pollutant, the value of cia is about 0.130 mem ) (-0.200 mg/m3 as
NO2) while the value of Tr is about 200 mem'. Then axe<II ; and
therefore the pollution penalty factor f p becomes:
f
P.
—
Table 3. Optimal results without pollution penalty costs
(baseline combined plant).
Decisico variables (optimal values)
ma ints moment=
1400 1C
pressure naio
1612
capital eons
10.63 eV,
Objective function (total pat)
55.22 ects
mg/Nm'p,.s
AR Ps
140
02052
195.76
CO
0.142
135.6
COI
68.97
65793
emmt ;censure
14.0 Mn
fuel cost
44.59 eV'
thermal efficiency • 53.82%
proud 15% Om
84:8 fuel
8.163
146.2
5.654
108.5
2744
33507
Table 4. Optimal results with pollution penalty costs (baseline
combined plant).
(12)
Decisica variables (optimal values)
pa =pat=
1291.5 •C
immure ratio
11.47
capital costs
10.032 efis
NO poll. cost ((peel .78)
15.36 cSis
Objective fimcnce (tail COST) '19. 81 ads
nw4m'rç.s
PAP Ps
NO
119.24
0.118
CO
0.286
287.8
COI
65.39
65793
In accordance with the emission limits imposed by EU, the selected
harmfulness limits for NO and CO are:
NO 50 ppm volume dry, 15% 02 as NO (about 75 ppmvd as NO2)
CO 80 ppm volume chy, 15% 02
The above value for NOx has been changed during the calculations in
order to analyze the influence of the harmfulness limit.
TOLL
5
Downloaded From: http://proceedings.asmedigitalcollection.asme.org/ on 12/29/2014 Terms of Use: http://asme.org/terms
Pam teem=
14 /tea
fuel cost
46.31 Ws
CO P011. cast (Ikea al)
2. 10 454
demo] efficiency 4 5113%
ika ant
4.973
12.00
2744
nand 1 554 02
89.07
230.3
33507
Table 5. Optimal results for case a): A) plant with SCR; B) plant
with steam injection.
Decision variables (optimal values)
ma gas tanpainze
1400°C
pressure ratio
16.69
gas =bac
4.1734/.
IIRSG
1.133 Ws
SCR
2392 40
0.076 el/a
PC
Fuel cost
44.63 Ws
Acta cost
0.216 412
Objective fuorlice (total cca)
60.62 eS/s
Emissions
rela sa
IttaNne gas
NO
0.02714
26.025
0.133
127.52
CO
CO2
68.61
65793
NO before abatement 4 0.213 p/kg gm
nerm pressure
NO degree of Mauna
num turbine
candmsa
GT Manna
steam plant alternator
NO poll cost (.s0.39)
CO poll. cost (feed:U.27)
thermal efficiency 4 53.77
cad
it/kg fial
1.085
19.4
5.32
102
2744
33507
Table 6. Optimal results for case b) (4=1 when less than 1): A)
plant with SCR; B) plant with steam injection.
Decision variables (optimal mates)
1400°C
max. gas temperature
pressure ratio
16.6
gas turbine
4.143 Ws
MO
1.137 als
SCR
3.129 Ws
0.076 Ws
oanps
Fuel cost
44.63 c5/s
Corea cost
0.221 44
steam pressure
NO degree of abatement
steam Patine
Censer
GT altaretor
steam plant altanator
NO poll oast (Eppel)
CO poll. cog (fPco4 1.3)
Objective fimnicm (total cost)
61.684/s Jamul effitheary = 53.77
Emissions
g/kg gas
cid
mgftirn3 PS
was fuel
NO
0.0209
20.02
0.835
15
5.39
0.135
129.34
103.5
CO
2744
CO2
68.7
65793
33507
NO before abscess = 0.211 grkg gas
13.99 MPa
0.872
3.257 Ws
0.589 40
0.9625 c.50
0.555 40
0.70544/8
1-533 40
Decision variables (optimal values)
ma gas temperanae
1359.8 °C
steam pressure
14 MI0
press= ratio
20.87
steam/fuel mass flow rat aria I 1.726
gas turbine
5.042 Ws
0.4804/s
*mean
HUG
1.1434/s
1.0614/s
GT alternator
0.077 ars
0.453 Ws
steam plant actor
Pima
stain =bine
2.8'29 4/s
Fuel cost
45.59 4/s
NO polleost (fpa0.41)
0.484 Ws
0.1967 4/s
Water cost
CO poll.cost (fpe0=1.174)
1.328 ars
Cbj. amnion (total cost)
58.69 Ws
drama efficimcy 4 5264
Emissicm
grkg WU
gikg ha
arid
Melleig13
NO
0.0773
6.80
0.284
5.1
CO
0.127
117.4
4.90
94
CO;
70.95
65793
2744
33507
14.0 MPs
0.901
3.263 Ws
0.5914/s
0.961 Ws
0.557 ads
1.396 Ws
1.577 40
Deediall variables (optimal values)
max gas tempace
Treasure ratio
gm turbna
FiRSG
Puma
co =bine
Fuel cost
Water cost
Obj, n=6011 total cost)
Emissions
ekg gas
NO
0.01705
CO
0.124
CO)
70.11
OPTIMIZAT ON RESULTS AND COMMENTS
Table 3 reports the results of the combined plant optimization,
without NOx abatement systems, and without taking into consideration the pollution penalty costs due to the NO and CO emissions
plant). It is interesting to observe how the high maximum
gas temperature involves a high NO emission (about 150 ppm).The
plant draws near to the optimum efficiency condition; this occurs in
order to limit the fuel consumption whose cost is about four times
greater than the total capital costs.
Instead, in Table 4, the results of the plant including the pollution
penalty costs are reported. In comparison with the previous case, the
NO), emission is lower (89 ppm), but the CO emission increases
considerably (230 ppm). This is due to the strong difference between
the two unit pollution penalties cm and cce (7.5 S/kg against 1.014
$/kg). The high influence of the NO pollution cost has brought the
decision variables towards values which decrease the NO emission:
in particular the maximum gas temperature and the pressure ratio are
decreased. In the table, the selected cases (a+c) for the pollution
penalty fitztor are not distinguished since fpNox and fpdo do not reach
values less than 1 (i.e. both the emissions are greater than the emission limits imposed by EU, 50 ppm for NO and 80 ppm for CO).
The next step regards the study of the optimum plant with SCR
or steam injection in the three decision cases (a, b and c) set for the
pollution penalty factor fp.
Table 5 (case a) and Table 6 (case b) report the main results of
the analyzed plants (independent decision variables, costs, emissions,
etc.). The second case involves greater values for the NO degree of
abatement and the steam/fuel mass flow rate ratio: this is caused by
the higher penalty policy obtained setting fp=1 when it is less than
one.
Unlike the previous case (Table 4), the introduction of NO%
abatement techniques results in an increase of the optimum value of
the maximum gas temperature and drives the optimal design towards
1400°C
20
4.8974/s
1.135 Ws
0.077 °Ls
2.931 Ws
45.334/s
0.146 4/s
58.111 44
matte gas
15.996
116.6
65793
1 14 MP.
cm pressure
steara/fuel mass fiow rate ntio 1 1.217
0.505 elk
<Cana
1 038 Ws
GI altermaar
0.476 Ws
steam plant Manatee
NO pal. cost (fpace0.24)
0.2707 c5/s
CO poll. cost (tirco41.166)
1.302 Ws
Mama efficiasey 4 5194
_Oa Nei
pgamf
0.667
11.9$
4.86
93.3
2744
33507
the optimum efficiency conditions again. Therefore, during the calculations, the maximum gas temperature and the steam pressure have
reached optimal values near to their upper limits (1400 °C and 14
MPa respectively), probably for the following reasons: i) in order to
respect the inequality constraint on the flue gas stack temperature
(>100 °C); in order to increase the thermal efficiency of the systems and therefore decrease both the fuel consumption and the mass
flow rate of the NO and CO emissions.
Obviously the SCR capital cost has a great influence on the total
capital costs (around 15-20°A), and the system with SCR has a total
cost greater than the system with steam injection; however, the difference between the plant with SCR and the plant with steam injection is not so great. This is due to the following considerations:
• The steam injection inside the combustor chamber causes an increase of the fuel consumption (whose energy cost has been fixed
equal to 4.0.106 $/U) and a decrease of the thermal efficiency even
if the power of the gas turbine increases. It must be remembered
that the steam is no easily available in nature: it must be produced
downstream the gas turbine with a process of thermal recovery.
Therefore, the steam injection improves bath the power and the efficiency of a gas turbine (STIG plant), but for a combined plant, the
injected steam must be considered removed from the expansion into
the steam turbine. In this case the following points occur the gas
turbine power increases while the steam turbine power decrease;
although not in proportion to the removed steam because the HRSG
operates with a greater flue gas mass flow rate, generating more
steam; the combined effect is that the combined cycle efficiency decreases. On the contrary, the introduction of the SCR inside the
IIRSG does not change the efficiency of the plant (if a pressure loss
is taken into consideration, the efficiency decrease is very small).
• The water consumption, if considerable, causes supplying problems
in certain localities; water must be treated and demineralized carefully, with an increase of the plant and management costs. A cost of
6
Downloaded From: http://proceedings.asmedigitalcollection.asme.org/ on 12/29/2014 Terms of Use: http://asme.org/terms
20
1.8 1.6
''''''''
0.75
' 0.70
.
In
NO unit pollution
penalty 1S/kg1
-
7.6
NO unit pollution
9611221' M91
26
0.8 -
as
0.65
0.625
0.5136
0.600
0.595
0.815
NO unit pollution
Panal1YIV8111
0.584
a5 0.583
0.810
0.605
20.582
0.603 •
o- 0.581
0.595
7.6
0.590
• 26
t
I
.
20
40
60
80
100
NO harmfulness limit (ppmvd 16%021
0.579
0.585
20
40 • 60
80
100
NO harmfulness limit 1ppmvd 16%021
120
Fig. 4. Influence of crou and a m:, on the plant with SCR - case
a): f, keeps Its value when less than 1.
120
Fig. 5. Influence of 0Nes and allo, on the plant with steam injection - case a): A, keeps its value when less than 1.
0.5 S/m3 has been set here for the deminanlized make-up water
but, by changing this value, different results can be obtained;
• The steam injection causes an increase of CO emissions in order to
keep these last at a low level, the combustor chamber pressure must
be increased (see Eqs. 2 and 3); therefore, in order to avoid a strong
CO emission increase, the optimal values of the pressure ratio for
the plants with steam injection are slightly greater than those for
the plants with SCR (see Table 6); this causes an increase of the
capital cost, in particular of the compressor (Agazzani and Massardo, 1996).
Figures 4,5,6 and 7 show the influence of the unit pollution
penalty ctcce, and the harmfulness limit a t,c, on the two plants. The
NO abatement should increase with the NO harmfulness limit reduction and the NO unit pollution penalty increase, causing higher total
costs particularly for low atm, values. This happens in particular for
the case (a) (Figs. 4 and 5), while a different behaviour is obtained
for the case (b) In this case, the plant with steam injection does not
seem to depend on the NO limits (see Fig. 7). Evidently, very low
NO emissions (less than 10 ppm) are reached also with high limits
(i.e.100 ppm); therefore fpNo x is always set equal to one. Instead, for
the plant with SCR, the same behaviour is found only for high cNct,
and ct n. values, while for low Ctme and allo. values, the degree of
abatement increases in order to limit the NO pollution penalty costs.
The numbers between brackets, near the points of Figs. 6 and 7, indicate the NO and CO emissions, respectively, (in ppm).
Increasing the NO unit pollution penalty or reducing its harmfulness limit, the steam injected (and therefore the water consumption)
increases a lot (see Figs. 5 and 7). This would cause a further increase of CO production and so the optimum pressure ratio would
increase some more. Therefore, if a greater penalty is imposed for the
CO (i.e. increasing the pollution penalty or decreasing the harmfulness limit), the plant with steam injection could not achieve the imposed harmfulness limit, with a strong increase of the CO damage
0.9
I
0.8
ICI
0.7
as
0.630
•
0.615
• 7.6
PC unit pollution
0.610
penalty WSW
0.605
215
0.680
0
20
40
so
80
100
NO harmfulness limit (ppmvd 16%021
120
Fig. 6. Influence of Crgoz and a Nox on the plant with SCR - case
b): fel when less than 1.
cost term. In this case, joined with also an increase of the make-up
water cost and the unit fuel cost, the cost difference between the
plant with SCR and the plant with steam injection could become
lower.
7
Downloaded From: http://proceedings.asmedigitalcollection.asme.org/ on 12/29/2014 Terms of Use: http://asme.org/terms
A) 0.95
1.8
(4.6, sot 12-6
-•
(140=5/3, C040)
•
1.7 -
0.84 -
ac 1.6
1.5
-
S 1.4
Tr
1.3
•
f
z 0.'-
•
NO unit pollution
penalty PAW
1.2
0.65
•
(9.5, 83)
NO4152
9488
CCe8D
• 24
140•152
-0.65
Pam
CO-75rem - 0.64
N0415 ten
ppm
0.63
puss
a7
- 0.52
-0.61 0/1
t0.80
!
0.92
0.91
1.1
- 0.59
1.0
0.591
0.593
0.589
0.588
0.90
0.58
1.82
0.66
•
M0.53 ppm _ 0 66
4-7 C0.70 rpm
•
p24
0.64
.
1.80
14045.3 rpm
CO•79 Mai #. 41
'5 0.587
0.586
0.585
•
2
•
04228
• 7.6
ft
•
NO unit pollution
Pellakii9/41)
0.554
0.583
e
•
0.582
0
20
•
43
1 1.74
80
1W
-0.61 .1
s- 0.03
you
1.72
• 2.6
so
-0.62
NO45.1 rem
C0494 ppn
-
120
0
NO harmfulness limit (ppm/eft 15%04
0.59
0.58
1.70
1
2
3
4
5
6
CO unft pollution penalty Ma]
Fig. 7. Influence of No. and allo, on the plant with steam injection - case b): f1 when less than 1.
Fig. 8. Influence of cco (a,=8O ppm, a No, =50 ppm, cNoe--7.5
S/kg): A) plant with SCR; B) plant with steam injection.
Up to this point, results show that CO emissions are always
higher than the limit of 80 ppm: the cause is that the value selected
for the CO unit pollution penalty (1.014 S/kg), is too low in comparison with c-No . Therefore a parametric study has been carried out for
the case (b), by increasing the cco value (see Fig. 8) and keeping always aco --280 ppm (the values of cNcr, and a m% are set equal to 7.5
S/kg and 50 ppm, respectively). In such a case, the CO emission decreases below the harmfulness limit of 80 ppm, causing greater values of the pressure ratio (see the value of near the points of Fig. 8).
On the contrary, the increase of this last involve also a short increase
of the NO pollution abatement
In case (c), in which the pollution penalty is similar to an inequality constraint, the optimintion code brings the decision variables, and in particular the NO L degree of abatement and the mass
flow rate ratio between injected steam and fuel, towards values
which allow the harmfulness limits to be respected (see Table 7).
Their values, and so the total costs, are considerably lower than those
obtained in the two previous cases, in particular for high NO limits.
The values of 8Noz, steam/fuel mass ratio and total cost (see Fig. 9)
do not depend on the NO unit pollution penalty, since the optimum
plant achieves NO emission values always lower than the limits imposed by regulations (i.e. the harmfulness limit a go,).
Table 7. Optimal results for case c) (441 when less than 1):
A) plant with SCR; B) plant with steam injection.
Decision variables (optimal values)
ma gas tempest=
1400°C
pressure ratio
18.66
capital costs
12.991 cS/s
Fuel cost
44.70 cS/s
Ammonia oust
0.212 cS/s
thenctal efficiency s• 53.69
Emissions
glect as
mg/14m3
NO
0.06831
66.914
CO
0.102
99.97
68.61
65793
CO2
NO before abatement v 0.246 reks ps
steam pressure
NO degree of abatement
NO pollution cost ((Nee)
CO pollution cost (fpcpe0)
Objective nation (total cost)
as
Oa fuel
2.79
4.169
2744
ppcowl
50
so
33507
Decision variables (optimal values)
ma ps temperance
1399.7°C
steam press=
pressure zatio
20
Scam/fuel mm flow nte ratio
agate) cans
11.32 cfr/s
NO pollution °est (fpame,,v0)
Fuel cost
45.04 cS/s
00 pollution cost (fiseoe0)
Water cost
0.0704 cS/s
Objective fimetice (weal cost)
thermal efficiency v 53.28
Emissiom
gekg gas
ntir/Nm3 gas
RAN fuel
Pluourl
NO
0.0589
66.66
2.78
50
CO
0.103
99.96
4.166
80
CO,
68.05
2744
65793
33507
CONCLUSIONS
Combined plants with SCR or steam injection as NO abatement
techniques have been analyzed from an environomic point of view.
CO and NO emissions have been taken into consideration; in particular for these last, the influence of the unit pollution penalty and of
the limit imposed by regulations on the system has been revealed.
With the introduction of the environomic approach, several scenarios can be analyzed for the same plant
13.98 MPa
0.722
0
0
57.903 cS/s
14 MPa
0.625
0
0
56.43 cS/s
The results obtained here must be considered not as an absolute
solution. They are valid under the economic assumptions made in
this work with no claim to general applicability. Furthermore, the
emissions are strongly influenced by the assumptions done for the
combustor chamber. However, the numerical examples presented
here MINE to clarify the methodology and can be interpreted as a
starting point for an environomic comparison of different plants.
8
Downloaded From: http://proceedings.asmedigitalcollection.asme.org/ on 12/29/2014 Terms of Use: http://asme.org/terms
Rangspoulos CA vas Spalcovsky MS., 1993. "The Environomic Analysis and Optimimica of Envy System (Part 1 and lir. Proceeding: of the International COliatita on
Energy Systems and Ecology: ENSEC '93, ASME, Caton, Poland, July, Vol. 1, pp. 123144.
Frey RC.. Rubin us., 1991, 'Probabilistic Evaluation of AdvaccedSOWNCts Cannel
Technology% The Journal of Air Waste Management Amociation vol. 41. no. IL December. pp. 1383-1393.
Gonvann D.Y., 1993, 'Soler Energy at the Enviscament", bacznationtl Cadaence am
Energy Systems and Ecology: ENSEC 93, J. Sawn and G. Tsatimmis eds.,
ASME.Cncow, Poland. July.
Lams A., Maces A_ 1994, -Mandnal and Average Cats in Engineering Functional
Amlysis% Florence Wald Erener Rewards Symp.:cam, FLOWERS '94. Flamm July 68.
Massada A., Sam A. Marini M., 1990, "Aeial Flow Compressor Design °Oman.
lion% teelh4E l21.201. 109111L2t.Daksombionz vol 112, OP. 399-410. July.
Risk NL, RC. Mcogia, 1993, -Semi analytical Correlations for NO% CO and IJHC
Embskas- agraEglaliSM. LMEIBLfal111/132200Lf2LQ132116MLIDELES5af. VeL
113, July, pp. 612-619.
Rosabag RS., Oxley JR, Hama RE, 1992, 'Selective Catalytic Roduaion far NQc
Como/ it Cogeneration Nom% IGTI-V017. ASME
EN-TURBO '92, pp. 409-417.
von Spakmaky MR., Evans RR_ 1993, 'Engin:aim Fmcsional Andpis - Part 1 at
Pan fr..5242-TXIMAR1233. icsollsflamigamommlitchmlogs. VoL 115. la PP
26-99.
vas Spakovsky MR. Framosoulos CA, 1994, "The Envinammic Analysis and °prim/swim of. CM Turbine Cycle with Cogeneration% ASME-WAM. AES-Vol 33.
Wcba a Sdrnidt D.. 1991. "ffigh at Low Dist SCR Precepts", Salpkur Diced&
and Nitrogen Oxides in Industrial Wane Gamy Emission Legislation and Abatement,
D.
Van Vet= ed. Fe. 223-234. ECSC, EEC.FAEC, Bnmeb and Latembours.
0.605
A) 0.95
0.90 0.85
- 01100
-
aeo
- 0.595
0,75
0.70
40 0.65
0.60
0.55
0.50
0.45 •
0.40
B) 1.8
- cusso
-
-
0.585
3
0.5 3
0
- 0.575 1- 0.570
0.565
0.569
1.4
0.56E1
1.2
1.0 -
In.
0.567
! 0.6
Go
•
i 0.4
0.565 1-
0.2
0.0
20
40
60
80
100
ND harmfulness lImit ppmed 15%CM
0.565
120
Flg. 9. Influence of Cwra and a Nce, - case c: f4) when less than
1. A) plant with SCR; B) plant with steam injection.
APPENDIX
The following equation has been used to evaluate the annualized
capital cost of each component of the plant (Frangopoulos, 1991a):
Also the effect of the variations of the unit pollution penalty and
the imposed limit for the CO emission, could prove to be interesting.
A ftutha development of the model can be carried out with the introduction of the unburned hydrocarbons emission (UHC). These last
two points will be analyzed in the near future.
FCR- (D i CU(3.6.10s • N)
vitae, in order not to complicate the analysis, the maintenance cost
has been taken into consideration by a constant factor Ct. FCR is the
annual fixed charge rate, obtained by the addition of several items:
interest or return on investment, depreciation, interim replacements,
property insurance, taxes, etc..
The values utilized for the analysis arc:
FCR=18.7%
d+,=1.06
N=8000 h/year
The purchased cost Cs of each component can be estimated by
means of equations written in terms of geometrical and manufacturing variables or in terms of performance and stream variables.
Further information about the component purchased cost equations are reported in Frangopoulos (1991a), Agazumi et al. (1995),
Agazzani and Massardo (1996), Boehm (1987), Lazzaretto et al.
(1994). The cost equation for the SCR, in particular, has been provided in the text
ACKNOWLEDGEMENT
The authors wish to acknowledge the funding from the National
Research Council of Italy (CNR), under contract n° 94.00038.CT07.
REFERENCES
Apzzani A.. Mmardo A., Sam A.. 1995, "Themmeconomic Analysis of Complex
Steam Pleas% ASME CCGEN TURBO POWER, Wien, 93-C7P-311.
Apr:rani A., Mansard° A., 1996, "A Tool for Thamocesnomic Analysis and Optimization of Cles. Steam and Combined Plants", accepted for publication in Mhakmmticm,
Journal ofEasimainnfor On Turbine and Power.
Amami A. 1996, "Thennacenemic Analysis at Enemy Symms% PkD. Thesis, University of Pin.
Boehm R.F 1987. 'Design Analysis of Thennal Systems", kiln Wiley and Sem. New
York.
Eskinazi 0.. Cichanowicz LE. Linak W.P. Hall I.E., 1989, "Stationary Combustion
NOx Control - A Summary of the 1989 Symposium% As Journal op:. Wane Manage:nen tessoclatlon, vol. 39. no. 8. August, pp. 1131-1139.
Frangepoulos CA. 1991a, 'Comparison of Tbertnoeconcmic and Thermodynamic Op.
Sal Designs of a Combined Cycle Plant", in Analyst, of Thermal and Envy Systems,
ATHENS I. Attens, (ham, lime 34 pp. 305-318.
Frangopoulos CA. 199 lb, introduction to Envircoomics% Second Law Analysts - Industrial mai Envirorunental Apo:mations, GM Reistad et aL. eds., AES-Vol. 25/117D.Vol.
191. ASME, N.Y., pp. 49.34.
Frangopodee CA. 1992, An busabaction to Envisonomic Analysis and Optimizatim
of &law/as:wive Symms'. in International Symposium on &friary, Cosa Optimization and Simulation of Entry Systems. ECOS 92, Zaragoza, Spain. 13-18 June, pp. 231239. ASME, New York.
9
Downloaded From: http://proceedings.asmedigitalcollection.asme.org/ on 12/29/2014 Terms of Use: http://asme.org/terms