Beruflich Dokumente
Kultur Dokumente
, Daniel Favrat
Laboratory for Industrial Energy Systems, Ecole Polytechnique F ed erale de Lausanne
CH-1015 Lausanne
ABSTRACT
Exergy concept combined with pinch based approach are used for studying the optimal integration
of energy conversion systems. The analysis rst considers the representation of the hot and cold
composite curves of the process and denes the energy and the exergy requirements. The basic as-
sumption of the Tmin required for the pinch analysis is represented as an exergy loss that increases
the exergy requirement of the process. The exergy composite curves put the focus on the opportuni-
ties for heat pumping in the process. The optimal integration of the utility system is then realised by
extracting the energy conversion system conguration from a superstructure using a Mixed Integer
Linear Programming formulation in which the heat cascade denition and the combined heat and
power production balances are introduced as constraints and where the exergy losses minimisation
is used as an objective function. The balanced exergy composite curves are used to visualise the
exergy losses in the heat exchanges and the analysis of the results is made using a denition of the
exergy efciency that accounts for the pinch point location, the process exergy and the exported
energy services.
Keywords: process integration, energy conversion integration, exergy analysis, second law analysis
NOMENCLATURE
CHP Combined heat and power
T
min
Minimum heat recovery approach temperature
e
f uel
specic exergy of the fuel (kJ/kg)
f
f uel
fuel owrate (kg/s)
E
cool
exergy available below the pinch point (kW)
E
f rg
exergy required for the refrigeration(kW)
E
heat
exergy required above the pinch point (kW)
E
res
exergy resource consumed by the process(kW)
GT Gas turbine
HP Heat pump system
LHV Lower Heating Value (kJ/kg)
MER Minimum Energy Requirement
MILP Mixed Integer Linear Programming
W Electrical power (kW)
INTRODUCTION
When considering the energy efciency of a pro-
cess, pinch based approaches [1] target the identi-
cation of the possible energy recovery by heat ex-
change and dene this way the minimum energy
requirement (MER) of a process. The integration
w=1
(C2
w
f
w
) +Cel
i
Wel
i
Cel
s
Wel
s
) t
+
n
w
w=1
(C1
w
y
w
)
+
i(1+i)
n
years
(1+i)
n
years
1
n
w
w=1
(ICF
w
y
w
+ICP
w
f
w
) (1)
subject to
Heat balance of the temperature intervals
n
w
w=1
f
w
q
w,r
+
n
i=1
Q
i,r
+R
r+1
R
r
= 0
r = 1, ..., n
r
(2)
Electricity consumption:
n
w
w=1
f
w
w
w
+Wel
i
Wc 0 (3)
Electricity exportation
n
w
w=1
f
w
w
w
+Wel
i
Wel
s
Wc = 0 (4)
Existence of technology w
f min
w
y
w
f
w
f max
w
y
w
w=1,...,n
w
y
w
{0,1}
(5)
Thermodynamic feasibility of the heat recovery and util-
ity systems
Wel 0,Wel
s
0 (6)
R
1
= 0, R
n
r+1
= 0, R
r
0 r = 1, ..., n
r+1
(7)
R
r
[kW] (r=1,n
r
+1) is the heat cascaded from the
temperature interval r to the lower temperature in-
tervals, n is the number of process streams, Q
ir
[kW]
is the heat load of the process stream i in the tem-
perature interval r (Q
ir
> 0 for hot streams), Wel
s
and Wel
i
are respectively the net production and net
import of electricity [kW]; C1
w
and C2
w
are respec-
tively the xed and the proportional cost of using
the technology w; ICF
w
and ICP
w
are respectively
the xed and the proportional cost related to the in-
vestment of using technology w. Although the for-
mulation is linear, it allows this way to account for
size effect in the investment; n
w
is the number of
technologies proposed in the utility system super-
structure; q
wr
is the heat load of the technology w
in the temperature interval r for a given reference
owrate, (q
wr
> 0 for a hot stream) ;w
w
is the me-
chanical power produced by the reference owrate
of technology w; (w
w
> 0 for a mechanical power
producer); Cel
s
and Cel
i
are respectively the selling
price and the import price of electricity [e/kW]; t is
the total annual operation time [s/year] ; Wc [kW] is
the overall mechanical power needs of the process;
(Wc < 0 if the overall balance corresponds to a me-
chanical power production).
This method may be applied to any kind of energy
conversion technologies. It is based on the assump-
tion that the operating conditions have been dened
for each of the equipment concerned and that only
the owrates are unknown. This is a limiting as-
sumption but it allows to solve most of the problems
of energy conversion integration mainly because non
linearities may usually be solved by discretising the
search space. The method has been further adapted
to compute the optimal integration of steam net-
works [12], to incorporate restricted matches con-
straints [13], to integrate refrigeration cycles [14]
and Organic Rankine Cycles[15] as well as heat
pumps [16].
The grand composite curve resulting from the opti-
misation is the balanced composite curves. An ex-
ample of such curves is given on gure 3, left. This
representation is characterised by a number of pinch
points, one being the process pinch point, the oth-
ers corresponding to the maximum use of the cheap-
est utilities to satisfy the process requirement. In
the Carnot representation (gure 3, right) the area
between the curve and the Y-axis represents the ex-
ergy losses in the system heat transfer. Two curves
are presented, the external one represents the over-
all exergy losses, while the internal one that feature
pinch points represents the one computed with the
corrected temperatures. The area between the two
curves represents therefore the exergy losses relat-
ing to the assumptions of the T
min
/2
j
contributions
of the streams. Analysing the inner curve allows to
identify further improvement for the energy conver-
sion by modifying for example the operating condi-
tions (pressures or temperatures) in the energy con-
version technologies. Integrated composite curves
[17] may be used to analyse the integration of sub-
systems in the process. The gure 4 shows for exam-
ple the integration of the steam network in the sys-
tem under study. The energy balance of the hot and
cold streams of the steam network is the net mechan-
ical power production. When appropriately placed,
it corresponds to a supplement of energy to be sup-
plied to the system. The fact that it appears on the
left of the temperature axis proves that the steamnet-
work characteristics are appropriate for the optimal
production of mechanical power. In the Carnot fac-
tor ordinate, the area between the two curves gives
an indication of the quality of the exergy use in the
corresponding sub-system.
One heuristic rule resulting from the exergy analysis
is to try to reduce to a minimum the area between
the hot and cold composite curves of the integrated
systems including the energy conversion system.
200
300
400
500
600
700
800
900
1000
1100
1200
0 1000 2000 3000 4000 5000 6000 7000 8000 9000
T
(
K
)
Q(kW)
Grand composite curve
Grand composite
Grand composite cor.
-0.2
0
0.2
0.4
0.6
0.8
1
0 1000 2000 3000 4000 5000 6000 7000 8000 9000
(
1
-
T
0
/
T
)
Q(kW)
Carnot Grand composite
Grand composite
Grand composite cor.
Figure 3: Balanced Grand composite curves of the
integrated system
200
300
400
500
600
700
800
900
1000
1100
1200
-1000 0 1000 2000 3000 4000 5000 6000
T
(
K
)
Q(kW)
Integrated composite curves of the steam network
Others
(STMNET)
Mech. power
-0.2
0
0.2
0.4
0.6
0.8
1
-1000 0 1000 2000 3000 4000 5000 6000
1
-
T
0
/
T
Q(kW)
Carnot integrated composite curves of the steam network
Others
(STMNET)
Mech. power
Figure 4: Integrated composite curves of the steam
network
Using the exergy losses as an objective func-
tion
Due to the linear nature of the problem, the use of
the energy cost as an objective function may reveals
some difculties [16]. When the cost of fuel and
electricity is such that the electrical efciency of
a cogeneration unit is attractive without the use of
heat (i.e. when the electrical efciency of the unit
el
=
W
el
LHV
f uel
is greater than
C
LHV
(e/kJ)
C
el
(e/kJ
e
)) there is an
economical interest to produce electricity even with-
out cogeneration). In this case, the linear program-
ming procedure leads to a situation where the cogen-
eration unit is used at its maximum. This situation
usually does not occur when the investment cost are
properly considered or when the cost of the differ-
ent forms of energy are coherent with respect to the
electrical efciency. Nevertheless, the relative price
of the different forms of energy will inuence the
technology selection and their level of usage in the
integrated solution. When the target is the maximi-
sation of the system efciency, alternative formula-
tions that take into account the value of energy in
the objective functions have to be considered. The
minimisation of the exergy losses (eq. 8) is an alter-
native way of formulating the objective function.
Min
R
k
,y
w
, f
w
n
w
w=1
( f
w
(Ex
w
n
k
k=1
ex
wk
+w
w
)) (8)
In this relation, Ex
w
is the exergy consumed to
produce the hot and cold streams and the electricity
of the conversion unit w, ex
wk
is the heat-exergy
supplied by the ns
w
hot and cold streams of the con-
version unit w in the temperature interval k. ex
wk
is
given by (9).
ex
wk
=
ns
w
s=1
q
sk
(1
T
a
ln(
T
k+1
+T
min
/2
s
T
k
+T
min
/2
s
)
T
k+1
T
k
) (9)
Using this formulation, it is possible to dene the
set of energy conversion technologies that minimises
the exergy losses of the system. It is even possible
to introduce the aspects related to the investment by
adding the grey exergy into the Ex
w
term.
EXAMPLE
Let us consider the system requirements dened on
table 1. These result from the hot and cold compos-
ite curves of gure 1 and the Grand composite curve
of gure 2. For the calculations, we assumed that all
the possible process improvements were already im-
plemented before analysing the energy conversion
technologies integration.
Table 1: Minimum energy and exergy requirements
of the process
Energy Exergy
Heating (kW) +6854 +567
Cooling (kW) -6948 - 1269
Refrigeration (kW) +1709 + 157
Several optional energy conversion system cong-
urations are studied, the results are summarized in
table 5 where the energy consumption of the energy
conversion sub-systems are presented. The simplest
solution (option 1) is to integrate a boiler using nat-
ural gas (with a LHV of 44495 kJ/kg) and to cool
the process with cooling water. The refrigeration
needs will be supplied with a refrigeration cycle us-
ing ammonia (R717). The operating conditions of
the refrigeration cycle (table 2) have been obtained
by simulation considering the temperature levels in
the composite curve and the T
min
to be reached
in the heat exchangers. The integrated composite
curves presenting the results of the optimisation are
presented on gure 5 (left). The refrigeration cy-
cle consumption is of 314 kW corresponding to an
exergy efciency of 50 %. It should be noted that
the energy consumption is higher than the MER due
to the losses at the boiler stack (398 K). The solu-
tion accounts for the possibility of air preheating to
valorise the energy excess available in the process.
The heat load of air preheating is of 131 kW. In or-
der to valorise the exergy potential, a steam network
has been integrated (Option 2). The steam network
headers are given on table 3, the isentropic efciency
of the turbines are assumed to be of 70 %.
Table 2: Refrigeration cycle characteristics
Refrigerant R717 Ammonia
Reference owrate 0.1 kmol/s
Mechanical power 394 kW
P T
in
T
out
Q Tmin/2
(bar) (K) (K) kW (K)
Hot str. 12 340 304 2274 2
Cold str. 3 264 264 1880 2
Applying the rules of the appropriate placement of
heat pumping devices, 3 heat pumping cycles have
been proposed and simulated (table 4).
The high values of the COP are explained by the
very small temperature raise to be obtained from the
heat pump when considering small T
min
/2 values
for the heat exchangers. Using the optimisation tool,
the optimal owrates in the three cycles have been
computed together with the new value of the fuel in
the boiler house (Option 4). In the example consid-
ered, this leads to a situation where the whole heat
requirement may be provided by the heat pumps.
When the steam network is considered together with
the heat pumps (Option 5), the results are slightly
-0.2
0
0.2
0.4
0.6
0.8
1
-2000 0 2000 4000 6000 8000 10000 12000
1
-
T
0
/
T
Q(kW)
Option 1 : Carnot composite curves
Process composite curve
Utility composite curve
-0.2
0
0.2
0.4
0.6
0.8
1
-2000 0 2000 4000 6000 8000 10000 12000
1
-
T
0
/
T
Q(kW)
Option 5 : Carnot composite curves
Process composite curve
Utility composite curve
Figure 5: Carnot integrated composite curves of the
energy conversion system for options 1 (left) and 5
(right)
different since in this case, an additional amount of
energy is required by the system to balance the me-
chanical power produced by expansion in the steam
network. The solution of heat pumping is then com-
pared with a combined heat power production using
a gas turbine (Option 3). In this situation, the two
options are conicting.
Energy and Exergy efciency
The summary of the energy conversion system inte-
gration is given on table 6. It is shown that a MER
of 6854 kW for the heating requirement and of 1709
kW for the refrigeration requirement is nally sup-
plied with an equivalent 893 kW of fuel when con-
sidering the possibility of heat pumping and when
converting the exergy content of the process streams.
Compared to the boiler house solution, the new situ-
Table 3: Steam cycle characteristics
Header P T Comment
(bar) (K)
HP2 92 793 superheated
HP1 39 707 superheated
HPU 32 510 condensation
MPU 7.66 442 condensation
LPU 4.28 419 condensation
LPU2 2.59 402 condensation
LPU3 1.29 380 condensation
DEA 1.15 377 deaeration
Table 4: Characteristics of the heat pump system,
based on R123 as working uid
P
low
T
low
P
high
T
high
COP kWe
(bar) (K) (bar) (K) -
Cycle 3 5 354 7.5 371 15 130
Cycle 2 6 361 10 384 12 323
Cycle 0 6 361 7.5 371 28 34
Table 5: Results of the energy conversion system
integration for different options
Opt Fuel GT CHP Cooling HP
kW
LHV
kWe kWe kW kWe
1 7071 - - 8979 -
2 10086 2957 9006 -
3 16961 5427 2262 9160 -
4 - - - 2800 485
5 666 - 738 2713 496
ation corresponds to a reduction by a factor 8 of the
fuel consumption. These data have been computed
by considering a fuel equivalence of 55% for the
electricity production (column Total 1). The order of
the solutions will be different if we consider the Eu-
ropean mix (38.7%) for the fuel equivalence (Table
6, column Total 2). In order not to rely on the de-
nition of a fuel equivalence, an exergy efciency
ec
of the energy conversion system will be computed
considering the exergy of the process. In this deni-
tion, we consider that the energy services delivered,
i.e. the process exergy requirement (E
heat
+E
f rg
)
and the export of electricity (Wel
s
), will be satised
with an efciency of
ec
leading to an exergy con-
sumption of
E
heat
+E
f rg
+Wel
s
ec
, while the exergy excess
(E
cool
) will be converted with an efciency of
ec
.
The balance (eq. 10) is equal to the energy resources
(E
res
= f
f uel
e
f uel
+Wel
i
) converted in the energy
conversion system. Solving (10) gives the denition
of the exergy efciency of the system (eq. 11). In
Table 6, it can be seen that the options that do not
convert the exergy excess (E
cool
) have smaller ex-
ergy efciencies. The best solutions are the one that
realise heat pumping that pump the excess of ex-
ergy from below to above the pinch point.
E
res
=
E
heat
+E
f rg
+Wel
s
ec
E
cool
ec
(10)
ec
=
E
res
(E
res
)
2
+4E
cool
(E
f rg
+E
heat
+Wel
s
)
2E
cool
(11)
CONCLUSIONS
The application of the exergy concept combined
with pinch based methods for analysing the optimal
integration of energy conversion system of industrial
processes has been studied. The exergy compos-
ite curves is used to compute the minimum exergy
requirement of the process, considering the pinch
point location. The exergy requirement is obtained
by rst considering an exergy loss resulting from
the denition of the T
min
. The remaining exergy
requirement is divided into three contributions: the
exergy required above the pinch point, the exergy
produced as energy excess between the pinch point
and the ambient temperature and the exergy required
for refrigeration. Starting with an energy conversion
system superstructure, a linear programming formu-
lation is used to extract the optimal energy conver-
sion system conguration that supplies the process
energy requirement and that integrates the combined
heat and power production and the heat cascade. In
this formulation, it is possible to use either the en-
ergy cost or the exergy losses as an objective func-
tion. The exergy balanced composite curves and the
exergy integrated curves are used to visualise the ex-
ergy losses in the system. The exergy efciency of
the conversion system is dened by comparing the
exergy resource consumption with the exergy export
and the process exergy, making the distinction be-
tween the exergy available and the exergy required.
This efciency denition sets the focus on the ex-
Table 6: Energy and exergy efciency of the differ-
ent options
Opt Fuel Net El. Total 1 Total 2
ec
kW
LHV
kWe kW
LHV
kW
LHV
%
1 7071.0 371.0 7745.5 8030 9.6
2 10086.0 -2481.0 5575.1 3675 30.6
3 16961.0 -7195.0 3879.2 -1630 45.16
4 0.0 832.0 1512.7 2149 49.6
5 666.0 125.0 893.3 989 50.5
ergy available in the process, promoting solutions
that integrate heat pumps, thereby converting the ex-
ergy excess into useful energy.
REFERENCES
[1] Boland P., Hewitt G.F., Thomas B.E.A.,
Guy A.R., Marsland R.H., Linnhoff B.,
Townsend D.W. A user guide on process in-
tegration for the efcient use of energy. The
Institution of Chemical Engineers, 1982.
[2] Grossmann I.E., Papoulias S.A. A struc-
tural optimization approach in process synthe-
sis - i. utility systems. Comp. Chem. Engng.,
7(6):695706, 1983.
[3] B. Kalitventzeff, Marechal F. Heat and power
integration : a milp approach for optimal
integration of utility systems. proceedings
COPE91 (Barcelona), 1991.
[4] Favrat D., Staine F. Energy integration of
industrial processes based on the pinch anal-
ysis method extended to include exergy fac-
tors. Applied Thermal Engineering, 16:497
507, 1996.
[5] Lavric V., Baetens D., Plesu V., De Ruyck, J.
Broadening the capabilities of pinch analysis
through virtual heat exchanger networks. En-
ergy Conversion and Management, 44:2321-
2329,2003.
[6] Linnhoff B., Dhole V.R. Total site targets
for fuel, co-generation emissions, and cooling.
Comp. Chem Engng,17:s101-s109, 1992.
[7] Ohba T., Ishida M. A new approach to energy-
utilization diagrams for evaluation of energy of
chemical systems. In ECOS 04 Proceedings,
845852., 2004.
[8] Kalitventzeff B., Marechal F. Identication
of the optimal pressure levels in steam net-
works using integrated combined heat and
power method. Chemical Engineering Science,
52(17):29772989, 1997.
[9] Linnhoff B., Townsend D.W. Heat and power
networks in process design. part 1: Criteria for
placement of heat engines and heat pumps in
process networks. AIChE Journal, 29(5):742
748, 1983.
[10] Mason D., Linnhoff B., Cerda J., Wester-
berg A.W. Minimum utility usage in heat
exchanger network synthesis. a transporta-
tion problem. Chemical Engineering Science,
38(3):378387, 1983.
[11] Kalitventzeff B., Marechal F. Process integra-
tion : Selection of the optimal utility system.
Comp. Chem. Engng., 22:S149S156, 1998.
[12] Kalitventzeff B., Marechal F. Targeting the op-
timal integration of steam networks,. Comp.
Chem. Engng., 23:s133s136, 1999.
[13] Kalitventzeff B., Marechal F. Restricted
matches and minimum cost of energy require-
ments: tools and methodology for produc-
ing practical solutions. PRES99 proceedings,
433438, 1999.
[14] Kalitventzeff B., Marechal F. A tool for op-
timal synthesis of industrial refrigeration sys-
tems. Computer Aided Chemical Engineering
series, Ed. Elsevier, 90:457463, May 2001.
[15] Kalitventzeff B., Marechal F., Favrat D. A
methodology for the optimal insertion of or-
ganic Rankine cycles in industrial processes.
2nd International Symposium on Process Inte-
gration, Halifax, 2001.
[16] Kalitventzeff B., Marechal F. Targeting the
integration of multi-period utility systems for
site scale process integration. Applied Thermal
Engineering, 23:17631784, 2003.
[17] Kalitventzeff B., Marechal F. Targeting the
minimum cost of energy requirements : a new
graphical technique for evaluating the integra-
tion of utility systems. Comp. Chem. Engng.,
20(Suppl.):S225S230, 1996.