Professional Documents
Culture Documents
Optimization of Energy Plants Including Water/lithium Bromide Absorption Chillers
Optimization of Energy Plants Including Water/lithium Bromide Absorption Chillers
)
generated in the absorption process is rejected to cooling water, which will be later used too as
a medium temperature sink in the condenser. The liquid solution rich in refrigerant, leaves the
absorber and starts again the solution circuit.
The refrigerant leaving the generator follows the same path as the refrigerant of a compression
machine. It enters the condenser where the steamis condensed by rejecting heat (Q
) to a medium
temperature sink (cooling water). The liquid refrigerant is then throttled to the evaporator
pressure. This expansion is usually accompanied by a small amount of #ashing due to the high
latent heat content of water. The evaporation of refrigerant takes place at low pressure using the
heat (Q
`
) released by the water to be chilled. The steam generated #ows to the absorber to dilute
the poor solution.
In the model used in this work, single-e!ect chillers have been modelled assuming that the
refrigerant and liquid solution leaving the generator are at equilibrium conditions and assuming
that pressure changes are only signi"cant in valves and pumps, and the heat losses to the
surroundings are negligible. The states at the outlet of the throttle or expansion valves are
calculated as isoenthalpic expansion. For a pure #uid, as is the case in the refrigerant loop, only
two properties are needed to completely de"ne the state of a #uid. Thus, the enthalpy and the
pressure at the valve outlet being known, other properties such as, temperature and vapour
quality can be determined using property functions. The throttle of a water/lithium bromide
solution is a special case of #ashing in which the "nal vapour composition is known as equal to 1.
Known parameters are the global phase mass #ow rate, concentration and enthalpy, and also
"nal pressure. To solve the set of equations an iterative process is required. An outlet condition is
assumed and calculated all the unknown properties, the procedure is repeated until the energy
balance is satis"ed with su$cient accuracy. The properties for water/lithium bromide solutions
are obtained from Patterson and Perez-Blanco (1988) and ASHRAE Handbook fundamentals
(1977). The complete set of modelling equations for this absorption cycle can be found in Bruno
(1999) and will not be given here for obvious reasons of space.
3. DOUBLE-EFFECT WATER/LITHIUM BROMIDE ABSORPTION CHILLER
A double-e!ect cycle uses the input heat twice within the cycle to generate steam, and require the
use of higher temperature input heat than a single-e!ect cycle. First, steam is raised in the
702 J. C. BRUNO, J. MIQUEL AND F. CASTELLS
Copyright 2000 John Wiley & Sons, Ltd. Int. J. Energy Res. 2000; 24:695}717
Figure 3. Schematic representation of a double e!ect parallel #ow absorption chiller.
high-pressure generator and #ows to the low-pressure generator to be used as input heat to
generate more steam. Compared to single-e!ect cycles, double-e!ect cycles are more e$cient the
COP being usually in the range of 0.9}1.25. These cycles can be viewed as three pressure
machines: one of the generators works at high pressure, the lower pressure generator and
condenser work at medium pressure, and the absorber and evaporator at low pressure.
Depending on the connection of the units present in the cycle, two types of double-e!ect cycles
are available: series #ow and parallel #ow. In the series #ow type the solution is "rst directed to
the high-pressure generator the liquid solution leaving the high-pressure generator being used as
strong solution entering the low-pressure generator. Another option is a series #ow con"guration
with the strong solution being sent "rst to the low-pressure generator. This option is technolo-
gically similar to the previous con"guration, but a bit lower in e$ciency, thus in this work, the
series #ow is designed with the strong solution being sent "rst to the high pressure generator.
In the parallel #ow type (Figure 3) the strong solution leaving the absorber is divided into two
streams, one of them is led to the high-pressure generator (9) and the other one, to the low
pressure generator (14). Solution heat exchangers are used for the same purpose as in the
single-e!ect cycle. Parallel #ow o!ers thermodynamic bene"ts achieving a bit higher COP, but to
obtain this relative small bene"t requires more control complexity. The state conditions in
double-e!ect cycles are quite similar to those of single-e!ect cycles. The same hypothesis
discussed for single-e!ect cycles apply here. The modelling equations for series and parallel
#ow double-e!ect chillers are given in detail in Bruno (1999).
OPTIMAL INTEGRATION OF WATER/LITHIUM BROMIDE ABSORPTION CHILLERS 703
Copyright 2000 John Wiley & Sons, Ltd. Int. J. Energy Res. 2000; 24:695}717
4. INTEGRATION OF WATER/LITHIUM BROMIDE CHILLERS IN A CHP PLANT
In industrial facilities, refrigeration demands between ambient temperatures and 53C can be
satis"ed using di!erent types of refrigeration cycles: single- and double-e!ect water/LiBr absorp-
tion chillers, and also compression chillers. Cycle selection is conditioned for a great number of
factors, which can be divided into two groups: internal and external factors. Internal factors are
due to the own features of each cycle. For example, to drive double e!ect chillers a steam of higher
pressure and temperature than that used for single e!ect is required, however they will be more
e$cient. Another example: compression chillers are the most e$cient, but nevertheless high-
quality energy, that is, electrical energy, will be consumed, with the potential increase in operation
costs. External factors are not related to the characteristics of the given refrigeration cycle. These
factors may include availability of electricity, coming from the general grid or locally generated,
the availability of steam at di!erent pressures and temperatures raised in a conventional or
cogeneration plant, and the availability of waste heat. In this group may be included also the
in#uence on the refrigeration system design of the retro"t, renewal or enlargement of the existing
equipment units.
Therefore, due to the variety and great number of parameters involved in many cases in the
integration of absorption cycles it is very di$cult to generalise about the selection of the best
refrigeration cycle for a given application. According to this, to select the most suitable refriger-
ation cycle for a given refrigeration load, it is not only necessary to model the performance of each
cycle, but also to take into account the interactions between the energy systemand the considered
cycles, optimizing the performance of the global plant.
Once the features of the energy plant and the features of the required refrigeration load
and its temperature are clearly stated, candidate cycles are proposed. Technical cycle limitations
should be checked at this point. For example, in the case that there is no heat available at
a temperature higher than 1503C, double-e!ect water/LiBr chillers have to be discarded
for this case, and only single e!ect and compression cycles have to be considered for energy
integration.
This integration is based on the simulation and further optimization of the operating condi-
tions of CHP plants and has been implemented in the code XV to be applied in an industrial
environment. The particular plant topology, that is, the arrangement and type of units con-
sidered, including the selected refrigeration cycle, and the operation parameters, such as, pres-
sures, e$ciencies in some devices, maximum#ow rates, and so on, are de"ned through a standard
data "le. Based on this "le, the program automatically build a linearized set of equations
describing the material and energy balances, and also equipment constraints involved in the
considered energy system. In a "rst step, the program performs the CHP plant simulation, giving
a "rst set of variables (#ow rates, enthalpies, temperatures, fuel consumption etc.) that satisfy the
set of equations. These results are used as an initial point to perform the optimization of #owrates
and enthalpies minimizing the energy plant operating costs by using a successive linear program-
ming (SLP) algorithm (Miquel, 1991). The SLP algorithm is fast, robust and accurate enough
for problems with a low nonlinearity as is the case of energy plants. This software provides the
optimal operating conditions of any unit that is usually present in a CHP plant, including the
most typical congeneration system consisting of a gas turbine and a heat recovery steam
generator. All the commercially available absorption chillers based on the water/lithium bromide
technology are modelled and included in the program as a module that require input variables
and supplies the output results of the chiller. The header providing the required steam to the
704 J. C. BRUNO, J. MIQUEL AND F. CASTELLS
Copyright 2000 John Wiley & Sons, Ltd. Int. J. Energy Res. 2000; 24:695}717
absorption cycle has to be chosen according to the selected cycle and has to be speci"ed in the
input data along with the necessary cycle operating parameters.
In this integration method, the candidate refrigeration cycle may be viewed as a bottoming
cycle, which produce the required refrigeration load (Q
`
) at the speci"ed temperature level. In
each iteration of the SLP algorithm, the optimization program supplies the currently &optimal'
values of #ow rate and enthalpy of the steam used to drive the absorption cycle. A subroutine
performs the simulation of the corresponding absorption cycle. This subroutine containing the
corresponding chiller simulation module returns all the chiller operation conditions including the
characteristics of the condensate leaving the generator that is returned to the condensate header
of the energy plant. Then a new iteration is carried out until convergence of the integrated
problem (CHP plant plus refrigeration system) is achieved. This strategy avoids the problem that
could be caused by nonlinearities caused by the inclusion of the absorption cycle streams in the
optimization strategy.
Plant operation conditions and therefore the features of the external input heat to the cycle are
calculated by the optimization algorithm using the global energy plant operation costs as the
objective function ( f
'
). This function is given by the following expression:
f
'
"(Q'"`'
'
#Q'
'
) C
'
#(Q
`
#Q
`
) C
`
#F
`
C
`
#(=
#=
!='
!=''
)C
(1)
As seen in the previous expression, when an electrically driven compression chiller is selected the
electricity and cooling water consumption are included in the cost function. But for this type of
cycle, the operating conditions inside the cycle are not simulated, as there is no interaction with
the steam network being optimised. Applying this approach, operation costs and performance of
the energy plant, using each of the candidate cycles, are obtained. These results are screened in
order to select the cycles, which o!er the best results. The trade o! here is between the operation
cost and the investment cost. When high capital cost cycles o!er an attractive operation cost, an
economic analysis including investment cost need to be performed to determine the viability of
the project. It is obvious that if an energy plant that includes a given refrigeration cycle, that o!ers
the worse operation cost among the considered cycles, also have the highest capital cost, there is
no need to calculate the economic viability of the alternative. Realistic investment costs for these
cycles were calculated according to some sources available in the literature (Dorgan et al., 1995;
SahuH n and Valle, 1994; Garret, 1989) or may be obtained directly from some manufacturers
(York and Tuma Turbomach).
5. A CASE STUDY OF A CHEMICAL PROCESS PLANT IN TARRAGONA
The proposed methodology for absorption chillers integration in CHP plants is applied to the
process plant of Bayer in Tarragona (Catalonia, Spain) where refrigeration at 73C is required.
Other refrigeration demands required in other plants of this facility will not be covered in this
case study. This plant generates an important amount of process heat, which is recovered to raise
steam at di!erent pressure levels. The steam at very low pressure (132 kPa) is used to drive
a single e!ect water/lithium bromide chiller that provides 2500 kW of chilled water at 73C. For
this temperature, water/lithium bromide is the most appropriate working #uid for absorption
systems. Eventually when the #ow of process heat is not enough or it is not available, steam is
OPTIMAL INTEGRATION OF WATER/LITHIUM BROMIDE ABSORPTION CHILLERS 705
Copyright 2000 John Wiley & Sons, Ltd. Int. J. Energy Res. 2000; 24:695}717
Figure 4. Schematic diagram of the global energy plant including a single e!ect absorption chiller.
#ashed from the low-pressure steam header of the energy plant. A diagram of this plant is
presented in Figure 4.
Given the features of the refrigeration load, single- and double-e!ect water/LiBr cycles are
considered. Both parallel and series #ow double-e!ect cycles could be selected to supply the
required refrigeration load. The data used for the modelling of absorption cycles is given in
Table I. Using this data for both cycles a very similar performance is obtained although a slight
better COP for the parallel #ow cycle is reached, 1.17, compared to the series #ow cycle, 1.12.
Regarding this, reducing the great range of options is to be considered, the parallel #ow-type cycle
has been used as double-e!ect cycle throughout this case study. The economic analysis will be
performed using the following utility costs:
Cost of electricity: $ 0.053 kWh (8 ptas kWh)
Cost of natural gas: $ 0.0032 MJ (2 ptas te)
Cost of cooling water: $ 0.033 m` (5 ptas m`)
Cost of demineralized water: $ 0.267 m` (40 ptas m`)
The same price is assumed for buying and selling electricity from or to the external grid. To
calculated the PBP the capital recovery factor is taken as 0.12. According to manufacturers and
706 J. C. BRUNO, J. MIQUEL AND F. CASTELLS
Copyright 2000 John Wiley & Sons, Ltd. Int. J. Energy Res. 2000; 24:695}717
Table I. Data for absorption cycles.
Single e+ect
Evaporator temperature (3C) 4.5
Condenser temperature (3C) 46
Generator temperature (3C) 100
Solution heat exchanger e$ciency 0.75
Solution pump e$ciency 0.8
Solution circulation ratio 16
Double e+ect
Solution circulaton ratio 12
Mass concentration at the absorber inlet 0.63
Heat recovery ratio at high pressure 0.7
Heat recovery ratio at low pressure 0.8
Drain heat exchanger e$ciency 0.7
Pressure in the HPG (Pa) 73000
Condenser pressure (Pa) 5900
Evaporator temperature (3C) 4.5
Distribution ratio of solution 0.5
(only in the parallel #ow con"guration)
some literature references the following capital costs for the units involved in this study
are used:
Refrigeration cycles
Single-e!ect absorption chiller: $ 42700 (64 MMptas).
Double-e!ect absorption chiller: $ 540000 (81 MMptas).
Compression chiller: $ 400000 (60 MMptas).
Other units
Cogeneration system: $733 kW of power generator (110 000 ptas kW)
Back pressure steam turbine: $247 kW (37 000 ptas kW)
5.1. Analysis of the current situation
The single-e!ect absorption cycle is simulated according to the model proposed in Bruno (1999)
and using the performance data supplied by the chemical company. Using the available waste
heat the operation bene"t of the global plant is $ 220 h. This includes the cost of all the
resources consumed in the global plant, to produce heat, power and refrigeration including
the CHP plant plus the considered refrigeration cycle. A compression cycle produces an opera-
tion bene"t of $ 211 h with a COP of 6.0 and applied to meet the same refrigeration load.
A much higher e$ciency in favour of the compression cycle is not enough to compensate for the
lower operation cost of the absorption cycle. The payback period (PBP) of the absorption
alternative is lower than one year. This proves once more the suitability of absorption systems
whenever adequate waste heat is available.
Another technically possible option for this plant would be the use of a double-e!ect chiller.
For this option, low-pressure steam(602 kPa) from the steam network is directly used to drive the
OPTIMAL INTEGRATION OF WATER/LITHIUM BROMIDE ABSORPTION CHILLERS 707
Copyright 2000 John Wiley & Sons, Ltd. Int. J. Energy Res. 2000; 24:695}717
Figure 5. Schematic view of the global energy plant including a double e!ect absorption chiller.
cycle (See Figure 5). The operation bene"t is $ 196 h using a double e!ect chiller. This bene"t is
lower than the other cycle alternatives, due to the fact that there is no waste heat available to
generate steam at the required pressure to drive this cycle and this heat has to be generated in the
CHP plant.
During short periods of time the #ow of process waste heat is not enough to supply all the heat
required by the cycle and steam is #ashed from the CHP plant. The study of this situation shows
that when the waste heat decreases, the compression cycle o!ers a higher operation bene"t. If the
waste heat reduction is quite important the gap between the single- and double-e!ect operation
bene"ts decreases. The reason is that now heat needs to be generated too for the single-e!ect
cycle, and this cycle is less e$cient than the double e!ect counterpart.
Another situation appears when for any reason (temporarily shut down of a process plant,
maintenance, etc.) there is no waste heat available to generate very low-pressure steam. The
results for this situation show that a double-e!ect chiller o!er better operation bene"t than
a single-e!ect chiller ($ 171 h). But for this hypothetical situation, compression chillers will be
the preferred option without discussion under the economic point of view, as this is also the
option that o!ers a lower capital cost.
Another interesting hypothetical case is the presence of waste heat to generate low-pressure
steam. Assuming the existence of this process heat and using absorption chillers, there are two
708 J. C. BRUNO, J. MIQUEL AND F. CASTELLS
Copyright 2000 John Wiley & Sons, Ltd. Int. J. Energy Res. 2000; 24:695}717
options. One is to use the steam directly in the double-stage chiller, and the other is to expand
it to very low pressure and use it in a single-e!ect chiller. The heat required to drive the
double-e!ect machine to supply the refrigeration required by the process plant is 2127 kW. The
operation bene"t for this case is $ 227 h, which is higher than the one o!ered by single e!ect or
compression systems. Logically when low-pressure steam raised using waste heat is available, to
use it directly in a double-e!ect chiller is more attractive economically than #ash it to very low
pressure and use it in a single-e!ect machine or use a compression chiller.
Let us assume now the presence of waste heat to generate high pressure steam (3100 kPa). For
this hypothetical case, which would be a real situation in other similar plants, knowing that the
steam is consumed at low (double e!ect) or very low (single e!ect) pressure, this steam may be
initially used in a steam turbine to generate mechanical power or to drive an electric generator.
Once this steam has been expanded it may be used to drive a single- or double-e!ect absorption
cycle. To consider this option it must be necessary that the existing steam turbine (1100 kW) will
be able to handle the increased #ow rate of steam through the steam network due to the presence
of the absorption cycle, or its substitution must be considered with a higher capacity turbine.
The double e!ect chiller o!ers a higher e$ciency, and logically the operation bene"t will turn
out to be higher. The drawback for this particular case will be a lower consumption of steam,
therefore the power generated in the steam turbine is expected to be lower. The point here is
which of these two e!ects have a greater in#uence.
The power generated in the steam turbine is 1400 kW when a single-e!ect chiller is used. The
operation bene"t for this case is $ 232 h. The bene"t using a double-e!ect cycle is $234 h and
the electric power generated in the steam turbine is only 1290 kW. The operation bene"t being
very similar in both cases, the cycle selection has to be made attending to other factors, such as,
investment costs of the chiller and steam turbine unit, in the case the latter has to be replaced,
availability of cooling water, possibility of taking pro"t of the extra power generated, and some
other existing plant constrain.
5.2. Increase of the cogeneration capacity
The inclusion of an absorption chiller in the steam network increases the steam demand.
Advantage can be taken from this situation by including a gas turbine of higher capacity to
increase simultaneously the production of power and heat used to supply steam simultaneously
to process plants and to the absorption cycle.
When the absorption cycle is included the proposed cogeneration cycle has to be sized. When
only compression chillers are considered the existing GT-1 of 5 MW is used. Gas turbines are
supplied in standard sizes, thus the size of the plant is constraint to the commercially available gas
turbine cycles. The viability of using gas turbines of 6.6 MW (TurboMach TGC-650-CT) and
9 MW (TurboMach TGC-880) is considered. For the plant under study, higher capacities are not
selected, because it has been seen that they lead to excess of heat in the system. Also if a steam
turbine is present in the steam network (combined cycle), as it is the case here, the potential
increase of power capacity in the steam turbine needs to be calculated.
When a 6.6 MW gas turbine is used, the heat contained in the exhaust gases is not high enough
to supply all the required heat, and supplementary "ring fuel has to be provided. Three situations
are compared:
A. The existing steam turbine (1100 kW) is not replaced.
B. A new steam turbine is purchased to take advantage of the potential increase of power.
OPTIMAL INTEGRATION OF WATER/LITHIUM BROMIDE ABSORPTION CHILLERS 709
Copyright 2000 John Wiley & Sons, Ltd. Int. J. Energy Res. 2000; 24:695}717
Table II. Results using a gas turbine of 6.6 MW.
Single e!ect Double e!ect
Situation Operation PBP Operation PBP
bene"t ($/h) (years) bene"t ($/h) (years)
A 225 4.2 250 3.7
B 239 4.2 260 3.8
C 239 3.8 260 3.5
Table III. Results using a gas turbine of 9 MW.
Single e!ect Double e!ect
Situation Operation PBP Operation PBP
bene"t ($/h) (years) bene"t ($/h) (years)
A 305 4.1 315 4.0
B 318 4.1 325 4.1
C 318 3.9 325 3.8
C. The existing steam turbine is assumed to be currently operating below its maximum load,
and is able to supply the potential increment of power.
The results for each situation and each refrigeration alternative are given in Tables II and III,
using a new gas turbine of 6.6 and 9 MW, respectively.
Using a gas turbine of 6.6 MW, it is seen in situation A, that the operation bene"t is increased
by 11 per cent when a double-e!ect cycle is used. This is due to the fact that the consumption of
supplementary "ring fuel when a double-e!ect is selected is only a 54 per cent of the supple-
mentary "ring fuel required when a single-e!ect cycle is used. In situations B and C this increment
is a bit lower (9 per cent) due to the greater quantity of steam that #ows through the steam
turbine, that is, due to the higher steamturbine power generated when a single-e!ect cycle is used.
Using a gas turbine of higher capacity, 9 MW, the di!erences between single and double e!ect
are less signi"cant. The double-e!ect cycle o!ers a 3.3 per cent higher bene"t than single e!ect for
situation A, and only 2.2 per cent for situations B and C. The reason is that for this cogeneration
system there is no need for supplementary "ring using single e!ect. When double e!ect is used
there is a small excess of heat. Thus, the di!erence is reduced to a lower consumption of cooling
water and demineralized water when the option of double e!ect is chosen.
The operation bene"t is $211 h using a compression cycle with a COP of 6.0 and a gas
turbine of 5 MW. For this alternative the payback period is of 3.5 and 3.2 years, in the case that
a new steam turbine of 1100 kW needs to be purchased or in the case that it is not necessary,
respectively.
As seen in Tables II and III, the preliminary design including absorption cycles present
a reasonable payback period quite similar to those obtained using conventional compression
chillers. The lower e$ciency and higher investment costs of absorption cycles are compensated by
710 J. C. BRUNO, J. MIQUEL AND F. CASTELLS
Copyright 2000 John Wiley & Sons, Ltd. Int. J. Energy Res. 2000; 24:695}717
the bene"t obtained selling the additional electricity generated. The situation is even more
attractive when a steam turbine can be used to take advantage of the power generated by
the expansion of the additional steam used to drive the absorption cycle, as commented in the
previous section. In spite of the extra investment required to purchase the steam turbine, this
option has got a great negative impact on the payback period, as may be concluded comparing
situations B and C.
There is a wide range of possible situations when absorption cycles are included increasing the
cogeneration system capacity, but depending on the required level of investment, the payback
period is reasonable, ranging from 3.5 to 4.2. From the economic point of view, the best
refrigeration alternative is a double-e!ect cycle integrated in a cogeneration system with a 9 MW
gas turbine. The optimization results for this alternative are given in Tables IV and V. Absorption
chillers require a higher capital investment and o!er a higher PBP than compression chillers, but
its operation bene"t is clearly higher. Thus, the net present value (NPV) of absorption alternatives
turns to be more attractive. The NPV (with a capital recovery factor"0.12) corresponding to the
best absorption alternative (double e!ect with a 9 MW gas turbine) is $15 320 000, while the NPV
for the compression alternative is $10 440 000. The energetic e$ciency of both alternatives is
graphically showed in Figure 6, which represents how the energy input, the energy contained in
the fuel consumed, is used in the global energy plant. In this "gure, the "rst bar corresponds to the
best absorption cycle alternative. For each 100 kW of input heat to the global energy plant,
42 kW is used for heating, 35 kW for electricity, 5 kW for refrigeration and the rest 18 kW are
assumed to be losses. The last bar corresponds to the compression alternative. The bars in the
middle of the graph show the e$ciency corresponding to the compression plus the use of an
external power station to supply the extra electricity that is not generated when this alternative is
chosen. In this case it is considered that the internal and external consumption of primary energy,
that is, the external energy consumed by a national power plant generates the additional
electricity that could be obtained in the case that the best absorption alternative was used. The
electric e$ciency assumed for the national power plant is 0.35, 0.45 and 0.55, respectively. Taking
into account only the energy consumed by the global energy plant, the compression system is the
most e$cient. The absorption alternative is as e$cient as a compression system and an external
power plant with an electric e$ciency of 0.45, regarding the consumption of natural resources
under the point of view of a whole country. The electricity generated by the external power plant
is needed to supply the electricity not given by the cogeneration system when a compression cycle
is selected.
6. CONCLUSIONS
A modelling and optimization tool to study the economic viability of absorption chillers
integration in energy systems (CHP plants) has been proposed. With this approach the best
refrigeration cycle, including the option of using a compression cycle, is selected among the
technically and commercially available cycle alternatives, constraint to the features of the
required refrigeration load and temperature level and also the features of the combined heat and
power plant under study. The methodology presented in this work is useful for:
(a) the evaluation of operation costs coming from the integration of refrigeration cycles in
combined heat and power plants based on cogeneration systems,
OPTIMAL INTEGRATION OF WATER/LITHIUM BROMIDE ABSORPTION CHILLERS 711
Copyright 2000 John Wiley & Sons, Ltd. Int. J. Energy Res. 2000; 24:695}717
Table IV. Results for the optimal global energy plant (Figure 5).
State point Flow rate Pressure Temperature Enthalpy
(kg h) (kPa) (3C) (kJ kg)
1 135 545 101 15 15
2 135 545 1327 394 408
3 137 646 1278 1000 1133
4 137 646 103 471 505
5 137 646 101 153 160
6 100 621 101 15 15
7 100 621 1079 371 383
8 106 640 1249 945 1066
9 106 640 103 449 481
10 106 640 101 180 188
11 4 500 1818 278 2977
12 158 15 3100 325 3055
13 54 602 159 2753
14 161 37 131 107 450
15 322 3100 236 1017
16 268 602 159 671
17 115 73 3100 325 3055
18 39 602 159 2753
19 118 09 131 107 450
20 236 3100 236 1017
21 197 602 159 671
22 2329 131 107 450
23 2329 3100 325 3055
24 0 * * *
25 0 * * *
26 24 162 3100 325 3055
27 4702 1818 278 2977
28 19 460 602 195 2836
29 5555 3100 325 3055
30 0 * * *
31 0 * * *
32 203 1818 278 2977
33 203 602 260 2977
34 908 602 196 2837
35 15 704 602 196 2837
36 26 263 150 90 377
37 38 131 107 2686
38 30 275 131 107 450
39 3144 602 196 2837
40 3144 602 95 398
41 3144 602 95 398
Electric power generated Consumption of natural gas
GT-1 9015 kW GT-1 2701 Nm` h
GT-2 6991 kW GT-2 1953 Nm` h
Steam turbine 1264 kW Boiler HRSG-1 0 Nm` h
712 J. C. BRUNO, J. MIQUEL AND F. CASTELLS
Copyright 2000 John Wiley & Sons, Ltd. Int. J. Energy Res. 2000; 24:695}717
Table V. Example of results of the double e!ect parallel #ow water/LiBr absorption chiller
(Figure 3) integrated in the CHP plant as shown in Figure 5.
State point Flow rate Pressure Temperature Enthalpy Composition
(kg s) (kPa) (3C) (kJ kg) (% LiBr)
1 1.06 5.9 3.6 149.8 *
2 1.06 0.9 4.5 149.8 *
3 1.06 0.9 4.5 2508 *
4 12.72 0.9 40 104.5 57.75
5 12.72 5.9 40 104.5 57.75
6 12.72 5.9 68 159.7 57.75
7 6.36 5.9 68 159.7 57.75
8 6.36 73.0 68 159.7 57.75
9 6.36 73.0 118 260.2 57.75
10 5.80 73.0 154 342.5 63.33
11 5.80 73.0 94 232.3 63.33
12 5.80 5.9 89 232.3 63.33
13 6.36 5.9 68 159.7 57.75
14 6.36 5.9 86.5 197.2 57.75
15 5.86 5.9 87 216.9 62.67
16 11.66 5.9 90 224.5 63.00
17 11.66 5.9 57 164.2 63.00
18 11.66 0.9 51 164.2 63.00
19 0.56 73.0 154 2783 *
20 0.56 73.0 91 381.8 *
21 0.56 5.9 36 381.8 *
22 0.50 5.9 87 2683 *
23 0.87 602 196 2837 *
24 0.87 602 156 671.2 *
25 0.87 602 95 398.1 *
Power consumption of the low-pressure pump 0.05 kW
Power consumption of the high-pressure pump 0.34 kW
Cooling water #ow rate 569 m` h
COP 1.17
(b) the study of the integration of absorption cycles in cogeneration systems. This integration
increases the heat demand and consequently the electric power is also increased, as
a consequence of the fact that a higher capacity cogeneration system is needed to supply
additional heat to drive the absorption cycle. For this case, the program XV, allows for the
sizing and determination of the operating parameters of the most adequate gas turbine and
absorption cycles. If the energy plant includes a combined cycle, then the size and operating
parameters of the steam turbine are also determined.
The optimisation of the global energy plant has been implemented in the computer program
XV, and tested using the real plant data of the petrochemical facility of Bayer in Tarragona. In
this plant, waste heat is recovered from the process plant to raise very low-pressure steam to drive
a single-e!ect water/LiBr chiller. Other refrigeration cycle alternatives are considered for this
plant. When a double-e!ect water/LiBr chiller is used, the steam needs to be generated by using
OPTIMAL INTEGRATION OF WATER/LITHIUM BROMIDE ABSORPTION CHILLERS 713
Copyright 2000 John Wiley & Sons, Ltd. Int. J. Energy Res. 2000; 24:695}717
Figure 6. E$ciency of the energy system. (Note: PS"external power station, in parenthesis
it is given its electrical e$ciency).
additional supplementary "ring fuel, and a worse operation bene"t is obtained. Using a com-
pression cycle the result is also worse than the one obtained by the single-e!ect cycle drived by
waste heat.
Other options assuming the hypothetical presence of waste heat are able to generate low- or
high-pressure steam, and in the latter case, also generate additional power in a steam turbine,
which show that the double-e!ect chiller is the best option if waste steam at the required pressure
is available. Also the option of using a higher capacity steam turbine prove to be economically
attractive.
Finally, it was seen that when a higher capacity cogeneration system is selected, the additional
heat could be used to drive absorption cycles. In this situation, a higher investment required is
compensated by the operation bene"t. Absorption alternatives o!er a reasonable payback period
and better NPV than the conventional compression systems. In summary, this methodology is
useful for the assessment of absorption chillers integration in combined heat and power plants
in order to save energy. Also the applications made show the bene"cial e!ect coming from the
integration of absorption chillers and cogeneration systems.
NOMENCLATURE
ARC "absorption refrigeration chiller
BW "boiling water header
C "cost per unit, $ unit
CHP "combined heat and power plant
COND "condensate header
COP "coe$cient of performance
CW "cooling water
714 J. C. BRUNO, J. MIQUEL AND F. CASTELLS
Copyright 2000 John Wiley & Sons, Ltd. Int. J. Energy Res. 2000; 24:695}717
D "deaerator
DW "demineralized water
F "#ow rate, kg h
GT "gas turbine
HP "high-pressure steam header
HPG "high-pressure generator
HRSG "heat recovery steam generator
HTA "high-temperature absorber
HTE "high-temperature evaporator
HTHX "high-temperature heat exchanger
LP "low-pressure steam header
LPG "low-pressure generator
LTA "low-temperature absorber
LTE "low-temperature evaporator
LTHX "low-temperature heat exchanger
MP "medium-pressure steam header
PBP "payback period
Q "heat, kW
ST "steam turbine
VLP "very low-pressure steam header
= "power, kW
WHB "waste heat boiler
Superscripts
a "absorption cycle
c "compression cycle
eg "electricity grid
gt "gas turbine
h "high pressure
HRSG "heat recovery steam generator
l "low pressure
pp "process plant
st "steam turbine
Subscripts
a "absorber
co "condenser
cw "cooling water
dw "demineralized water
e "electricity
ev "evaporator
F "fuel
FG "#ue gas
g "generator
h "high
OPTIMAL INTEGRATION OF WATER/LITHIUM BROMIDE ABSORPTION CHILLERS 715
Copyright 2000 John Wiley & Sons, Ltd. Int. J. Energy Res. 2000; 24:695}717
l "low
m "medium
m "mechanical power
pp "process plant
r "refrigerant
WHB "waste heat boiler
ACKNOWLEDGEMENTS
The authors would like to fully appreciate the collaboration of the Bayer technical sta! in Tarragona
(Catalonia, Spain) in testing the methodology proposed in this work.
REFERENCES
Alvarado S, Gherardelli C. 1994. Exergoeconomic optimization of a cogeneration plant. Energy 19:1225}1233.
ASHRAE. 1977. Fundamentals Handbook. American Society of Heating, Refrigerating and Air Conditioning Engineers:
1977. New York.
Balestieri JAP, Correja PB. 1997. Multiobjective linear model for pre-feasibility design of cogeneration systems. Energy,
22:537}548.
Bruno JC, FernaH ndez F, Castells F. 1996. Absorption chillers integration in a combined heat and power plant.
Proceedings of the International Absorption Heat Pump Conference, Montreal, Canada, vol. II:759}767.
Bruno JC, FernaH ndez F, Castells F, Grossmann IE. 1998. A rigorous MINLP model for the model for the optimal
synthesis and operation of utility plants. ransactions of the Institution of Chemical Engineers 76:246}258.
Bruno JC. 1999. IntegracioH de cicles d'absorcioH en xarxes d'energia en plantes de proceH s. Ph.D. hesis, Universitat Rovira
i Virgili, Tarragona.
Chang CT, Hwang JR. 1996. A multiobjective programming approach to waste minimization in the utility systems of
chemical processes. Chemical Engineering Science 51:3951}3965.
Cheung K, Hwang Y, Judge JF, Kolos K, Singh A, Radermacher R. 1996. Performance assessment of multistage
absorption cycles. International Journal of Refrigeration 19:473}481.
Chou CC, Shih YS. 1987. A thermodynamic approach to the design and synthesis of plant utility systems. Industrial
Engineering Chemical Research 26:1100}1108.
Colmenares TR, Seider WD. 1989. Synthesis of utility systems integrated with chemical process. Industrial Engineering
Chemical Research 28:84}93.
Diaz MS, Bandoni JA. 1996. A mixed integer optimization strategy for a large scale chemical plant in operation.
Computers and Chemical Engineering 20:531}545.
Dorgan CB, Leight SP, Dorgan CE. 1995. Application Guide for Absorption Cooling/Refrigeration ;sing Recovered Heat.
ASHRAE Publications: Atlanta.
Elmasri, M. (1993). Gas turbine performance. Design of Gas urbine Combine Cycle and Cogeneration Systems, CIDESPA,
Madrid.
Erickson DC, Papar RA, Anand G. 1996. Basic GAX cycle prototype results. International Absorption Heat Pump
Conference, Montreal, Canada, vol. II:717}724.
Frangopoulos CA, Lygeros AI, Markou CT, Kaloritis P. 1996. Thermoeconomic operation optimization of the hellenic
aspropyrgos re"nery combined-cycle cogeneration system. Applied hermal Engineering 16:949}958.
Garret DE. 1989. Chemical Engineering Economics. Van Nostrand Reinhold: New York.
Grossman G, Wilk M. 1994. Advanced modular simulation of absorption systems. International Journal of Refrigeration
17:231}244.
Grossman IE. 1996. Mixed-integer optimization techniques for algorithmic process synthesis. Advances in Chemical
Engineering 23:171}246.
Hart DR, Rosen MA. 1996. Environmental and health bene"ts of district cooling using utility-based cogeneration in
Ontario, Canada. Energy 21:1135}1146.
Henning D. 1997. Modest*an energy-system optimisation model applicable to local utilities and countries. Energy
22:1135}1150.
Herold KE, Radermacher R, Klein SA. 1995. Absorption Chillers and Heat Pumps. CRC Press: New York.
Heui-Seok Y, Yeong-Koo Y, Jin-Kuk K, Myung K, Sun SK. 1998. A rule-based steam distribution system for
petrochemical plant operation. Industrial & Engineering Chemistry Research 37:1051}1062.
716 J. C. BRUNO, J. MIQUEL AND F. CASTELLS
Copyright 2000 John Wiley & Sons, Ltd. Int. J. Energy Res. 2000; 24:695}717
Heyen G, Kalitventze! B. 1997. Methodology for optimization of operation to reduce site-scale energy use in production
plants. Applied hermal Engineering 17:1005}1014.
Hua B, Chen QL, Wang P. 1997. A new exergoeconomic approach for analysis and optimization of energy systems.
Energy 22:1071}1078.
Hui CW, Natori Y. 1996. An industrial application using mixed-integer programming technique: a multi-period utility
system model. Computers and Chemical Engineering 20(Suppl.):S1577}S1582.
Iyer RR, Grossmann IE. 1997. Optimal multiperiod operational planning for utility systems. Computers and Chemical
Engineering 21:787}800.
Kalitventze! B. 1991. Mixed integer non-linear programming and its application to the management of utility networks.
Engineering Optimization 18:183}207.
Koeppel EA, Klein SA, Mitchell JW. 1995. Commercial absorption chiller models for evaluation of control strategies.
Ashrae echnical Data Bulletin 11:71}80.
Maia LOA, de Carvalho LAV, Qassim RY. 1995. Synthesis of utility systems by simulated annealing. Computers and
Chemical Engineering 19:481}488.
Maia LOA, Qassim RY. 1997. Synthesis of utility systems with variable demands using simulated annealing. Computers
and Chemical Engineering 21:947}950.
Marechal F, Kalitventzef B. 1997a. E!ect modelling and optimization, a new methodology for combined energy and
environment synthesis of industrial processes. Applied hermal Engineering 17:981}992.
Marechal F, Kalitventze! B. 1997b. Identi"cation of the optimal pressure levels in steam networks using integrated
combined heat and power method. Chemical Engineering Science 52:2977}2989.
Meloche N, Larsson I, Snoek C. 1996. Integration district cooling with combined heat and power. Proceedings of
International Absorption Heat Pump Conference, Montreal, Canada, vol. II:795}802.
Miquel J. 1991. AplicacioH d'algoritmes d'OptimitzacioH a xarxes de vapor de plantes de proceH s. PhD. hesis, Universitat de
Barcelona.
Mohanty B, Panda H. 1993. Integrated energy system of industrial complexes. Part I: a linear programming approach.
Applied Energy 46:317}348.
Mohanty B, Panda H. 1995. Integrated energy system for industrial complexes. Part II: a case study using the LP model.
Applied Energy 51:19}38.
Mohanty B, Paloso G. 1995. Enhancing gas turbine performance by intake air cooling using an absorption chiller. Heat
Recovery Systems and CHP 15:41}50.
Najjar YSH. 1996. Enhancement of performance of gas turbine engines by inlet air cooling and cogeneration system.
Applied hermal Engineering 16:163}173.
Papalexandri KP, Pistikopoulos EN, Kalitventeze! B, Dumont MN, Urmann K, Groschluter J. 1996. Operation of
a steam production network with variable demands modelling and optimization under uncertainty. Computers and
Chemical Engineering 20(Suppl.):S763}S768.
Papoulias SA, Grossmann IE. 1983. A structural optimization approach in process synthesis*I, utility systems.
Computers and Chemical Engineering 7:695}706.
Patterson MR, Perez-Blanco H. 1988. Numerical "ts of the properties ol lithium bromide}water solutions. ASHRAE
ransactions 94(2):2059}2077.
Puigjaner L. 1997. Process integration with combined heat and power (CHP). Applied hermal Engineering 17:1015}1034.
SahH n J, Vale F. 1994. Economic Aspects of Absorption Systems. Course on Energy Management by Absorption Cycles,
Chapter 7. Ciemat: Madrid.
Tsatsaronis G. 1993. Thermoeconomic analysis and optimization of energy systems. Progress in Energy Combustion
Science 19:227}257.
Valero A, Lozano MA, BartolomeH JL. 1996. On-line monitoring of power plant performance using exergetic cost
techniques. Applied hermal Engineering 16:933}948.
Wardono B, Nelson RM. 1996. Simulation of a double-e!ect LiBr/water absorption cooling system. ASHRAE Journal
38:32}38.
Xu GP, Dai YQ, Tou KW, Tso CP. 1996. Theoretical analysis and optimization of a double-e!ect series-#ow-type
absorption chiller. Applied hermal Engineering 16:975}987.
Xu GP, Dai YQ, Tou KW, Tso CP. 1997. Theoretical analysis and optimization of a double-e!ect parallel-#ow-type
absorption chiller. Applied hermal Engineering 17:157}170.
OPTIMAL INTEGRATION OF WATER/LITHIUM BROMIDE ABSORPTION CHILLERS 717
Copyright 2000 John Wiley & Sons, Ltd. Int. J. Energy Res. 2000; 24:695}717