Design and Experimental Performance Evaluation of An Absorption Heat Transformer With Self-Circulation
Design and Experimental Performance Evaluation of An Absorption Heat Transformer With Self-Circulation
Design and Experimental Performance Evaluation of An Absorption Heat Transformer With Self-Circulation
257-272, 1995
Pergamon
0890-4332(94)E0029-J
Akstract--A 10 kW experimental absorption heat transformer unit operating with self-circulation has been
thoroughly tested. The self-circulation is obtained according to the thermosyphon principle. The pressure
difference in the unit is achieved through a difference in hydrostatic pressures. Theoretical relationships
for the pressure profiles within the different components of the heat transformer have been derived.
Stabilising the self-circulation has been the primary objective in this work, rather than the optimisation
of the efficiency of heating and cooling areas. A satisfactory stable operation with self-circulation has been
achieved.
A reference heat transformer plant, delivering 100 kW, has been designed and installed in a major pulp
and paper mill. This unit is directly incorporated with one of the evaporation plants of the mill. Plant
operation data obtained under real industrial conditions are presented.
1. I N T R O D U C T I O N
In 1922, the two Swedish students Carl Munters and Baltzar von Platen designed an absorption
refrigerator working with self-circulation [1]. The refrigerator was operating with N H 3 - H 2 0 as the
working pair. The total pressure was equal throughout the apparatus, but the partial pressure
difference between the generator/condenser and evaporator/absorber was maintained with an inert
gas, such as H2, to achieve a lower pressure over the ammonia and the water in the absorber and
the evaporator. The refrigerator had no pumps, so circulation was maintained by the density
difference between the strong and weak solutions and utilisation of the thermosyphon effect in the
generator.
In 1986, an absorption heat transformer (AHT) was built at the Dept of Chem. Eng. 1, based
on self-circulation and operating with H z O - N a O H as the working pair [2]. This A H T works with
heat input to the evaporator and the generator and with no electricity input at all. The pressure
difference between the components is achieved by the difference in hydrostatic pressures. The
reason to construct an A H T with self-circulation can be explained by considering an A H T
operating as shown in Table 1. In an AHT, the largest stream is the strong solution from the
absorber to the generator. This stream is also warmer than the weak solution stream from the
generator. The temperature of the strong solution leaving the absorber is 135C and is saturated.
After passing through the heat exchanger, the temperature of the strong solution is at 94C and
it becomes subcooled. The saturation temperature for the strong solution is 80C at 6.5 kPa, so the
strong solution will expand through the valve and a two-phase flow will be formed with about 1.5%
of steam, if the pressure is assumed to be 6.5 kPa directly after the valve. Even if an infinite solution
heat exchanger is assumed, the strong solution can never be cooled to the same temperature as the
weak solution leaving the generator. So the strong solution will always flash after the valve in an
A H T cycle.
Conversely, consider an absorption heat pump (AHP) operating at the temperature and pressure
levels shown in Table 2. In the AHP, the largest stream is the strong solution from the absorber
to the generator, but the weak solution stream is warmer than the strong solution stream in this
type of cycle. The temperature of the weak solution leaving the generator is 144C and is saturated.
After passing the heat exchanger, the temperature of the weak solution is 100C and is unsaturated.
The saturation temperature of the weak solution is 101C at 14.3 kPa, so the weak solution will
257
K. ABRAHAMSSON el al.
258
Evaporator:
Absorber:
Generator:
Condenser:
Heat exchanger:
External
heat/cool
letup
(C)
Internal
outgoing
temp
(C)
90
130
90
36
88
135
85
38
Internal
pressure
(kPa)
UAvalue
(kW/ C)
Temp
difference
(C)
64.9
64.9
6.5
6.5
64.9
500
192
192
500
110
2.0
5.0
5.0
2.0
7.0*
Cone. of
outgoing
solution
(%-H20)
100.0
48.6
44.9
100.0
*LMTD.
not flash after the valve. To ensure that the weak solution will always flash after the valve, in this
type of cycle, a much smaller solution heat exchanger has to be incorporated, but this results in
a lower COPh (coefficient of performance for heating).
Applying the self-circulation principle to an A H T will eliminate the need for solution pumps.
2.
Several improvements in the A H T design have been made since the first prototype [2], in order
to obtain stable circulation of the liquid streams and to stabilise the pressure in the evaporator and
the absorber. The most important modifications of the A H T were the reconstruction of the
absorber, part of the generator and the solution heat exchanger and the incorporation of control
devices. The concept of stabilising the self-circulation has been the primary objective in this design,
rather than the optimisation of the efficiency of heating and cooling areas.
2.1. Design aspects of the pilot-plant absorption heat transformer
The total pressure between the components varies in the AHT. The pressure difference is achieved
by the difference in hydrostatic pressures. The generator and the condenser are located at a higher
level than the evaporator and absorber. The distance between the condenser and evaporator is
determined by the desired pressure difference and the choice of the working pair used in the AHT,
as is the distance between the generator and the absorber. The pilot plant is constructed to maintain
a pressure difference of approximately 1 bar with the working pair H20--NaOH. Other solutions
which have a density close to that of H 2 0 - N a O H , i.e. approximately 1.4 times the density of water,
for example, H2 O - K O H , could also be used in the same unit. A pressure difference of 1 bar leads
to a fairly tall AHT, which is 14 m between the lowest and the highest point. The evaporator and
absorber are located at 10 m and 7 m below the condenser and generator, respectively. The heat
exchanger is located under the absorber, in order to avoid flashing before heat exchanging weak
and strong solutions. Figure 1 shows the A H T and the component orientation at the different floors
in the pit of the equipment hall.
The components included in the laboratory unit of the AHT are made fairly simple. To avoid
corrosion problems, most of the parts are built of Sanicro 28, a nickel and chrome metal alloy,
but constructions of stainless steel also exist. There are no moving parts inside the AHT, which
also minimises the risk of corrosion problems.
The evaporator is a jacketed vessel with a total heat transfer area of 0.65 m:. Steam is used as
the heating medium in the jacket. The refrigerant condensate from the condenser enters at the
Table 2. Simulated data for an absorption heat pump ( W P : H 2 0 - N a O H )
Evaporator:
Absorber:
Generator:
Condenser:
Heat exchanger:
*LMTD.
External
heat/cool
temp
(C)
Internal
outgoing
temp
(C)
55
90
149
90
53
95
144
92
Internal
pressure
(kPa)
UAvalue
(kW/C)
Temp
difference
(C)
14.3
14.3
77.0
77.0
77.0
500
266
282
538
124
2.0
5.0
5.0
2.0
6.9*
Conc. of
outgoing
solution
(%-H20)
100.0
49.5
45.9
100.0
259
naer
Conde:
Genel
Condel
T]
vi
Steamin ~
Condensateout ~
. . . . . . . . . . . . . . . . .
-~ ~ Evaporator
]
i. . . . . . . .
Fig. 1. The pilot-plant absorption heat transformer working with self-circulation(scale 1:60).
bottom, is evaporated and leaves at the top. A U-tube is connected to the bottom and the top of
the evaporator, together with a conductive measuring rod, so the volume in the evaporator is
continuously registered on a printer during operation. The volume of water on the inside of the
evaporator is normally 2/3 full during operation. Variations in the concentration of the working
solution can be indirectly read out from the printed graph when the volume of working medium
in the evaporator is changing. In this type of construction, the heat transfer area is related to the
volume of working medium in the evaporator.
The absorber is designed as a falling-film absorber. It is a jacketed vessel with a total heat transfer
area of 1.18 m 2. The film flows down the inside wall of the absorber and the cooling medium flows
upward in the jacket. Vapour from the evaporator enters the absorber at the top and fills it totally
with vapour. The weak solution is spread out on the absorber wall at the top, the vapour condenses
into the falling-film and the strong solution leaves the absorber at the bottom. The volume of strong
solution in the bottom of the absorber can be viewed in a glass-tube. From the lower part of the
absorber, a vent pipe is connected from the inner part of the absorber to the thermosyphon pipe
above the throttling valve. De-aerating the absorber will contribute to an additional pumping effect
260
K. ABRAHAMSSON et al.
in the thermosyphon pipe. The cooling system can either be arranged so the absorber is chilled
by a glycol stream or by saturated water, in order to produce high pressure steam.
Part of the generator is constructed as a climbing-film generator with the other part as a jacketed
vessel at the top. The total area is 0.84 m 2, where the climbing-film area is 0.39 m 2 and the constant
volume part has an area of 0.45 m 2. The climbing-film generator is a one-tube heat exchanger
constructed of a 21 mm tube. The same steam is used as heating medium in both sections of the
generator. The strong solution enters the inside tube at the bottom of the climbing-film section and
the tube ends above the surface of the weak solution in the jacketed vessel. The weak solution leaves
the generator from the bottom of the jacketed vessel via a U-tube, down to the solution heat
exchanger and the absorber. The vapour produced in the generator leaves from the top to a cyclone
which is located after the generator, in order to separate entrained drops of working solution. The
top of the generator consists of two jacketed vessels. The lower one, mentioned above, is for
heating. The one on the top is an extra condenser (previously the old condenser) that may be used
to prevent the weak solution being too weak. If two volatile media are used, it may be used as
a rectifier.
The condenser is a vessel which contains a 10 mm copper-spiral with an area of 0.63 m 2. Water
is used as the cooling medium in counter-current mode. The vapour from the generator-separator
enters at the top and the condensate leaves at the bottom. At the top of the condenser, a pipe is
connected to a vacuum pump to secure the removal of all air before the start-up procedure and
removal of air and non-condensables during the operation.
The solution heat exchanger is a plate heat exchanger (Alfa-Laval Type P01-VB). The heat
exchanger consists of 26 plates in lncoloy 825. The total heat transfer area is 0.77 m 2. Strong and
weak solutions are split in two streams, which flow in counter-current mode through the solution
heat exchanger five times each. Before installation in the AHT, the plate heat exchanger was first
tested against design data in a separate pump-rig, where it was revealed that the plate heat
exchanger was close to design data [3].
2.2. Pressure profiles
A vital detail in the design of the A H T with self-circulation is the importance of taking the
pressure dependence into consideration. The total flow is very much dependent on the absolute
pressure in the evaporator and the condenser, where the highest and the lowest pressure prevail,
respectively. To make a qualitative analysis of the flow in the A H T with self-circulation, three
equations can be stated. Equation (1) gives the requirement for the strong solution to flow upward
from the absorber to the generator. Equation (2) gives the corresponding requirement for the weak
solution to flow downward from the generator to the absorber and, finally, equation (3) gives the
requirement for the condensate to flow downward from the condenser to the evaporator.
p ~.o,a. . . . .. . Ap oc
AO + AP AG
I. . . .j_. . AU gr,,ity
toss< Pevaporator -- AP iEA
os
(!)
GC
GA
Pcond ..... + A P ,oss + m e gravity - - P
(2)
GA
,oss > Pevaporator -
P~o.d. . . . . -~ AP CE
gr,~itr - AP CE
Ios~> Pevaporator,
m e ,EosA
s
(3)
where P = absolute pressure; APgr~vity= gravitational pressure difference between different components; AP~oss= flow-induced pressure drop between different components.
The pressure in the condenser, Po,a...... depends on the temperature and flow rate of the cooling
medium to the condenser and the heating medium to the generator. Similarly, the pressure in the
evaporator, P~v~por~,or,depends on the temperature and flow rate of the heating medium to the
evaporator and the cooling medium to the absorber.
The gravitational pressure differences were assigned separate terms in the above equations, due
to their influential importance in the total cycle. The flow-induced pressure drops can be divided
into pressure drops due to friction, A P f , and pressure drops due to acceleration, AP~~. The term
APtoss can thus be written as:
APtoss = A P f q- AP,~c.
(4)
261
solution heat exchanger is the most dominating pressure loss. The only term where the pressure
lOSS due to acceleration is important is the AP A~
~o~sterm. Even here, the pressure drop in the solution
heat exchanger is an important factor.
As the heights of the liquid streams in the different pipes are a direct result of the pressure
difference between evaporator and condenser, they are balanced due to the density of each stream
and the gravity, according to:
Ah = APheight.
(5)
Pg
When calculating the hydrostatic pressure of the strong solution after the throttling valve, the
distribution of the different densities for the liquid phase and the vapour phase has to be calculated.
The definition of the two-phase density can be written as a function of the volumetric void [4, 5] Rv:
p = Rvp~+ (1 - gv)'p~,
(6)
where Rv = Vv/V for a small height segment. There are different empirical correlations for
calculating the volumetric void. The most well-known is the Lockhart-Martinelli correlation [5].
When the strong solution leaves the absorber it first passes through the solution heat exchanger
where it is subcooled. In an A H T with self-circulation, the design of the solution heat exchanger
is an important part of the total design. To achieve high COPts, the heat transfer in the solution
heat exchanger has to be as effective as possible, but to achieve stable operation of the selfcirculation, the pressure drop in the solution heat exchanger has to be as low as possible. The
quality and velocity of the flow of the strong solution, after the throttling valve, depends on the
temperature and concentration of the strong solution, after the solution heat exchanger and
the pressure drop over the throttling valve. The higher the temperature, the more vapour will be
produced and the more effective the thermosyphon effect will be. The thermosyphon effect is
supposed to occur directly after the throttling valve and to avoid expansion before the throttling
valve, the orifice of the throttling valve has to be as small as possible. When a climbing-film
generator is used, the upward flow will be even more effective if the heat flux to the generator is
maximised. The throttling valve is located right after the solution heat exchanger and the pipe after
the throttling valve is extended straight up into the generator.
2.3. Control system
A very simple control system has been implemented, with satisfactory results. Two PI-regulators
are used to control the pressure of the steam to the evaporator and the generator. The steam pipes
are arranged so that it is possible to regulate the steam supply to both modules with only one
PI-regulator. It is also possible to disconnect the evaporator and introduce the steam directly into
the absorber instead. During this operation, the condensate from the condenser flows via a
condensate drum to waste.
When glycol is used as the cooling medium in the absorber, a pump is used to circulate the glycol
between a storage tank and the absorber. The absorber is chilled in a counter-current mode, with
the inlet of the glycol located at the bottom of the absorber. The inlet temperature of the glycol
is controlled by a thermostat, which controls the flow of cooling water through a spiral tube inside
the glycol storage tank. When water is used as the cooling medium, the heat produced in the
absorber will produce high pressure steam. An external self-circulation system of the water ensures
that the water at the inlet is always saturated. The pressure of the steam is controlled using an
external dump condenser.
The flow of cooling water to the condenser is controlled by a thermostat working according to
the principle of closed liquid-expansion. An increase of the saturation temperature of the vapour
flowing from the generator expands the liquid, to open the regulation valve for cooling water, and
the reversed action will occur for a decrease of the saturation temperature, to ensure that the correct
condenser pressure is maintained.
The internal flow of weak solution into the absorber is controlled by a regulation valve located
just before the inlet to the absorber. Under a given set of operating conditions, the circulation ratio
can be changed, either by changing the temperature of the heating medium to the generator or
evaporator, or by changing the cooling effect of the condenser or absorber.
262
K. ABRAHAMSSONet al.
,/" /
130~-
vapour from ~
u
Evaporator
II0~'~
St*maW
I0(~
.~. .
'
~
-
. . . . . . . .
.
Generator
~:K/~I~-~T
- . . . . .
80
Weak coin. frojm.~
-~
- - -', ~ - Generator- 70
/'
Weak soln. to.__ 6_~0~601
Heat exch.
50401
~_ ]
^ ,L~
~e~~Absorber
" "~"--~Glycol
I Absrber
~
Condensate
l f r o m
Condenser
301
20,:11:1:11111111111111111111tll
0 ,5 10 15 20 2`5 30 35 40 4`5 `50 5`5 60 6,5 70 7,5
&
Time (min.)
Sample 1
into
Sample 2
Design a n d e v a l u a t i o n o f a h e a t t r a n s f o r m e r
263
Sample I
Sample 2
58.0
54.1
11.8
57.8
54.1
12.4
It can be seen that the AHT is operating using vapour at approximately 98C in the generator
and the absorber is chilled with glycol at an incoming temperature of 115C and an outgoing
temperature of 120C. The temperature of the vapour to the evaporator was 105C. The condenser
pressure, about 10 kPa, was read off on a pressure-gauge at the condenser during the operation.
Since the heating and cooling areas were not optimised, the internal temperatures are the ones that
should be taken into consideration. It can thus be observed that the temperature of the weak solution
in the generator is 88C and the temperature of the outgoing vapour from the evaporator and the
outgoing strong solution from the absorber are 98 and 127C, respectively. Consequently, a
temperature lift of approximately 30C can be achieved. Since the pilot plant is a small unit, and
not completely insulated, the C O P t is difficult to calculate. The heat losses have been estimated to
about 50% of the heat input [6], which result in very low C O P t values of the total cycle; this is
clearly indicated in Fig. 2 by the temperature difference between the weak solution leaving the
generator and the weak solution entering the solution heat exchanger. The temperature difference
is about 11C. The decreasing temperature of the outgoing condensate (40-27C) corresponds to
a vapour saturation pressure of 7.5-3.5 kPa, but the total pressure was kept constant at about 10 kPa
by regulating the cooling water stream. The measured concentrations of strong and weak solutions
are shown in Table 3, at the different sample times marked in Fig. 2. The concentrations are analysed
by standard acid titrations.
2.5. Simulation results
How the pilot-plant AHT would operate without any heat or pressure losses can be revealed by
a simulation of the cycle. The input data used for this simulation are displayed in Table 4.
Using saturated steam at 105C and 98C to the evaporator and the generator, respectively,
together with the same temperature differences as measured in the experimental work, the simulation
results are close to the experimental results, as can be seen in Table 5. The U-value of the evaporator
was assumed to be 1.5 kW m -2 C-1 and the evaporator area was fixed to 2/3 of the total available area.
It may be observed that the absorber has a slightly higher temperature of the outgoing solution
than measured in the experimental work; this can be partly explained by the fact that the measured
heat loss of about llC of the weak solution from the generator has not been taken into
consideration in the simulation. Consequently, the temperatures of both solution streams leaving
the solution heat exchanger are also slightly higher. The pressure in the condenser was increased
in the simulation to 14 kPa to obtain the right concentrations similar to the measured values. The
lower pressure in the experimental work can be explained by the pressure loss between the generator
and the condenser.
3. R E F E R E N C E
ABSORPTION
HEAT TRANSFORMER
INCORPORATED
WITH AN
EVAPORATION PLANT
A simulation study [7] revealed that an AHT installed on-line with one of the evaporation plants
would reduce the amount of live steam consumption by 18.5%, corresponding to 3.09 MW. As a
Table 4, Input data for the absorption heat transformer pilot-plant simulation
Variable
U-value ( k W m -2''C t)
Area (m 2)
Heat/cool temp. ( C )
ATmi~ CC)
Int. pressure (kPa)
Circulation ratio (--)
*Steam.
tCooling water.
HRS 15/3--C
Evaporator
Generator
1.5
0.40
105"
7.0
---
.
0.84
98*
10.0
-12.0
Absorber
.
.
1.18
115t
15.0
---
Condenser
H. ex.
.
0.63
15t
5.0
14.0
--
0.77
-7.0
---
264
K. ABRAHAMSSON et al,
Table 5. Output data for the absorption heat transformer pilot-plant simulation
Variable
Effect (kW)
U - v a l u e ( k W m z C ~}
LMTD ( C )
Int. pressure (kPa)
Conc. (wt% H 2 0 )
Soln. temp. out ( C )
Soln. flow out (kg s t)
Evaporator
Generator
Absorber
4.2
1.5
7.0
94.3
100
98
0.0017
3.8
0.38
12.(}
14.0
54.1
88
0.018
3.8
0.21
15.3
94.3
58.0
131
0.020
Condenser
H. ex.
4.2
(}.41
16.2
14.0
100
53
0.0017
2.2
0.33
8.8
99, 124
direct response of this result, an A H T has been built and incorporated with one of the evaporation
plants at the STORA Papyrus Nym611a AB's pulp and paper mill in Nym611a. To avoid any
operational disturbances to the evaporation plant itself, the size of the A H T was limited to
approximately 200 kW of heat input, corresponding to approximately 100 kW of heat output.
Liquor streams
'1
Thick liquor
58 w t %
Pre-evaporated
thin liquor
20 w t %
265
IA
130"
120
11o
101)"
~
"
"
90"
80-
7o-~
6
A
-5
10
15
20
25
30
35
40
Time
(rain.)
heat exchanger
Fig. 5. Schematic view of the reference absorption heat transformer plant (scale 1:150).
266
K. ABRAHAMSSONet
al.
Fig. 7. The absorber component together with the feed water tank (scale 1: 50).
267
Fig. 8. The generator component together with the separator and the condenser component (scale I : 50).
268
K. ABRAHAMSSON et al.
Solution into and
from Ab~rber
140T
--7
3 , "0 ~ ~ T
- ........
- . . . . . . . . . . . .
-- ~ ~
Produced s t o a m - - ~
-,
. . . . . . . . . . . . . . .
._. 1 0 0 ~ ~
. . . .
~ ...................
120~
- -~ _ _ ~W
| eaJtsou
" l ti ona fte r
-/~
Heat exchanger
o~110 ~ / -
801 ~ / _
:f.
70i
60
50: ~
[~
402
301
20:
10
,-,
. . . . .
_WA-k ~ l Z o n _ a t ~ _ / ~ / ~ r _ ~ _ _ ~ _
~Vapour from
S t U n -~lution a~er.
Evaporator
Heat exchanger
~"
~? i i ~ i - i-_i-.-_-._-_~_-_-_ . . . .
_--_-_-." _ " - _ - ~
Condensate from
"-Condenser
J I I I I I I I I I I q I I I I I I I I I I I
0
20 40 60 80 100 120 140 160 180 200 220 240
Time (rain.)
Fig. 10. S o m e t e m p e r a t u r e profiles m e a s u r e d in the reference absorption heat transformer plant.
The solution heat exchanger component is a tube heat exchanger in a horizontal orientation. It
has three shells connected in series. The strong and weak solutions flow in a counter-current mode,
with the strong solution on the shell side. Figure 9 shows schematically the layout of the solution
heat exchanger.
3.3. Preliminary results
Temperatures are measured continuously in all external and internal streams of the AHT by
thermocouples of Chrom-Alumel. Pressures of incoming and outgoing external vapour streams are
also measured with electronic pressure transmitters. The pressures in the absorber and separator
are similarly measured. Liquid levels in the evaporator and absorber are measured using magnetic
level indicators. A PI-regulator keeps the liquid level in the feed water tank constant. Flow rates
are continuously measured for the strong solution to the generator and steam produced from the
absorber. A mass flowmeter is used to measure the flow rate of the strong solution and a differential
pressure meter is used for the steam flow. The external flow rates of the condensate and cooling
water streams are measured using the bucket-and-watch method. The only automatic control device
installed in the system is a thermostat for the control of the flow rate of cooling water to the
condenser, which is similar to the control loop used in the pilot-plant AHT. Due to accumulation
of non-condensable gases in the condenser, a venting procedure is carried out automatically every
hour for a time period of 1 min maximum.
Currently, the AHT plant is still in the first test period. Consequently, the operating data
available is not sufficient to accurately determine the efficiency of the AHT. The first tests were
performed with the AHT operated in the forced circulation mode of the working solution around
the absorber. A set of measured temperature profiles and other measured data are shown in Fig. 10
and Table 6, respectively.
It can be seen, from the steady-state part of the temperature profiles, that the AHT used steam
at 100C in both the generator and evaporator and that the absorber produced steam at 123C.
It should be pointed out that the feed water used for the absorber, not shown in the figure, had
a temperature of 55C. The temperature of the vapour leaving the evaporator was 97C, which
results in a temperature difference of 3C in the evaporator. Due to the forced circulation of liquid
Table 6. Measured data corresponding to the steady-state period in
Fig,
I0
Parameter
Pressure in absorber (kPa)
Pressure in separator (kPa)
Cone. of strong solution (wt% H.,O)
Cone. of weak solution (wt% H,O)
Flow rate of strong solution (kgs i)
Flow rate of produced vapour (kg s ~)
Value
82.5
7.6
55.0
51.6
0.53
0.03
269
around the absorber, the incoming and outgoing solutions had nearly the same temperature at
125C, this gives a temperature difference of 2C in the absorber. The strong solution was cooled
down to 95C in the solution heat exchanger before entering the generator. The weak solution
leaving the generator had a temperature of 79C, which results in a temperature difference of 21C
in the generator.
By executing WPData, using a pressure of 7.6 kPa in the separator, as seen in Table 6, and the
measured temperature of the weak solution, the theoretical saturation concentration is 51.5 wt%
H20. This value agrees very well with the measured concentration of 51.6 wt% H20. A similar
calculation can be done for the strong solution leaving the absorber. Together with the pressure
in the absorber and the temperature of the outgoing strong solution, the theoretical concentration
is 58.9 wt% H:O. This value does not agree so well with the measured concentration of 55.0 wt%
H20. The difference is attributed to the presence of non-condensable gases in the absorber. If the
measured concentration is correct, it would correspond to a saturation pressure of 67.2 kPa in the
absorber. Based on the measured concentrations of strong and weak solutions, the circulation ratio
is calculated to f = 14.2.
If the pressure in the condenser is assumed to be equal to the saturation pressure corresponding
to the condensate leaving the condenser, the pressure in the condenser would be 6.8 kPa.
Accordingly, the pressure drop between the generator and the condenser is less than 1 kPa.
The measured flow rate of steam from the absorber is 0.03 kg s-' which corresponds to 75 kW.
Unfortunately, it seems rather difficult to measure the flow rates of the condensate from the
evaporator and the generator.
As shown in Fig. 4, the boiling temperatures are increased in the evaporation plant as a result
of the liquor shift procedure. The response of the AHT unit during this liquor shift is exemplified
in Fig. i 1, which depicts the temperature of the strong solution from the absorber, together with
the temperature of the steam from effects 3 and 1A/lB.
It can be seen that it takes about 60 min for the strong solution in the absorber to reach a
temperature in the absorber that is high enough to produce steam. The average solution
concentrations were the same before and after this temperature change in the AHT. The speed of
response differs between different liquor shifts. As indicated in Fig. 10, the response at this time
is very fast, where the strong solution is all the time warmer than the vapour from effect 1A/lB.
Unfortunately, the way of executing the liquor shifting procedure seems to depend on the shift
personnel responsible for plant operation.
An economic evaluation was performed to determine the economic feasibility of incorporating
AHT systems in this type of evaporation plants. Based on the information obtained for the
Hoogovens plant [9] and existing cost functions [10, 11], which are presented in the Appendix, the
installed equipment cost was assumed to be 410 $ kW -~ of delivered heat. The pay-off period of
this type of investment mainly depends on the local steam cost, AHT efficiency and also on how
much of the existing equipment in the evaporation plant can be utilised as components in the AHT.
140
.-~ 1301
~Steam
~110 ........
. - : .
,,_''"
, ooZE--- -'~'
40
-20
fromeffect1A/1B
.
"CAbsorber solution
~,
20
40
60 80
Time (min.)
100
120 140
Fig. 11. Temperature response of the strong solution leaving the absorber during liquor shift,
270
K. ABRAHAMSSONet al.
In this work, the third factor was not considered in the economic evaluation and a COP of
approximately 45% was assumed. The annual plant operation time and steam cost were taken as
8000 h and 0.024 $ kWh-~, respectively. The calculated pay-off period for the investigated A H T
configuration would be 4.4 years. It should be noted, however, that this steam cost, corresponding
to 14 $ ton-~, is considered to be relatively low, compared to a comparable Scandinavian pulp and
paper mill having the same capacity. For a steam cost of 20 $ ton-t, the pay-off period for the
considered configuration would be reduced to 3.1 years.
3.4. F u t u r e w o r k
No electronic control systems are currently installed in the A H T , since the reference plant was
deliberately designed initially for manual operation. Together with another research group working
in our department with Knowledge-based Risk Management ( K R M ) [8], a database will be
developed containing the operational know-how of the A H T on-line with the evaporation plant.
This database will be valuable to process engineers not familiar with this type of energy
conservation systems. Based on careful analysis of the operational behaviour during the test period
of both the evaporation plant and the A H T unit, some changes in the A H T might be suggested.
For instance, an appropriate control system will be designed and installed at a later stage of the
project.
REFERENCES
1. K. W. Gullers and S. Rydberg, Meeting with Munters (in Swedish), GuUers Pictorial AB, Stockolm (1983).
2. K. Eriksson and /~,. Jernqvist, Heat transformers with self-circulation: design and preliminary operational data,
Int. J. Refrig. 12 (1), 15-20 (1989).
3. W. Isacsson, An investigation of heat exchangers for use in a heat transformer, Report, LUTKDH/(TKKA-7004).
Dept. of Chem. Eng. 1, University of Lund (1991).
4. C. Engman, Adiabatic one- and two-phase flow (in Swedish), Licenciate thesis, LUTKDH/(TKKA-1002). Dept. of
Chem. Eng. 1, University of Lund (1983).
5. H. Kockum, Correlations for two-phase flow and heat transfer--A literature study (in Swedish), Report
(in preparation). Dept. of Chem. Eng. I, University of Lund (1993).
6. A. Gidner, Evaluation of a heat transformer with self-circulationand a fallingfilm absorber (in Swedish), M.Sc. Thesis,
LUTKDH/(TKKA-5032). Dept. of Chem. Eng. 1, University of Lund (1991).
7. K. Abrahamsson, G, Aly and A. Jernqvist, Heat transformer systemsfor evaporation applications in the pulp and paper
industry, Nordic Pulp & Paper Research J. 6, 9-16 (1992).
8. R. Wennersten and A. Gr~infors, KRM--An integrated system for process analysis and operator support,
Plant~Operation Progress (submitted).
9. Rinheat Oy, Finland, private communication (1990).
10. RCG/Hagler-Bailly,Inc., Opportunities for industrial chemical heat pumps in process industries. Final Report prepared
for the US DOE Office of Industrial Technologies, August 2 (1990).
11. J. Berghmans, High temperature industrial heat pumps. Final report prepared for the lEA Advanced Heat Pumps
Annex IX (1990).
12. M. S. Peters and K. D. Timmerhaus, Plant Design and Economics for Chemical Engineers, McGraw-Hill,Chemical and
Petroleum Eng. Series, 4th Edn. McGraw-Hill, New York (1991).
13. R. Krom, and P. Oostendorp, Heat transformer finds success in the steel industry, IEA Heat Pump Centre Newsletter
I1 (1), 17-19 (1993).
A P P E N D I X : C O S T F U N C T I O N S F O R AHTs AND AHPs
A capital investment is required for any industrial process, and determination of the necessary
investment is an important part of a plant-design project. Some economic data have recently been
published [10, 11] concerning absorption heat cycles. It is, however, rather difficult to evaluate these
data owing to the lack of a sufficient number of commercial full-scale plants.
The Lang multiplication factor is used quite frequently to obtain order-of-magnitude cost
estimates which usually result in a probable accuracy of estimate of about + 30% [12]. This factor
recognises that the cost of a process plant may be obtained by multiplying the basic equipment
cost by some factor, to approximate the capital investment. These factors vary depending upon
the type of process being considered.
Consisting of conventional heat transfer equipment, it is rather easy to estimate the basic
equipment cost of an absorption heat cycle, taking into consideration the additional cost necessary
to account for appropriate materials of construction. Conversely, it is more difficult to estimate
271
the Lang multiplication factors for this type of plant. The little cost information reported in the
literature for single-stage AHTs and AHPs is briefly described in this Appendix
Heat transformers
(A.l)
kt Q -0.3 [$ kW output-I].
(A.2)
(A.3)
where: C = system initial cost ($); H C = system initial heat cost ($ kW output J); I H C = system
investment heat cost ($ kW output-~); Q = thermal output (kW); k~ = correlation constant = 3750
In another report released the same year by the U.S. Department of Energy, DOE [10], the
industrial heat pump market opportunities in the U.S. process industries were analysed. The capital
costs for different heat pumps and construction materials were compiled for different temperature
lifts. The capital cost was expressed as heat cost in $ kW-t output and a number of curves were
given for different materials of construction. The plotted cost information for AHTs in high-grade
steel was curve-fitted for an approximate temperature lift of 30C. The resulting function may be
written as:
IHCAHx = k2 Q 03 [$ kW output- 1],
(A.4)
There was no cost information concerning AHPs in the above mentioned report by Berghmans
[11]. The construction of an A H T and an AHP is in principle identical, however an AHP delivers
twice the amount of useful energy compared to an AHT, thus the equipment cost of an AHP can
800\
...... I E A (equip.)
"" - - I E A (capita])
D O E (capital) i
" Hoogoven AHT,
700"
~,~
600-
5oo-
4o0.
.
'--..O.oo
~ 300200.
100
Delivered
MW
10
272
K. ABRAHAMSSONet al.
soo I
'1.,50"
~d
O, 3501
3001
250 ~
o
900
o~
1501
100
50
'
10
15
20
25
I
30
Delivered heat, MW
Fig. A2. Absorption heat pump cost as function of delivered heat.
be assumed to be half the equipment cost of an AHT; this leads to the following capital heat cost
for an AHP:
IHCAHp = k~ Q -o3 [$ kW output -~ ]
(A.5)
In the U.S. D O E report [10], the capital heat cost for AHPs was found to fit the following
correlation:
IHCAHv = k 3 Q
(A.6)