SYNOPSIS

The use of ammonia as a refrigerant is widespread in vapor absorption refrigeration system.

Ammonia is generally not used in low capacity applications due to lack of economically
available components. For this reason, the objective of this study is to design an economical shell
and tube condenser for the use in ammonia absorption refrigeration system. The work presents
designing and fabrication of a shell and tube condenser.

1. INTRODUCTION TO CONDENSERS IN REFRIGERATION SYSTEM

Vapor absorption refrigeration system based on ammonia-water is one of the oldest refrigeration
systems. In this system ammonia is used as refrigerant and water is used as absorbent. The
system constitutes of a number of heat exchangers employed for different purposes. Hence the
design of these heat exchangers plays an important role in determining the COP of the system.

Condensers are basically heat exchangers in which the refrigerant undergoes a phase change.
Hence proper design and selection of condensers is very important for satisfactory performance
of any refrigeration system. In condensers the refrigerant vapor condenses by rejecting heat to an
external fluid, which acts as a heat sink. The refrigerant enters the condenser in a superheated
state. It is first de-superheated and then condensed by rejecting heat to an external medium. The
refrigerant may leave the condenser as a saturated or a sub-cooled liquid, depending upon the
temperature of the external medium and design of the condenser. Figure1.1 shows the variation
of refrigeration cycle on T-s diagram. In the figure, the heat rejection process is represented by 2-
3’-3-4. The temperature profile of the external fluid, which is assumed to undergo only sensible
heat transfer, is shown by dashed line. It can be seen that process 2-3’ is a de-superheating
process, during which the refrigerant is cooled sensibly from a temperature T 2 to the saturation

temperature corresponding condensing pressure, T3’. Process 3’-3 is the condensation process,

during which the temperature of the refrigerant remains constant as it undergoes a phase change
process.
 Fig.1.1 Refrigeration cycle on T-S diagram

Based on the external fluid, these condensers can be air cooled , water cooled or evaporative
type. The air cooled condenser costs 2 to 3 times more than water cooled condenser, whereas
evaporative condensers are employed generally for large capacity, hence water cooled condenser
was selected.

1.1 Selection of shell and tube type heat exchanger

The advantage of shell and tube condenser is that a large condensing surface area can be installed
in a small space, compared to other heat exchangers. The shell and tube exchanger can be
reasonably easily cleaned. Finally, good design methods exist, and the expertise and shop
facilities for the successful design and construction of shell and tube exchangers are available
throughout the world. The shell and tube condenser is basically arranged in a counter flow
arrangement. They are usually constructed so that refrigerant condenses on the shell side or the
outside of the tubes while cooling water is pumped through the inside of the tubes. They are
normally mounted in the horizontal position to facilitate condensate drainage.

2. LITERATURE SURVEY

The major areas of focus of literature survey were related to the constructional and designing
features of the shell and tube condenser and its modeling by computational fluid dynamics
(CFD). Journals, books and manufacturing standard details of various parts on the website were
referred for this purpose. It was through literature survey that the kern method for theoretical
designing was adopted. The previous works done on the computational fluid dynamics analysis
on shell and tube exchanger revealed the difficulties and complexities involved in the 3D
analysis of the shell side flow hence focus on 2D analysis was done.

3. DESIGNING AND MODELING OF SHELL AND TUBE CONDENSER

3.1 Sizing procedure

The basic sizing procedure for the designing of the heat exchanger followed is shown in fig. 3.1.
In the sizing or designing procedure, the Kern method is used for calculating the shell side
parameters and heat transfer coefficient on shell side where condensation takes place. Initially
the sizing begins with a guess value of overall heat transfer coefficient and has to be performed
number of times so, a program in C++ language was coded to make the job easier and faster.
After obtaining overall heat transfer coefficient value greater and closer enough to initial
assumed value, the pressure drops on both shell and tube side were calculated and found within
limits. The design parameters involved are shown below in fig .3.2

Fig. 3.2 Design parameters taken as input in C++ program

3.2 Modeling Details
The 2-D computational fluid dynamic analysis has been done in order to validate theoretical
design . The working fluid of the shell side is ammonia and the fluid flowing through tubes is
water. Ammonia undergoes condensation i.e. its phase change occurs at a temperature of 35 °C
and under pressure of 14 Bar. The vapor ammonia which enters at the top is condensing on the
tube surface through which cold water flows at inlet temperature of 28 °C. It is observed that thin
film of liquid ammonia is formed on the tube surface. A 2-D analysis is performed separately for
the shell side flow and the tube side flow. GAMBIT 2.3.16 is used for creating geometry and
FLUENT 6.3.26 for the analysis.

a) Shell side
Governing equations: Time dependent momentum equation, energy equation and the continuity
equation has been solved.

Boundary Conditions: The desired velocity and temperature was assigned to the inlet. The
temperature of ammonia vapor at the inlet was taken as 360K and the inlet velocity was taken to
be 0.85 m/s. Zero gauge pressure is assigned to the outlet nozzle in order to obtain the relative
pressure drop between inlet and the outlet. The zero heat flux boundary condition is assigned to
the outer wall, assuming shell is perfectly insulated outside.The constant wall temperature of
302K is assigned to the tube walls.

Mesh selection: Mesh generation is performed using Gambit as shown in fig. 3.3

Turbulence model : Since the flow in this study is turbulent, turbulence effects should be taken
into account using turbulence modeling. The type of k–ε turbulence model tried was standard.

Other modeling choice : Multiphase model was selected for mixtures. Ammonia vapor phase was
assigned as the primary phase whereas ammonia liquid was assigned as secondary phase. In the
first order discretization, the standard scheme is selected for pressure, and the first order upwind
scheme is selected for momentum, turbulent kinetic energy and dissipation rate.
 Fig. 3.3 View of geometry in GAMBIT

b) Tube side
Compared to the shell side analysis it is relatively simpler.

Governing equations: In this model we solve equations like Continuity equation, Momentum
equation, Energy equation.
Boundary Conditions: The inlet velocity was given as 0.205 m/s after calculating according to
the inlet diameter. A thin film of ammonia is formed on the tube wall hence the tube wall
temperature has been taken as 305K. Zero gauge pressure is assigned to the outlet in order to
obtain the relative pressure drop between inlet and the outlet.

Mesh selection: Mesh generation is performed using Gambit. The surfaces of the model are
meshed using rectangular elements. The tube was taken to be of 0.9 m long and having diameter
of 16 mm. In the gambit the top and bottom edge were taken to be walls whereas the inlet was
taken as velocity inlet and outlet was taken as pressure outlet. And in specifying continuum types
water was taken as fluid.

Turbulence model: This is same as the shell side flow.

Fig. 3.1Flowchart for sizing procedure
4. RESULTS AND DISCUSSIONS
4.1 Computational Results

The computational analysis was performed in Gambit and Fluent software to predict the shell
and tube side temperature contour and the heat transfer coefficients on the both side. The
analysis was done on a 2D model. The contours for heat transfer coefficient for shell and tube
side are shown below in fig. 4.1 and 4.2

Fig. 4.1 Contour plot of surface heat transfer coefficient shell side

From the fig.4.4 it can be observed that the scale on the left side shows variation from
0 to – 15000 W/m2k. The heat transfer coefficient depicted in the contour shows negative values.
The reason for this is the fluent evaluates heat transfer coefficient by considering the temperature
difference ∆T as ( T wall −T reference ). In the case solved here, T reference >T wall since the through the
tube walls the cold water flows, whereas the reference here refers to the higher temperature of
ammonia to be condensed which is on shell side. Hence, the maximum value of heat transfer
coefficient on the shell side achieved is 15000W/m2k. The condensing coefficient of ammonia is
high, this results in higher value of heat transfer coefficient compare to other refrigerants.
 Fig. 4.2 Magnified heat transfer coefficient contour along tube side

The contour showed that there was highest heat transfer coefficient of 1630W/m 2k at the wall
and it got reduced to lower values like 400 W/m 2k.As there was a flow of water inside the tube
so no heat transfer coefficient was visible in that region. The heat transfer coefficient was
gradually increasing as we approached near to the wall. Hence, the maximum heat transfer
coefficient obtained from the analysis is found out to be 1630W/m 2k. Temperature and pressure
contours were also obtained

4.2 Theoretical Results (Kern Method)

The designing procedure of the shell and tube condenser for our purpose was coded in the C++
programming language. Developer (Dev C++) software was used for this purpose. For guess
value of overall heat transfer coefficient of 700 W/m 2k, finally the outcome of 729 W/m2k was
achieved. This value was greater and closer to the assumed value hence basic designing
procedure was completed. At this design parameters pressure drops for both shell and tube side
were calculated and found to be well within the limits. Hence, the operating conditions were
decided to be safe. Figure 4.3 shows the results obtained from the programming software.
 Fig. 4.3 Program output screen

4.3 Comparison of the Computational and Theoretical Results

The values of heat transfer coefficients obtained from CFD analysis were found to be close to the
theoretical results. The comparison was made separately for shell and tube heat transfer
coefficients between CFD and theoretical results and is shown in the table.
 Table 4.1 Comparison of CFD and theoretical results

Heat transfer coefficients (W/m 2

Flow region k) % difference
CFD
Theoretical results
results
1630 1385.56
Tube side 15

Shell side 15000 17241.2 13

For the tube side flow the result obtained by computational method for heat transfer coefficients
depicts a higher value compared to analytical result.CFD result is about 15 % higher. For the
shell side flow analytical result gives higher heat transfer coefficient and it is greater by 13%
compared to the CFD results. The reason for this higher value can be inaccuracy in predicting the
temperature of the tube walls which is taken to be constant for modeling of both the sides.
However, this difference has to be considered while moving onto the fabrication of the
condenser.
From CFD results higher overall heat transfer coefficient of 790 W/m 2K compared to 726 W/m 2
K of theoretical results was obtained.

4.4 Fabrication of Shell and Tube Condenser

The last step of fabricating the shell and tube condenser was implemented taking lower value of
overall heat transfer coefficient into account. Based on the designing calculations, less
complexity involved and the economics a simple fixed tube sheet type of shell and tube
exchanger was chosen for fabrication. Since ammonia reacts with copper and is compatible with
steel, it was chosen as the material for the construction. The tubes are mainly responsible for the
heat exchange between shell side and tube side fluids. So, low carbon steel tubes ASTM A 179
especially for heat exchangers were brought. Galvanized steel was used for shell while baffles
and other components were made of mild steel. The pictures below show some of the parts
during fabrication.
Fig. 4.4 (a) 4.4 (b)

Fig. 4.4 (c)

 Fig. 4.4 (d)

The fig. 4.4 (a) shows the tube sheet with equilateral triangular arrangement of holes, fig. 4.4 (b)
shows the baffle cut given on the right and also trimmed base of the baffles for condensate flow
can be observed on the left. Fig. 4.4( c) shows the tube bundle assembly with the baffles in
position and fig. 4.4 (d) shows tube bundle assembly inserted into shell.

4.5 Economic Analysis

Compared to the market the self fabrication of the condenser came out to be very cheap. The
table below lists the expenditure of the major components and their processing.

S.NO ITEMS AMOUNT (in Rs.)

1 ASTM A 179 Tubes 1500
2 Galvanized steel (Shell) 750
3 Baffles 225
4 Mechanical processing 1500
5 Gas welding 500
6 Conveyance charges 400
TOTAL 4875

The cost of the tubes is high since low carbon steel tube especially for heat exchangers is used.
The mechanical processing includes the labor costs which can be brought down by using the
laboratories but in order to obtain accuracy work was carried outside. Also the availability of gas
welding in college would have reduced the cost. Still, the total cost including the conveyance
charges comes out to be very economical compared to the quotes received from the industry for
the same condenser capacity design.

Thus, the cost of manufacturing vapor absorption refrigeration can be brought down by the self
fabrication of the individual components.

5. CONCLUSION

The design, modeling and fabrication of shell and tube type condenser have been successfully
completed. The first step taken was theoretical designing of the exchanger based on Kern’s
procedure. To make the job easy the design was coded in Developers (Dev C++) programming
software to generate sizing results. 2D computational fluid dynamics analysis is done by
modeling the geometry in GAMBIT 2.3.16 and analyzing the results in FLUENT 6.3.26. The
comparison of the heat transfer coefficients of both shell and tube side obtained from CFD and
theoretical results was made. The value of overall heat transfer coefficients for theoretical results
was found to be little less than the CFD results. The final fabrication was implemented by taking
lower value of overall heat transfer coefficients to remain on the safer side. The economics
involved in the self fabrication was found out to be very economical compared to the market
price for the same sizing. Hence, the cost of the vapor absorption refrigeration system which is
generally high can be brought down by self fabrication of as much components as possible.

Finally, we would consider this project a success if the ideas described in the report can become
a useful reference for future work on the subject. Also building a 3D simulation on shell and tube
condenser and the testing of the component which could not be realized in this short time are the
major areas left for the future work on the subject.