Design Ref 2

University of Cantabria
Ghent University
Faculty of Engineering
Department
Flow, heat and combustion mechanics
Head: Prof. Dr. Ir. Erik Dick
Refrigerating Plant Design

By
Wouter De Muynck & Davy Poelman
Tutor: Prof. Dr. Ir. Michel De Paepe

Thesis supervisors: Valentin Castaera & Delfin Sili
Thesis is presented to obtain the Diploma of the Second Cycle of Electromechanical

Engineering.
Study year 2005-2006

Ghent University
Department
Flow, heat and combustion mechanics

By
Wouter De Muynck & Davy Poelman


Engineering.

Preface
In the past a lot of scientist, engineers and other men have devoted their interest and energy to
freezing problems, heat transfer and pressure drops in ducting systems. We have challenged
faith to find our way in a huge amount of information to select the proper information needed
to design our own refrigeration plant. We do not want to overwhelm you with facts and
formulas so we made a brief selection of the correlations necessary to understand and to
follow the leitmotiv of our work.
We, Davy Poelman and Wouter De Muynck both Erasmus students at the University of
Santander are proud to present a thesis, which hopefully extends your interest in this subject
or helps you finding solutions for problems you will encounter.
Expression of gratitude
In the first place we would like to thank our parents for their endless help and support. Great
appreciation and gratitude for their perpetual support and patience goes out to our local tutors
of the University of Cantabria, Delfin Sili and Valentin Castaera. We also want to thank
our Belgian tutor Prof. Dr. Michel De Paepe and colleague assistant Christophe TJoen who
also did great effort to make this work possible. Furthermore we owe much gratitude and
beers to Carlos Torrontegui, maintenance worker of Barandica, for the time and practical
information he offered us. We extend our thanks to Filip Pauwels and Willem De Jongh for
their important contribution to this work and to our Spanish friend Nuria who helped us a lot
despite of the language barrier at the beginning. We may not forget to thank Willem De
Muynck and Jim Van Durme for their experience and knowledge (even though they are
Biological engineers). And last but not least we thank our friends for their understanding
during the past busy period.
The author and the promoters give the authorization to consult and to copy parts of this work
for personal use only. Every other use is subject to the copyright laws, more specifically the
source must be extensively specified when using results from this thesis.
Santander, june 2006
Tutor: Auteurs:
Prof. Dr. Ir. Michel De Paepe De Muynck Wouter
Poelman Davy
Refrigerating plant design
by
Wouter De Muynck and Davy Poelman

Engineering.

Ghent University
Department of Flow, heat and combustion mechanics

Summary
We design a refrigerating plant to freeze fish. Therefore we use two installations: a blast
freezer and a brine tunnel. The blast freezer freezes tuna and mackerel and the brine tunnel
freezes sardine. We give the construction of both freezers. We calculate the refrigerating load
and search the most appropriate thermodynamic cycle that works with ammonia. Afterwards
we select all its machinery and devices. Freezing times are also calculated. So we give all
information necessary for the good functioning of the refrigeration installations.
Our work is grouped into four parts. The first part explains the thermodynamic cycle
selection. The second is devoted to the blast freezer while the third part contains the brine
tunnel. In the last part we enumerate the machinery and devices and calculate pressure drops
in the ducting and freezing times.
Key words: thermodynamic cycles, heat transfer, two-phase flow, refrigeration

Davy Poelman & Wouter De Muynck
Supervisor(s): Delfin Sili & Valentin Castaera
Abstract: A refrigerating plant existing of a blast load for the blast freezer is found to be 292 kW and
freezing chamber and a brine tunnel need to be designed. 660 kW for the brine tunnel.
The blast freezing room with postulated dimensions has to
be able to freeze a maximum charge of 60000 kg of Once the refrigerating load is known the evaporators
bonito tuna or mackerel in 24 hours. The brine tunnel
also with postulated dimensions will freeze 7000 kg of
can be determined. In the blast freezer, as well in the
sardines in one hour. The time in which one fish passes brine tunnel overfeed evaporators are selected. The
through the brine tunnel amounts to 26 minutes. As evaporator circulation number is 4 and means that
refrigerant we use ammonia. The fish enters both only one quarter of the liquid fed into the evaporator
installations at about 8-11C. The bonito tuna and is evaporated. The blast freezer evaporators are
mackerel leave the blast freezing room at -18C. The standardized products. Nevertheless standard
Sardines leave the brine tunnel at -12C ventilators are removed and more powerful fans are
mounted to overcome high pressure drop caused be
Keywords: Thermodynamic cycles, heat transfer, the fish arrangement in the freezing space.
two-phase flow, refrigeration Automated gas defrosting of the evaporators has to
prevent a thick layer of frost and thus deteriorating
An elaborate study about the thermodynamic cycles heat exchange in the evaporators. Brine tunnel
is made to select the most appropriate cycle for the evaporator dimensions are calculated. The interior
refrigerating plant. A two-stage refrigerating cycle convection coefficient calculation is based on two-
with flooded evaporators and open intercooler is phase flow correlations of Zrcher et al. The exterior
found to be the most efficient solution. Both convection coefficient is determined with formulas
installations, blast freezing chamber and brine tunnel, for double pipe heat exchangers with several internal
will not run simultaneously but use the common pipes.
compressors, open intercooler, low-pressure receiver,
high-pressure receiver and condenser. Evaporator With the refrigerating load and working temperatures
temperatures in the blast freezer and the brine tunnel the compressors can be dimensioned. The best
are 45C respectively 25C. compressor choice for present refrigerating plant is
screw compressors because of high volume flow
The required freezing space in the blast freezer is rates and varying compression ratios. Appropriate
calculated to make sure that an average air velocity compressors are selected by York Refrigeration.
of about 3.5 m/s is reached. This value will lead to a
freezing time of 24 hours and reduced energy Two air-cooled condensers are selected. They are
consumption. Agitators in the brine tunnel will also standardized and run at 35C condensing
circulate the brine at a velocity of 0.15 m/s, which is temperature. Both condensers will operate together
sufficient to freeze Sardine in 26 minutes. Using for blast-freezer functioning, only one will work for
modifications to the Plancks equations it is possible brine tunnel functioning.
to calculate the freezing time for the fish in the blast
freezer. Another method is applied for the fish in the Pressure drops in the brine tunnel evaporators are
brine tunnel. calculated to dimension the pumps that will circulate
the cold refrigerant fluid through the evaporators.
Refrigerating loads are calculated. For the blast Pressure drops are obtained with the Friedel
freezer the product load, transmission load, air load, correlation and the flow pattern map method of Taitel
internal load and infiltration load are considered. For and Dukler, which are valid for two-phase flow. A
the brine tunnel product load, transmission load and pressure drop of 15.7 kPa is found in the brine tunnel
internal load are taken into account. The insulation evaporators.
of blast freezing room and brine tunnel is calculated
on recommended R-values by ASHRAE. Measures Auxiliary circuit instrument such as the low-pressure
are taken to prevent frost heave of the blast freezer receiver or separator, intercooler, high-pressure
floor and the brine tunnel foundation. Refrigerating receiver, purge unit, filters and oil separators are
selected based on refrigerating capacity needs. All
components are standardized.
Table of Contents
Part 1 1
Chapter 1: Introduction
1 Justification ........................................................................................................................ 2
2 Objective ............................................................................................................................ 2
3 General description of the brine tank ................................................................................. 2
4 General description of the blast freezer ............................................................................. 4
Chapter 2: Thermodynamic cycle selection

1 Introduction: Multistage Vapour Compression.................................................................. 6
2 The Plant ............................................................................................................................ 7
3 Comparison ........................................................................................................................ 8
4 Conclusion ....................................................................................................................... 32
5 Two-stage cascade system ............................................................................................... 33
6 Final Conclusion .............................................................................................................. 39
Part 2 :The blast freezing chamber........................................................................................... 40
Chapter 3: The floor construction

1 Introduction ...................................................................................................................... 41
2 Floor construction ............................................................................................................ 42
3 Heat flux calculation ........................................................................................................ 47
Chapter 4: The floor ventilation

1 Introduction ...................................................................................................................... 51
2 Calculations...................................................................................................................... 51
3 Ventilator selection .......................................................................................................... 53
Chapter 5:Refrigerating space calculation

1 Introduction ...................................................................................................................... 56
2 Calculations...................................................................................................................... 56
Chapter 6:Air circulation in the blast freezer

1 Introduction ...................................................................................................................... 59
2 Air flow rate calculation................................................................................................... 60
3 Ventilator selection .......................................................................................................... 61
Chapter 7: Refrigerating load

1 Product load ..................................................................................................................... 64
2 Transmission load ............................................................................................................ 65
3 Air load............................................................................................................................. 75
4 Internal load ..................................................................................................................... 77
5 Infiltration load................................................................................................................. 78
6 Equipment related load .................................................................................................... 78
7 Total refrigerating load .................................................................................................... 78
Chapter 8:Evaporators
1 Introduction ...................................................................................................................... 79
2 Selection........................................................................................................................... 82
Chapter 9:Evaporator defrosting

1 Introduction ...................................................................................................................... 84
2 Defrost system selection .................................................................................................. 85
Part 3:The brine tunnel............................................................................................................. 89
Chapter 10: The brine tank

1 Introduction ...................................................................................................................... 90
2 What do we have to specify? ........................................................................................... 90
3 Type of steel..................................................................................................................... 90
4 Thickness of the steel plates............................................................................................. 91
5 Construction ..................................................................................................................... 91
Chapter 11:Brine selection

1 Introduction ...................................................................................................................... 92
2 Criteria.............................................................................................................................. 92
3 Possibilities ...................................................................................................................... 92
4 Solution ............................................................................................................................ 92
Chapter 12:The agitators

1 Introduction ...................................................................................................................... 94
2 Selection of the agitator type............................................................................................ 94
Chapter 13: The conveyer belts

1 Introduction ...................................................................................................................... 96
2 Selection........................................................................................................................... 96
Chapter 14: The refrigeration load and cooling time of the brine
1 Introduction ...................................................................................................................... 99
2 Refrigeration load............................................................................................................. 99
3 Total load ....................................................................................................................... 104
4 Safety factor ................................................................................................................... 104
5 Refrigeration load calculation ...................................................................................... 1055
6 Time necessary to cool the brine tank: 25 -20C......109
7 Time necessary to cool the brine tank: ? -20C.....110
Chapter 15: The evaporator

1 Introduction .................................................................................................................. 1111
2 Description ................................................................................................................... 1111
3 What do we have to specify? ....................................................................................... 1122
4 Intensive study of the pressure drop and heat flux in the evaporator........................... 1177
Chapter 16: The brine basin

1 Introduction .................................................................................................................. 1222
2 Selection......................................................................................................................... 122
3 Solution .......................................................................................................................... 122
Part 4 ...................................................................................................................................... 123
Chapter 17: Screw compressors

1 Introduction .................................................................................................................. 1244
2 Compressor Selection................................................................................................... 1277
3 Oil management ........................................................................................................... 1288
Chapter 18: Condensers

1 Introduction ................................................................................................................ 13030
2 Selection....................................................................................................................... 1322
Chapter 19: Piping

1 Introduction .................................................................................................................. 1344
2 Recommended Material ............................................................................................... 1344
3 Pipe Location ............................................................................................................... 1344
4 Pipe Sizing ................................................................................................................... 1355
Chapter 20: Auxiliary circuit instruments

1 Introduction .................................................................................................................. 1399
2 Oil separator ................................................................................................................. 1399
3 High-pressure receiver ............................................................................................... 14040
4 Low-pressure receiver.................................................................................................. 1411
5 Purge Unit .................................................................................................................... 1433
6 Filters ........................................................................................................................... 1433
7 Intercooler .................................................................................................................... 1455
8 Refrigerant liquid pumps.............................................................................................. 1466
9 Check valves ................................................................................................................ 1477
Chapter 21: Refrigeration circuit controls

1 Introduction .................................................................................................................. 1544
2 Pressure switches ......................................................................................................... 1544
3 Temperature switches................................................................................................... 1577
4 Liquid level indicators.................................................................................................. 1588
5 Regulating and throttling valves .................................................................................. 1599
6 Testing safety devices .................................................................................................. 1622
7 Automation................................................................................................................... 1633
8 Refrigerant circuit ........................................................................................................ 1644
Chapter 22: Pressure drops

1 Introduction .................................................................................................................. 1655
2 Single-phase flow......................................................................................................... 1655
3 Two-phase flow............................................................................................................ 1666
4 Flow pattern map transition data.................................................................................. 1755
5 Flow pattern map transition calculation....................................................................... 1766
6 Pressure loss calculations............................................................................................. 1766
Chapter 23: Thermal properties of tuna and mackerel

1 Introduction .................................................................................................................. 1799
2 Thermal properties of food constituents ...................................................................... 1799
3 Thermal properties of food......................................................................................... 18080
4 Calculations of the thermal properties of tuna and mackerel and their food components
...................................................................................................................................... 1844
Chapter 24: heat transfer between air and tuna in the blast freezer
1 Introduction .................................................................................................................. 1877
2 Heat convection coefficient.......................................................................................... 1877
3 Heat convection coefficient calculation. ...................................................................... 1899
Chapter 25: Freezing time for tuna and mackerel

1 Introduction .................................................................................................................. 1922
2 Calculation methods..................................................................................................... 1923
3 Freezing time calculations............................................................................................ 1977
Chapter 26: The freezing time of Sardine

1 Introduction .................................................................................................................. 1999
2 Freezing time calculation ............................................................................................. 1999
3 Freezing time of sardine............................................................................................... 2022
Chapter 27: Ammonia (R-717)

1 Gas Properties .............................................................................................................. 2033
2 Material compatibility.................................................................................................. 2044
3 Health effects ............................................................................................................... 2055
4 First Aid ....................................................................................................................... 2065
5 Summary ...................................................................................................................... 2076
Chapter 28: Safety considerations
Chapter 29: Storage of frozen fish

1 Introduction ................................................................................................................ 21010
2 Composition ............................................................................................................... 21010
3 Temperature ............................................................................................................... 21010
4 Humidity ...................................................................................................................... 2111
5 Packaging and glazing.................................................................................................. 2111
6 Maximum Storage life.................................................................................................. 2111
Chapter 30: Average weather condition in Santander
Reference list.. 213

Part 1
The consumption and diversity of frozen products has increased considerable in developed
countries due to the advantages compared to fresh products. The longer conservation time is
the best example. Due to these benefits, there is an increasing interest for freezing
installations and equipment. Knowledge is very important to reply the request. That is why all
existing possibilities have to be studied.
1 Justification
The client cleans, freezes, packages and conserves. They want to extend the working place
with two freezing facilities:
- Blast freezer: for big fish quantities.
- Brine tank: for fast freezing of individual small fish.
2 Objective
Our goal consists of two objectives. The blast freezing room with postulated dimensions has
to be able to freeze a maximum charge of 60000 kg of bonito tuna or mackerel in 24 hours.
The brine tunnel also with postulated dimensions will freeze 7000 kg of sardines in one hour.
The time in which one fish passes through the brine tunnel amounts 20 minutes. As
refrigerant we use ammonia. The fish enters both installations at about 8-11C. The bonito
tuna and mackerel leave the blast freezing room at -18C. The Sardines leave the brine tunnel
at -12C.
3 General description of the brine tank
3.1 Dimensions
Dimensions: Length: 18 m
Width: 2.67 m
Height: 1.7 m
3.2 Arrangement
Fig. 1.1 An overview of the brine tank equipped with two agitators and an evaporator.
2
The brine tank has to cool sardine fillets. They are immersed in a cold salty liquid (brine).
This liquid is cooled when passing the submerged evaporator, also situated in the brine tank.
Sardine fillets enter and leave the brine tank with the aim of three conveyer belts:
- The first one, situated on the front side of the tank lifts (approximate 1.7 m) and
deposits the fillets in the brine tank.
- The second conveyer belt, half-submerged in the brine controls the immersion and
transport of the fish (approximately 30 minutes). The fish meet this conveyer belt with
the aid of two agitators that forces the brine stream.
- A third one situated on the backside of the tank controls the exit of the fish out of the
brine tank.
The third conveyer belt passes the fish through to the glazing conveyer belt, which in turn
transports the fish to the one where it is packaged.
The several stages necessary before conservation are illustrated in figure 1.1. Figure 1.2
shows an overview of the tank with its accessories.
Lifting conveyer
belt (from the
ground to the Principal conveyer belt
Fish receipt (immerses and transports the
8C board of the tank)
8C fish)
-12C
2 lifting conveyer belt (lifts

the fish out of the tank)
-12 C
Packaging Glazing
Conservacion in the conveyer conveyer belt
storage room belt -12 C
-18 C -12C
Fig. 1.2 The sardine cycle diagram (in every phase the fish temperature is specified).
3
The evaporator (Evaporador) cools the brine, which is circulated with the aid of two agitators
(Agitador). This circulation provides a uniform brine temperature and a higher heat
transmission.
4 General description of the blast freezer
4.1 Dimensions
Internal dimensions: Length: 18 m

Width: 5 m
Height: 3 m
Useful refrigerating space: Length: 18 m
Width: 3 m
Height: 2 m
4.2 Arrangement
useful refrigerating
space
Fans
evaporators
Fig. 1.3 Blast freezing room arrangement.
The fish to be frozen is hung up on an internal metallic construction mounted in the useful
refrigerating space. Depending on the size of the fish they are hung up in one or two layers.
To achieve an uninterrupted vapour retarder/insulation envelope a total encapsulation

of the structural system by an exterior vapour retarder/insulation system under the floor, on
the outside of the walls, and over the roof deck is placed (Fig. 1.4)
This method offers the least number of penetrations through the vapour retarder, as well as the
lowest cost.
4
Fig. 1.4 Total exterior vapour retarder system
Next sketch (Fig. 1.5) shows the air circulation in the refrigerated facility, the fan-coil units
are draw-through (i.e., room air is drawn through the coil and discharged through the fan).
Fig. 1.5 Air circulation in the blast freezer room.
5
1 Introduction: Multistage Vapour Compression.
Multistage vapour compression refrigeration is used when several evaporators are needed at
various temperatures such as in a supermarket or when the temperature of the evaporator
becomes very low. Low evaporator temperature indicates low evaporator pressure and low
refrigerant density into the compressor. Two small compressors in series have a smaller
displacement and usually operate more efficiently than one large compressor that covers the
entire pressure range from the evaporator to the condenser. The same can be said for
compressors with partial or total injection at intermediate pressure. This is especially true in
refrigeration systems that use ammonia because of the large amount of superheating that
occurs during the compression process. So the choice for multistage vapour compression is
obvious.
When compressors are connected in series, the vapour between stages should be cooled to
bring the vapour to saturated conditions before proceeding to the next stage of compression.
Intercooling usually minimizes the displacement of the compressors, reduces the work
requirement, and increases the COP of the cycle. If the refrigerant temperature between stages
is above ambient, a simple intercooler that removes heat from the refrigerant can be used. If
the temperature is below ambient, which is the usual case, the refrigerant itself must be used
to cool the vapour. This is accomplished with a flash intercooler. A common assumption is to
operate the intercooler at about the geometric mean of the evaporating and condensing
pressures. This operating point provides the same pressure ratio and nearly equal volumetric
efficiencies for the two compressors.
As pressure ratios increase, single-stage ammonia systems encounter problems including (1)
high discharge temperatures on reciprocating compressors causing the lubricant to deteriorate,
(2) loss of volumetric efficiency as high pressure leaks back to the low-pressure side through
compressor clearances, and (3) excessive stresses on compressor moving parts. Thus,
manufacturers usually limit the maximum pressure ratios for multicylinder reciprocating
machines to approximately 7 to 9. For screw compressors, which incorporate cooling,
compression ratio is not a limitation, but efficiency deteriorates at high ratios.
When the overall system pressure ratio (absolute discharge pressure divided by absolute
suction pressure) begins to exceed these limits, the pressure ratio across the compressor must
be reduced. This is usually accomplished by employing a multistage system. A properly
designed two-stage system exposes each of the two compressors to a pressure ratio
approximately equal to the square root of the overall pressure ratio. In a three-stage system,
each compressor is exposed to a pressure ratio approximately equal to the cube root of the
overall ratio. When screw compressors are used, this calculation does not always guarantee
the most efficient system.
Depending on the commercially available refrigerant feed method used, both the capacity
ratings of the cooler unit and its refrigerant flow rates vary. The following feed methods are
used:
6
Dry Expansion.
In this system, a thermostatically controlled, direct-expansion valve allows just enough
liquid refrigerant into the cooling coil to ensure that it vaporizes at the outlet. In
addition, 5 to 15% of the coil surface is used to superheat the vapour. The direct
expansion (DX) coil flow rates are usually the lowest of all the feed methods.
Recirculated Refrigerant.
This system is similar to a dry expansion feed except it includes a recirculated
refrigerant drum (i.e., a low pressure receiver) and a liquid refrigerant pump connected
to the coil. It also has a hand expansion valve, which is the metering device used to
control the flow of the entering liquid refrigerant. The coil is intentionally overfed
liquid refrigerant by the pump, such that complete coil flooding eliminates
superheating of the refrigerant in the evaporator. The amount of liquid refrigerant
pumped through the coil may be two to six times greater (overfeed: 1 to 5) than that
passed through a dry DX coil. As a result, this coils capacity is higher than that of a
dry expansion feed.
Flooded.
This system has a liquid reservoir (surge drum or accumulator) located next to each
unit or set of units. The surge drum is filled with a subcooled refrigerant and
connected to the cooler coil. To ensure gravity flow of the refrigerant and a completely
wet internal coil surface, the liquid level in the surge drum must be equal to the top of
the coil. The capacity of gravity-recirculated feed is usually the highest attainable, in
part because large coil tubes (25 mm OD) are required so that virtually no evaporator
pressure drop exists. In flooded gravity systems, the relative position of the surge
drum to the air cooler, as well as their interconnecting piping and valves, are all
important for proper operation.
2 The Plant
The plant consists of a blast freezing room and a saltwater tunnel with ammonia (R-717) as
refrigerant. The temperature of operation differs for both installations. The air of the blast
freezing room is kept at -40C during operation. This means that the ammonia has to be
cooled a few grades centigrade colder than -40C. Lets say -45C. At the other hand the brine
of the saltwater tunnel is maintained at -15C out of operation. During operation the brine is
cooled to -20C. This means that the evaporator temperature has to be lower than -20C. We
choose 25C. The evaporator of the blast freezing room will be called the low-stage
evaporator, the one of the brine bath, high-stage evaporator. The operation pressure of the
low-stage evaporator measures 0,5450 bar, the operation pressure of the high-stage evaporator
amounts to 1,5150 bar.
There are two general cycle possibilities to investigate. A cycle can be designed to let work
the blast freezing room and the saltwater tunnel together or separately. Or we could design
two different cycles, one for the blast freezing room and one for the saltwater tunnel. If we
design the cycle to make enable that both installations can run together we should control
what happens if one is out of operation. If turns out that it is not possible to run the cycle only
for one of the two installations we should make two different cycles.
7
We prefer to choose a cycle that let work together both plant because of the more stable
cooling load. If for example the blast freezing room is working then the saltwater of the
saltwater tunnel would warm up if both installations can not work together. And because the
freezing time of one batch 60000 kg of fish in the blast freezing room is more or less 24
hours the warming up of the saltwater can be reasonable.
The maximum cooling load of the blast freezing room amounts to 292 kW.
The maximum cooling load of the saltwater tunnel amounts to 660 kW.
Nevertheless these values do not influence the temperature after compression nor the COP
(Coefficient Of Performance). The COP does not change with cooling load because the
compressor work changes with mass flow and mass flow changes with cooling load. If for
example the cooling load halves, the mass flow halves too and so the quotient of cooling load
and compressor work stays the same.
We choose the condenser temperature 35C. Average summertime temperature of Santander

is 22C (see average weather conditions Santander), so there is an average margin of 13C.
We will assume an ambient temperature for design of 25C. This should make it possible to
condense the ammonia with ambient air forced with ventilators through the condenser
mounted on the roof of the factory. The condensing pressure corresponding with the
condenser temperature is 13,5038 bar.
Following results are obtained with the freeware program coolpack, this program was made
by the Technical University of Denmark.
In the plant-diagrams there will only be drawn one evaporator even though there are more in
reality. For thermodynamic calculations it is not necessary to take this into account. Still we
have to take into account that there exist different systems to feed the evaporators. The two
main types are called dry system and flooded system. In a liquid ammonia recirculation
system, a pump circulates the ammonia from a low-pressure receiver to the evaporators. The
low-pressure receiver is a shell for storing refrigerant at low pressure and is used to supply
evaporators with refrigerant, either by gravity or by a low-pressure pump. It also takes the
suction from the evaporators and separates the gas from the liquid. Because the amount of
liquid fed into the evaporator is usually several times the amount that actually evaporates
there, liquid is always present in the suction return to the low-pressure receiver. Frequently,
three times the evaporated amount is circulated through the evaporator.
3 Comparison
Pressure losses in suction and discharge lines can be neglected with regard to the comparison
of different cycles.
We posit that the isentropic efficiency of the compressor is 0.8. Which is a realistic number
for screw compressors.
Furthermore, at first sight we neglect the compressor heat loss. Afterwards we look how much
influence there is on the superheated vapour temperature.
To make sure that there will be no flash gas formed in the line between the condenser and the
expansion valve we propose a 1C subcooling.
Unuseful superheat high stage: 1C
Unuseful superheat low stage: 1.5C
8
These last two values depend on the length and the insulation of the suction lines between
evaporator and compressor.
3.1 Both installations can run simultaneously with use of common condenser
3.1.1 Two one-stage refrigeration cycles with direct expansion evaporators and a common
condenser.
Each refrigeration system consists of a compressor, a condenser, an expansion valve, and an

evaporator. The use of a suction gas heat exchanger (SGHX) in the high temperature cycle is
optional. Subcooling of the liquid for the low temperature evaporator can be obtained in a
liquid subcooler.
Superheat high stage: 1C

Superheat low stage: 1C
These values are sufficient to avoid that unevaporated ammonia would enter the compressor.
As you can see a suction gas heat exchanger is integrated into the plant-cycle. For the thermal
efficiency we take 0.3. The liquid subcooler integrated in the plant-cycle has a thermal
efficiency of 0.5. These should be reasonable values. The thermal efficiency of a heat
exchanger expresses how efficient the heat exchanger utilises the temperature difference
(driving potential) between, for example, the liquid- and suction line in a suction gas heat
exchanger. The thermal efficiency has a value between 0 and 1. An efficiency of 0 means that
no heat-exchange take place. An efficiency of 1 means that the whole driving potential in the
heat exchanger is utilised. For example in a suction gas heat exchanger the suction gas is
heated to the same temperature as in the liquid line inlet.
Fig. 2.1 Two one-stage refrigeration cycles with direct expansion evaporators and a
common condenser.
9
Description of cycle process:

From state point (1) the suction gas is compressed by the high stage compressor and
discharged (2). At (3) the refrigerant is mixed with gas from the low stage compressor and
lead through the discharge line to the common condenser (4). The condensed and subcooled
refrigerant in the condenser outlet (5) is then lead to the liquid inlet of the suction gas heat
exchanger, if this has been selected, or directly to the inlet of the expansion valve. If a SGHX
is included the exit condition (6) will be different from condition (5). From the liquid outlet of
suction gas heat exchanger (6) the subcooled refrigerant flow is split into two parts:
The major part is expanded (7) into the high stage evaporator.
A small part is expanded (10) into the subcooler where it will evaporate and thereby
further subcool the liquid refrigerant for the low stage evaporator.
The refrigerant from the high stage evaporators is mixed (8) with the evaporated refrigerant
form the liquid subcooler and lead through the suction line (9) to the gas side inlet of the
suction gas heat exchanger, if this has been selected, or directly to the compressor inlet (1).
The highly subcooled refrigerant leaving the liquid subcooler (11) is expanded (12) into the
low stage evaporator. From the low stage evaporator (13) the refrigerant is lead through the
suction line to the inlet of the low stage compressor (14). From the outlet of the low stage
compressor (15) the refrigerant is mixed with gas from the high stage compressor (3).
In the suction lines the gas may experience both a pressure drop and a temperature increase.
In the common discharge line the gas may experience a pressure drop only.
Fig. 2.2 Mollier diagram with SGHX
10
As we notice in figure 2.2 we see that superheated vapour temperatures are extremely high.
Even with intensive cooling with an economizer and oil-cooler the temperatures stay too high
to be acceptable.
The coefficient of performance (COP) of the cycle with suction gas heat exchanger is 2.14.
The COP is defined as follows:
Q& evaporator
COP =
W& compressor
If we consider the plant diagram without suction gas heat exchanger, we get following result:
Fig. 2.3 Mollier diagram without SGHX
Comparing both Mollier diagrams shows us that the superheated vapour temperature after the
high compression is much more lower than with suction gas heat exchanger. T2 is now 172.2
C instead of 199.8 C. Also the high stage compressor work is lower: 272.5 kW instead of
282.5 kW. The COP is 2.189. Thus, the suction gas heat exchanger is harmful for current
operating conditions, which is logic because of the superheating at the high stage cycle.
Liquid subcooling is always advantageous with regard to the COP and thus compressor work.
3.2 Both installations can work simultaneously using a common high pressure compressor.
3.2.1 Two-stage refrigeration cycle with direct expansion evaporators and subcooling of
suction gas by liquid injection.
The refrigeration system consists of two one-stage compressors, a condenser, four expansion
valves, two evaporators, a suction gas heat exchanger and a liquid subcooler. The use of a
suction gas heat exchanger (SGHX) in the high temperature cycle is optional.
Superheat high stage: 1C
11
Superheat low stage: 1C

These values are sufficient to avoid that unevaporated ammonia would enter the compressor.
As you can see a suction gas heat exchanger is integrated into the plant-cycle. For the thermal
efficiency we take 0.3. The liquid subcooler integrated in the plant-cycle has a thermal
efficiency of 0.5.
The superheated discharge gas from the low stage compressors is mixed with the superheated
refrigerant from the high stage evaporators. To reduce the superheat of this refrigerant before
compressing it in the high stage compressors, liquid refrigerant is injected into the suction line
to cool the refrigerant gas by evaporation of the liquid.
Fig. 2.4 Two-stage refrigeration cycle with direct expansion evaporators and subcooling
of suction gas by liquid injection.
Description of cycle process

discharged into the high stage discharge line (2) that will lead the refrigerant to the inlet of the
condenser (3). The condensed and subcooled refrigerant in the condenser outlet (4) is then
lead to the liquid inlet of the suction gas heat exchanger, if this has been selected, and to the
liquid inlet of the subcooler. If a SGHX is included the exit condition (5) will be different
from condition (4). From the liquid outlet of the suction gas heat exchanger (5) the subcooled
refrigerant flow is split into three parts:
The major part is expanded (6) into the high stage evaporator.
further subcool the liquid refrigerant for the low stage evaporator.
12
A minor part is expanded (17) into the suction line of the high stage compressors to
cool the suction gas.
The refrigerant from the high stage evaporators is mixed (7) with the evaporated refrigerant
form the liquid subcooler and lead through the suction line (8) to the gas side inlet of the
suction gas heat exchanger, if this has been selected. From the gas outlet of the suction gas
heat exchanger (9) the refrigerant is mixed with the discharge gas from the low stage
compressors and the liquid refrigerant injected to cool the suction gas before it is drawn into
the high stage compressor (1).
The highly subcooled refrigerant leaving the liquid subcooler (11) is expanded (12) into the
low stage evaporator. From the low stage evaporator outlet (13) the refrigerant is lead through
the suction line to the inlet of the low stage compressor (14). From the outlet of the low stage
compresor (15) the refrigerant is lead through the low stage discharge line (16) to the high
stage suction line (1).
Fig. 2.5 Mollier diagram
The temperature of the superheated vapour after the high stage compression is found to be
approximately 170 C. This would exclude the system to be a possibility for our application
if there would not exist something like compressor oil cooling. Next we will define the heat
loss factor fQ.
fQ is the ratio between the heat loss of the compressor to the surroundings and the energy
consumption of the compressor. The value of fQ does not have any influence on the capacity
or efficiency of the compressor. It only determines the discharge temperature of the
compressor.
13
A realistic heat loss ratio for screw compressors is accepted to be 20%. If we introduce this
number in the calculations we obtain a vapour saturation temperature after the second
compressor of 136.7 C. This temperature is still too high to work with.
The conclusion is that even with oil cooling this cycle cannot be applied for our plant.
3.2.2 Two-stage refrigeration cycle with flooded evaporators and an open intercooler.
The refrigeration system consists of two one-stage compressors, a condenser, two expansion
valves, two evaporators, an open intercooler and a liquid separator.
Fig. 2.6 Two-stage refrigeration cycle with flooded evaporators and an open intercooler.
Description of cycle process:

condenser (3). The condensed and subcooled refrigerant in the condenser outlet (4) is
expanded (5) into the open intercooler.
A part of the refrigerant from the intercooler (6) is lead through the high stage evaporator,
where it partly evaporates, and back to the intercooler (7). Another part is expanded into the
low stage liquid separator. From the intercooler (8) the refrigerant is drawn through the high
stage suction line and into the high stage compressor (1).
From the liquid separator (10) the refrigerant is lead through the low stage evaporator, where
it partly evaporates, and back to the separator (11). From the separator the refrigerant gas is
led through the low stage suction line to the low stage compressor, and from there trhought
the low stage discharge line back to the open intercooler.
14
In a liquid overfeed system, the circulating number or rate is the mass ratio of liquid pumped
to amount of vaporized liquid. The amount of liquid vaporized is based on the latent heat for
the refrigerant at the evaporator temperature. The overfeed rate is the ratio of liquid to vapour
returning to the low-pressure receiver. When vapour leaves an evaporator at saturated vapour
conditions with no excess liquid, the circulating rate is 1 and the overfeed rate is 0. With a
circulating rate of 4, the overfeed rate at full load is 3; at no load, it is 4.
Minimum circulation ratios in flooded evaporators can vary from 2 to 4 for downfeed
systems (largediameter tubes) and vary from 6 to 7 for upfeed systems (small-diameter
tubes) . These values are recommended by ASHRAE. Downfeed is also called topfeed and
upfeed is sometimes called bottom feed. Many evaporator manufacturers specify
recommended circulating rates for their equipment.
Earlier in this work we stated that the evaporators of the blast freezing room are defrosted
with hot gas. The selected evaporators are topfeed. We choose ncirc = 4, for both
installations.
Without compressor oil-cooling the ammonia temperature during the second compression
rises to 170.6 C. This is unacceptable for a ammonia refrigeration plant. If we take a heat
loss factor of 20 % into account, temperature lowers to137.3 C. This temperature is still to
high to let the refrigeration cycle be functional.
3.2.3 Two-stage refrigeration cycle with flooded evaporators and a closed intercooler.
valves, two evaporators, a close intercooler and a liquid separator.
The liquid subcooling coil has a thermal efficiency of 0.5, this is supposed to be a reasonable
value although it is difficult to know.
15
We choose again ncirc = 4.
Fig. 2.8 Two-stage refrigeration cycle with flooded evaporators and a closed intercooler.

condenser (3).
A part of the condensed and subcooled refrigerant in the condenser outlet (4) is expanded (5)
into the intercooler. The refrigerant from the intercooler (6) is lead through the high stage
evaporator and back to the intercooler (7). From the intercooler (8) the refrigerant is drawn
through the high stage suction line and into the high stage compressor (1).
The other part of the condensed and subcooled refrigerant in the condenser outlet is lead
through a liquid subcooling coil placed in the in the intercooler (9) and expanded into liquid
separator (10). From the separator (11) the refrigerant is lead through the low stage evaporator
and back to the separator (12). From the separator (13) the refrigerant is drawn through the
low stage suction line and into the low stage compressor (14). From the low stage compressor
the refrigerant is discharged (15) and lead through the low stage discharge line and into the
closed intercooler (16).
16
The same conclusion as in part 3.2.2 can be made regarding to the superheated vapour
temperature. This means that both installations cannot work parallel with previous plant
cycles. So we have to find a solution whereby both installations will not be able to work at
the same time.
3.3 Both installations cannot work at a time.
The compressors and condenser are dimensioned to serve only one installation at a time. That
means that they are dimensioned for the heaviest cooling load of both installations, which is
660 kW of the brine tunnel.
3.3.1 Two-stage refrigeration cycle with a direct expansion evaporator and a two-stage
compressor.
Subcooling of the liquid for the evaporator can be obtained in a liquid subcooler. Suction gas
for high stage can be cooled by liquid injection.
Each refrigeration system consists of a two-stage compressor, a condenser, three expansion
valves, an evaporator, and a liquid subcooler. The use of the liquid subcooler is optional.
When both refrigeration systems use the two-stage compressor, condenser, expansion valves
and liquid subcooler in common, then they are not able to operate together.
17
Fig. 2.10 Two-stage refrigeration cycle with a direct expansion evaporator and a two-
stage compressor.
From state point (1) the suction gas is compressed by the high stage part of the two-stage
compressor and discharged into the high stage discharge line (2) that will lead the refrigerant
to the inlet of the condenser (3). The condensed and subcooled refrigerant in the condenser
outlet (4) is split into three parts:
The major part is led through the liquid side of the liquid subcooler to the expansion
valve (7).
further subcool the liquid refrigerant for the evaporator.
A minor part is expanded (13) into the suction line of the high stage part of the two-
stage compressor.
The refrigerant from the liquid subcooler (6) is mixed with the superheated refrigerant from
the low stage part of the two-stage compressor (12) and the liquid refrigerant injected to cool
the suction gas before it is drawn into the high stage part of the two stage compressor (1).
The highly subcooled refrigerant leaving the liquid subcooler (7) is then expanded (8) into the
low stage evaporator. From the low stage evaporator outlet (9) the refrigerant is lead through
the suction line to the inlet of the low stage part of the two-stage compressor (10). From the
outlet of the low stage part of the two-stage compressor (11) the refrigerant is lead through the
18
low stage discharge line (12) to be mixed with the superheated refrigerant gas from the liquid
subcooler.
In the discharge line the gas may experience a pressure drop only.
In this case we opt that it is not required that the two installations operate together. The plant-
diagram above is drawn only for one installation. The low stage evaporator low stage
expansion valve combination as shown in the plant-diagram should be doubled and drawn
parallel to each other. One represents the blast freezer room evaporator, the other represents
the evaporator of the brine tunnel. Both parallel branches are connected by means of
selection valves mounted on points 7 and 9 in the plant-diagram above. These valves serve to
select which installation will be supplied by cooled ammonia and thus which installation is in
operation. The plant components apart from the evaporators are designed for the heaviest of
both cooling loads (of the two different installations).
We make a summary of the cycle conditions for the blast freezing room:
3.3.1.1 Blast freezing room
Cooling capacity: 292 kW

Evaporator:
o Temperature: -45C
o Pressure: 0.5450 bar
o Low stage superheat: 1C
Condenser:
o Temperature: 35C
o Subcooling: 1C
Suction line:
o Low stage unuseful superheat: 1,5C
o High stage unuseful superheat: 0C
Compressors performance:
o Isentropic efficiency: 0.8
o fQ: 0%
Pressure losses in the evaporator and condensor: 0 kPa
Liquid subcooler:
o Thermal efficiency: 0.5
Following parameters are variables and are selected to have the maximum COP and minimal
high-stage compression temperature. Look further on for explanation.
Volume ratio: Vs,ls / Vs,hs = 3.8
High stage compressor inlet temperature: -9.1C
We propose at least 1C of subcooling in the condenser to avoid the production of flash gas in
the pipe that leads to the high-stage expansion valve. Flashing of subcooled liquid can occur
when there is enough pressure drop in a pipe to form saturated liquid.
19
Fig. 2.11. Mollier diagram for blast-freezer operation
There exist two variables: the volume ratio and the high stage compressor inlet temperature.
In table 2.1 we illustrate numerical what the effects on the COP and on the high stage
compression temperature are.
Afterwards we give a physical explanation of the conclusions we made.
Table 2.1 Influence of the volume ratio on the COP and superheated vapour
temperature.
Volume ratio T1 (C) TE,HS (C) COP T2 (C)

2.6 -15.7 -15.8 1.734 155.7
2.6 -17.0 -17.2 1.737 154.0
2.6 -18.2 -18.4 1.740 152.5
2.6 -19.4 -19.5 1.743 150.9
3 -15.5 -15.7 1.750 139.5
3.2 -13.7 -14.0 1.751 134.5
3.4 -12.1 -12.3 1.753 129.6
3.6 -10.6 -10.7 1.753 125.1
3.8 -9.1 -9.2 1.753 120.9
4 -7.6 -7.7 1.752 117.0
4.2 -6.2 -6.3 1.751 113.3
If we have a closer look at the first 4 rows of table 2.1 then we can conclude that the COP
increases when the high stages compressor inlet temperature T1 decreases till it is almost
equal to the intermediate saturation temperature TE,HS. We explain this as following. The less
superheat there is, the less compression work is needed because the specific volume of
20
ammonia increases with rising temperature. The less compression work required, the higher
the COP will be.
A second remark we do is that the temperature after the second stage compression T2
decreases with reducing the high stage compressor inlet temperature T1 closer to the
intermediate saturation temperature TE,HS.
That means that in the next lines we can confine ourselves to high stage compressor inlet
temperatures T1 chosen close enough to intermediate saturation temperature.
The higher the volume ratio the higher the intermediate saturation temperature TE,HS. This is
logic because the more compression in the first stage the smaller the volume to be compressed
in the second stage and the higher the temperature after the first compression.
The aim is to have a COP that is as big as possible. In the table above we see that there is a
zone between volume ratio of 3.4 and 3.8 with a COP of 1.753. We choose a volume ratio of
3.8 because of a lower high-stage compressor outlet temperature T2. This temperature
measures 120.9C, which is acceptable but still high. With saturation temperature of -9.2C
corresponds a pressure of 3.0033 bar. Resulting pressure ratios of low-stage and high-stage
compression are respectively: 5.51 and 4.50.
Considering a heat loss factor of 20% of both compressor stages gives us following results:
Saturated vapour temperature: T2 = 96.3C
COP: 1.793
Intermediate temperature: TE,HS = -7.6C
Intermediate pressure: 3.2024 bar.
Pressure ratios: 5.88 and 4.22 for respectively low-stage compression and high-
stage compression.
3.3.1.2 Brine tunnel
Evaporator:
o Temperature: -25C
o Low stage superheat: 1C
Condenser:
o Temperature: 35C
o Subcooling: 1C
Suction line:
o Low stage unuseful superheat: 1,5C
o High stage unuseful superheat: 0C
Compressors performance:
o Isentropic efficiency: 0.8
o fQ: 0%
Pressure losses in the evaporator and condensor: 0 kPa
Liquid subcooler:
o Thermal efficiency: 0.5
21
Following parameters are variables and are elected to have the maximum COP and minimal
high-stage compression temperature. Look further on for explanation.
High stage compressor inlet temperature: 2.3C
Fig. 2.11. Mollier diagram for brine tunnel operation
Table 2.2 Influence of the volume ratio on the COP and superheated vapour
temperature.
Volume ratio T1 (C) TE,HS (C) COP T2 (C)

2.0 -3.2 -3.3 2.733 105.6
2.2 -0.3 -0.4 2.738 98.8
2.3 1 0.9 2.739 95.7
2.4 2.3 2.2 2.739 92.7
2.5 3.5 3.6 2.738 89.9
2.6 4.8 4.7 2.737 87.2
3 9.4 9.3 2.727 77.6
As we have to choose the COP as high as possible and the high-stage compressor outlet
temperature T2 as low as possible, we obtain for a volume ratio of 2,4 with a corresponding
high-stage compressor inlet temperature T1 of 2.3C. The COP turns out to be 2,739 and the
high-stage compressor outlet temperature T2 is 92.7C, which is acceptable.
With saturation temperature of 2.3C corresponds a pressure of 4.6761 bar. Resulting
pressure ratios of low-stage and high-stage compression are respectively: 3.09 and 2.89
Just like the cycle for the blast freezing chamber the COP will increase and the superheated
vapour temperature will decrease if we take the heat loss of the compressor into account.
22
3.3.1.3 Conclusion
We can conclude that this cycle is useful for our application. Although an overfeed system is
preferable to a direct expansion system since the installation consist of multiple evaporators.
A steady refrigerant supply is required, which is obtained with overfeed evaporators.
3.3.2 Two-stage cycle with open intercooler
A possible plant diagram could look like this:
Fig. 2.12 Plant diagram of two-stage direct expansion cycle with open intercooler.
With following thermodynamic cycle:
23
Fig. 2.13 Mollier diagram of direct expansion cycle with open intercooler.
Another option is the two-stage refrigeration cycle with flooded or overfeed evaporators and
an open intercooler.
The circulation pump has not been drawn between the liquid separator and the low stage
evaporator. We choose the circulation rate ncirc = 4.
valves, two evaporators, an open intercooler and a liquid separator.
24
Fig. 2.14 Overfeed system of two-stage cycle with open intercooler.
Description of cycle process.

condenser (3). The condensed and subcooled refrigerant in the condenser outlet (4) is
expanded (5) into the open intercooler.
The refrigerant is expanded into the low stage liquid separator. From the intercooler (8) the
refrigerant is drawn through the high stage suction line and into the high stage compressor (1).
From the liquid separator (10) the refrigerant is lead through the low stage evaporator, where
it partly evaporates, and back to the separator (11). From the separator the refrigerant gas is
led through the low stage suction line to the low stage compressor, and from there trough the
low stage discharge line back to the open intercooler.
25
3.3.2.1 Blast freezing chamber
Fig. 2.15 Mollier diagram for blast-freezer operation.
We let vary the intermediate temperature to maximize the COP. We start with the temperature
at the geometric mean of the evaporating and condensing pressures: -11.7 C. This value
brings us close to the optimum but if we higher the intermediate temperature TE,HS a little the
COP increases a little too and the degree of superheating decreases reasonably. When TE,HS
reaches 7.5 C the COP is 1.838 and the temperature after second compression is 117.6
(table 2.3). So this means that the temperature is low enough to avoid lubricant dissociation
in the compressor. The intermediate pressure measures 3.2152 bar. Corresponding pressure
ratios are 5.90 for the low-pressure stage and 4.20 for the high-pressure stage.
Table 2.3 Influence of the intermediate temperature on the COP and the superheated
vapour temperature for blast-freezer operation.
TE,HS (C) COP T2 (C)

-11.7 1.835 129.1
-10 1.837 124.4
-8 1.838 119.0
-7.4 1.838 117.4
-7 1.837 116.3
If we include the compressors heat loss, with a heat loss factor of 20% we get next results:
Saturated vapour temperature: T2 = 95.4C
COP: 1.878
Intermediate temperature: TE,HS = -6.5C
Intermediate pressure: 3.3452 bar.
26
Pressure ratios: 6.14 and 4.04 for respectively low-stage compression and high-stage
compression.
Fig. 2.16 Mollier diagram for brine tunnel operation.
We do the same for the cycle of the brine tunnel. Optimal COP of 2.833 is reached at an
intermediate temperature of 4C with a superheat temperature after second compression of
89.8 C (table 2.4). The intermediate pressure is 4.9750 bar. Corresponding pressure ratios
are 3.28 for the first stage and 2.71 for the second stage.
vapour temperature for brine tunnel operation.

1.4 2.831 95.7
2 2.832 94.3
3 2.833 92.1
4 2.833 89.8
4.5 2.832 88.7
5 2.832 87.6
If we take the compressors heat loss into account the COP will increase and the saturated
vapour temperature will decrease.
27
3.3.2.3 Conclusion
This thermodynamic cycle is found suitable for our application. We will compare results with
those of the two-stage cycle with closed intercooler. The cycle with the highest COP will be
preferred.
3.3.3 Two stage cycle with closed intercooler
A possible plant diagram could look like this:
Fig. 2.17 Plant diagram of two-stage direct expansion cycle with closed intercooler.
With following thermodynamic cycle:
28
Fig. 2.18 Mollier diagram of direct expansion cycle with closed intercooler.
And again another option is the two-stage refrigeration cycle with flooded or overfeed
evaporators and a closed intercooler.
valves, two evaporators, a close intercooler and a liquid separator.
The circulation pump has not been drawn between the liquid separator and the low stage
evaporator. Like before we choose ncirc = 4.
29
Fig. 2.19 Overfeed system of two-stage cycle with closed intercooler.

condenser (3).
A part of the condensed and subcooled refrigerant in the condenser outlet (4) is expanded (5)
into the intercooler. From the intercooler (8) the refrigerant is drawn through the high stage
suction line and into the high stage compressor (1).
The other part of the condensed and subcooled refrigerant in the condenser outlet is lead
through a liquid subcooling coil placed in the in the intercooler (9) and expanded into liquid
separator (10). From the separator (11) the refrigerant is lead through the low stage evaporator
and back to the separator (12). From the separator (13) the refrigerant is drawn through the
low stage suction line and into the low stage compressor (14). From the low stage compressor
the refrigerant is discharged (15) and lead through the low stage discharge line and into the
closed intercooler (16).
30
3.3.3.1 Blast freezing chamber
Fig. 2.20 Mollier diagram for blast-freezer operation.
Exactly like in the two-stage cycle with open intercooler we try to find the maximum COP in
combination with the lowest superheating. We find this for TE,HS = 7.5 C, with
corresponding COP of 1.752 and T2 = 117.6 C (table 2.5). Whereby T2 is low enough to
avoid lubricant dissociation in the compressor.
vapour temperature for blast-freezer operation.

-13 1.751 132.8
-11.7 1.752 129.1
-10 1.752 124.4
-9 1.752 121.6
-7.4 1.752 117.4
-7 1.751 116.5
-6 1.751 113.7
31
Fig. 2.21 Mollier diagram for brine tunnel operation.
We do the same for the brine tunnel and we find: TE,HS = 3.5 C, COP = 2.737, T2 = 90.9 C.
(Table 2.6)
vapour temperature for brine tunnel operation.

0 2.735 98.9
1.4 2.737 95.7
2 2.737 94.3
3 2.737 92.1
3.5 2.737 90.9
4 2.736 89.8
Just like in the cycle with the open intercooler, heat loss in the compressors will increase the
COP and decrease the superheating after compression.
4 Conclusion
Comparing both overfeed cycles with open and closed intercooler we see that the one with the
open intercooler is the most efficient.
The two-stage refrigeration cycle with a direct expansion evaporator and a two-stage
compressor could be a possibility too but has is less efficient and uses direct expansion
evaporators which are not favourable in a multiple-evaporator plant.
32
5 Two-stage cascade system
5.1 Introduction
A two-cascade refrigeration system consists of two separate refrigeration systems that are
connected to each other in such a way that the evaporator of the higher stage system serves a
condensing medium of the lower stage system. A two-stage cascade staging incorporates two
individual refrigeration systems that use different refrigerants and have a closed heat
exchanger to achieve low operating temperatures and reasonable condensing pressures. For
some industrial applications, which require moderately low temperatures, single stage vapour
compression refrigeration cycles become impractical. Therefore, cascade systems are
employed to obtain high temperature differentials between the heat source and heat sink.
These systems are applied for temperatures ranging from) 70 to 100 C. The important
guidelines that have to take into consideration in the design of low temperature cascade
refrigeration systems can be found in the study presented by Stegmann.
The low temperature circuit contains a refrigerant suitable to obtain the desired low
temperature. Standard refrigerants cannot operate at very low temperatures because their
saturation pressure at the low temperature is too low. If the saturation pressure is less than
about 21 in. Hg vac/4 psi (28 kPa), very little refrigerant vapour is drawn into the compressor.
Vapour density is also extremely low at these pressures, so the mass low of refrigerant
through the system is very low.
Refrigerant used for the low temperature circuit of cascade systems generally have a
saturation pressure at the low temperature condition above atmospheric pressure to help keep
air from being drawn into the system. The higher pressure cascade refrigerant, because of its
density, will require a much smaller compressor to provide the needed system capacity then if
a standard refrigerant were used.
5.2 Cycle description
We will examine two different cycle diagrams:

A two-stage cascade system with one evaporator in the high-stage cycle. This
evaporator will serve at the same time as the low-stage condenser.
A two-stage cascade system with two evaporators in the high-stage cycle. One is used
as condenser for the low temperature system, the other as normal evaporator for the
high temperature.
Both systems are drawn below.
Each refrigeration system consists of a compressor, a condenser, expansion valves, and an

evaporator. The use of a suction gas heat exchanger (SGHX) in both cycles is optional.
33
Fig. 2.22 Two-stage cascade system with one evaporator in the high-stage cycle.
Fig. 2.23 two-stage cascade system with two evaporators in the high-stage cycle.
Description of cycle process.
High stage cycle:

From state point (1) the suction gas is compressed by the compressor and discharged into the
discharge line (2) that will lead the refrigerant to the inlet of the condenser (3). The condensed
and subcooled refrigerant in the condenser outlet (4) is then lead to the liquid inlet of the
suction gas heat exchanger. From the liquid outlet of the suction gas heat exchanger (5) the
subcooled refrigerant is divided into two separate flows. One part is expanded (6) and leads
into the high stage evaporator #1, which is also the low stage condenser. The other part is
expanded (6) and leads in to the high stage evaporator #2.
The flows of evaporated and superheated refrigerant from both evaporators (7) are assumed to
have the same superheat. These flows are mixed and lead through a suction line to the inlet of
the suction gas heat exchanger (8). From the gas outlet of the suction gas heat exchanger the
refrigerant is drawn into the compressor (1).
Low stage cycle:

From state point (1) the suction gas is compressed by the compressor and discharged into the
discharge line (2) that will lead the refrigerant to the inlet of the condenser (3). The condensed
34
and subcooled refrigerant in the condenser outlet (4) is then lead to the liquid inlet of the
suction gas heat exchanger. From the liquid outlet of the suction gas heat exchanger (5) the
subcooled refrigerant is expanded (6) and lead into the low stage evaporator. The evaporated
and superheated refrigerant leaving the evaporator (7) is then lead through a suction line to the
inlet of the suction gas heat exchanger (8). From the gas outlet of the suction gas heat
exchanger the refrigerant is drawn into the compressor (1).
In the discharge lines the gas may experience a pressure drop only.
5.3 Analysis
5.3.1 Refrigerants
Next we will analyse both systems on performance starting with the cascade system with one
evaporator in the high-stage cycle. Analysis will be made for different refrigerant
combinations of which one of them will be ammonia. Afterwards we analyse the cascade
system with two evaporators in the high-stage circuit.
Next list is giving most important refrigerants that are compatible with ammonia in a cascade
refrigeration cycle.
Methane series:
R-12 : dichlorofluoromethane
R-22 : chlorodifluoromethane
R-23 : trifluoromethane
Ethane series:
R-134a : 1,1,1,2-tetrafluoroethane
R-170 : ethane
R-290 : propane
Zeotropic blends (% by mass):
R-404A : [R-125/143a/134a (44/52/4)]
R-407A : [R-32/125/134a (20/40/40)]
R-407C : [R-32/125/134a (23/25/52)]
R-410A : [R-32/125 (50/50)]
Azeotropic blends (% by mass):
R-502 : [R-22/115 (48.8/51.2)]
R-507A : [R-125/143a (50/50)]
Inorganic refrigerants:
R-717: ammonia
R-744 : carbondioxide
Hydrocarbon Refrigerants:
R-600a : isobuthane
R-1150 : ethylene
R-1270 : propylene
35
5.3.2 A two-stage cascade system with one evaporator in the high-stage cycle.
In the next study we will find out which refrigerant is most appropriate in combination with
ammonia for the cascade system mentioned above, applied on both installations: the brine
tunnel and the blast freezing room. Remind that with this type of circuit both installations do
not work at the same time.
As we stated above, the pressure in the low-stage cycle should be higher than atmosphere
pressure to avoid air being drawn into the system. This excludes ammonia to be the
refrigerant of the low temperature cycle. Indeed, saturation pressure of ammonia at 45C is
0.545 bar. Evaporator temperature is or 45 C or 25 C depending on which installation
runs. Because 45 C is the most critical situation we made list of those refrigerants, which
have a saturation pressure at 45C above atmosphere pressure (1.013 bar).
All results are obtained with coolpack, simulation software developed by the Technical
University of Denmark. Following assumptions are made:
1K superheating in the evaporators
1K subcooling in the condensers
Isentropic efficiency of both screw compressors: 0.8.
The heat exchanger temperature difference: 5K.
Unuseful superheat in the low-stage suction line: 1.5K
Unuseful superheat in the low-stage suction line: 1K
Pressure losses are ignored
Compressor heat loss: assumed as negligible, in reality heat loss can be up to 20% of
the energy consumption of the compressor.
Table 2.7 Comparison of different refrigerant on the basis of their COPLT.
Refrigerant Saturation pressure (bar) at 45C COPLT

R-502 1.033 3.550
R-404A 1.058 3.469
R-507A 1.102 3.493
R-1270 1.144 3.614
R-410A 1.396 3.612
R-23 5.877 3.125
R-170 6.578 3.183
R-744 8.336 3.124
R-1150 12.465 2.676
The highest recorded atmospheric pressure, 1.086 bar recorded in Tosontsengel, Mongolia, on
19 december 2001. This means that we better exclude refrigerants R-502 and R-404A to keep
a small margin between system pressure and atmosphere pressure. The most suitable
refrigerant will be this one with the best coefficient of performance in the low temperature
circuit (COPLT).
The COPLT is considered at a low-stage condenser temperature TC,LT of 2.4C. The high-
stage evaporator temperature TE,HT is 7.4C. The COPHT, thus, coefficient of performance of
the high-stage and thus the COP of the ammonia cycle is 4.206.
A closer look to the table above shows us that refrigerants R-1270 and R-410 are most
appropriate for the low-stage cycle.
36
The global COP is defined as the ratio between cooling load and total compressor work.
The maximum global COP for R-1270 and R-410A is reached at TE,HT = 7.4C and amounts
to 1.719. The temperature superheated ammonia vapour is 117.4C. Remember that these
results are obtained for blast freezer functioning. The proper Mollier-diagram for R-1270 is
shown in figure 2.24.
Fig. 2.24 Mollier diagram for R-1270 during blast-freezer operation.
Note that in these calculations the thermal efficiency of heat exchanger is 0, what means that
this one is dropped out of circuit and that the heat transfer Q4-5 is 0 kW.
The Mollier diagram for R-1270 in brine tunnel functioning is shown in figure 2.25. The
maximal global COP is 2.564 on condition that TE,HT = 0.0C .
37
Fig. 2.25 Mollier diagram for R-1270 during brine tunnel operation.
5.3.3 A two-stage cascade system with two evaporators in the high-stage cycle.
As we found in previous paragraph, the optimal refrigerant compatible with ammonia is

propylene or R-1270. Next we look if there is a possibility if both installations, brine tunnel
and blast freezer, can run simultaneously.
Fig. 2.26 Mollier diagram for simultaneous operation of blast freezer and brine tunnel.
38
As expected, exceeds the superheated vapour the maximum allowable temperature (130C)
and makes the cycle useless for our application.
6 Final Conclusion
A two-stage cascade system with ammonia as high-stage refrigerant does not provide a more
economical solution than a two stage refrigerating cycle with open intercooler. Without doubt
there exist two-stage cascade refrigerating systems with other refrigerants than ammonia and
propylene (R-1270), which have a better efficiency but this is not the aim of this work.
39
Part 2 :
The blast freezing
chamber
40
1 Introduction
Low temperature stores built directly on the ground may require special precautions to
prevent the build-up of ice below the freezing chamber floor. The ice formation causes
distortion known as "frost heave" and in particularly bad cases, it can lead to the complete
destruction of the chamber and structure of the building (Figure 3.1).
Fig. 3.1 Ice formation resulting in the frost heave of the freezing chamber.
The conditions that give rise to frost heave are rather complex, since they are related to the
type and texture of the soil, the insulation properties, the availability of moisture, the
dimensions of the store, seasonal climatic variations and other factors.
Two methods of preventing frost heave are commonly used. The ground below the store can
be heated either by a low voltage electrical mat in the freezing chamber foundation or by
circulating a heated liquid such as glycol through a pipe grid built into the foundation (Figure
3.2). The heat for the glycol is usually obtained from the compressor hot gas through a heat
exchanger.
Fig. 3.2 Frost heave prevention using floor heating
41
Another method of preventing frost heave is to leave a space below the store for ventilation
(Figure 3.3).
The provision of a vapour barrier and the prevention of frost heave are probably the two most
important requirements in the construction of a freezing chamber.
Fig. 3.3 Frost heave prevention using underfloor ventilation
2 Floor construction
Consider following recommended (by ASHRAE) insulation R-values for a blast freezer built
within an unconditioned space. The values are given for a temperature range of -40C to -
46C.
Table 3.1 Insulation R-values
Construction Thermal Resistance R (m2K/W)

Floors 5.3 7.0
Walls/Suspended Ceilings 7.9 8.8
Roofs 8.8 10.6
Source: ASHRAE HandbookCD
We assume a maximum soil temperature of 15C. However in normally the soil temperature
will be about 12C. The aim is to find a suitable floor construction with the appropriate
thermal resistance, which is also designed to prevent moistening of the insulation layer and
frost heaving. We suggest following combination. First of all we place a layer of dry sand,
which is at least 300 mm thick. On the top of that we place the ventilation layer, which will
be a kind of vaulting. Above the ventilation layer comes the insulation layer. On both sides
of the insulation layer an impervious foil is placed. On the insulation layer a concrete slab of
200 mm will be moulded.
Because the blast freezer temperature is about -40C the maximum temperature difference
between soil and cooled air amounts to 55C. This means that the heat flux through the floor
42
based on the recommended R-values given by ASHRAE has to be in the range of 7.86 W/m2
to 10.38 W/m2.
2.1 Construction materials
2.1.1 Dry sand layer
The dry sand layer is applied to achieve fix subsoil. The thickness of the layer can vary. For
the relevant construction calculations we choose a thickness of 300 mm.
Thermal properties of dry sand:

Specific density = 1600 kg/m3
Specific heat Cp = 794 J/kgK
Thermal conductivity k = 0.35 W/mK
2.1.2 Ventilation layer
We consider two different concrete blocks that can form the ventilation layer. The two blocks
have the same shape but the thickness differs. Sketches are given below. As you can see the
dimension of the block in longitudinal direction is only 400 mm. In reality the vault will be
longer but for our calculations the length is of no importance. We will call the two different
structures, block 120 and block 160, corresponding with their dimension. These concrete
blocks are placed next to each other to form ventilation ducts in the longitudinal direction of
the blast-freezing chamber.
Thermal properties of concrete:

Block 120
Fig. 3.4 Ventilation vault, type block 120
43
Block 160
Fig. 3.5 Ventilation vault, type block 160
Next we will calculate the convection coefficient in the ventilation layer in order to estimate
the heat transmission through the floor afterwards.
2.1.2.1 Determination of the convection coefficient in the ventilation layer.
2.1.2.1.1 Forced convection in a rectangular tube
Symbols:
A: Flow area of a ventilation duct
O: Wetted perimeter of the cross section of a ventilation duct
D: Hydraulic diameter of a ventilation duct
vair: air velocity in the ventilation ducts
Tair: average air temperature in the ventilation ducts
: Specific density of the air
: Dynamic viscosity of the air
k: Thermal conductivity
Cp: Specific heat
L: Length of the duct
Reynolds number:
v D
Re = air

Prandtl number:
Cp
Pr =
k
Nusselt number:
44
hD
Nu =
k
Laminar forced convection in tubes
For laminar flow in tubes (Re < 2300) following correlations apply:
For fully developed flow (hydrodynamic and thermal) ( L > 0.0575 ReD):
1/ 3
1/ 3

3

Re Pr D
Nu L = 3.66 3 + 0.7 3 + 1.615 0 .7
L

For short tubes, a non-developed flow can arise. The equation becomes for 0.5 < Pr < 500:
Re Pr D 1 / 2
Nu L = 0.664 Pr 1 / 3 ( )
L
Turbulent forced convection in tubes
For turbulent flow (developed and non-developed) (Re > 2300 and 0.5 < Pr < 2000) the
equations of Gnielinski apply:
f
(Re 1000) Pr D 2 / 3
NuT = 2
1/ 2 1 +
f L
1 + 12.7 (Pr 2 / 3 1)
2
f = (1.58 ln (Re ) 3.28)

2
The ducts of floor 1 and floor 4 measure 50 mm by 140 mm. They are composed of blocks
120. The ducts of floor 2 and 3 measure 70 mm by 140 mm. They are composed of blocks
160. Now we will find out in which range the convection coefficient will vary with varying
air velocity and average air temperature. Because the only thing we know is that the
ventilation air will enter at ambient temperature of maximum 25C we have to make an
estimation of the average temperature in the ducts. We assume the air velocity between 1 m/s
and 3m/s. We take average air temperatures between 10 and 20 degrees Celsius. The length
of the blast-freezing chamber is 18 m. Because the ventilation vaults cross the whole width of
the building the characteristic length L is 40 m. In the formulas above you can see that the
bigger L the smaller the Nusselt number will be and consequently the convection coefficient
be smaller too.
2.1.2.1.2 Convection coefficient in the ventilation layer constructed with block type
block 120
A = 0.050 0.140 = 0.0070 m 2

O = 2 0.050 + 2 0.140 = 0.380 m
A
D=4 = 0.0737 m
O
45
Table 3.2 Convection coefficient calculation results for block 120
vair Tair Re k Cp Pr f Nu h
(m/s) (K) (kg/m3) (Pa.s) (mW/m.K) (kJ/kg.K) (W/m2.K)
1 10 1.247 17.63 5212 24.95 1.006 0.711 0.00953 17.63 5.97
1 15 1.226 17.87 5055 25.31 1.006 0.710 0.00962 17.14 5.89
1 20 1.204 18.10 4903 25.66 1.006 0.710 0.00971 16.67 5.80
3 10 1.247 17.63 15635 24.95 1.006 0.711 0.00697 43.42 14.70
3 15 1.226 17.87 15166 25.31 1.006 0.710 0.00703 42.38 14.55
3 20 1.204 18.10 14708 25.66 1.006 0.710 0.00708 41.36 14.40
2.1.2.1.3 Convection coefficient in the ventilation layer constructed with block type
block 160
A = 0.070 0.140 = 0.0098 m 2

O = 2 0.070 + 2 0.140 = 0.420 m
A
D=4 = 0.0933 m
O
Table 3.3 Convection coefficient calculation results for block 160
vair Tair Re k Cp Pr f Nu h
(m/s) (K) (kg/m3) (Pa.s) (mW/m.K) (kJ/kg.K) (W/m2.K)
1 10 1.247 17.63 6602 24.95 1.006 0.711 0.00887 21.71 5.80
1 15 1.226 17.87 6403 25.31 1.006 0.710 0.00895 21.15 5.74
1 20 1.204 18.10 6210 25.66 1.006 0.710 0.00904 20.59 5.66
3 10 1.247 17.63 19805 24.95 1.006 0.711 0.00655 52.32 13.99
3 15 1.226 17.87 19210 25.31 1.006 0.710 0.00661 51.07 13.85
3 20 1.204 18.10 18630 25.66 1.006 0.710 0.00666 49.85 13.70
If we take a closer look at the tables above, we can conclude that the average air temperature
does not have an important influence on the convection coefficient. We assume that the air
velocity applied in soil ventilation is in the range of 1 m/s to 3 m/s. This means that the
convection coefficient will be in the range of 5 W/ m2K to 15 W/m2K.
2.1.3 Insulation layer
Just like for the walls we take polyurethane as insulation material.
Thermal properties of polyurethane:

46
We will consider 2 different standard sizes: insulation panels which are 100 mm thick or
panels of 150 mm thick. In next paragraph we will calculate the required thickness of the
insulation layer.
2.1.4 Concrete slab
The concrete slab is 200 mm thick which is supposed to be enough to withstand external
forces.
Thermal properties of concrete:

With the two different ventilation blocks and the two different insulation panels we can form
4 different combinations for the floor construction. They are given in following table.
Table 3.4 Floor construction
Thickness Concrete slab Polyurethane Ventilation block Dry sand

Floor 1 200 mm 100 mm 120 mm 300 mm
Floor 2 200 mm 100 mm 160 mm 300 mm
Floor 3 200 mm 150 mm 160 mm 300 mm
Floor 4 200 mm 150 mm 120 mm 300 mm
For these 4 combinations we have calculated the heat flux with the restraints described below.
3 Heat flux calculation
To make an estimation of the heat flux entering the blast freezing room through the floor we
use the computation program Cosmos Works. First we made a model of a part of the floor in
Solid Works afterwards we applied the material properties and restraints for the model. The
model of Floor 1 is drawn below.
Fig. 3.6 Solid Works floor model
47
An air velocity of 3.7 m/s in the blast freezer gives us a convection coefficient hfloor for the
concrete slab of 14.17 W/m2K. If the air velocity would have been 5 m/s, hfloor would be 18.0
W/m2K. We will use these values to examine the influence of the internal air velocity on the
heat transmission through the floor.
The convection coefficients in the ventilation layer will be in the range of 5 to 15 W/ m2K as
calculated before. We let vary the average air temperature Tair in the ventilation layer
between 10C and 20C. As we assume that air enters at ambient temperature and cools
down while passing through the ducts, we are quite sure that Tair will be between 10C and
20C
The soil temperature Tsoil is the temperature of the ground under the sand layer. Because we
dont know exactly what temperature we can expect we will compare heat fluxes for Tsoil =
12C and Tsoil = 15C. We will do this for Floor 1.
An example of graphical results of Cosmos Works is given below.
Fig 3.7 Temperature plot of Floor 1 for hfloor = 14.2 W/m2K, Tsoil = 15C,
hventilation = 15 W/m2K and Tair = 288 K.
In figure 3.8 the heat flux in vertical direction is plotted. Heat fluxes through the faces of the
model in horizontal directions are set to be 0 W/m2.
48
Fig 3.8 Heat flux plot of Floor 1 for hfloor = 14.2 W/m2K, Tsoil = 15C,
hventilation = 15 W/m2K and Tair = 288 K.
Calculation results are reported in table 3.5.
Table 3.5 Heat transmission through Floor 1
hventilation (W/m2K) Tair (K) Q(W/m2)

Tsoil = 15C, Tsoil = 12C, Tsoil = 15C,
hsurface = 14.2W/m2K hsurface = 14.2 W/m2K hsurface = 18 W/m2K
5 278 9.993 9.948 10.03
5 283 11.01 10.97 11.05
5 288 12.03 11.99 12.08
15 278 10.05 10.04 10.09
15 283 11.15 11.14 11.19
15 288 12.24 12.23 12.29
As you can see by comparing the results at 12C and 15C soil temperature, there is no
important influence on the heat flux. Furthermore by comparing results for surface
convection coefficients 14.2 and 18 W/m2K we see that the air velocity in the blast freezer
does not have a great impact on the heat flux either. If we compare the heat transmission for
equal temperatures but for a different hventilation we remark that the effect is really small too.
So the only important factor on heat transfer through the floor is the average temperature of
the air flowing in the ducts. From the table above we remark that the minimum heat flux for
one specific floor is attained for hventilation = 5 W/m2K and Tair = 278 K and the maximum for
49
hventilation = 15 W/m2K and Tair = 288 K. That means when we want to compare the 4 different
floor constructions it will be sufficient to do calculations only for these values. Results are
collected in following table.
Table 3.6 Heat transmission through the different floor constructions.
Floor construction hventilation (W/m2K) Tair (K) Q(W/m2)

1 5 278 9.993
15 288 12.03
2 5 278 9.948
15 288 12.22
3 5 278 6.837
15 288 8.355
4 5 278 6.865
15 288 8.365
Comparing the results of Floor 1 with Floor 2 and Floor 3 with Floor 4 shows that the
thickness of ventilation layer not really matters for the heat flux.
We remind that the heat flux, as derived from the recommendations in ASHRAE, should be
between 7.86 W/m2 and 10.38 W/m2. Table 3.6 makes clear that Floor 1 and Floor 2 do not
fulfil the requirements. Floor 3 and Floor 4 do fulfil the requirement and insulate even better
than required.
Because the less material is used the cheaper the floor, we would prefer Floor 4 above Floor
3. So far we have not considered the pressure losses into the ducts due to friction. As we
know, pressure drop increases with increasing air velocity and with decreasing flow area.
Therefore we will subsequently compare pressure losses in both types of ventilation ducts
with air velocities of 1 m/s and 3 m/s. Once we know the pressure drop we can select an
appropriate ventilator out of the Sodeca catalogue.
50
Chapter 4: The Floor ventilation
1 Introduction
Because natural circulation would not occur in a 40-meter long duct with a flow area less than
1 dm2, we have to apply forced circulation with ventilators placed outside the building. The
ducts must be sloped for drainage to remove condensed moisture.
To estimate the ventilation power needed to overcome the pressure losses that arise in the
ducts we will use the Darcy equation. This equation calculates the friction loss for fluid flow
in conduits.
1000 fL v 2
p f =
D 2
p f = friction losses in terms of total pressure, Pa
f = friction factor, dimensionless
L = duct length, m
where
D = hydraulic diameter, mm
= density, kg/m 3
v = velocity, m/s
A simplified formula for calculating friction factor, developed by Altshul (Altshul et al. 1975)
and modified by Tsal, is
68
0.25
f ' = 0.11 +
D Re
if f' 0.018 : f = f '
if f' < 0.018 : f = 0.85 f' + 0.0028
The conduits are made of concrete. Corresponding roughness factor is taken between 1.3
mm and 3 mm. For our calculations we consider = 2 mm.
2 Calculations
As we know the blast freezing room measures 18.4 m long, external length. Nevertheless the
ducts will be longer because they run from one side to the other side of the building in which
the blast freezing room is located. We assume that the duct length is approximately 40 m,
which is the width of the building. For the average air temperature in the ducts we take 15
C. Corresponding air density is 1.226 kg/m3.
We make different calculation for ducts constructed with ventilation blocks block 120 and
block 160. Obtained results are showed in following table.
51
Table 4.1 Friction loss and volume flow rate in the ventilation ducts
Duct type v (m/s) f p f (Pa) V& (m3/h)

Block 120 1 0.0494 16.43 680.4
3 0.0464 138.89 2041.2
Block 160 1 0.0465 12.23 952.56
3 0.0437 103.43 2857.68
We remind that the blast freezer outer dimensions are 5.4 m to 18.4 m. This means that with
the ventilation blocks we have got, there need to be placed 9 blocks next to each other to
cover the whole width of the blast freezer. It follows that there will be 27 ducts which will
guarantee the removal of moisture. With this number we are able to calculate the volume
flow rate as following:
V& = 27 v A 3600
where
v: air velocity
A: flow area of ventilation duct
Results are given in table 3.1.
We also have to take into account that there will be dynamic losses, since the airflow
produced by the ventilators needs to be split up over the various ducts. The end openings
should be screened to keep out rodents, insects, and any material that might close off the air
passages. This leads to an extra pressure drop, calculated with following equation:
v2
p = C
2
where
C : local loss coefficient, dimensionless
C = 1 for the screens
C = 0.5 for the inlet
Next table shows the calculation results of the existing dynamic pressure losses.
Table 4.2 Dynamic pressure losses.
v (m/s) p screen (Pa) p inlet (Pa) p total , dynamic (Pa)

1 0.61 0.31 1.53
3 5.52 2.76 13.79
After considering pressure losses it is now possible to select the required ventilators out of the
Sodeca catalogue. Because of material savings and thus cost savings, we prefer constructing
the ventilation layer with the Block 120-type blocks. The air velocity will be 3 m/s. This
means that the total pressure drop we have calculated is 152.68 Pa, which is low. We will
take a safety margin of 100% to make sure that the ventilators are powerful enough to
overcome the pressure losses.
We conclude:
Total pressure drop p = 305.36 Pa
52
& = 2041.4 m3/h

Total air flow rate V
3 Ventilator selection
We decide to install 3 ventilators to improve airflow distribution over the 27 ducts.
Selected ventilator type: low-pressure centrifugal fan (single inlet).

Model: CB-820-4T
3.1 Series description
Reinforced steel sheet housing.

Galvanized steel sheet turbine.
Max. air flow temperature: 120 C.
Asynchronous motors, with squirrel-cage rotor.
Standard models are supplied with three-phase 230/400 V, 50 Hz and single-phase 230
V, 50 Hz motor.
Class F insulation and IP-55 protection. (IP-54 models 1428-4M and 1733-6M)
Epoxy resin powder anticorrosive finishing, zinc phosphate treated and kiln-dried at
200 C.
3.2 Technical specification
Speed: 1410 rpm

Max. electrical supply: 230 V: 1.33 A
400 V: 0.77 A
Power: 0.25 kW
Maximum air-flow: 2100 m3/h
Sound pressure level (Lp): 64 dB(A)
Approximated weight: 12.2 kg
53
3.3 Dimensions
Fig.4.1 Ventilator
A 321 mm E 138 mm
B1 374 mm H1 135.5 mm
B2 222 mm I 184 mm
C 339 mm J 213 mm
C1 268 mm J1 94.5 mm
C2 71 mm K 130 mm
D1 200 mm K2 160 mm
DE 247 mm L 160 mm
DE1 230 mm O 9 mm
DE2 M,6 4 x 90
54
3.4 Performance curve
Fig. 4.2 Ventilator performance curve
.
Selected working point: Flow rate: V = 680 m3/h
Pressure drop P = 300 Pa
.
Working point: Flow rate: V = 687 m3/h
55
Chapter 5: Refrigerating space calculation
1 Introduction
In this chapter we will calculate the space required to freeze the tuna and mackerel in the blast
freezing chamber. The data we depart from is the maximum batch mass which amounts
60000 kg of bonito tuna or mackerel.
2 Calculations
To know the volume of that mass of fish we need the density. ASHRAE provides us the
unfrozen composition data of blue fin tuna and Atlantic mackerel. We assume that the
composition of Bonito tuna is quite similar to the one of blue fin tuna.
Table 5.1 Fish composition
Components Tuna, Bluefin Mackerel, Atlantic

Moisture content xwo(%) 68.09 63.55
Protein xp(%) 23.33 18.60
Fat xf(%) 4.90 13.89
Carbohydrate xc(%) 0.00 0.00
Fiber xfb(%) 0.00 0.00
Ash xa(%) 1.18 1.35
Other components xo(%) 2.50 2.61
Source: ASHRAE
ASHRAE offers thermal property models for food components between 40C and 150C.
Density (kg/m)
Protein : = 1.3299 103 5.1840 101t
Fat: = 9.2559 102 4.1757 101t
Carbohydrate: = 1.5991 103 3.1046 101t
Fiber: = 1.3115 103 3.6589 101t
Ash : = 2.4238 103 2.8063 101t
Water : = 9.9718 102 + 3.1439 103t 3.7574 103t 2
Ice : = 9.1689 102 1.3071 101t
With t the food temperature.
As we dont know the exact composition of the other components of tuna and mackerel we
posit that they have a weighted average density of all existing components in the fish
(moisture, protein, fat and ash).
Fish enters the blast freezing room between 8C and 11C and leaves at 18C.
As usual we take the most extreme case: t = 11C. In next table we list the densities of all
components of the fish.
56
Table 5.2 Component densities at 11C and 18C
Density (kg/m) 11C -18C

Water 996.76
Ice 919.24
Protein 1324.20 1339.23
Fat 921.00 933.11
Carbohydrate 1595.68 1604.69
Fiber 1307.48 1318.09
Ash 2420.71 2428.85
Now we can calculated the density of both fish species with following formulas:
Unfrozen fish:
water x wo + protein x p + fat x f + carbohydrate xc + fiber x fb + ash x a

fish =
x wo + x p + x f + xc + x fb + x a
Frozen fish:
ice x wo + protein x p + fat x f + carbohydrate xc + fiber x fb + ash x a

fish =
x wo + x p + x f + xc + x fb + x a
The results are collected in the table below.
Table 5.3 Fish density
Density (kg/m) 11C -18C

Bonito Tuna 1088.54 1038.71
Mackerel 1068.23 1022.36
Table 5.4 shows the volume taken by the fish. The volume is found by dividing the mass by
the density.
Table 5.4 Fish volume
Volume (m) 11C -18C

Bonito Tuna 55.12 57.76
Mackerel 56.17 58.69
We postulated before a refrigerating chamber with following dimensions of the useful

freezing space. That means the space we have to store the fish.
Length: 18 m
Width: 3m
Height: 2m
57
Thus the available volume measures 108 m. If we now take a closer look to percentage of
the freezing space volume filled with fish (table 5.5) we can conclude that there is still enough
free space left to have sufficient air circulation.
Table 5.5 Percentage of the freezing space volume filled with fish
Filled volume (%) 11C -18C

58
Chapter 6: Air circulation in the blast freezer
1 Introduction
Air is the secondary refrigerant that removes heat from the produce and its surroundings and
carries it to the evaporator where it is cooled and discharged again into the refrigerated space.
Heat, both sensible and latent due to water vapour condensation, is absorbed by the
evaporator surface.
Air movement inside the freezing chamber serves two main functions: first, the atmospheric
temperature and relative humidity are homogenized to keep them reasonably uniform; second,
evaporator efficiency is improved, in that the heat transfer coefficient is increased as air speed
rises above it.
Air movement may be enhanced by differences in air density, known as natural convection, or
may be promoted by mechanical action, which is called forced convection. Fans or blowers
are used to control the air speed in the chamber and also over the evaporator. An air
circulation pattern is established inside the freezing room, the evaporator being the focal
point, while the rest is governed by smaller forces that direct the air as it moves through the
stacking and recirculation spaces.
The uniform quality of air circulation is determined by measuring the differences in
temperature, for instance, in distinct zones inside the freezing chamber itself. The smaller the
difference the better the air distribution.
When using forced convection systems, uniform atmospheric conditions depend on the air
being dispersed with reasonable consistency, and if the circulation pattern involves much
horizontal travel the product stacking design and the air flow pattern should be compatible.
Stacked rows of produce sometimes serve to regulate the airflow by creating a uniform
resistance all across the room without interfering with circulation.
The air circulation pattern is dependent on several factors: fan air flow; devices for
distribution and suction; quantity of stored produce and stacking patterns; insulation
efficiency; dimensions and shape of the chamber; external ambient temperature; pressure
differences due to air flow; operating condition of the evaporator and frost accumulation.
From this it can be seen that the circulation pattern changes according to the moment of the
operating cycle and the state of the evaporator surface.
Air distribution in the cold store is achieved by using long-throw fans. The throw of the air
delivery must be sufficiently far-reaching to ventilate adequately all points of the cold room.
It is clear that the most important factor in air distribution, no matter which system is used, is
the stacking pattern. The air circulation in the freezing room must be correct no matter the
state of loading. If the stacking pattern requirements are all met and the temperature still
remains too high in certain positions, the reason is most probably an insufficient supply of
cold air. This can be overcome with higher air velocity flows.
For blast freezing chambers the mean air speed in the zone where the product is placed is the
characteristic factor, which defines air distribution. In blast freezers the air speed must be
59
about 2 m/s or higher, measured in the empty section. The air velocity through the evaporator
should be about 3m/s to achieve high overall heat exchange coefficients.
The air circulation must be permanently maintained, as a sufficient supply of cold air to the
product surface is necessary to sustain adequate freezing rates, independently of air or
evaporator temperatures.
2 Air flow rate calculation
With following formula out of Tratado prctico de refrigeracin automtica by Jose Alarcn
Creus it is possible to estimate the airflow rate that has to be accomplished in the freezing
chamber, with a certain refrigerating load given.
. Q
V=
T air Cp
.
V = flow rate (m 3 h)
Q = refrigeration load (kcal h)
Where T = temperature difference incoming air - leaving air (K)
air = air density (kg m 3 )
Cp = specific heat of air (kcal kgK)
Before, we found the refrigeration load: Q = 291.3 kW (250473 kcal/h). We assume the air
heats 3.3C when refrigerating the fish. This means T = 3.3 K. At -40C the air density
air = 1.5156 kg/m3 and the specific heat Cp = 0.240 kcal/kgK.
Substituting these values in the above equation we get a total airflow rate:
.
V = 208697 m3/h.
We provide one ventilator for every 1.5 m wide section of the 18 m long blast freezing
chamber to achieve an adequate air distribution. This implies that there will be 12 ventilators
installed.
The airflow rate for each ventilator:
.
V = 17391 m3/h
This is the airflow rate blown through a section of 1.5 meters wide and 2 meters high.
Because we calculated the degrees of filling of the blast freezing chamber in chapter 5 we are
able to estimate the air velocity.
So, the section surface S = 1.5 x 2 = 3 m2. We define the free section S as the free surface
that will let pass the air. S= S x (1-degree of filling). In the table below we collect the
results.
60
Table 6.1 Free section per 1.5 m length of the blast-freezing chamber
Free section S (m2) 11C -18C

The approximate air velocity vair is calculated as follows:
.
V
v air = m s
S '3600 s h
Table 6.2 Air velocity in the useful blast freezer zone.
vair (m/s) 11C -18C

Mackerel 3.36 3.53
Different sources exist which are recommending certain values of the air velocity:
Tratado prctico de refrigeracin automtica by Jose Alarcn Creus states air
velocities between 2.5 m/s and 16 m/s.
Manual on meat cold store operation and management by Dr G. Cano-Muoz states an
air velocity around 3 m/s.
ASHRAE HandbookCD R18: Static pressure in these rooms is considerable, and air
velocities average between 2.5 and 7.5 m/s, with 6 m/s being common. Air velocities
between 2.5 and 5 m/s give the most economical freezing. Lower air velocities slow
down product freezing, and higher velocities increase unit freezing costs
considerably.
ASHRAE HandbookCD R42 states that air velocities up to 2.5 m/s over the product
are typical for most freezer applications. Moreover, higher velocities require
additional fan power and, in many cases, only slightly decrease the cooling time.
We can conclude that a maximum air velocity of 3.53 m/s is acceptable.
From here of it will be possible to select the appropriate ventilators on condition that we have
a good estimation of the pressure drop over the fish and the evaporator. Because the latter is
impossible to know without experiments we have to find an other way. This will be
explained in next paragraph.
3 Ventilator selection
.
With an air flow rate V of 17391 m3/h for each ventilator and the assumption that a power of
3 kW will be sufficient to provide the required air distribution, we are able to select an
appropriate ventilator. We will do that based on the ventilator catalogue provided by Sodeca.
Selected ventilator type: axial flow fan.

Model: HCT-80-4T-4
3.1 Series description
61
Rolled steel sheet casing.

IP-55 protected external connection box (HCT).
Impeller: variable pitch blades, which can be supplied in 2 versions:
Version PL: reinforced fibreglas polyamid 6 hub and blades (Aluminium hub in
models 80, 90 y 100).
Version AL: aluminium casting impellers.
HCT-40-2T and HCT-45-2T only available in aluminium.
Standard motors 230/400 V, 50 Hz, IP-55, class F.
(IP-54 models 45-4M-0,5 and 56-4M-0,75)
Epoxy resin powder, zinc-phosphated and kilndried at 200C.
On request:
impellers with air-flow direction: impeller to motor.
Reversible impellers 100%.
EExe or EExd - proof motor with special voltages or two speed motors.
3.2 Technical specification
Speed: 1450 rpm

Max. electrical supply: 230 V: 11.3 A
400 V: 6.5 A
Power: 3 kW
Maximum air-flow: 37000 m3/h
Sound pressure level (Lp): 83 dB(A)
Approximated weight: 69.7 kg
3.3 Dimensions
A : 900 mm
B : 860 mm
D : 800 mm
E : 430 mm
J : 12 mm
N: 16x2230
Fig. 6.1 Dimensions of the selected ventilator
62
3.4 Performance curve
Fig. 6.2 Performance curve of the selected ventilator
.
Selected working point: Flow rate: V = 17391 m3/h
.
Working point: Flow rate: V = 17388 m3/h
63
1 Product load
1.1 Introduction
To calculate the heat load produced by the fish we need following data:
Maximum mass: m = 60000 kg
Freezing time: t = 24 h = 86400 s
Maximum fish temperature when entering the blast freezing room: tin = 11C
Fish temperature when leaving the blast freezing room: tout = -18C
Initial freezing point: tf
Specific heat above freezing: Cp1
Specific heat below freezing: Cp2
Latent heat: Lh1
ASHRAE provides us the initial freezing point, the specific heats and the latent heat of blue
fin tuna and Atlantic Mackerel. We assume that the characteristics of blue fin tuna are similar
to those of Bonito tuna.
Following formula calculates the heat load:
mCp1 (t in t f ) + mLh1 + mCp 2 (t f t out )

QFISH =
t
1.2 Tuna
tf = -2.2C
Cp1 = 3.430 kJ/kgK
Cp2 = 2.190 kJ/kgK
Lh1 = 233 kJ/kg
QTUNA = 213.110 kW
1.3 Mackerel
tf = -2.2C
Cp1 = 3.330 kJ/kgK
Cp2 = 2.230 kJ/kgK
Lh1 = 212 kJ/kg
QMACKEREL = 202.215 kW
64
2 Transmission load
2.1 Introduction
First we will give a theoretical expansion with the formulas and correlations we will need to
calculate the heat transfer through a wall. We have to take into account two mechanisms of
heat-transfer: convection and conduction. Radiation is negligible small.
2.1.1 Convection
Symbols:
vair: air velocity

Tf: film temperature
Text: ambient temperature
Tint: blast freezing room temperature
Ts: wall temperature
: Air density
: Air dynamic viscosity
k: Air thermal conductivity
Cp: Air specific heat
L: Significant length
: Air coefficient of thermal expansion
g: Gravitational acceleration
T: Temperature difference Text - Tf
2.1.1.1 Internal convection
Flow parallel to plane surfaces (forced convection):
Reynolds number:
v air L
Re =

Prandtl number:
Cp
Pr =
k
Nusselt number:
The boundary layer flow regimes are laminar for Rex < 2 x 105 and turbulent for 3 x 106 < Rex
For the laminar range
1 1
Nu x = 0.332 Re x 2 Pr 3
65
and
1 1
Nu L = 0.664 Re L 2 Pr 3
For the turbulent range

4 1
Nu x = 0.0288 Re x 5 Pr 3
and
4 1
Nu L = 0.036 Re L 5 Pr 3
Fluid properties should be evaluated at the film temperature. The film temperature Tf is
defined as the arithmetic mean between the wall and the bulk temperature:
Ts + Text
Tf =
2
Internal convection coefficient:
kNu L
hint =
L
We assume the Nusselt number to be constant and evaluated at x = L.
2.1.1.2 External convection
Natural convection for horizontal plates:
Grashof number:
g 2 L3 T
Gr =
2
Rayleigh number:
Ra = Gr Pr
The correlations suggested by McAdams are well accepted for this geometry. A distinction is
made regarding whether the surface faces up or down. It is clear that the induced buoyancy
will be much different for a hot surface facing up than down. McAdamss correlations are,
for a hot surface facing up or cold surface facing down
1
10 5 < Ra L < 2 10 7 Nu L = 0.54 Ra L 4
1
2 10 7 < Ra L < 3 1010 Nu L = 0.14 Ra L 3
66
and for a hot surface facing down or a cold facing up
1
3 10 5 < Ra L < 1010 Nu L = 0.27 Ra L 4
In each of these correlating equations the film temperature Tf, should be used for the air
property evaluation. The length scale, L, is the ratio of the plate-surface area to the perimeter.
kNu L
hext =
L
Natural convection for vertical plates:
Transition from laminar to turbulent flow in natural convection boundary layers adjacent to
vertical plane surfaces has been determined to occur at, or near
Grt Pr = Rat = 10 9
where t, indicates transition.
Churchill and Chu have correlated a large amount of experimental data for natural convection
adjacent to vertical planes over 13 orders of magnitude of Ra. They propose a single equation
for NuL that applies to all fluids. This powerful equation is
2
1

0.387 Ra L 6
Nu L = 0.825 +
[
1 + (0.492 Pr )
9 16
] 8
27

Churchill and Chu show this expression to provide accurate results for both laminar and
turbulent flows. Some improvement was found for the laminar range (Rat < 109) by using the
following equation:
1
0.670 Ra L 4
Nu L = 0.68 +
[1 + (0.492 Pr ) ] 9 16 4 9
2.1.2 Heat resistance
The insulation sandwich panels are composed of 2 stainless steel sheets covering a
polyurethane board at both sides. The stainless steel sheets have a certain thickness dss, for
our calculations we assume dss = 2 mm. Nevertheless this will not affect the heat resistance of
the panel because of the high thermal conductivity of stainless steel.
Thermal conductivity of Stainless Steel: kss = 16 W/mK

Thermal conductivity of Polyurethane: kpu = 0.024 W/mK
We define the external thermal resistance Rss,ext for a stainless steel sheet as follows:
d ss
Rss ,ext =
k ss S ext
67
where Sext : external surface, in contact with ambient air.
Analogous we can define the internal thermal resistance Rss,int (K/W) for a stainless steel
sheet:
d ss
Rss ,int =
k ss S int
where Sint : internal surface, in contact with cooled air.
The thermal resistance for a plane surface due to convection Rconv (K/W) is defined as
follows:
External convection
1
Rconv ,ext =
hext S ext
Internal convection
1
Rconv ,int =
hint S int
For the polyurethane panel we will define the thermal resistance Rpu (K/W) using the
geometric mean surface:
d pu
R pu =
k pu S int S ext
where dpu: thickness of polyurethane panel.
In following calculations we will calculate the thickness of the polyurethane panel so that the
total heat resistance of the walls accomplishes the recommendations we find in ASHRAE.
The total heat resistance Rtotal of the insulation panel is the sum of the heat resistance due to
internal convection, conduction in the stainless steel sheets, conduction in the polyurethane
panel and external convection. Thus:
Rtotal = Rconv ,int + Rss ,int + R pu + Rss ,ext + Rconv ,ext

1 d ss d pu d ss 1
= + + + +
hint S int k ss S int k pu S int S ext k ss S ext hext S ext
Consider following recommended (by AHSRAE) insulation R-values for a blast freezer built
within an unconditioned space. The values are given for a temperature range of -40C to -
46C.
68
Table 7.1 Thermal resistance recommended by ASHRAE
Construction Thermal Resistance R (m2K/W)

Floors 5.3 7.0
Walls/Suspended Ceilings 7.9 8.8
Roofs 8.8 10.6
These numbers are given for 1 m2. If we want to calculate the thickness dpu out of those we
have to divide them by the geometric mean surface of the wall we consider. As you can see
in above table there are 2 values given. These are the borders of the interval where the
thermal resistance should be in. It follows that there will be two values for dpu, namely dpu,min
and dpu,max corresponding with the minimum and maximum values for R.
The wall temperatures are obtained by first calculating the heat transmission Q through the
walls:
Text Tint
Q=
Rtotal
and with that we can calculate the temperatures of the interior and exterior wall surface:
Q
Tw ,ext = Text
hext S ext
Q
Tw ,int = Tint +
hint S int
where
Tw,ext: exterior wall surface temperature
Tw,int: interior wall surface temperature
2.2 Calculations
As air velocity in the blast freezing room we take 3.7 m/s. The origin of this value is given
farther on in this work. Assume Text = 25C. This is 3 degrees Celsius more than the average
maximum temperature in Santander (see Average weather conditions in Santander). The
temperature inside the blast freezing room reaches -40C.
2.2.1 Ceiling
Because we dont know the film temperature at the interior wall surface nor at the exterior
wall surface we make an estimation: Tf,int = -40C and Tf,ext = 22C. Because of the forced
convection inside the freezing chamber the convection coefficient at the interior wall surface
will be higher then that at the exterior wall. And as we know, the bigger the convection
coefficient the smaller the difference between the bulk temperature and wall temperature. It
follows that the estimations we made will be a good approximation of the real temperatures.
In spite of this we will do 3 iterations in our calculations to make sure our results are accurate.
The interior dimensions of the ceiling:
Width: 5 m
69
Length: 18 m
Sint = 90 m2
The exterior dimensions of the ceiling:

Width: 5 m + dwall x 2
Length: 18 m + dwall x 2
where dwall = wall thickness
This means that for the internal convection the significant length L = 5 m. For the external
convection the significant length L is calculated like mentioned before. Hence:
S ext
L=
2 Width + 2 Length
.
To calculate the Nusselt number of the exterior convection we will use the McAdamss
correlation for a cold surface facing up.
We start the calculations with dwall = 0 mm
70
Table 7.2 Calculations for the heat transfer through the ceiling
Heat transfer variables Iteration 1 Iteration 2 Iteration 3

Internal convection Tf (C) -40 -39.71 -39.72
(kg/m3) 1.5156 1.5138 1.5139
Cp (kJ/KgK) 1.006 1.006 1.006
(Pa.s) 15.11 15.12 15.12
k (mW/mK) 21.270 21.292 21.291
Pr 0.71465 0.71463 0.71463
ReL 1855632 1851625 1851748
NuL 3330.13 3324.33 3324.51
hint (W/m2K) 14.17 14.16 14.16
External convection Tf (C) 22 20.25 20.87
T (C) 3 4.75 4.13
dwall (mm) 0 170 180
K (mW/mK) 25.860 25.722 25.771
Pr 0.70921 0.70965 0.70950
g 2
( )
144534000 148813028 147295558
1 K m3
2
Length (m) 18 18.34 18.36

Width (m) 5 5.34 5.36
Sext (m2) 90 97.9356 98.4096
L (m) 1.957 2.068 2.074
Gr 3247471212 6255462228 5432576440
Ra 2303147177 4439194556 3854388247
NuL 59.15 69.69 67.27
Hext (W/m2K) 0.7818 0.8667 0.8358
Heat resistances Rss,ext (K/W) 1.389.10-6 1.276.10-6 1.270.10-6
Rss,int (K/W) 1.389.10-6 1.389.10-6 1.389.10-6
Rconv,ext (K/W) 0.014213 0.011779 0.012158
Rconv,int (K/W) 0.784330 10-3 0.784897 10-3 0.784880 10-3
Rmax,recomm (K/W) 0.097778 0.093733 0.093507
Rmin,recomm (K/W) 0.087778 0.084146 0.083944
dmax,recomm (mm) 178.8 182.9 182.0
dmin,recomm (mm) 157.2 161.3 160.4
dpu (mm) 170 180 180
Length (m) 18.34 18.36 18.36
Width (m) 5.34 5.36 5.36
Sext (m2) 97.9356 98.4096 98.4096
Rss,ext (K/W) 1.276 10-6 1.270 10-6 1.270 10-6
Rss,int (K/W) 1.389 10-6 1.389 10-6 1.389 10-6
Rconv,ext (K/W) 0.013061 0.011722 0.012158
Rconv,int (K/W) 0.784330 10-3 0.784897 10-3 0.784880 10-3
Rmax,recomm (K/W) 0.093733 0.093507 0.093507
Rmin,recomm (K/W) 0.084146 0.083944 0.083944
dmax,recomm (mm) 180.0 182.9 182.0
dmin,recomm (mm) 158.4 161.3 160.4
dpu (mm) 170 180 180
Rpu (K/W) 0.075448 0.079693 0.079693
Rtotal (K/W) 0.089296 0.092203 0.092639
Heat transfer Q (W) 727.9 705.0 701.6
Wall temperatures Tw,ext (C) 15.49 16.74 16.47
Tw,int (C) -39.49 -39.45 -39.45
71
where
2
8 .8 m K
Rmax,recomm = W
S int S ext
and
2
7 .9 m K
Rmax,recomm = W
S int S ext
Conclusion:
To accomplish to the recommendations of ASHRAE a wall thickness of at least 160.4 mm is

necessary. We chose 180 mm because that is a standard size (see Technical details sandwich
panels). After three iterations we know the heat loss through the ceiling and the interior and
exterior wall temperatures, respectively 701.6 W, -39.45 C and 16.47 C.
2.2.2 Walls
Calculations for the walls are nearly the same excepting. In stead of the McAdams
correlation we should use Churchill and Chu for natural convection for vertical plates.
There are two types of walls regard to thermal calculations. The sidewalls and the back and
front wall. We assume the back and front wall to be equal however the front wall has a door.
We can do this assumption because the door is constructed with the same materials as the
walls. Just like in the calculations for the ceiling we start the iterations with Tf,int = -40C and
Tf,ext = 22C.
The interior dimensions of the sidewalls:

Height: 3 m
Length: 18 m
Sint = 54 m2
This means that for the internal convection the significant length L = 3 m.
The exterior dimensions of the sidewalls:

Height: 3 m + dwall
Length: 18m + dwall x 2
The interior dimensions of the back and front wall:

Width: 5 m
Height: 3 m
Sint = 15 m2
This means that for the internal convection the significant length L = 5 m.
The exterior dimensions of the back and front wall:

Width: 5 m + dwall x 2
Height: 3 m + dwall
72
For the external convection the significant length L is equal to the exterior height of the wall.
In what follows we will only report the results of the third iteration because other results are
insignificant for further analysis.
Table 7.3 Calculation results of the heat transfer through the sidewall and front wall.
Heat transfer variables Sidewall Front/Back wall

Internal convection Tf (C) -39.75 -39.71
(kg/m3) 1.5140 1.5138
Cp (kJ/KgK) 1.006 1.006
(Pa.s) 15.12 15.12
k (mW/mK) 21.29 21.29
Pr 0.71463 0.71463
ReL 1111248 1851594
NuL 2209.59 3324.29
hint (W/m2K) 15.68 14.16
External convection Tf (C) 22.89 22.93
T (C) 2.11 2.07
dwall (mm) 190 190
k (mW/mK) 25.93 25.93
Pr 0.70899 0.70898
g 2
( )
142353968 1422263004
1 K m3
2
Length (m) 18.38 5.38

Height (m) 3.19 3.19
Sext (m2) 58.6322 17.1622
L (m) 3.19 3.19
Gr 9734467031 9556083394
Ra 6901631370 6775070246
NuL 224.38 223.07
hext (W/m2K) 1.82 1.81
Heat resistances Rss,ext (K/W) 2.132 10-6 7.283 10-6
Rss,int (K/W) 2.315 10-6 8.333 10-6
Rconv,ext (K/W) 9.351 10-3 0.032131
Rconv,int (K/W) 1.181 10-3 4.709 10-3
Rmax,recomm (K/W) 0.156393 0.548467
Rmin,recomm (K/W) 0.140398 0.492374
dmax,recomm (mm) 197.0 197.0
dmin,recomm (mm) 175.4 175.4
dpu (mm) 190 190
Rpu (K/W) 0.140695 0.493413
Rtotal (K/W) 0.151231 0.530269
Heat transfer Q (W) 859.6 122.6
Wall temperatures Tw,ext (C) 20.98 21.06
Tw,int (C) -39.49 -39.42
73
Conclusion:
To accomplish to the recommendations of Ashrae a wall thickness of at least 175.5 mm is

necessary. We chose 190 mm in the calculations. After three iterations we know the heat
transfer Q through the sidewalls and the interior and exterior wall temperatures Tw,int and
Tw,ext. Respectively 859.6 W, -39.49 C and 20.98 C. If we do the same for the back and
front wall we get Q = 122.6 W, Tw,int = -39.42C and Tw,ext = 21.06C. We could have chosen
dwall = 180 mm too, but then the margin with the minimum wall thickness is small. Because
190 mm is a not standard size (see Technical details sandwich panels) we will have to change
our choice to dwall = 200 mm. A summary of the results adapted to standard is given in a next
paragraph.
2.2.3 Floor
In previous chapter about the floor construction we calculated that the heat transfer through
the wall will be maximum 8.365 W/m2. This means that the floor fits to the recommendations
given by ASHRAE. Indeed, it follows that the minimum thermal resistance R = 6.58 Km2/W.
We assumed the earth temperature about 15C.
We assume that the ventilation blocks we chose have standard sizes.
The dimensions of the floor are:

Width: 5 m
Length: 18 m
Sint = 90 m2
This means that for the internal convection the significant length L = 5 m. The internal
convection coefficient hint is also calculated with the Nusselt number obtained from the
formula for flow parallel to plane surfaces.
Results of the calculations we made are reported in next paragraph.
2.3 Transmission load calculated with standard insulation size
Following results are obtained with the same formulas like before. The only difference is that
we have assumed a wall thickness of 200 mm as well for the ceiling as for the side, back and
front walls.
Table 7.4 Calculation results of the heat transfer through the ceiling, sidewalls, front
wall, back wall and floor with standard insulation size.
variables Ceiling Sidewall Back/Front wall Floor

dpu (mm) 200 200 200 150
hint (W/m2K) 14.16 15.68 14.16 14.16
hext (W/m2K) 0.82 1.80 1.79
Rtotal (K/W) 0.101 0.158 0.555 0.073
Q (W) 642.5 410.4 117.2 752.9
Tw,ext (C) 17.12 21.13 21.22
Tw,int (C) -39.50 -39.52 -39.49 -39.41
74
2.4 Total transmission load
The heat absorbed through the ceiling, walls and floor of the blast freezing chamber is quite
small as we see. This is the result of good insulation and small wall surfaces.
Qtatal = Qceiling + 2 Qsidewall + Q frontwall + Qbackwall + Q floor

= 642.5W + 2 410.4 W + 117.2 W + 117.2 W + 752.9 W
= 2450.5W
3 Air load
3.1 Introduction
When the freezing process starts the blast freezing room is filled with air at ambient
temperature. This air has to be cooled before the freezing process can start.
3.2 Calculations
When the room is maximum charged, that means that the maximum mass of fish is in the
freezing room, the existing air volume in the freezing space is more or less the half of the
useful freezing space volume. In addition to that we have the air volume in the evaporator
and in the ventilation space behind the ventilator and above the freezing space. Lets assume
that the air volume in the evaporator is the half of the total volume of the proper evaporator.
So, approximately:
3m 2m 18m 1m 2m 18m
Vair = + + 1m 2m 18m + 5m 1m 18m = 198m 3
2 2
Ambient temperature: tamb = 25C

Ambient pressure: p = 1 atm = 101.325 kPa
Blast freezer operating temperature: tc = -40C
Air density at 25C: = 1.18435 kg/m3
Air mass: Mair = Vair x = 234.50 kg
Average relative humidity in Santander : =77.5%
Saturation pressure at 25C: pws = 3.1692 kPa
Partial pressure of water vapour at 25C: pw = x pws = 2.4561 kPa
The humidity ratio W is given by:
Mw pw
W= = 0.62198
M da p pw
where
Mw : mass of water vapour
Mda : mass of dry air
75
It follows that W = 0.015452

The dew point temperature for absolute pressure pw = 2.4561 kPa is 20.8C. That means that
condensation starts when room temperature gets below 20.8C. Because the coils of the
evaporator are constantly at 45C, frost will be formed immediately on these coils.
With
M air = M da + M w
and with previous formula we get:
1
M da = M air
1+W
W
Mw = M air
1+W
Thus,
Mda = 239.93 kg air
Mw = 3.57 kg water
At freezing chamber temperature of 40C the humidity ratio Ws of saturated air amounts to
0.0000793 kg water/kg dry air. Multiplying this with Mda gives us the quantity of water
vapour that remains in the cooled air: 0.019 kg water. We neglect this in further calculations.
The energy needed to cool down the air and to freeze the water vapour in it can be calculated
with next formula:
0
q = M da Cp air (t amb t c ) + M w Cp w (t amb 0 ) + M w Lw + M w Cp ice dt
t eva
where
Cpair: heat capacity of air = 1006.5 J/kgK
Cpw: heat capacity of water = 4220 J/kgK
Cpice: heat capacity of ice = 2062.3 + 6.0769 t J/kgK
Lw : Latent heat of water congelation = 333.8 kJ/kg
teva: Evaporator temperature = -45C
Substitution of the data in above formula gives: q = 16.985 MJ.
We assume that the time taken to cool down the air to -40C is about 1 hour. That means that
the evaporator capacity will be slightly higher in the first hour of operation. We obtain the
heat load by dividing the total energy needed by the time needed to cool down:
Qair = q / 3600 s = 4.718 kW.
76
4 Internal load
4.1 Introduction
All electrical energy dissipated in the refrigerated space (from lights, motors, heaters, and
other equipment) must be included in the internal heat load. Heat equivalents of electric
motors are listed in the table below.
Table 7.5 Heat equivalents of electric motors.
Source: ASHRAE
77
4.2 Calculations
The air in the blast freezing chamber is circulated by 12 ventilators of 3 kW. Because a high
pressure drop caused by the type of arrangement of the fish could be expected we opt for quite
heavy fans. Remind that the fish is hung up on a internal metallic construction. The motor
and the driven equipment are located in the blast freezing room. If we look in the table
below, extracted from ASHRAE, we can make a estimation of the heat load due too
ventilation.
Because electrical motors in refrigerated room can work in overload the heat gain is bigger
than motor power.
Corresponding to a 3 kW electrical motor the heat gain in the blast freezer per ventilator is
about 3711 W. We assume the lights to be switched off when the freezer is in operation. So
that involves that there is no extra heat load due to lighting.
Qinternal = 12 x 3711 W = 44.532 kW
5 Infiltration load
The door is well sealed, so we can conclude that air infiltration load is negligible.
6 Equipment related load
Heat gain due to evaporator defrosting is difficult to estimate. We will opt for hot gas
defrosting which will give us the smallest heat gain possible regarding to defrosting. As you
will see in the next paragraph we take a 10 percent safety factor for the total heat load
excluding defrosting. We assume that this will be sufficient to cover the heat gain due to
defrosting.
7 Total refrigerating load
We recapitulate what the maximum total refrigerating load consists of:

Maximum Product load, which will be this one of tuna: 213.110 kW
Transmission load: 2.451 kW
Air load: 4.718 kW
Internal load: 44.532 kW
Defrost load: included in the safety factor
Safety factor: 10%
Qtotal = 291.291 kW
78
Chapter 8: Evaporators
1 Introduction
Evaporator is the name given to any heat exchanger where the refrigerant is evaporated at low
temperature and therefore at low pressure, but usually above atmospheric pressure to prevent
gas and/or water vapour leakages into the low-pressure circuit. The evaporator is the element
of the refrigerating circuit through which heat is absorbed from the environment that is being
cooled. It can absorb the heat necessary for refrigerant vaporization either from the air (direct
expansion systems), or from a liquid (usually water), or from a solution, which in the case of
the fish industry is used as a secondary refrigerant, being circulated through the air cooler
inside the refrigerated room.
Only direct air cooling through liquid refrigerant vaporization will be described, as secondary
refrigerant plants are installed when it is necessary to accumulate reserves of cold, as in
multipurpose refrigerating installations. Direct expansion involves less initial investment and
power consumption is generally lower, as temperature differences between primary
refrigerant and air are less than in secondary fluid systems. Air-cooling evaporators can
operate under natural convection, where air movement is governed by differences in air
density, or under forced or mechanical convection when fans or blowers are employed to
expedite air movement over the cooler and to facilitate air distribution inside the blast
freezing room.
Natural convection systems have two main advantages: no energy is needed for air circulation
and the desiccation of the produce is much less because air velocities are much lower and
relative humidity is high. There are obviously some drawbacks, first, in their defrosting and,
second, and more important, in their low overall heat exchange coefficient owing to low air
velocities over the evaporator that lead to large exchange surface areas, making them bulky
and very expensive. Their cost, including installation in the blast-freezing chamber, is from
three to four times that of a forced convection evaporator because the exchange surface is
larger, the tube manufacturing cost is high and erection in the chamber is difficult and time-
consuming.
Forced convection or forced draught evaporators have been developed because the higher
overall heat transfer coefficient achieved by rapid air circulation over the coils permits a
drastic reduction in evaporator surface areas. Further, they have great refrigerating capacity.
Air velocities through the evaporator must be in the range of 1.5 to 3 m/s to achieve a high
heat-transfer coefficient, but it should be borne in mind that above the upper limit (3 m/s) the
heat created by the electric fan motor will surpass the increase in cooling capacity. Also,
velocities above 3 m/s will tend to carry moisture deposited on the coils toward the produce.
The throw draught must be sufficiently far-reaching to achieve uniform air circulation and an
even temperature distribution in the cold store.
Evaporator coils are finned tubes built to produce a large heat exchange area in a compact
element. Good thermal contact between the tube and the fins must be assured (soldering,
biting into the tube surface by tube expansion, securing by straightening fin flares). Fin size
and spacing depend on the operation and fundamentally on the operating temperature.
79
Fin height should usually be equal to the diameter of the tube and the fin should be thick
enough to provide mechanical resistance (0.7-1 mm). Fin spacing is between 6 and 15 mm or
more, to minimize the risk of ice block. Sometimes fins are spaced wider at the coil air inlet to
act as a frost catcher to avoid obstruction of the airflow.
Forced convection evaporators are placed for easy access to the units themselves and also to
ancillaries like the expansion valve, defrosting equipment and particularly the fan motor. The
position is decided by the capacity of the cooler. The high-capacity coolers are installed on
the floor.
Evaporators fall into two categories, depending on the way they are fed with liquid
refrigerant: dry or direct expansion and flooded.
In dry expansion operation the evaporator receives the fluid from thermostatic expansion as a
mixture of liquid and a small proportion of vapour. The refrigerant leaves the evaporator
totally evaporated and slightly superheated, about 24 C above the evaporation temperature;
this avoids any risk of liquid reaching the compressor. Liquid refrigerant expansion takes
place in the thermostatic expansion valve, a fluid-metering device fitted with a thermometric
sensor that is able to maintain the superheating of a moderate level of refrigerant vapour.
Its major drawbacks are that evaporation temperature is not constant but decreases with the
temperature of the refrigerated space during the operation of the plant and this decreases
relative humidity, and there is some risk of throbbing or faulty feeding of the evaporator,
which is affected by the condensation pressure. The temperature difference between the air
and the boiling refrigerant is usually kept in the range of 5 to 7C, which should be used to
calculate the correct surface area of the evaporator.
Flooded evaporators operate by a continuous pure liquid flow circulating through the
evaporator, fed by gravity or by pump. The fluid leaves the evaporator as a mixture of liquid
and vapour as only a part of the refrigerant volume delivered to it is evaporated. The mixture
is directed to a liquid separator or surge drum from where the unevaporated liquid is fed back
to the evaporator inlet. The high-pressure refrigerant is released into the liquid separator at
low pressure in such a way that the level is kept fairly constant. The expansion of the fluid can
be controlled by a low-pressure float-valve, a hand-operated valve, a constant flow rate or
thermostatically, and today by an electronic expansion valve allowing easy adjustment of the
liquid level.
One of the main advantages of the flooded system is a great heat-transfer coefficient because
boiling liquid completely wets the internal surface of the evaporator; the lower the vapour
content at the exit of the evaporator the greater this coefficient, so it is advisable to employ
pump circulation to increase the liquid/vapour ratio.
The temperature difference between air and boiling refrigerant is kept in the range of 4 to
5C, which is possible because the high heat-transfer coefficients and flooding operation
assure high efficiency over the entire evaporator surface.
Flooded systems are usually employed in high-capacity ammonia plants and where a
changeable thermal system is likely. When flooded evaporators are fed by gravity this can be
achieved from a single liquid separator or accumulator that serves all the evaporators, which
are run in parallel. Another gravity design has a liquid separator for every evaporator, so the
number of expansion valves is equal to the number of separators. This system is used when a
80
single separator is not enough for correct feeding to all the evaporators, as when pressure
losses are excessive, and in large cold chambers and freezing and chilling tunnels.
Flooded evaporators can be fed using a mechanical pump to circulate liquid refrigerant in the
evaporator-separator circuit at a rate high enough to return the fluid to the separator in a wet
vapour state. The flow rate recommended to achieve a high overall heat-transfer coefficient is
in the range of three to eight times the rate of refrigerant vaporization, dependent on the heat
load of the chamber.
The low-pressure liquid separator or surge drum is placed in the machine room and the
pressure losses in the refrigerant circuit are counteracted by the pump, not by the compressor,
which is useful when operating at low temperatures. All the evaporators in a single circuit
should be fed in parallel using at least two pumps, one operational and one standby, isolating
each evaporator with a solenoid valve under the control of the room thermostat. When these
valves are closed the liquid refrigerant must be recirculated to the liquid separator.
The main disadvantages of flooded evaporators are that they are usually bulky, and they
require a relatively high refrigerant charge.
Some care should be taken when designing a low-pressure circuit. The suction pressure
should be increased, as the evaporator temperature must be kept as high as possible. This is
achieved by choosing the correct size for the evaporator transfer area, limiting pressure losses
between evaporator and compressor by using the right pipe diameter and short pipes between
them, and lastly, avoiding too frequent or too little defrosting; too little producing too thick a
frost deposit.
When suction pressure is unduly low it may be because of poor feeding of the evaporator or
that it is not fed at all. There are several reasons for both irregular situations: the expansion
valve may be obstructed, the solenoid valve may not be in operation (closed when out of
order), too high a compressor capacity for the evaporator capacity (the capacity control of the
compressor is not in operation or there is too much frost on the evaporator), the constant
pressure valve placed downstream to the evaporator may not be correctly adjusted, or the
lower pressure limit for valve closing may be too high so that suction pressure is excessively
reduced.
There is an ideal refrigerant feed and flow system for each evaporator design and
arrangement. An evaporator designed for gravity-flooded operation cannot always be
converted to an overfeed arrangement, and vice versa, nor can systems always be designed to
circulate the optimum flow rate. When top feed is used to ensure good distribution, a
minimum quantity per circuit must be circulated, generally about 30 mL/s. Distribution in
bottom-feed evaporators is less critical than in top or horizontal feed because each circuit fills
with liquid to equal the pressure loss in other parallel circuits. Circuit length in evaporators is
determined by allowable pressure drop, load per circuit, tubing diameter, overfeed rate, type
of refrigerant, and heat transfer coefficients. The most efficient circuiting is determined in
most cases through laboratory tests conducted by the evaporator manufacturers. Their
recommendations should be followed when designing systems.
Top Feed Versus Bottom Feed System design must determine whether evaporators are to be
top fed or bottom fed, although both feed types can be installed in a single system. Each feed
type has advantages; no best arrangement is common to all systems.
Advantages of top feed include:
Smaller refrigerant charge
81
Possibly smaller low-pressure receiver

Possible absence of static pressure penalty
Better oil return
Quicker, simpler defrost arrangements
Bottom feed is advantageous in that
Distribution considerations are less critical
Relative locations of evaporators and low-pressure receivers are less important
System design and layout are simpler
2 Selection
Several types of evaporators are used in ammonia refrigeration systems. Fan-coil, direct
expansion evaporators can be used, but they are not generally recommended unless the
suction temperature is 18C or higher. This is due in part to the relative inefficiency of the
direct-expansion coil, but more importantly, the low mass flow rate of ammonia is difficult to
feed uniformly as a liquid to the coil. Instead, ammonia fan-coil units designed for
recirculation (overfeed) systems are preferred. Typically in this type of system, high-pressure
ammonia from the system high stage flashes into a large vessel at the evaporator pressure
from which it is pumped to the evaporators at an overfeed rate of 4 to 1. This type of system
is standard and very efficient. The evaporators are top feed.
In a liquid ammonia recirculation system, a pump circulates the ammonia from a low-pressure
receiver to the evaporators. The low-pressure receiver is a shell for storing refrigerant at low
pressure and is used to supply evaporators with refrigerant by a low-pressure pump. It also
takes the suction from the evaporators and separates the gas from the liquid. Because the
amount of liquid fed into the evaporator is 3 times the amount that actually evaporates there,
liquid is always present in the suction return to the low-pressure receiver.
The total cooling capacity of the evaporators in the blast freezing room has to be
approximately 292 kW. Helpman is specialized in industrial shock coolers. These are blast
coolers and freezers for tunnel applications (11 kW 114 kW). Helpman is located in the
Netherlands.
The Helpman catalogue is added. We follow catalogue instructions to select the required
evaporators.
We supposed earlier that the air is 3.3 K heated while flowing over the fish. That means that
the air-on temperature amounts to about -37C. The air-on temperature is the air temperature
at the intake side of the coil block. The evaporating temperature is -45C. Thus, the
temperature difference between evaporation and entering air is about 8 K. Because tables are
composed for an evaporating temperature of 8C for refrigerant R-404A and R-507A and
with temperature difference DT1 = 8 K, a correction factor has to be taken into account.
For the relevant evaporators (R-717 pumped system) the correction factor is 0.94. This
number is not in the table but is obtained through extrapolation.
The modified total capacity of the evaporators is obtained by multiplying the correction factor
with the total cooling capacity of the evaporators. That means that the required total modified
cooling load is calculated as follows: 292 kW / 0.94 = 310.6 kW.
We will choose 6 equal evaporators of each at least 52 kW. The total length of the blast
freezer is 18 meters. That means that the evaporator length has to be smaller than 3 m. The
only suitable model fabricated by Helpman is: TYR-T 368-7
82
Evaporator characteristics:
Type: TYR-T 368-7

Capacity:
o R-404A and R-507A at evaporation temperature to = -8C, DT1 = 8 K
dry conditions: 61.4 kW
frosted conditions: 70.6 kW
o R-717 at evaporation temperature to = -45C, DT1 = 8 K,
correction factor: 0.94
dry conditions: 57.7 kW
frosted conditions: 66.4 kW
Fin spacing: 7 mm
Coil surface: 432.6 m2
Internal Volume: 148 dm3
Volume flow rate: 35400 m3/h
Dimensions:
o Weight: 602 kg
o Length: 2920 mm
o Width: 1040 mm
o Height: 1830 mm
Defrost system: Hot gas coil in drip tray
One evaporator consist of an evaporating coil and 6 fans. Because of the blast freezing room
construction we replace the 6 ventilators that are mounted directly on the evaporating coil by
2 ventilators mounted on the top of the coils. So the air circulation is a draw-through system.
That means that the room air is drawn through the coil and discharged through the fan. The
ventilator selection is given is a previous chapter. The capacity of each of them is 3 kW. So
instead of 3.3 kW ventilator power per evaporator coil we have 6 kW. The goal is to have a
bigger external pressure, namely: 364 Pa. The volume flow rate produced by the two
ventilators is 34776 m3/h. Which means that the capacity of the evaporator coil will remain
the same as with the original fans.
Not all 6 evaporators will be defrosted at the same time. This would lead to a high-pressure
receiver with a volume of at least 6 x 148 dm3 = 888 dm3. As we will see in chapter 20, the
high pressure receiver is selected based on other criteria. The defrosting is started when
pressure drop over the evaporator is getting to high. Defrosting can be started automatically
or manually. Each evaporator will be defrosted
For further information about the evaporators: see catalogue or http://www.helpman.nl
83
Chapter 9: Evaporator defrosting
1 Introduction
Whenever the evaporator operates at temperatures below 0C a layer of frost is deposited on

the exchanger surface, increasing its thickness with time. Frost, a solid phase, is the result of
moisture condensation and solidification which is deposited by the air circulating in the
freezing space.
Moisture originates from produce water evaporation; from atmospheric moisture load as the
outside air has a higher moisture content (closing and/ or sealing devices must be periodically
and thoroughly checked).
Frosting-up of the evaporator is inevitable despite all the normal precautions (adequate air
circulation rate to minimize desiccation of produce; efficient thermal insulation and water
vapour barrier; good quality gaskets on the door). Frost deposit on the evaporator first
reduces the overall heat exchange coefficient because of the thermal resistance of ice and
secondly hinders air circulation. Both cause deterioration in the performance of the evaporator
and of the refrigerating installation. The decrease in the heat exchange makes the compressor
work for longer periods, and also lowers the refrigerant boiling temperature. Both factors
increase energy consumption. Further, the temperature of the frosted surfaces diminishes,
increasing the relative humidity in the chamber.
Regular defrosting of the evaporator is necessary because equipment performance deteriorates

with increasing thickness of ice, primarily affecting the free passage of air through the
evaporator coils. First the method to supply heat to the evaporator to melt the ice has to be
decided, then the optimum defrosting frequency must be found.
There are several ways to defrost an evaporator. Occasionally more than one can be used
simultaneously.
Mechanical defrosting. Frost is removed by scraping the evaporator tubes- used only
on exposed tubes.
Thermal defrosting. Heat to melt frost is supplied from either inside or outside the coil
tubes.
Internal heating by hot gases, hot liquid refrigerant or hot brine. Heat is transmitted to
the frost by the metallic mass of the evaporator, achieving defrosting without total
melting of the ice.
External heating by air, water, electricity or antifreeze solution. Heat is transmitted to

the outside layers of ice, melting progressing from the outside to the coil tube surface.
84
2 Defrost system selection
Gas defrosting is the fastest and most efficient method because an adequate supply of hot gas
is available. Besides performing the defrost function, the hot refrigerant discharge gas
internally clears the coil and drain pan tube assembly of accumulated compressor oil. This
aids in returning the oil to the compressor. Gas defrosting is used for small, commercial single
and multiplex units, as well as for large, industrial central plants; it is broadly used on most
low-temperature applications. Hot-gas defrosting also increases the capacity of a large,
continuously operating compressor system because it removes some of the load from the
condenser as it alternately defrosts the multiple evaporators. The compressor must be in
operation but the fans are stopped. This method of defrost puts the least amount of heat into
the enclosure ambient.
Defrost Control. For the most part, defrosting is done with the fan turned off. Inadequate
defrost time and over-defrosting both can degrade overall performance; thus, a defrost cycle is
best ended by monitoring temperature. A thermostat is mounted within the cooler coil to
sense a rise in the temperature of the finned or tube surface. A temperature of at least 7C
indicates the removal of frost and automatically returns the unit to the cooling cycle. The
same thermostat delays fan operation, until the coil surface temperature approaches its normal
operating level.
This practice prevents unnecessary heating of the enclosure after defrost. It also prevents
drops of defrost water from being blown off the coil surface, which avoids icing of the fan
blade, guard, and orifice ring.
Initiation of defrost will be automated by air pressure differential controls. Time clocks,
running time monitors, or by monitoring the air temperature difference through the coil
(which increases as frost accumulation reduces the airflow) are other possibilities. Adequate
supplementary heat for the drain pan and condensate drain lines are considered. Two methods
will run simultaneously (hot-gas and electric) to simplify drain pan defrosting and shorten the
defrost cycle. Drain lines should be properly pitched, insulated, and trapped outside the
freezer, preferably when traversing a warm area.
An optimum defrosting frequency must be established. If the frequency is too low the heat
transfer coefficient and air circulation deteriorate and equipment efficiency decreases. If the
frequency is very high the thermal load increases and the total efficiency of the system is
reduced. Optimum frequency must be established by trial. One easy way is to measure the
duration of the cut-in of the thermostat; when the evaporator has no frost this time is minimal.
Defrosting is necessary when the cut-in time increases about one third, all other conditions
being equal. The trial should be conducted when the room is in thermal equilibrium, for
instance when the plant is closed.
The periods between defrosting may be relatively long in dry climates but they will be much
shorter in equatorial regions. They are also longer in summer than in winter.
The forced convection units with finned coils should be defrosted at least once every three to
six hours. Defrosting will be conducted automatically. A clock-timer or a programmer that
functions during fixed periods at regular intervals controls automatic defrosting. The clock-
timer method is simple and accurate; the only inconvenience being that the frequency has to
be modified according to external conditions.
85
The soft hot-gas defrost system is particularly well suited to large evaporators and is used on
all coils of 50 kW of refrigeration or over. The system eliminates the valve clatter, pipe
movements, and noise associated with large coils during hot-gas defrost. Soft hot-gas defrost
can be used for upfeed or downfeed coils; however, the piping systems differ. Figure 9.1
shows the piping system of downfeed or also called top feed coils. Coils operated in the
horizontal plane must be orificed. The soft hot-gas defrost system is designed to increase coil
pressure gradually as defrost is initiated. This is accomplished by a small hot-gas feed having
a capacity of about 25 to 30% of the estimated duty with a solenoid and a hand expansion
valve adjusted to bring the pressure up to about 275 kPa (gage) in 3 to 5 min. (Refer to
Sequence of Operation in Figure 9.2) After defrost, a small suction line solenoid valve is
opened so that the coil can be brought down to operation pressure gradually before liquid is
introduced and the fans started. A pressure switch initiates the system. The soft hot-gas
defrost system eliminates check valve chatter and most, if not all, liquid hammer (i.e.,
hydraulic problems in the piping).
The following are features of Soft Hot-Gas Defrost System.

Regulating the hot gas to approximately 700 kPa (gage) in the equipment room gives
the gas less chance of condensing in the supply piping. Liquid in hot-gas systems may
cause problems due to the hydraulic shock created when the liquid is forced into an
evaporator (coil). Coils in hot-gas pans may rupture as a result.
Draining the liquid formed during defrost with a float drainer eliminates hot-gas blow-
by normally associated with pressure-regulating valves installed around the wet
suction return line pilot check valve.
Returning the ammonia liquid to the intercooler or high-stage recirculator saves
considerable energy in a system. This assumes that only liquid is drained and is the
saving when liquid is drained to the intermediate point, not the total cost to defrost. If
a pressure-reducing valve is used around the pilot check valve, this rate could double
or triple because hot gas flows through these valves in unmeasurable quantities.
When coils defrost, condensate that has formed as ice or frost on the coils melts. This new
condensate collects in a pan beneath the coil and flows into collection drains outside of the
freezer space. Because the space is cold, condensate pans are heated with hot gas and
electrical resistance to prevent ice formation. Likewise, all condensate drain lines must be
wrapped in heat-tracing tape and trapped outside of the refrigerated space to ensure that
condensate can drain unrestricted.
86
Fig 9.1 Hot gas defrost cycle for bottom feed coils actuated by
airside pressure switch at predetermined setting (approximately 250 Pa).
87
Fig 9.2 Sequence of operation
88
Part 3:
The brine tunnel
89
1 Introduction
The brine tank is not just an ordinary reservoir. Just like the evaporator it has to be designed
and approved. It has several functions that have to be considered during the design process.
First of all it has to support the fluid pressure. Because this pressure is negligible other
restraints are more important. For example if we weld the tank the thickness is primordial. So
we see that some things have to be specified before ordering material.
2 What do we have to specify?
Like considered in the introduction the plate thickness is important. We also need to specify
the material type and fabrication method.
3 Type of steel
We have to consider some facts:

- The tank will be exposed to an aggressive environment (water with NaCl).
- The temperature changes between the environmental temperature (25C) and the
working temperature of the brine (-20C).
- The need for a steel that is easy to weld, to guarantee the quality of the unions.
The most appropriate steel is austenitic stainless steel 316 or 317 thanks to:
- Because it is stainless and with the addition of chromium it obtains excellent
resistance to corrosion.
- Because it is austenitic, it is easy to weld. That is why it is mostly used in the food
industry.
Table 10.1 Chemical composition (%) of the austenitic stainless steels
Grate 316 317

S31600 S31600
Carbon (C) max. 0.08 0.08
Magnesium (Mn) max. 2.00 2.00
Phosphorus (P) max. 0.04 0.04
Sulphur (S) max. 0.03 0.03
Silicon (Si) max. 0.75 0.75
Chromium (Cr) max. 16.0 to 18.0 18.0 to 20.0
Nickel (Ni) 10.0 to 14.0 11.0 to 14.0
Molybdenum (Mb) 2.0 to 3.0 3.0 to 4.0
90
4 Thickness of the steel plates
The fluid pressure is negligible and the steel plates are extra supported by the insulation. That
is why weld performance and costs are more important. We opt for a stainless steel plate of 2
mm.
5 Construction
First we consider two long plates of 17.65 x 1.55 m. We bend these plates in the length at
2.35 m. In the work place we weld them to form the vertical sides of the tank. Afterwards the
bottom (18 x 2.35 m) is welded too.
Because the long dimensions of these plates another solution can be taken. We just cut the
long plates in halve and once in the workplace we weld them together again.
We connect the two long sides of the tank with two bars to give extra support.
The tank also has an opening to remove the brine when necessary. Dirt can be one of the
reasons. Figure 10.1 shows the opening on the front side of the tank.
Fig. 10.1 Opening to remove the brine in the tank
91
Chapter 11: Brine selection
1 Introduction
The cooling medium in which the sardine is submerged is very important. It has to fulfil a lot
of criteria because the food industry regulation is very strict. In this chapter we give a brief
summary of the criteria and a possible solution.
2 Criteria
An important consideration in immersion freezing is selecting a suitable freezing medium.

The medium should be non-toxic, acceptable to public health regulatory agencies, easy to
renew, and inexpensive; it should also have a low freezing temperature and viscosity. It is
difficult to obtain a freezing medium that meets all these requirements.
3 Possibilities
Water: Water isnt appropriate for this application. It is obvious it cannot be used
when the working temperature (-20C) is lower than its freezing point. For these
applications brines are used.
Calcium brine: It dehydrates and gives a bitter taste to the immersed products. That is
why it cannot be in direct contact with food. The minimum working temperature is
55C.
Sodium brine: Used when there is physical contact with the food. Sodium chloride
brine and a mixture of glucose and salt in water are acceptable media. The glucose
reduces salt penetration into the fish and provides a protective glaze.
4 Solution
We use sodium chloride brine with a mass fraction NaCl of 23 %. It fulfils all the criteria
mentioned above. We choose a NaCl fraction of 23 % because it is the eutectic solution. With
this eutectic solution we reach lower temperatures without precipitation. With a smaller NaCl
fraction the solution precipitates ice crystals at a higher temperature. With a higher NaCl
fraction the solution precipitates salt crystals.
92
Fig. 11.1 Equilibrium diagram of sodium chloride (E is the eutectic point)
93
Chapter 12: The agitators
1 Introduction
Agitators force a fluid so that it acquires a circulatory movement. In our case the objective of
agitation is to improve the heat transfer.
2 Selection of the agitator type
The agitators can be vertical or horizontal, fixed or detachable. In our case we will select two
vertical agitators to produce a circulatory flow in the tank. The agitators will be fixed, since
there is no necessity to disassemble them. Both installation and agitators have to work
simultaneously to prevent ice forming and malfunctioning of the system.
Sentido de circulacin
de la salmuera
Cinta elevadora
Cinta principal (extrae el pescado del tanque)
(sumerge y transporta el pescado)
Entrada de Salida de
pescado pescado
Agitador Agitador
Serpentn enfriador
(evaporador)
Lmina separadora
Fig. 12.1 Overview of the brine tank with brine circulation due to agitator work.
2.1 Selection of the vane type
Fig. 12.2 A screw propeller
We opt for a screw propeller because this type is suitable for simple movement and
applications.
94
2.2 Recommendations of the manufacturer
Place two vertical agitators, one on each end side of the tank. Due to its form a
continuous circulation of the brine is guaranteed.
The vanes of the agitators will be screw propellers, since we solely try to maintain a
constant circulation of brine at low speed and low viscosity.
The calculation of the power of each one of the agitators is very complex because an
infinity of variables (dimensions and form of the tank, height of the liquid, form of the
helix and their distance to the bottom, speed of turn of the agitator, viscosity and
density of the fluid, etc) interfere. The manufacturer, based on experience in similar
facilities, advises a power of 1.5 kW per agitator.
A turning speed of 700 r.p.m. is recommended for freezing in brine baths.
2.3 Selection of the commercial agitator model (technical characteristics)
Agitator S.S.C: 1-2-3, commercial company Agitaster.

Turbo-agitator provided with a truss, a bearing and rigid connection to the axis of the
motor.
Direct motor speed of 3000, 1500, 1000 or 750 rpm. By means of a digital frequency
converter the speed can be changed between 300 and 3000 rpm.
The motor can be electrical, with all the variants of protection, or air driven.
Power: up to 1,5 kW.
Axis and helix are in stainless steel or carbon. A detachable axis is optional. Axis and
helix covered with polyester, PP, PVCcan be provided.
We will use two agitators of the described model above (see figure 12.3), specifying that all
the parts in direct contact with the brine will have to be made in stainless steel due to the
corrosive action of the brine. A conventional electrical motor will be used.
Fig. 12.3 Agitator with screw propeller

95
1 Introduction
The installation needs five conveyer belts. Two are situated in the brine tank and three are
posted outside the tank.
2 Selection
Two conveyer belts come into contact with the brine. That is why they are fabricated in
stainless steel. We can order these conveyer belts in the length adjusted to our needs because
the manufacturer offers them in modules.
The other three, outside the tank, are made of P.V.C. Approximately 15 % of their surface is
free to evacuate the excess of water or brine.
The manufacturer Tecnofish delivers all the conveyer belts.

Lifting conveyer belt Tecnofish, with funnel (see figure 13.1).
Width: 800 mm
Material: P.V.C.
Length: 3.3 m
Chassis: stainless steel AISI-316
Modular conveyer belt Tecnofish (see figure 13.2).
Width: 1000 mm
Material: stainless steel AISI-316
Length: 15.3 m
Modular lifting conveyer belt Tecnofish (see figure 13.3).
Material: P.V.C.
Width: 800 mm
Length: 4 m
Chassis: Stainless steel AISI-316
Glazing conveyer belt Tecnofish
Material: P.V.C.
Width: 800 mm
Length: 4 m
Conveyer belt Tecnofish.
Material: P.V.C.
Width: 800 mm
Length: 4 m
96
Fig. 13.1 Lifting conveyer belt
Fig. 13.2 Lifting conveyer belt (left) and modular conveyer belt (right)
97
Fig. 13.3 Modular lifting conveyer belt (left) and modular conveyer belt (right)
98
Chapter 14: The refrigeration load
Chapter 14: The refrigeration load and cooling time

of the brine
1 Introduction
The segments of the total refrigeration load are (1) transmission load, which is heat
transferred into the refrigerated space through its surface; (2) product load, which is heat
removed from and produced by products brought into and kept in the refrigerated space; and
(3) equipment-related load.
The first two segments of load constitute the net heat load for which a refrigeration system is
to be provided; the third segment consists of all heat gains created by the refrigerating
equipment. Thus, net heat load plus equipment heat load is the total refrigeration load for
which a compressor must be selected.
When the blast freezer works the brine temperature increases. That is why we first use the
total capacity of the installation, which approximately equals the refrigeration load to decrease
its temperature to the working temperature (-20C). We also calculate the cooling time when
the brines initial temperature equals the surrounding temperature (25C). In this time interval
no fish is cooled.
2 Refrigeration load
2.1 Transmission load
2.1.1 Sensible heat gain through the walls and floors
Sensible heat gain through walls and floor is calculated at steady state as
Qtrans = UAT (14.1)
where
Qtrans = heat gain, W
A = area of section, m2
T = difference between outside air temperature and air temperature of the
refrigerated space, C
The overall coefficient of heat transfer U of the wall, floor, or ceiling can be calculated by the
following equation:
1
U= (14.2)
1 xj 1
+ +
hi k h
j o
99
where
U = overall heat transfer coefficient, W/m2K
x = wall thickness, m
k = thermal conductivity of wall material, W/mK
hi = inside surface conductance, W/m2K
ho = outside surface conductance, W/m2K
In our case, the walls are made of more than one material; therefore, the value x/k represents
the composite heat resistance of the materials. From the inner to the outer border the tank
consists of a plate of resin, polyurethane, another plate of resin and a thin layer of stainless
steel.
The area of section through witch the heat is transported is not just the outside or inside area.
Because both areas are different an equivalent area is used:
Aeq = Ain Aout (14.3)
2.1.2 The external heat convection coefficient
The brine tank is situated in a normal workplace without extra facilities to circulate air. Heat
transport results from free convection. The Nusselt correlation is given by:
2

1
0.825 + 0.387( Ra 6 )
Nu = (14.4)
8
27
9
50.492 16
1 + Pr

where Ra (Rayleigh) = Gr.Pr
Note that the properties are taken at film temperature:
T + Ts
Tf = (14.5)
2
where
T = the air temperature, K
Ts = the wall temperature,K
Now the external convection coefficient can be calculated with formula (14.6).
Nu.k
ho = (14.6)
L
where
k = the fluid conductivity, W/mK
100
L = the length of the particular part, m
We have one problem. All the variables in the equations above are known except for one. We
do not know the wall temperature in advance. What we do know is that we want to keep the
heat flux under 7 W/m2.
2.1.2.1 Algorithm (1) for external heat convection coefficient estimation
The external heat convection coefficient can be estimated with the aid of the following
algorithm:
Step 1: Estimate an outside wall temperature.
Step 2: Using equations (14.4) through (14.6), determine ho.
Step 3: Using equation (14.1) and (14.3), determine Ts.
Step 4: Repeat step 2 and 3 until Ts and ho converge.
2.1.2 The internal heat convection coefficient
The brine in the tank is circulated with the aim of two propellers of each 1.5 kW. This gives
rise to forced convection. The Nusselt correlation is given by:
1 1
Nu L = 0.664 Re 2
Pr 3
, for Re < 105, 0.6 < Pr < 2000 (14.7)
1 1
NuT = 0.664 Re 2
Pr 3
, for 105 < Re < 107, 0.6 < Pr < 2000 (14.8)
Nu = Nu L2 + NuT2 , for 105< Re <5.105, 0.6 < Pr < 2000 (14.9)
vL
Re (Reynolds) = (14.10)

where
v = velocity of the fluid, m/s
l = the characteristic length, m
= the specific density, kg/m3
= the dynamic viscosity, Pa.s
Now the internal convection coefficient can be calculated with formula (14.11).
Nu.k
hi = (14.11)
L
Now we fail the inside wall temperature. Similar to the procedure mentioned in algorithm 1
we find the inside wall temperature and internal convection coefficient.
101
2.1.2.2 Algorithm (2) for internal heat convection coefficient estimation
Step 1: Estimate an inside wall temperature.
Step 2: Using equations (14.7) through (14.11), determine hi.
Step 3: Using equation (14.1), determine Ts.
Step 4: Repeat step 2 and 3 until Ts and hi converge.
2.1.3 Conductivity of the walls
Heat transported through the sides of the brine tank encounter resistance. This is expressed by
the heat conductivity k (W/mK). Each element of the composite wall has his specific k-value
and thickness. In our case, the wall consists of four layers: from outside to inside, respectively
a layer of resin, of insulation (polyurethane), another of resin and one of stainless steel. Here
we fail the insulation thickness. We estimate one and together with the heat calculated
convection coefficients we value the overall heat transport coefficient U with formula (14.2).
Formula (14.1) gives the heat flux. This value has to be smaller than the limit mentioned
above (7 W/m2). Not to small because higher costs are involved with extra insulation.
Because the practical insulation thickness is standardized, the heat loss will differ of 7 W/m2
(considered as maximum in part 2.1.2). The final thickness and heat transfer must be inserted
in the calculations above to obtain the precise convection coefficient and outside wall
temperature.
2.1.4 Sensible heat gain through the free surface
For this we compare the brine tank with a swimming pool. The losses (Qu) are due to
radiation, superficial convection, evaporation, minus the solar radiation. Because the brine
tank is situated in a hall with a roof, we can
( )
Qu = 0 S p Tm4 Teq4 + hS p (Tm Ta ) + r
dM
dt
as S p I s (14.12)
where
Ta = air temperature, C
Tm = brine temperature, C
Teq = equivalent temperature, C
Sp = the free surface area of the swimming pool
Is = The solar radiation intensity falling into such a horizontal surface
as = The corresponding absorption coefficient (80%)
The heat convection coefficient h (W/m2K) is given by:
h = 7 + 3.3v (14.13)
where
v = air velocity, m/s
102
The latent feat of evaporation r is estimated using the formula of Regnault:
MJ
r = 2.50 2.36 10 3 T (14.14)
kg
The rate of evaporated water mass is approximately calculated with the Dalton law:
dM p pa
= zS p s (14.15)
dt H
where
ps = the vapour saturation pressure at surrounding temperature, mmHg
pa = partial vapour pressure in the air close to the water surface, mmHg
H = atmospheric air pressure, mmHg
z = (3.36+2.96v).10-3, kg/m2s
The equivalent temperature:
Teq = 0.0552Ta1.5 (14.16)
2.2 Product load
The quantity of heat to be removed can be calculated as follows:
1. Heat removed to cool the fish from the initial temperature to the freezing point of the
product:
Q1 = mc1 (t1 tf ) (14.17)
2. Heat removed to freeze the product:
Q2 = mhif (14.18)
3. Heat removed to cool the fish from the freezing point to the final temperature below the
freezing point:
Q3 = mc3 (tf t2) (14.19)
where
Q1, Q2, Q3 = heat removed, J
m = mass of product, kg
c1 = specific heat of product above freezing, kJ/kgK
t1 = initial temperature of product above freezing, C
t2 = final temperature of product below freezing, C
tf = freezing temperature of product, C
hif = latent heat of fusion of product, kJ/kg
c3 = specific heat of product below freezing, kJ/kgK
103
The refrigeration capacity required for products brought into storage is determined from the
time allotted for heat removal and assumes that the product is properly exposed to remove the
heat in that time. The calculation is
Q1 + Q2 + Q3
Q prod = (14.20)
3600n
where
Qprod = average cooling load, W
n = allotted time, h
2.3 Internal load
All electrical energy dissipated in the refrigerated space (from lights, motors, heaters, and
other equipment) must be included in the internal heat load. Heat equivalents of electrical
motors are listed in Table 14.1. The brine tank consists of three conveyer belts: a large one to
conduct the fish and two short ones to let the fish enter and exit. Also two agitators situated in
the tank provide brine circulation.
Table 14.1 Overview of the heat equivalents of 3-phase electric motors (1750 rpm)
Location of motor and driven equipment with respect to

conditioned space or air stream
Motor Full load motor Motor in, driven Motor out, driven Motor in, driven
rated, kW efficiency, % equipment in, W equipment in, W equipment out, W
0.75 75 993 747 249
1.50 79 1887 1491 396
2.20 81 2763 2238 525
3 Total load
When all the calculations are done in the parts above, the total load is:
Qtot = Qtrans + Q prod + Qu + Qint (14.21)
4 Safety factor
Generally, the calculated load is multiplied by a factor of 10% to allow for possible
discrepancies between the design criteria and actual operation.
Q final = 0.1Qtot (14.22)
104
5 Refrigeration load calculation
Calculations of the transmission load were based upon the dimensions of an existing brine
tank (Barandica, Santander, Spain). The tank had the following dimensions:
length = 18 m,
width = 2.67 m,
height = 1.7 m.
The brine has a temperature (=Tbrine) of 253 K and reaches a liquid level of about 1.45 m.
The floor temperature (=Tfloor) and the temperature around the tank (=Toutside) are respectively
293 and 298 K.
The next two tables 14.2 and 14.3 list the properties of air and brine at two different
temperatures. We picked these temperatures with the aim of covering the total working
temperature range.
Table 14.2 Properties of air at 280 and 300 K.
Property Temperature, K
280 300
3
Density , kg/m 1.26 1.18
Dynamic viscosity , Pa.s 1.75 1.84
Thermal conductivity k, W/mK 2.47 2.62
Prandtl, Pr 0.708 0.713
Gr, g(T)L3/2 1.82 1.33
Table 14.3 Properties of brine at 258 and 253 K.
Property Temperature, K
258 253
3
Density , kg/m 1187 1190
Dynamic viscosity , Pa.s 5.7 7
Thermal conductivity k, W/mK 0.42 0.41
Prandtl, Pr 13.6 13.6
5.1 Transmission load calculation
5.1.1 Sensible heat gain through the walls and floors
5.1.1.1 Calculation of the external heat convection coefficient
We calculate the properties Pr and Gr interpolating the values using table 14.2. Table 14.4
gives the results of the different parameters calculated according to the algorithm 1.
105
Table 14.4 Overview of the different calculation steps and results of algorithm 1.
Ts, K Tfilm Pr Gr, 109 Ra, 109 Nu L, m ho, W/m2K Ts, K

288 293 0.710 7.36 5.22 205.57 1.7 3.11 295.74
295.74 296.87 0.709 1.55 1.10 126.53 1.7 1.93 294.38
294.38 296.19 0.709 2.52 1.79 147.08 1.7 2.24 294.88
294.88 296.44 0.709 2.17 1.54 140.26 1.7 2.14 295.01
295.01 296.51 0.709 2.07 1.47 138.36 1.7 2.11 294.97
5.1.1.2 Calculation of the internal heat convection coefficient
We calculate the properties , and k by interpolating the values mentioned in Table 14.3.
We propose to divide the calculation of the internal heat convection coefficient into two parts
(a and b) due to the fact that the specific length (L) differs around the circumference of the
tank.
a) The specific length L is the length of the tank (18 m) minus the two times the wall
thickness. Because we do not know the wall thickness yet, we propose a thickness of 15 cm.
We begin with Ts = -15C or 258K. Table 14.5 lists all the results of the formulas used in
algorithm 2.
Table 14.5 Overview of the different calculation steps (algorithm 2) and internal heat
convection coefficient along the length of the tank.
Ts, K Tfilm,K , ,10-3 k,10-1 Pr Re, Nuturb Nulam Nu L,m hi,W/m2K Ts, K
kg/m3 Pas W/mK 105
258 255.50 1188.50 6.35*10-3 4.15*10-1 13.6 4.97 3098.71 1117.27 3293.98 17.7 77.23 253.09
253.09 253.05 1189.97 6.99*10-3 4.10*10-1 13.6 4.52 2873.01 1065.69 3064.29 17.7 71.00 253.10
253.10 253.10 1189.97 6.99*10-3 4.10*10-1 13.6 4.52 2870.75 1065.17 3061.99 17.68 71.01 253.09
In the last iteration step, L and Q are changed according to the insulation thickness
calculations. We see that the convection heat coefficient and the wall temperature are more or
less the same so further iterations are not necessary.
W
hi = 71.01
m2K
b) The specific length L is the width of the tank (2.67 m) minus the two times the wall
thickness. Like above we propose a thickness of 15 cm. We begin with Ts = -15C. Table 14.6
lists all the results of the formulas used in algorithm 2.
Table 14.6 Overview of the different calculation steps (algorithm 2) and internal heat
convection coefficient along the width of the tank.
Ts, K Tfilm,K , kg/m3 , 10-3 Pas k,10 W/mK Pr

-1
Re, 10
4
Nu L, m hi,W/m2K Ts, K
253 255.50 1188.50 6.35 4.15 13.6 6.65 408.83 2.37 71.59 253.10
253.10 253.05 1189.97 6.99 4.10 13.6 6.05 389.99 2.37 67.48 253.10
253.10 253.10 1189.97 6.99 4.10 13.6 6.00 388.36 2.35 67.77 253.09
106
In the last iteration step, L and Q are changed according to the insulation thickness
calculations. We see that the convection heat coefficient and the wall temperature are more or
less the same so further iterations are not necessary.
W
hi = 67.77
m2 K
5.1.1.3 Calculation of the insulation thickness
Insulation type: Polyurethane

Fabricant: poliuretanos, S.A.
The polyurethane is delivered with a 4 mm resin layer on each side. These layers are anti-
vapour barriers.
kpolyurethane = 0.024 W/mK
kresin = 0.027 W/mK
Now we have to find a xpolyurethane, that yields a Q less than the 7 W/m2 premised. xpolyurethane =
0.015 m satisfies this demand as well for both vertical sides as for the bottom (see table 14.7).
Table 14.7 List of conduction coefficients and thickness (x) of the elements composing
the tank walls.
resin polyurethane Stainless steel

k, W/mK 0.027 0.024 58
x, mm 4 150 2
Inserting hi, ho, the values of table 14.7 together with the temperatures Tout, Tfloor and Tbrine in
equation (14.1) and (14.2) gives the heat fluxes through the several surfaces.
Table 14.8 List of the overall heat coefficients of the walls, temperature differences over
the walls and heat fluxes through the walls.
U, 10-1 W/mK T, K Q, W/m2

Vertical side, length 1.42 45 6.40
Vertical side, width 1.42 45 6.40
Bottom side 1.52 40 6.10
Now we multiply the heat fluxes (Q) in table 14.8 with the appropriate surfaces. This gives us
the heat transfer through the walls and floor. The heat transmission for every side is listed in
table 14.9.
107
Table 14.9 List of the inside, outside and equivalent areas of the several sides; together
with their heat transmission.
Ain, m2 Aout, m2 Aeq, m2 Q, W

Vertical 58.02 70.28 63.85 408.52
Bottom 41.31 48.06 44.56 271.66
5.1.2 Heat loss through the surface.
The brine surface is in contact with the surrounding air. The length of the surface is the whole
length of the tank (18 m) minus two times the thickness of the walls. Likewise, the width of
the surface is given by the total tank width minus two times the thickness of the walls. All the
results and data necessary to solve equations (14.12) to (14.16) are listed in table 14.10.
Table 14.10 Overview of the data and partial solutions to determine the sensible heat
gain through the surface.
L, m 17.72
b, m 2.34
A, m2 41.31
vbrine, m/s 0.15
T, K 288.0
Teq, K 279.9
Tbrine, K 253.0
h, W/m2K 7.50
pa, mmHg 18.49
ps, mmHg 23.78
H, mmHg 760.00
z 3.80 10-3
dM/dt 1.09 10-3
r 2.54 106
Qu , W 16930.44
Qradiation, W 4.78 103
Qsuperficial convection , W 13.93 103
Qevaporation, W -2.79 103
Qsolar radiation, W 0
5.2 Product load
We list all data necessary for the formulas (14.16) through (14.19) with their results in table
14.11.
108
Table 14.11 Overview of the data and partial solutions to determine the product load.
c1, J/kgK 3518.17

c2, J/kgK 2019.15
hif, J/kg 239764.8
m, kg/s 7000
t1, K 281
t2, K 270.8
tif, K 261
Q1 , W 6.98 104
Q2 , W 4.66 105
Q3 , W 3.85 104
Qprod, W 5.74 105
5.3 Internal load
The brine tank contains two conveyer belts and two agitators inside. Both conveyer belts and
agitators are driven by an electric motor outside the tank.
Motor conveyer belt 1: 3-phase, P = 2 kWQ = 2238 W

Motor conveyer belt 2: 3-phase, P = 0.8 kWQ = 747 W
Motor agitator: 3-phase, P = 1.5 kWQ = 1491 W
Qint = 2238 + 747 + 1491 x 2 = 5967 W
5.4 Total load
Formula (14.20) gives Qtot = 597040.15 W.
5.5 Final load
Formula (14.21) gives Qfinal = 656744.17 W.
6 Time necessary to cool the brine tank: 25C -20C
We have the full refrigeration capacity at our disposal (660 kW). The agitators force the fluid
over the evaporator. Due to their functioning the final temperature is sooner reached. But
because of the work they dissipate also heat. Likewise the walls, surface and floor are a cause
of heat entrance. Formula (14.23) gives us the useful energy to cool the brine.
Quseful = Q final Qint Qtrans Qu (14.23)
Calculation of time necessary to cool the brine tank
Inserting the values calculated above in formula (14.23):

109
Quseful = 656744.14 5967 680.18 15930.44 = 634166.52 W
This cooling capacity is employed for the cooling of the brine. We transform formula (14.24),
which relates the temperature rise and the heat added, and put only the time in front (see
formula (14.25)).
Quseful t = c p mT (14.24)
c p mT
t = (14.25)
Quseful
where
cp = specific heat of the brine, J/kgK
m = the total mass of the brine, kg
T = temperature difference, K
We calculate with an average specific heat cp = 3300 J/kgK and specific density = 1186
kg/m3. The mass m = V(m3) x (kg/ m3) = 71450.10 kg.
Formula (14.25) gives t = 16731.65 s = 4.65 h
6.1 Solution
We need 4.65 hours to cool the tank from its initial temperature (25C) to its working
temperature (-20C).
7 Time necessary to cool the brine tank: ? -20C
First we have to know the temperature rise of the brine in 24 hours. For these calculations we
make an important assumption. We do not approach the process with differential equations.
We see in the tables above that the heat flux through the surface is the highest of all heat
gains. Also with no brine circulation it is the most important one (calculating the formulas
(14.12) through (14.16) give Qu (v = 0 m/s) = 15335.5 W). This is even much higher then the
heat flux through the walls and floor with brine circulation at 20C, Qtrans = 680.18 W.
We assume that the total heat flux equals Qu (v = 0 m/s) = 15335.5 W. Formula (14.24) gives
the temperature rise in 24 hours. We use the same average properties as above: cp = 3300
J/kgK, = 1186 kg/m3 T = 5.62C.
After the 24 hours we use full refrigeration capacity (660 kW) to reduce the brine temperature
again. Formula (14.25) gives t = 2007.55 s = 0.56 h
7.1 Solution
If we consider the assumptions made above we estimate that one hour will be more then
sufficient to cool the tank, which was not working for 24 hours, to its working temperature (-
20C).
110
1 Introduction
The evaporator cools the brine by means of evaporating the ammonia inside its tubes. So
when the brine flows along its surface, it cools down. It is necessary to equalize the total heat
load and the cooling capacity of the evaporator. In this chapter we determine the appropriate
evaporator for our special case.
2 Description
The evaporator is situated in the brine tank. It consists of four separate equal heat exchangers.
The heat exchangers are self-designed. So no commercial type will be considered. Our heat
exchanger is very easy and fundamental. It is a collection of parallel pipes, without fins,
through which the ammonia is pumped (see figure 15.1). The pipes are all submerged in the
brine in the tank. Agitators circulate the brine to provide a better heat transport from the
evaporator to the fish. Generally, a brine velocity of 0.15 m/s gives a good performance.
Fig. 15.1 Overview of a partially submerged evaporator in a brine tank.
111
sardine
brine
Fig. 15.2 Schematic overview of the brine tank with evaporator
3 What do we have to specify?
We specify the nominal diameter of the evaporator pipes, distances between them, quantity,
material, distributors and flow direction. Everything depends on the nominal diameter of the
evaporator pipes.
3.1 Nominal diameter of the evaporator pipes.
We choose a nominal diameter of 25 mm. The properties are listed in table 15.1.
Table 15.1 Properties of the steel pipe, nominal size 25 mm
U.S. Nominal Schedule Wall Inside Surface Area Cross Section Mass Working Pressure
Nominal size, mm Thickness Diameter ASTM A53 B to 200C
size, in. t, mm d, mm Outside, Inside, Metal Flow Pipe, Water, Mfr. Joint KPa
m2/m m2/m Area, Area, kg/m kg/m Proces Type (gage)
mm2 mm2
1 25 40 ST 3.38 26.64 0.105 0.084 318.6 557.6 2.50 0.558 CW T 1558
3.2 Type of material
We choose stainless steel AISI-316. It has a relative high heat conduction coefficient (k = 16
W/mK) and it can resist an aggressive environment.
3.3 The evaporator outside surface area
We want to specify the outside surface area. For this we need the outside heat convection
coefficient and the logarithmic temperature difference (see formula (15.1), (15.2) and (15.3)).
This coefficient depends in turn of the amount of evaporator pipes, which is related to the
outside surface area. We have to iterate.
112
Q
A= (15.1)
UTlm
1 1 d d d 1
= + o + o ln o (15.2)
U ho d i hi 2 d i k
T T2
Tlm = 1 (15.3)
T
ln 1
T2
We cannot solve these formulas because we have too much unknown variables. But with an
extra formula (15.4), which expresses the balance between the temperature difference of the
brine at entrance and exit and the total heat load, we are able to solve them. We first estimate
the overall heat transport coefficient. Then we estimate Tx and Ty (we already now their
difference (see formula 15.4)). This gives us the logarithmic temperature difference. With the
overall heat transport coefficient (U) and the logarithmic temperature difference formula
(15.1) we find the outside surface area. With this outside surface area we yield the outside
convection coefficient. Now formula (15.1) gives a new logarithmic temperature difference.
With this new value and with formula (15.4) we find the correct Tx and Ty.
Q
Tx Ty = (15.4)
c p vS
Fig. 15.3 Temperature change in and outside the evaporator
3.3.1 Calculation of the evaporator outside surface area
3.3.1.1 Overall heat transfer coefficient calculation
In formula (15.2), both hi and ho are unknown. Stainless steel gives a k = 16 W/mK.
113
Fig. 15.4 Convective heat transfer coefficient in of NH3 in function of

the vapour quality (G = 50 kg/m2s).
Figure 15.4 shows two different mean heat fluxes (experimental results already published by
Zrcher et al.). It can be shown that there is an important influence of heat flux on the local
heat transfer coefficient, and the typical peak is clearly appearing at a vapour quality of 75 %.
Note that at this value, the experimental heat transfer coefficient is scattered, which
demonstrates the sensitivity of the wetted angle in the region close to the annular flow pattern.
Nevertheless we expect a heat flux of 1.5 kW/m2, this figure is of importance because it
shows that the convective heat transfer on the ammonia side is much higher than that on the
brine side. The squares in the upper left side of the figure depict stratified and stratified wavy.
We assume hi = 2000 W/m2K.
Because the calculations for ho require the total surface of contact with the brine, which can
only be found knowing ho. We predict a ho = 300 W/m2K.
Now all parameters are known in formula (15.2), giving U = 238.31 W/m2K.
3.3.1.2 Brine entrance and exit temperature estimation
Inserting = 1190 kg/m3, cv = 3.29 103, Q = 660 kW, v = 0.15 m/s, S = 1.70 m2 (this is the
height of the liquid level multiplied by halve the width of the tank) in formula (15.4) gives:
660000
Tx T y = = 0.663K (15.5)
1190 3.29 10 3 0.15 1.70
Now we predict Tx: Tx = -19.34C, and using formula (15.5): Ty =-20C.
114
3.3.1.3 Estimation of the total evaporator area and amount of pipes
Formula (15.1) gives A = 520.13 m2. Dividing this surface by sixteen, for having four heat
exchangers each four meters long, we get the amount of pipes necessary per heat exchanger,
being 309.81( 324 = 182). We take 324 pipes giving a total contact area A = 543.95 m2.
Fig. 15.5 A cross section overview of the brine tank equipped with the 324 pipes.
The 324 pipes fit in the evaporator side of the tank. In vertical direction the pipe diameters are
80 mm separated. This leaves a flow space of 46.6 mm between two pipes. In the horizontal
direction the pipe diameters are 60 mm separated leaving a flow space of 12.27 mm between
the two pipes. All the pipes are submerged. On the left and right side there is an opening of 45
mm each.
3.3.1.4 Heat convection coefficient calculation on the brine side.
With the amount and dimension of the pipes we can calculate the convection coefficient on
the brine side. For this we need the free flow area, the heat-exchanging circumference and the
thermal properties of the brine.
The flow area of the brine:
d o2
Ac = hb n (15.6)
4
where
h = the height of the brine level, m
b = the width of the space through which the brine flows over the evaporator pipes, m
do = the outside diameter of the evaporator pipes, m
n = number of pipes
Formula (15.6) gives Ac = 1.67 m2.
The heat-exchanging circumference:
115
Pwo = n d o = 1.89m
The thermal diameter:
4 Ac
Dt = = 3.54m
Pwo
The thermal properties are listed in table 15.2.
Table 15.2 Brine properties at 20C
, kg/m3 1190
, Pa.s 7*10-3
k, W/mK 0.42
Pr 54.67
The Reynolds number:
vDt
Re d = = 90375.69

Now we can determine the Nusselt number using the Gnielinski correlation:
f
(Re d 1000) Pr
di 3
2
NuT = 2 1 + = 2663.75
f 12 2 3
1 + 12.7( ) (Pr 1)
L
2
f = (1.58 ln(Re d ) 3.28)

2
= 4.60 10 3
NuT k W
h= = 315.67 2
DT m K
This new convection coefficient doesnt differ a lot from the one estimated above. Even
because it is bigger we are sure the same amount of pipes is sufficient and correct.
3.3.1.5 Brine entrance and exit temperature.
Now we now the convection coefficient we can find the exact brine entrance and exit
temperature in the tank. Formula (15.1) gives us a new logarithmic temperature difference:
Tlm = 4.89 K
Combining formulas (15.1) and (15.3) gives:
116
(Tx (248) ) (T y (248) ) 660 10 3

= = 4.89 K
Tx (248) 248.09 543.96
ln
T (248)
y
T T = 0.663K
x y
Tx = 19.76C
T y = 20.42C
4 Intensive study of the pressure drop and heat flux in the evaporator
For the calculation of the evaporator outside area we assumed an internal heat flux hi = 2000
W/m2K. Because this internal heat transfer resistance is much smaller than the external one it
has little influence in the final result. Nevertheless, we want to give a detailed description of
the processes that occur in the evaporator. Our purpose is to show the pressure and internal
heat convection coefficient in function of the length inside the evaporator tubes. We calculate
the pressure losses with the Friedel correlation and Chisholms method (see chapter 22) for
two-phase flows and the convection coefficient with a correlation proposed by Ohadi (see
4.1).
The two concepts are linked because the convection coefficient depends on the pressure. And
the pressure decreases with the losses related to the vapour quality, which in turn depends on
the heat flux through the pipe walls. We wrote a maple (a mathematical program) file in
which we calculate pressure drops and local convection coefficients. We explain now the
method and formulas.
4.1 Correlation for two-phase flow heat transfer of ammonia in plain tunes
Ohadi bases on the widely referred Chart correlation of Shah. Shah's Chart correlation
employs four dimensionless parameters defined as
hTp
= (15.7)
hf
0.5
1 x
0.8
g
Co = (15.8)
x
f
q"
Bo = (15.9)
Gh fg
G2
Fr f = (15.10)
2f gD
where is simply defined as the ratio of the heat transfer coefficient of the two-phase flow
(hTp) to that of the superficial liquid phase heat transfer (hf). The quantity hf is calculated by
the Dittus Boelter equation as
117
G (1 x )D
0.8
kf
h f = 0.023 Pr f0.4 (15.11)
f D
where G is the mean mass flux on the cross-section of the tube, D is the internal diameter of
the tube, f is the dynamic viscosity of the liquid, x is the vapour quality, kf
is the thermal conductivity of the liquid, and Prf is the Prandtl number of the liquid. The ratio
is dependent on the other three dimensionless parameters, Co, Bo, and Frf .
The parameter Co in Equation (15.8) is defined as the convection number, with pg and pf as
densities of the gas phase and the liquid phase respectively. The larger the
parameter Co, the smaller the convective boiling influence. Bo is defined as the boiling
number, with q" as the heat flux, hfg as the latent heat, and G as the mean mass flux. Frf is
defined as the superficial Froude number. By considering the influence of the tube orientation,
an additional parameter N was introduced to modify the convection number:
Co for vertical tubes and horizontal tubes with Frf 0.04

N =
(15.12)
0.3
0.38 Fr f Co for horizontal tubes with Frf < 0.04
The parameter is defined for three specified regions of N i.e. N > 1.0, 0.1 < N 1.0 and N
0.1.
For N > 1.0
230 Bo 0.5 for Bo 0.3 10-4

(15.13)
nb =
1 + 46 Bo 0.5
for Bo < 0.3 10-4
1 .8
cb = (15.14)
N 0.8
and is taken as the higher of the values of nb and cb:
= max{ nb , cb } (15.15)
For 0.1 < N 1.0:
bs = FBo 0.5 exp(2.74 N 0.1 ) (15.16)
with then selected as
= max{ nb , cb } (15.17)
where cb is given by equation (15.14).

For N 0.1:
118
bs = FBo 0.5 exp(2.74 N 0.15 ) (15.18)
with then selected as
= max{ nb , cb } (15.19)
where again cb is given by equation (15.14).

The constant F in equations (15.16) and (15.18) is given as
14.7 for Bo 11 10
-4
(15.20)
F =
15.43
for Bo < 11 10-4
4.2 Short description of the algorithm
We divide each pipe in 20 parts. In each part we consider the vapour quality constant. We
calculate the convection coefficient (hTp) with the method given above. With this coefficient
we find the heat transfer through the pipe walls (formulas (15.1) and (15.2)). We assume a
constant brine temperature. The ammoniacs temperature is pressure dependent. So we also
calculate the pressure drop with the Friedel correlation to find the saturation temperature in
the next pipe part. Vapour quantity increases due to the heat transfer.
Q
m& vapour = xm& tot = (15.21)
hlv
with flv = latent heat of evaporation.
We do this for one pipe. At the end of the pipe we use Chisholms method to find the pressure
loss over the bend. Static pressure drops are neglected. Now the next pipe can be considered.
Figures 15.6 and 15.7 give respectively the pressure and the two-phase heat convection
coefficient.
119
p (bar)
l (m)
Fig. 15.6 The progress of the local evaporator pressure.
hTp (W/m2K)
l (m)
Fig. 15.7 The progress of the local two-phase heat transfer coefficient (hTp ) in the
evaporator.
120
4.3 Conclusions
Figure 15.6 shows clearly the 15.17 pressure drops over the returning bends. The total
pressure drop over the evaporator is 0.157 bar. This value coincides very well to the one
calculated in chapter 22.
Figure 15.7 shows a remarkable change in the heat convection coefficient at x = 0.0148. We
can relate this change to another flow pattern. However, we calculated the transition between
stratified and stratified-wavy in chapter 22 and found an x = 0.05.
We decide that the Ohadi-model inaccurate is for low vapour qualities.
At higher vapour qualities the convection coefficient shares the same interval as the one
calculated by Thome et al. in figure 15.4.
121
1 Introduction
We find it necessary to dispose of a brine container. With such a basin we dont need to buy
small amounts of brine to oppose the ever-existing losses. We also can store the brine in the
container when the tank has to be repaired for example.
2 Selection
We select a vertical cylindrical container with a conical base supported by four legs.
We list the commercial models of the manufacturer Secovisa in table 16.1.
Table 16.1 Dimensions of the basin with conical basis in function of the capacity
Capacity, litres Diameter, m Cylinder height, m Total height, m

2.000 1.250 1.500 2.400
2.500 1.225 2.000 2.900
3.000 1.340 2.000 2.920
5.000 1.500 2.700 3.640
7.500 1.735 3.000 3.970
10.000 1.835 3.600 4785
12.500 1.950 4.000 5.340
15.000 2.080 4.200 5.560
20.000 2.320 4.500 5.890
25.000 2.485 4.900 6.310
30.000 2.640 5.200 6.635
3 Solution
After going through the different possibilities and comparing with the need of the installation
we decide to take the container with following characteristics:
Capacity: V = 25000 dm3

Diameter: D = 2.485 m
Cylinder height: h = 4.900 m
Total height: H = 6.310 m
It will be made in stainless steel AISI-316, general indicated for products containing corrosive
characteristics.
122
Part 4
123
1 Introduction
The screw compressor is the active element of the refrigerating circuit. It has two functions:
first, to reduce the pressure in the evaporator until the liquid refrigerant evaporates to the
current temperature and maintain the pressure by drawing off the vapour produced through
evaporation of the liquid refrigerant, and second, to compress the vapour by raising its
temperature and pressure to the point at which the vapour can be condensed at the normal
temperature of the condensing media.
To estimate the correct size of the compressors the size of the evaporators and condensers
should be previously evaluated. No general rule for dividing the refrigerating compressor
capacity can be established. However it is obvious that cost criteria should be followed. To
help in deciding on the process some general factors must be borne in mind: compressors
should be interchangeable and preferably of a similar type so the stock of spare parts can be
reduced; the refrigerating plant should be provided with standby machines which are not used
for any other purpose than replacing machines that are out of order; the compressors are
equipped with capacity regulators and with automatic regulation. In spite of all these
requirements costs should be kept as low as possible.
Compressors, controls, condensers and their ancillaries are grouped in the machine room. The
location is carefully decided on to minimize pipe runs to all sections of the cold store and so
as not to impede any future extension of the refrigerating plant. Care must be taken not to
exceed the floor load limit. The position of the compressor in the machine room allows easy
access for maintenance, repairs and possible replacement. The environment should be cool
and dry in any weather conditions and be well ventilated. The machines must remain clean
and be safe from damage
When evaluating the total heat load of the refrigerating plant the operating conditions should
be such that they lead to peak heat loads. This means that if compressor capacity is higher
than the heat load on the evaporator the system is not in equilibrium and will not perform
satisfactorily.
To adjust the functioning of the compressor to the actual heat load it has to operate unloaded
or under a reduced capacity; this operating mode reduces compressor efficiency. Some
recommendations should be followed to avoid this situation. First, it is preferable to use small
compressors at full capacity rather than a large one at partial capacity. Second, especially for
centralized installations, compressors must be of a satisfactory size and work at full capacity
for the expected heat loads.
Modifying the suction volume through the action of an oil-pressure circuit controls screw
compressor capacity. This control is continuous from 100 to 10 percent, and the power
requirements are proportional to the refrigeration capacity necessary in each situation. As
well, there is no heating of the refrigerant fluid.
124
Because the refrigerating plant is operated at low temperature it is advisable and more
economical to use a two-stage compression system, especially in this high-capacity plant,
despite the complexity in their installation and operation, as energy consumption is much less.
Helical screw compressors are the choice for most industrial refrigeration systems. All
helicoidally screw compressors have a constant volume (displacement) design. The volume
index (Vi) refers to the internal volume ratio of the compressor. There are three types of screw
compressors:
Fixed Vi with slide valve

Variable Vi with slide valve and slide stop
Fixed Vi with bypass ports in lieu of slide valve
A variable Vi compressor makes compressor selection simpler because it can vary its volume
index from 2.0 to 5.0; thus, it can automatically match the internal pressure ratio within the
compressor with the external pressure ratio.
Figure 17.1 shows a typical capacity and volume index control system in which two four-way
control valves take their signals from a computer controller. One four-way valve controls the
capacity by positioning the slide valve in accordance with the load, and the other positions the
slide stop to adjust the compressor internal pressure ratio to match the system suction and
discharge pressure. The slide valve works the same as that on fixed Vi compressors. Adjusting
the slide stop on the discharge end of the compressor varies volume index.
Fig. 17.1 Typical positioning system for slide valve and slide stop
for variable Vi screw compressor
125
Variable volume ratio.

While operating, some twin-screw compressors adjust the volume ratio of the compressor to
the most efficient ratio for whatever pressures are encountered. In fixed volume ratio
compressors, the motion of the slide valve toward the inlet end of the machine is stopped
when it comes in contact with the rotor housing in that area. In the most common of the
variable volume ratio machines, this portion of the rotor housing has been replaced with a
second slide, the movable slide stop, which can be actuated to different locations in the slide
valve bore (Figure 17.2). By moving the slides back and forth, the radial discharge port can be
relocated during operation to match the compressor volume ratio to the optimum. This added
flexibility allows operation at different suction and discharge pressure while still maintaining
maximum efficiency. The comparative efficiencies of fixed and variable volume ratio screw
compressors are shown in Figure 17.3 for full load operation on ammonia refrigerant. The
figure shows that a variable volume ratio compressor efficiency
curve encompasses the peak efficiencies of compressors with fixed volume ratio over a wide
range of pressure ratio. Following are other secondary effects of a variable volume ratio:
Less oil foam in oil separator (no over compression)

Less oil carried over into the refrigeration system (because of less oil foam in oil
separator)
Extended bearing life; minimized load on bearings
Extended efficient operating range with economizer discharge port corrected for flash
gas from economizer, as well as gas from suction
Less noise
Lower discharge temperatures and oil cooler heat rejection
Fig. 17.2 View of fixed and variable volume ratio (Vi) slide valves from above
126
Fig. 17.3 Twin-screw compressor efficiency curves
2 Compressor Selection
Because selecting the right compressors out of a catalogue is quasi impossible, we asked a
sales engineer of York Refrigeration to calculate the two compressors we need for the plant.
York developed a computation program called comp1 that makes it possible to find the
most efficient compressor choice.
As we remember the selected cycle is a two-stage cycle with open intercooler. Out of
calculations we made in the chapter 2 we know that that the intermediate temperature varies
depending on which installation operates. Optimal intermediate temperature for blast freezing
room operation was 7.5C, for the brine tunnel was this 4C.
Both compressors are calculated based on the cooling load and volume flow rate of the blast-
freezing tunnel. Afterwards they checked if these compressors could fulfil the requirements
of the brine tunnel.
The low-stage compressor: SAB 233E.
Technical Data:
o Swept volume: 2258 m3/h
o Dimensions: 4266 x 1775 x 2445 mm
o Weight: 4536 kg (excluding electric motor)
o Sound pressure level: 86 dB(A)
The high-stage compressor: SAB 163 H male drive.
Technical Data:
o Swept volume: 635 m3/h
o Dimensions: 2630 x 1470 x 1725 mm
o Weight: 1800 kg (excluding electric motor)
o Sound pressure level: 86 dB(A)
For further information about the compressors: see added catalogue on the disc or
http://www.sabroe.com
For the blast freezer the low-stage compressor is working between 45C evaporator
temperature and 7.5C intermediate temperature. The high-stage compressor is working
between the intermediate temperature and 35C condenser temperature.
127
Once these compressors selected they checked if operating conditions of the brine tunnel can
be fulfilled. We remind you on these conditions:
Evaporator temperature: -25C

Intermediate temperature: 4C
Condenser temperature: 35C
As we can see in the calculation results for the brine tunnel, the high-stage compressor is
working at 100% at cooling load 660 kW. This is reached with lowest intermediate
temperature possible: 7.3C. This is not the intermediate temperature we mentioned before
by optimal COP.
There are two possibilities now. Or we select a bigger high-stage compressor, which lowers
the intermediate temperature to 4C. But which will also decrease the efficiency of the
compressor during blast freezer operation. Or we accept an intermediate temperature of 7.3C
that will cause a decrease of the COP. However this loss will be negligible small in
comparison with the bigger energy use due to a more powerful compressor.
Thus, we keep the two compressors that York selected before.
The simulations made with comp1, Yorks computation program, are done for following
operation conditions:
Blast freezing room:
o Total cooling capacity: 302.7 kW
o Swept volume: 2108.9 m3/h
o Evaporator temperature: -45C
o Intermediate temperature: -7.4C
o Condenser temperature: 35C
Brine tunnel
o Total cooling capacity: 660.4 kW
o Swept volume: 1742.1 m3/h
o Evaporator temperature: -25C
o Intermediate temperature: 7.3C
o Condenser temperature: 35C
The calculation results can be found on the disc.
3 Oil management
3.1 Introduction
Compressor lubricants should be those recommended by the compressor manufacturer. They

should also be compatible with the refrigerants and have a sufficiently low pour point to
permit recovery from the low side of low pour point to permit recovery from the low side of
the system at all expected temperature levels. HFC refrigerants will generally use a polyester
synthetic lubricant. For ammonia refrigerants, a hydrotreated paraffinic oil is recommended.
All compressors should have coalescing oil separators sufficient to limit oil carryover to no
more than 5 ppm so oil will not concentrate in the evaporators. Oil in the evaporator prevents
proper heat transfer and impedes oil recovery.
128
Most oil recovery is accomplished directly by the refrigerant flow through the evaporator coil
and suction line. The refrigerant tends to sweep the oil along. In flooded ammonia systems,
oil is recovered through the use of oil pots that are located below the refrigerant level. The
heavier oil drains into the oil pots and may be recovered either manually or automatically.
3.2 Selection
The oil is recommended by the compressor manufacturer and more details can be found on
the calculation document of York added with the disc.
129
1 Introduction
Condensers are basically heat exchangers in which the refrigerant vapour is cooled and
liquefied after compression. The evaporator heat load plus the heat of compression are
released into the atmosphere via the condenser by means of a fluid (normally water) or by air:
condensers can be cooled by water or air.
Comparing water-and air-cooled condensers, it can be said that they are similar in initial cost
and also in operating costs, except when the former work in an open-circuit system supplied
with cheap water and have no environmental restrictions. Since there is no source of cheap
supply water in the vicinity we will obtain for an air-cooled condenser.
The air-cooled condensers operate under forced convection, using either axial flow fans,
which are noisy, to move air at high speeds over the finned tube. The usual air speed is
between 2.5 and 5 m/s. However, as power consumption increases with the square of air
velocity, 3 m/s is considered a reasonable speed.
Air-cooled condensers have some advantages over water-cooled condensers. Water
consumption is nil, which is vital where water is in short supply or not available. Their
efficiency is indifferent to moisture content in the air, so they are appropriate for humid
climates. They are relatively simple to install and require little maintenance because cleaning
is simple and rapid.
However, they present some drawbacks, the most relevant being the high condensation
temperature, which is usually 1520C higher than the entering air temperature. The optimum
size of the condenser is computed on a condensation temperature of about 4245C. As the
heat transfer coefficients are low large exchange surfaces are necessary, with obvious
consequences in size and cost of equipment. Power consumption is another drawback, though
it can be considered comparable to the consumption of cooling towers or evaporative
condensers, as they are also equipped with fans and pumps.
Good air circulation is essential for satisfactory operation, therefore confined spaces must be
avoided. For this reason the general trend is to mount condensers at a high level, far from the
compressor, and in a cool location that is clean, dry and well ventilated, protected from the
heat of the sun to prevent air recirculation and sited to take advantage of prevailing winds,
particularly during periods of maximum load. They should be close to the evaporator, and
whenever possible slightly above rather than below it.
The condensers should be readily accessible for cleaning and maintenance, installed on anti-
vibration mountings to control noise and be safe from damage. Their efficiency depends on
the cleanliness air side of the exchanger surface. Therefore periodical cleaning is necessary.
Air condensers are cleaned by washing the tube bank with a hose and stopping the fans no
more than once a month for a very dirty atmosphere.
130
Furred condensers raise the condensation pressure, reducing the performance of the
installation. If the condenser becomes blocked there is a risk of breaking the refrigerating
circuit.
Efficiency also depends on the condition of the exchange surface on the refrigerant side, so
precautions should be taken to prevent oil accumulation. Purging non-condensable gases
which may accumulate in the circuit will improve the efficiency of the condenser. To ease
systematic maintenance it is advisable to install more than one condensing unit or, if a single
condenser is used, this should be divided into separate circuits. This will also help in case of
breakdown.
Condensation pressure is the parameter to control for efficient and safe operation in the high-
pressure circuit. This pressure should not be excessively high, first for safety considerations
and second because thermodynamically the refrigerating cycle operates less efficiently as the
pressure rises, influencing energy consumption and raising operating costs.
The condenser is protected against high-pressure failure by the high-pressure pressostat, a

device commanding compressor cycling which cycles-off whenever the pressure rises above
an established level.
It may also be necessary to regulate the condensation pressure to avoid it becoming too low,
as this does not maintain a high enough pressure differential across the refrigerant expansion
valve and the evaporator is fed with an insufficient refrigerant flow. To maintain a high
condensation pressure, and corresponding high temperature, it is necessary to control the
condenser capacity when the ambient temperature is low. This control is achieved either by
reducing the flow of air through the condenser or by diminishing the effective heat exchange
surface area or condensing area.
The condensing surface is modulated by retaining liquid refrigerant in the lower part of the
condenser, using a valve known as a pressure regulator.
For air-cooled condensers the air flow control system may cycle-off some of the fans, when
there are more than one, or the air flow can be reduced with pressostatic shutters activated by
the condenser pressure. When the pressure is low the shutters will close in relation to the
decrease, reducing the flow of air circulating through the condenser.
Thus it is possible to cut air circulation and so reduce the heat exchange and consequently
raise the condensation pressure, but it is not recommended because of surface scaling, unless
it is done for a long period in cold weather before condenser cleaning.
For vapour desuperheating and liquid subcooling (i.e., inlet and outlet regions of the coil) a
cross-counterflow arrangement of refrigerant and air produces the best performance. In the
condensing portion (i.e., at the center of the coil) the latent heat dissipates as a result of the
refrigerant pressure drop. That is, the refrigerant enters the condenser as superheated vapour
and cools by singe-phase convection to saturation temperature, after which condensation
starts. If the coils surface temperature is much lower than the saturation temperature,
condensation occurs close to the gas entry area of the condenser coil. Eventually, in the last
part of the condenser, all the refrigerant vapour condenses to liquid. Because of the change of
state, the velocity and volume of the refrigerant, as a gas, drops substantially as it becomes a
liquid. In general, most designs provide an adequate heat transfer surface area and
131
complementary pressure drop to cause the refrigerant to subcool an average of 1 to 3 K before

it leaves the condenser coil.
2 Selection
The condensers are selected on the basis of total heat rejection at maximum load.
According to calculations made with coolpack, the simulation program of the technical
university of Denmark, the total heat rejection at maximum load differs from this
corresponding to calculations made by York Refrigeration. York calculated a total heat
rejection at maximum load of 503 kW for the brine tunnel and 920 kW for the blast freezing
chamber. These values are much higher than those calculated with coolpack. We will use
the values given by York for the condenser selection to make sure its capacity is sufficient.
The condensers we choose are made by Helpman. This manufacturer of heat exchangers is
located in the Netherlands and provides condensers with a large range of capacity (9 1080
kW). Helpman is specialized in NH3-air cooled condensers. The catalogue is added on the
disc.
The basic capacity ratings (pages 4, 5, 8 and 9) are for NH3 at a td of 15 K and ambient
temperatures up to 25 C.
td = difference between condensing- and ambient temperature.
Capacities are directly proportional with other temperature differences between 10 and 20 K.
Assume that the ambient temperature is 25C, that means a td of 10K.
It follows that the basic capacity ratings have to be modified:
Brine tunnel: 15K/10K x 503 kW = 754.5 kW
Blast freezer: 15K/10K x 920 kW = 1380 kW
With the capacities given above match following condensers of the catalogue.
HTCN 090.254-680
Characteristics:
o Capacity: 762.2 kW
o Number of separated coils: 2
o Number of fans per coil: 5
o Number of tube rows in air direction: 4
o Fan speed: 680 rpm
o Air flow rate: 197600 m3/h
o Sound level: 62 dB(A)
o Internal volume: 154 dm3
o Surface: 2460 m2
o Weight: 2137 kg
o Dimensions: see catalogue
HTCN 090.243-850
Characteristics:
o Capacity: 621.4 kW
o Number of separated coils: 2
o Number of fans per coil: 4
o Number of tube rows in air direction: 3
132
o Fan speed: 850 rpm

o Air flow rate: 227800 m3/h
o Sound level: 66 dB(A)
o Internal volume: 94 dm3
o Surface: 1476 m2
o Weight: 2124 kg
o Dimensions: see catalogue
When the blast freezing chamber is in operation then the capacity of the first condenser
(HTCN 090.254-680) is sufficient to remove the heat. When the brine tunnel works then both
condensers will run. The total capacity of both condensers is 1383.6 kW.
For further information about the condensers: see added catalogue on the disc or
http://www.helpman.nl
133
Chapter 19: Piping
Chapter 19: Piping
1 Introduction
In this chapter we will choose the recommended material and size for all the pipes in our
system. We follow the recommendations given by ASHRAE and change sizes properly when
pressure drops are too significant.
2 Recommended Material
Because copper and copper-bearing materials are attacked by ammonia, they are not used in
ammonia piping systems. Steel piping, fittings, and valves of the proper pressure rating are
suitable for ammonia gas and liquid.
Ammonia piping should conform to ASME Standard B31.5, Refrigeration Piping, and IIAR
Standard 2, which states the following:
1. Liquid lines 40 mm and smaller shall be not less than Schedule

80 carbon steel pipe.
2. Liquid lines 50 through 150 mm shall be not less than Schedule
3. Liquid lines 200 through 300 mm shall be not less than Schedule
4. Vapor lines 150 mm and smaller shall be not less than Schedule
5. Vapor lines 200 through 300 mm shall be not less than Schedule
6. Vapor lines 350 mm and larger shall be not less than Schedule 10
carbon steel pipe.
7. All threaded pipe shall be Schedule 80.
8. Carbon steel pipe shall be ASTM Standard A 53 Grade A or B, Type E (electric
resistance welded) or Type S (seamless); or ASTM Standard A 106 (seamless), except
where temperature-pressure criteria mandate a higher specification material. Standard
A 53 Type F is not permitted for ammonia piping.
3 Pipe Location
Piping should be at least 2.3 m above the floor. Locate pipes carefully in relation to other
piping and structural members, especially when the lines are to be insulated. The distance
between insulated lines should be at least three times the thickness of the insulation for
screwed fittings, and four times for flange fittings. The space between the pipe and adjacent
surfaces should be three-fourths of these amounts. Hangers located close to the vertical risers
to and from compressors keep the piping weight off the compressor. Pipe hangers should be
placed no more than 2.5 to 3 m apart and within 0.6 m of a change in direction of the piping.
Hangers should be designed to bear on the outside of insulated lines. Sheet metal sleeves on
134
Chapter 19: Piping
the lower half of the insulation are usually sufficient. Where piping penetrates a wall, a sleeve
should be installed; and where the pipe penetrating the wall is insulated, it must be adequately
sealed. Piping, to and from compressors and to other components, must provide for expansion
and contraction. Sufficient flange or union joints should be located in the piping that
components can be assembled easily during initial installation and also disassembled for
servicing.
4 Pipe Sizing
Table 19.1 presents practical suction line sizing data based on 0.005 K and 0.01 K differential
pressure drop equivalent per meter total equivalent length of pipe.
Table 19.1 Steel nominal size (mm) for different saturation suction temperatures (C).
Steel nominal line Saturation suction temperature, C

size, mm
-40 -30 -20
T = 0.005 K/m T = 0.01 K/m T = 0.005 K/m T = 0.01 K/m T = 0.005 K/m T = 0.01 K/m
p = 19.2 Pa/m p = 38.4 Pa/m p = 29.1 Pa/m p = 58.2 Pa/m p = 42.2 Pa/m p = 84.4 Pa/m
10 0.35 0.51 0.58 0.85 0.91 1.33
15 0.65 0.97 1.09 1.60 1.72 2.50
20 1.41 2.08 2.34 3.41 3.66 5.31
25 2.72 3.97 4.48 6.51 6.98 10.10
32 5.71 8.32 9.36 13.58 14.58 21.04
40 8.64 12.57 14.15 20.49 21.99 31.73
50 16.89 24.50 27.57 39.82 42.72 61.51
65 27.13 39.27 44.17 63.77 68.42 98.23
80 48.36 69.99 78.68 113.30 121.52 174.28
100 99.50 143.84 161.77 232.26 249.45 356.87
125 181.16 261.22 293.12 420.83 452.08 646.25
150 294.74 424.51 476.47 683.18 733.59 1046.77
200 609.20 874.62 981.85 1402.03 1506.11 2149.60
250 1107.64 1589.51 1782.31 2545.46 2731.90 3895.57
300 1777.96 2550.49 2859.98 4081.54 4378.87 6237.23
Note: Capacities are in kilowatts of refrigeration resulting in line friction loss per unit equivalent pipe length (p in Pa/m), with
corresponding change in saturation temperature per unit length (T in K/m)
Table 19.2 lists data for sizing suction and discharge lines at 0.02 K differential pressure drop
equivalent per meter equivalent length of pipe, and for sizing liquid lines at 0.5 m/s.
135
Chapter 19: Piping
Table 19.2 Steel nominal line size (mm) in function of the type of line and temperature
Steel Suction lines (T = 0.02 K/m) Discharge lines Steel nominal Liquid lines
nominal T=0.02K/m, p=84.0Pa/m line size, mm
line Saturated Suction Temperature, C Saturated suction temp, C Velocity=0.5 p=450.0
size, m/s Pa/m
mm -40 -30 -20 -5 +5 -40 -20 +5
p=76,9 p=116.3 p=168.8 p=276,6 p=370,5
10 0.8 1.2 1.9 3.5 4.9 8.0 8.3 8.5 10 3.9 63.8
15 1.4 2.3 3.6 6.5 9.1 14.9 15.3 15.7 15 63.2 118.4
20 3.0 4.9 7.7 13.7 19.3 31.4 32.3 33.2 20 110.9 250.2
25 5.8 9.4 14.6 25.9 36.4 59.4 61.0 62.6 25 179.4 473.4
32 12.1 19.6 30.2 53.7 75.4 122.7 126.0 129.4 32 311.0 978.0
40 18.2 29.5 45.5 80.6 113.3 184.4 189.4 194.5 40 423.4 1469.4
50 35.4 57.2 88.1 155.7 218.6 355.2 364.9 374.7 50 697.8 2840.5
65 56.7 91.6 140.6 248.6 348.9 565.9 581.4 597.0 80 1536.3 8008.8
80 101.0 162.4 249.0 439.8 616.9 1001.9 1029.3 1056.9 - - -
100 206.9 332.6 509.2 897.8 1258.6 2042.2 2098.2 2154.3 - - -

125 375.2 601.8 902.6 1622.0 2271.4 3682.1 3783.0 3884.2 - - -
150 608.7 975.6 1491.4 2625.4 3672.5 5954.2 6117.4 6281.0 - - -
200 1252.3 2003.3 3056.0 5382.5 7530.4 12195.3 12529.7 12864.8 - - -

250 2271.0 3625.9 5539.9 9733.7 136119.6 22028.2 22632.2 23237.5 - - -
300 3640.5 5813.5 8873.4 15568.9 21787.1 35239.7 36206.0 37174.3 - - -
Notes : 4. Values in the table are based on 30C condensing temp. Multiply
1. Table capacities are in kilowatts or refrigeration, table capacities by the following factors for other condensing temp:
p=pressure drop due to line friction, Pa/m
T=corresponding change in saturation temperature,K/m Condensing temp, Suction lines Discharge lines
2. Line capacities for other saturation temp T and equivalent C
lengths Le 20 1.04 0.86
Line capacity = Table capacity 30 1.00 1.00
3. Saturation temp T for other capacities and equivalent lenghts Le 40 0.96 1.24
50 0.91 1.53
5. Liquid line capacities are based on 5C suction.
Table 19.3 presents line-sizing information for pumped liquid lines.
Table 19.3 Nominal steel size in function of the liquid overfeed ratio for pumped liquid
lines.
Nominal size, mm Pumped liquid overfeed ratio

3:1 4:1 5:1
40 513 387 308
50 1175 879 703
65 1875 1407 1125
80 2700 2026 1620
100 4800 3600 2880
125 - - -
150 - - -
200 - - -
Source: While (1977).
We choose a nominal size using table 1, 2, 3; and the steel type according the norm described
in the material selecting part.
4.1 Suction line (low pressure)
Distance: 7 m
Accessories: 1 stop valve, 1 check valve, 1 filter
136
Chapter 19: Piping
Fittings: six 90 standard elbows

Refrigeration capacity: 660 kW
Evaporation temperature: -25C
nominal size: 150 mm, schedule 40 ST
Schedule 40 carbon steel
4.2 Discharge line (low pressure)
Distance: 7 m
Accessories: 1 stop valve, 1 check valve
4.3 Suction line (high pressure)
Distance: 7 m
Accessories: 1 stop valve, 1 check valve, 1 filter
4.4 Discharge line (high pressure)
Distance: 7 m
Accessories: 1 stop valve, 1 check valve, 1 solenoid valve
Condensing temperature: 35C
4.5 Liquid line between condenser and high-pressure receiver
Distance: 7 m
Accessories: 1 stop valve
137
Chapter 19: Piping
4.6 Liquid line between high-pressure receiver and intercooler
Distance: 7 m
Accessories: 1 stop valve, 1 filter, 1 expansion valve, 1 solenoid valve
4.7 Liquid line between intercooler and liquid separator
Distance: 7 m
Accessories: 1 stop valve, 1 filter, 1 solenoid valve, 1 expansion valve, 1 check valve
4.8 Pumped liquid overfeed lines (starting from the liquid separator)
Distance: 38.5 m
Accessories: 2 stop valves, 1 filter, 1 check valve
Fittings: eight 90 standard elbows, two 45 long radius elbows, one tee branch
Evaporating temperature: -25C
Pumped liquid overfeed ratio: 4
4.9 Pumped liquid overfeed lines (starting from the evaporator)
Distance: 38.5 m
Accessories: 1 stop valve, 1 check valve, 1 solenoid valve
Fittings: eight 90 standard elbows, two 45 long radius elbows, one tee branch
Evaporating temperature: -25C
Pumped liquid overfeed ratio: 4
138
1 Introduction
Theoretically it would be sufficient having a compressor, condenser, evaporator and an

expansion valve connected by adequate tubing to form a refrigerating installation. However,
the installation realized would not accomplish the requirements for which it is designed.
All the instruments mentioned are no perfect devices, neither is the refrigerant used. This
leads to the use of complementary devices that are indispensable for correct functioning of the
installation.
Oil separator
High pressure receiver
Liquid separator or low-pressure receiver
Purge Unit
Liquid filters
Intercooler
Refrigerant liquid pumps:
o Low pressure pump
o High pressure pump
Check valves
2 Oil separator
2.1 Introduction
The lubrication of the refrigerating compressors is based on superabundance to make sure that
the minimum oil quantity reaches the lubrication point.
The oil is swept away with the compressed gases in extremely fine droplets and can also be
swept away in the form of vapour. It is in great interest to halt the sweep away at the
compressors outlet. An oil separator is added for several reasons:
Maintaining an adequate oil level in the reservoir of the compressor to ensure

abundant lubrication of the moving parts.
Suppressing unexpected accumulation of oil on certain points of the installation.
Maintaining the oil concentration in the refrigerant as low as possible. This results in
the absence of a thin oil layer on condenser and evaporator walls and with that the
extra thermal insulation of the oil layer disappears.
Indeed, if the lubricant is free to circulate through the whole installation it would accumulate
on the points with the lowest temperature. This is due to the increasing viscosity with
decreasing temperature which facilitates the deposit of oil in the evaporators where the
temperature is low. So the lubricant tends to occupy a part of the evaporators volume
reserved for the refrigerant which brings about a reduction of the utile part of the evaporator.
139
2.2 Selection
The selected oil separators of current installation form part of the compressors group of
Sabroe as selected before.
Corresponding to the low-stage compressor SAB 233 E next oil separator model is selected:
OHU 09135,I
Corresponding to the high-stage compressor SAB 163 H male drive following oil separator
is selected:
OHU 4123/2716D1
3 High-pressure receiver
3.1 Introduction
Industrial systems generally incorporate a central high-pressure refrigerant receiver, which

serves as the primary refrigerant storage location in the system. It handles the refrigerant
volume variations between the condenser and the systems low side during operation and
pumpdowns for repairs or defrost. Ideally, the receiver should be large enough to hold the
entire system charge, but this is not generally economical. An analysis of the system should
be made to determine the optimum receiver size. The receivers is equalized to the condenser
inlet and operates at the same pressure as the condenser.
3.2 Capacity calculation
The receiver volume depends on the design concept of the installation, and it is difficult to
give general rules. In each case a minimum liquid level has to be considered (approximately
20% of the total volume of the receiver). Furthermore there should be a certain fluid volume
that allows level fluctuations due to opening and closing of the separator level regulating
system. Theoretically this capacity is sufficient, nevertheless there will always exist
situations in which those fluctuations, normally linked to the separator, will have influence on
the receiver. When for some reason temporally a great amount of extra refrigerant returns to
the separator, its floating valve system stays closed during a certain period in that way that the
fluid can not leave the receiver. Extra available space must exist to store the liquid surplus
else too much refrigerant will be accumulated in the bottom part of the condenser.
It would be easy to calculate the required volume of the liquid receiver to contain the total
load of the installation, but that would lead to enormous receivers.
Because of this, an alternative procedure is proposed. Suppose that the liquid refrigerant
quantity stored in the separator is M. The volume of the high-pressure receiver must be
calculated in that way that it is adapted at quantity M. Because when the installation is started
up this liquid quantity will flow to the receiver. This situation will prevail till de evaporators
reach their normal fluid capacity. If the circulated refrigerant mass flow in the installation is
m kg/s and the refrigerant pumps are pumping nm kg/s, then (n-1)m kg/s will be used to fill
the evaporators and m kg/s of refrigerant vapour will flow through the low-stage compressor
to the open intercooler.
140
The filling time t of the evaporators will be:
M/(n-1)m seconds
The liquid quantity that has to be stored in the receiver is:
Mm/(n-1)m or M/(n-1) kg
With a minimal available liquid capacity of 20% in the receiver and another 20% free gas
volume, the required volume of the receiver is:
1.67M/(n-1) kg
When n = 4, this equations becomes:
0.555M kg or 0.555M/ m3
Note that in this calculation the value M is standing for al the evaporators that are not
working. Nevertheless, there exist evaporators that are constantly operating, even during
weekends. Theoretically, the load of such evaporators should be subtracted of the total
volume of refrigerant fluid in the system.
3.3 Selection
In the next paragraph we mention that the low pressure receiver available volume to store
ammonia is more or less 0.95 m3. The specific density of ammonia at 45 C is 694.44
kg/m3, at 25C it is 671.14 kg/m3. So that means that the maximum mass in the low-
pressure receiver is 660 kg.
We chose n = 4 before. It follows that the fluid quantity that has to be stored in the high-
pressure receiver is: 366 kg (807.21 lb) or 0.622 m3 (21.98 ft3), where the specific density at
35C is 588.24 kg/m3.
Suitable model: HR-24-136

This is a horizontal high-pressure receiver. Dimensions are given in the added catalogue page
on the disc.
4 Low-pressure receiver
4.1 Introduction
In a liquid ammonia recirculation system, a pump circulates the ammonia from a low-pressure
receiver to the evaporators. The low-pressure receiver is a shell for storing refrigerant at low
pressure and is used to supply evaporators with refrigerant, either by gravity or by a low-
pressure pump. It also takes the suction from the evaporators and separates the gas from the
liquid. Because the amount of liquid fed into the evaporator is usually several times the
amount that actually evaporates there, liquid is always present in the suction return to the low-
141
pressure receiver. Frequently, three times the evaporated amount is circulated through the
evaporator
Low-pressure receivers are also called liquid separators, suction traps, accumulators, liquid-
vapour separators, flash coolers, gas and liquid coolers, surge drums, knock-out drums, slop
tanks, or low side pressure vessels, depending on their function and the preference of the user.
The required liquid holdup volume and the allowable gas velocity determine the sizing of
low-pressure receivers. The volume must accommodate the fluctuations of liquid in the
evaporators and overfeed return lines as a result of load changes and defrost periods. It must
also handle the swelling and foaming of the liquid charge in the receiver, which is caused by
boiling during temperature increase or pressure reduction. At the same time, a liquid seal must
be maintained on the supply line for continuous circulation devices. A separating space must
be provided for gas velocity low enough to cause a minimum entrainment of liquid drops into
the suction outlet. Space limitations and design requirements result in a wide variety of
configurations (Miller 1971; Stoecker 1960; Lorentzen 1966; Niemeyer 1961; Scheiman
1963, 1964; Sonders and Brown 1934; Younger 1955).
4.2 Selection
M&M Refrigeration offers an elaborate range of vertical and horizontal recirculator bare
vessels. The factor on which the selection will be made is the refrigerating load of both
installations.
For the blast freezing room, operating at 40C with evaporator and thus low-pressure
receiver temperature at 45C, the refrigerating load is 292 kW.
For the brine tunnel, operating at 20C with evaporator and thus low-pressure receiver
temperature at 25C, the refrigerating load is 660 kW.
Both installations use the same recirculator. This means that it should be selected taking into
account both refrigerating loads at both temperatures.
We prefer to choose a horizontal low-pressure receiver which will be mounted at a certain

height above the ground close to the compressors.
The catalogue is based on tons of refrigeration. As we now is 1 ton of refrigeration equal to
3517.2 Watt.
Follows that:
660 kW = 187.65 tons of refrigeration
292 kW = 83.02 tons of refrigeration
Temperatures are given in F.
-25 C = -13 F
-45 C = -49 F
Suitable model: HRC-36-118.

We mention that the surge volume is 25.45 Ft3 (0.72 m3), this is the amount of area between
the normal operating level and high level cut-out.
The available volume to store refrigerant fluid is considered to be the halve of the total vessel
volume. That means approximately 0.95 m3. This number will be used to select the
appropriate high-pressure receiver for the system.
Dimensions are given in added catalogue page on the disc.
142
5 Purge Unit
5.1 Introduction
A non-condensable gas separator (purge unit) is useful in most plants, especially when suction
pressure is below atmospheric pressure. Purge units on ammonia systems are piped to carry
non-condensables (air) from the receiver and condenser to the purger, as shown in figure 20.1.
High-pressure liquid expands through a coil in the purge unit, providing a cold spot in the
purge drum. The suction from the coil should be taken to one of the low-temperature suction
mains. Ammonia vapour and non-condensable gas are drawn into the purge drum, and the
ammonia condenses on the cold surface. When the drum fills with air and other non-
condensables, a float valve within the purger opens and permits them to leave the drum and
pass into the open water bottle. Purge units are available for automatic operation.
Fig. 20.1 Purge unit and piping for non-condensable gas.
5.2 Selection
The purger is selected out of the catalogue of Hansen Technologies.

Suited model: AUTO PURGER AP
For more information about the purge unit see added catalogue on the disc or visit
http://www.hantech.com
6 Filters
6.1 Introduction
Despite all precautions and care token by all the material and device manufacturers, by
mounting and starting up the installation it would be exceptional that it stays uncontaminated
without efficient filters fitted into the system.
143
All refrigerants are very effective detergents or solvents able to dissolve the most persistent
dirt deposits in little time. Originally, mentioned deposits can be classified in 3 groups:
Foreign particles originated from machining

Foreign particles originated from mounting work
Deposit of impurities in the course of functioning.
Oxides, metallic salt and sludge tend to close off narrow orifices such as the orifices of
pressure equalizers of some automatic elements, the orifices of expansion valves, the orifices
of oil lead back of separators, with all resulting consequences possible. Therefore the circuit
should dispose of some efficient filters, independent of the existent, which are already
incorporated in the mechanical components and the automation devices (expansion valves,
magnetic valves, etc.).
Generally the filters are placed in the liquid tubing of the installation, what the use of suction
line filters not exclude. The filter should be mounted as close as possible to the expansion
valve. When the installation has a magnetic valve inserted in the liquid lines, the filter will be
placed on top of it.
6.2 Selection of filter size
Filters are selected out of the Danfoss catalogue, which is also added to this work. For more
information about the filters see http://www.Danfoss.com
The mesh aperture size of the filter must satisfy the requirements stated by the suppliers of the
equipment to be protected.
The following recommendations of aperture size apply in general to refrigeration installations:
6.2.1 All lines
First start up: 50

Insert should be removed after first 24 hours of operation.
6.2.2 Liquid lines
Ahead of the pumps: FIA DN50 mesh 38

After the pumps: FIA DN50 mesh 38
6.2.3 Protection of automatic regulation equipment
Ahead of the high-stage expansion valve: FIA DN25-40 mesh 72

Ahead of the low-stage expansion valve: FIA DN25-40 mesh 72
6.2.4 Suction lines
Ahead of the low-stage screw compressor: FIA DN150 mesh 72
144
Ahead of the high-stage screw compressor: FIA DN100 mesh 72
7 Intercooler
7.1 Introduction
An intercooler (subcooler/ desuperheater) is the intermediate vessel between the high stage
and low stage in a multistage system. One purpose of the intercooler is to cool the discharge
gas of the low-stage compressor to prevent overheating the high-stage compressor. This can
be done by bubbling the discharge gas from the low-stage compressor through a bath of liquid
refrigerant or by mixing liquid normally entering the intermediate vessel with the discharge
gas as it enters above the liquid level. The heat removed from the discharge gas is absorbed
by the evaporation of part of the liquid and eventually passes through the high-stage
compressor to the condenser. Disbursing the discharge gas below a level of liquid refrigerant
separates out any lubricant carryover from the low-stage compressor. If the volume of liquid
in the intercooler is to be used for other purposes, such as liquid makeup or feed to the low
stage, periodic lubricant removal is important. Another purpose of the intercooler is to lower
the temperature of the liquid being used in the system low stage. Lowering the refrigerant
temperature increases the refrigeration effect and reduces the low-stage compressors required
displacement, thus reducing its operating cost. Two types of intercoolers for two-stage
compression systems are shell-and-coil and flash intercoolers. Figure ?.2 depicts a shell-and
coil intercooler incorporating an internal pipe coil for subcooling the high-pressure liquid
before it is fed to the low stage of the system. Typically, the coil subcools the liquid to within
6 K of the intermediate temperature.
Fig. 20.2 Vertical shell-and-coil intercooler
Vertical shell-and-coil intercoolers with float valve feed perform well in many applications
using ammonia refrigerant systems. The vessel must be sized properly to separate the liquid
145
from the vapour that is returning to the high-stage compressor. The superheated gas inlet pipe
should extend below the liquid level and have perforations or slots to distribute the gas evenly
in small bubbles. The addition of a perforated baffle across the area of the vessel slightly
below the liquid level protects against violent surging. A float switch that shuts down the
high-stage compressor when the liquid level gets too high should always be used with this
type of intercooler in case the feed valve fails to control properly.
The flash intercooler is similar in design to the shell-and-coil intercooler with the exception of
the coil. The high-pressure liquid is flash cooled to the intermediate temperature. Caution
should be exercised in selecting a flash intercooler because all the high-pressure liquid is
flashed to intermediate pressure. Though colder than that of the shell-and-coil intercooler, the
liquid in the flash intercooler is not subcooled and is susceptible to flashing due to system
pressure drop. Two-phase liquid feed to control valves may cause premature failure due to the
wire drawing effect of the liquid/vapour mixture.
Accessories necessary are the liquid level control device and the high-level float switch.
Though not absolutely necessary, an auxiliary oil pot should also be considered.
7.2 Selection
The cycle we selected is a two-stage cycle with open intercooler because of the better COP.
M&M Refrigeration disposes of an assortment of Standard Vertical Intercoolers
corresponding with a range of different cooling capacities.
Intermediate temperatures are 7.5 C for the blast freezing tunnel and 7.3 C for the brine
tunnel. In grades Fahrenheit is that 18.5 F and 45.14 F.
Cooling capacities are 187.65 tons of refrigeration for the brine tunnel and 83.02 tons of
refrigeration for the blast freezer.
Out of the M&M Refrigeration catalogue we select following intercooler.

Intercooler model: VI-30-115
Dimensions are given in added catalogue on the disc.
8 Refrigerant liquid pumps
8.1 Introduction
The refrigerant liquid pumps force the refrigerant through the pipes. It increases its pressure to
withstand pressure losses due to friction or to put the refrigerant on a higher pressure in
general.
8.2 Low-pressure pumps
Low-pressure pumps force the refrigerant from the liquid separator through the liquid
overfeed lines, the evaporator and return lines. Their purpose is to withstand pressure losses in
the trajectory and to provide a liquid overfeed. These pressure losses are calculated in chapter
146
22 We make the sum of the losses in the liquid overfeed line to the evaporator, the losses in
the evaporator and the losses in the return pipe: p = 48411.41 Pa. This equals a head of 7.35
m.
8.2.1 Selection
Hansen Technology Corporation provides a great variety of hermetic liquid refrigeration

pumps. They are safer for personnel and the environment because they are sealless,
eliminating the potential for harmful refrigerant leakage. Also they require no lubricating oil
or grease, no replacement of shaft seals, are not affected by moisture or frost accumulation,
and provide many years of quiet, reliable operation. These pumps feature a low, required net
positive suction head (low NPSH), which makes them ideal recirculating pumps for liquid
overfeed evaporators.
For more information about the low-pressure pump see added catalogue on the disc or visit
8.3 High-pressure pumps
The high pressure pumps increase the ammoniac pressure to transport the ammoniac from the
liquid separator to the high-pressure receiver. This has several purposes: a) to empty the low
pressure part of the installation for maintenance; b) to empty the low pressure part of the
installation to start hot gas defrosting; c) to maintain a constant liquid level of the liquid
separator during operation
8.3.1 Selection
The pumps in the catalogue of Hansen Technology Corporation are also suited for liquid
transport.
For more information about the high-pressure pump unit see added catalogue or visit
9 Check valves
9.1 Introduction
Refrigerant check valves are normally used in refrigerant lines in which pressure reversals can
cause undesirable reverse flows. In compressor discharge lines, check valves are used to
prevent flow from the condenser to the compressor during the off. Although a 14 to 42 kPa
pressure drop is tolerable, the check valve must resist pulsations cause by the compressor and
the temperature of discharge gas. Also, In hot-gas defrost lines, check valves may be used in
the branch hot-gas lines connecting the individual evaporators to prevent crossfeed of
refrigerant during the cooling cycle when the defrost operation is not taking place. In addition,
check valves are used in the hot-gas line between the hot-gas heating coil in the drain pan and
the evaporator to prevent pan coil sweating during the refrigeration cycle. Tolerable pressure
drop is typically 14 to 42 kPa, seating must be nearly bubble-tight, and seat materials must
withstand high temperatures.
147
9.2 Selection
Check valves are selected out of the Danfoss catalogue.
9.2.1 Check valve between high-pressure receiver and intercooler
To select the desired check valve we follow the Danfoss catalogue instruction:
For blast freezer operation running conditions are as follows:
o Te = -45C
o Q0 = 292 kW
o Tliq = 35 C
o Max p = 0.3 bar
The capacity table is based on nominal conditions (p = 0.2 bar, Tliq = 30C).
The actual capacity must therefore be corrected to a nominal condition by
multiplication with correction factors.
Correction factor for p 0.3 bar fp = 0.82.
Correction factor for liquid temperature fTliq = 1.02.
Qn = Q0 fp fTliq = 292 0.82 1.02 = 244 kW.
From the capacity table a NRVA 15 is the correct selection for the application.
For brine tunnel operation running conditions are as follows:

o Te = -25C
o Q0 = 660 kW
o Tliq = 35 C
o Max p = 0.3 bar
Qn = Q0 fp fTliq = 660 0.82 1.02 = 552 kW.
This means that a NRVA 20 has to be selected for the application.
9.2.2 Check valve between intercooler and low-pressure receiver
o Te = -45C
o Q0 = 292 kW
o Tliq = -7.4 C
o Max p = 0.3 bar
148
Qn = Q0 fp fTliq = 292 0.82 0.87 = 208 kW.


o Te = -25C
o Q0 = 660 kW
o Tliq = 7.4 C
o Max p = 0.3 bar
Qn = Q0 fp fTliq = 660 0.82 0.91 = 492 kW.
9.2.3 Check valves between the pump and the evaporators in the recirculation circuit
o Te = -45C
o Q0 = 292 kW
o Circulation rate = 4
o Max p = 0.3 bar
The capacity table is based on nominal conditions (p = 0.2 bar, circulation rate = 4).
Correction factor for circulation rate frec = 1.
Qn = Q0 fp frec = 292 0.82 1 = 239 kW.

o Te = -25C
o Q0 = 660 kW
o Max p = 0.3 bar
Qn = Q0 fp frec = 660 0.82 1 = 541 kW.
149
9.2.4 Check valves in the wet suction lines
For the blast freezer running conditions for one single evaporator are as follows:
o Te = -45C
o Q0 = 50 kW
o Max p = 0.3 bar
Qn = Q0 fp frec = 50 0.82 1 = 41 kW.
For the brine tunnel running conditions for one single evaporator are as follows:
o Te = -25C
o Q0 = 165 kW
o Max p = 0.3 bar
Qn = Q0 fp frec = 165 0.82 1 = 135 kW.
9.2.5 Check valve in the dry suction line ahead of the high-stage compressor
o Te = -45C
o Q0 = 292 kW
o Tliq = -7.4 C
o Ts = 1C
o Max p = 0.3 bar
Correction factor for superheat fTs = 1.
Qn = Q0 fp fTliq fTs = 292 0.82 0.87 1 = = 208 kW.
150

o Te = -25C
o Q0 = 660 kW
o Tliq = 7.3 C
o Ts = 1C
o Max p = 0.3 bar
Correction factor for liquid temperature Tliq = 0.91.
9.2.6 Check valve in the dry suction line ahead of the low-stage compressor
o Te = -45C
o Q0 = 292 kW
o Tliq = -40 C
o Ts = 1C
o Max p = 0.3 bar
Qn = Q0 fp fTliq fTs = 292 0.82 0.70 1 = 168 kW.

o Te = -25C
o Q0 = 660 kW
o Tliq = -25 C
o Ts = 1C
o Max p = 0.3 bar
151
9.2.7 Check valve in the dry suction line after the high-stage compressor
o Te = -45C
o Q0 = 292 kW
o Tliq = -7.4 C
o Tdisch = 68 C
o Max p = 0.3 bar
The capacity table is based on nominal conditions (p = 0.2 bar, Tliq = 30C, Pdisch =
12 bar, Tdisch = 80 C).
Correction factor for discharge temperature fTdisch = 0.98.
Correction factor for discharge temperature fPdisch = 1.
Qn = Q0 fp fTliq fTdisch fPdisch = 292 0.82 0.87 0.98 1 = 204 kW.

o Te = -25C
o Q0 = 660 kW
o Tliq = 7.3 C
o Tdisch = 66 C
o Max p = 0.3 bar
9.2.8 Check valve in the dry suction line after the low-stage compressor
o Te = -45C
o Q0 = 292 kW
o Tliq = -40 C
152
o Tdisch = 62 C
o Max p = 0.3 bar

o Te = -25C
o Q0 = 660 kW
o Tliq = -20 C
o Tdisch = 65 C
o Max p = 0.3 bar
153
1 Introduction
The refrigerating plant is automatically operated to some extent, as manual control of the
various refrigerating functions is difficult and not very accurate. Moreover very skilled
operators or technicians must be employed for such tasks.
The following operations are more in need of automatic control than others: regulation of
temperature in the blast freezing chamber and the temperature of the brine in the brine tunnel;
feeding to the evaporators; adapting the refrigerating capacity of the plant to one heat load of
the freezing room and the brine tunnel at different ambient conditions; regulating condensing
temperature; starting the compressor operation sequence; and initiating the sequence for the
defrosting cycles in the blast freezer.
Where control is automatic safety devices must be installed to protect the equipment against
malfunction. These safety devices are not only for the refrigerating system but also for the
electric circuit, though the latter concerns other specialists and is not within the scope of this
work.
When a safety device is in operation it indicates there has been a breakdown or a malfunction.
It is necessary to search for the cause of the trouble and repair it before the system is restarted,
otherwise the safety device will operate again and the malfunction may cause further
problems.
Safety devices only operate occasionally and some do not operate for long running periods.
Consequently it is advisable to check periodically that they operate properly and adjust their
operating ranges. These checks should be scheduled once a month for high-pressure, low-
pressure and lubricating oil pressostats as these prevent any malfunction of controlled
elements and they undergo frequent on-off operation during the running cycles.
Repairs and maintenance must be entrusted to competent refrigeration mechanics because of
the risk to personnel, particularly in ammonia plants. Intensive training should be provided for
local mechanics and technicians.
The following section describes the main safety devices and the elements they control.
2 Pressure switches
Next list shows various types of pressure switches needed in the installation with their
corresponding function.
High-pressure cut-out (HPCO) : Stops compressor when excessive pressure occurs.

High-side low-pressure (HSLP) : Prevents compressor operation under low ambient or
loss of refrigerant conditions.
High-side fan cycling (HSFC) : Cycles condenser fan on and off to provide proper
condenser pressure.
Low-side low-pressure (LSLP) : Initiates defrost cycle; stops compressor when low
charge or system blockage occurs.
Low-side compressor cycling (LSCC) : Cycles compressor on and off to provide
proper evaporator pressure and load temperature.
Lubricant pressure differential failure (LPDF) : Stops compressor when difference
between oil pressure and compression space is too low for adequate lubrication.
154
2.1 High-pressure safety pressostat
High-pressure safety pressostats or cutouts are activated when the pressure in the condenser is
too high, protecting the system against overload and rupture of the piping. The element
controlled by this device is the compressor. The pressostat is fitted to the compressor
discharge line and wired in series with the compressor drive, which is cycled-off whenever an
increase in condensation pressure becomes excessive.
High condensation pressure can be caused by any of the following.
Faulty cooling of the condenser, which may be because of: the failure of fans in air-
cooled condensers and fouling of the condenser surface (scale, algae and dust in air-
cooled condensers).
The heat exchange surface area is reduced by refrigerant liquid accumulation in the
lower part of the condenser. This may be the result of an excessive refrigerant fluid
charge in the system or irregular operation of the pressure control.
Non-condensable gases present in the circuit, particularly with ammonia installations.
In this case a purge is necessary.
The automatic control of the condensation pressure is adjusted to an unusually high
level.
A fire on the premises, producing a quick rise in circuit pressure.
The high-pressure pressostat controls most of the risks derived from an unusual increase in
pressure. A high temperature level is usually considered to be in the 4060C range. However
this safety device is not enough to protect the refrigeration plant and some other pressure cut-
out device such as a safety pressure relief valve, a fusible plug, a rupture disk or a fuse are
necessary.
Whenever high-pressure vapour breaks out it must be evacuated from the environment
without any risk to personnel or people in the machine room or surrounding areas. Sometimes
the high pressure is released into the low-pressure circuit, which should be provided with a
leakage restrictor.
The cut-out pressure should be set well above the maximum condensing pressure when the
plant is operating at full load; 20.33 bar gauge pressure is recommended as the cut-out setting.
This is equivalent to 50C for a air-cooled ammonia plant.
The pressure differential should be established according to the operation and bearing in mind
that a small differential will result in compressor shortcycling that may cause electrical
damage, while large differentials will result in too long off-cycles, leading to an abnormally
high rise in storage temperature.
A high-pressure pressostat must be manually reset to prevent any risk of automatic on-cycling
before the malfunction has been found and repaired.
2.2 Low-pressure safety pressostat
The low-pressure safety pressostat or cutout operates when the pressure in the suction circuit
decreases below a set level. The most common risk is that of very low evaporation
temperatures with possible damage to the refrigeration plant.
Low-pressure pressostats break the circuit to the compressor motor when evaporating pressure
falls. There are several causes for low evaporating pressures:
The evaporator is not correctly fed or is not fed at all. This may be owing to the
blockage of the expansion valve or its filter, or the solenoid valve upstream of the
155
expansion valve is not in operation, generally because the coil is out of order and thus
the automatic valve operated by the solenoid valve is not working.
Compressor capacity clearly exceeds evaporator capacity. This may be because of
malfunctioning of the compressor capacity control device; the evaporator fans may
have stopped accidentally, reducing heat exchange; there may be excessive frost
deposit on the evaporator leading to the same heat exchange problem; or the constant
pressure valve regulating the evaporating temperature may not be properly adjusted. If
there is a very small difference between room and evaporator temperatures the suction
pressure may be too low.
The refrigerant fluid charge in the circuit is too small, which may be because of an
installation error or a small leakage difficult to discover.
Sudden and unusual oil dragging into the low-pressure circuit, which both diminishes
heat exchange and increases pressure loss in the suction line.
The thermostat is blocked in an open position, keeping the compressor in operation
while the thermal load of the evaporator decreases continually.
The pressostat must be set to cutout at a pressure well below the lowest normal evaporating
pressure when the evaporator is working at minimum load, but above a certain pressure to
avoid damage to the refrigerating plant.
Small differentials will result in compressor short cycling so they must be avoided to prevent
equipment damage. On the other hand large differentials will result in long off-cycles.
When the low-pressure safety device is in operation a limited risk is avoided, which does not
represent any great danger to the refrigerating plant. Therefore the pressostat may or may not
be manually reset, though for high-capacity machines manual resetting is recommended.
These safety devices can suffer from such mechanical failure as leakage of the pressure
bellows (they must be replaced), obstruction or partial blockage of the small connecting pipe
of the control element with pipe scale and/or sludge (isolate and remove the pipe, flush with
refrigerant R.11) or the linkages may be worn, damaged, stuck or loose and must be replaced
or readjusted.
2.3 Oil-pressure control
Faulty lubrication of the compressor can be caused by either of the following.

The oil pump is accidentally out of order, probably because of breakdown. The suction
and/or delivery oil pipes are obstructed and the pressure of the pump is not enough to
overcome this.
The oil-pressure control stops the compressor when the useful pressure developed by
the pump falls below a fixed minimum or the oil pressure does not rise to the
minimum safety level within an established period. Useful pressure is the difference
between total pressure and suction pressure. Therefore two pressure bellows to the
crankcase and the oil pump discharge must connect the oil-pressure switch, which is
activated by useful pressure.
A time delay relay is incorporated in the oil-pressure control to allow the compressor an
operating period of about 30120 seconds with oil pressure below safety level. If the control
starts operating as a result of malfunction in the compressor lubricating system, a lamp and/or
a hooter (visible audible alarm) will be activated. The latter can be manually disconnected.
After failure and before the compressor can be restarted the oil-pressure control must be
manually reset because of the high risk involved in its operation.
156
The compressor manufacturer's instructions usually set the cut-in and cut-out pressures for the
oil-pressure control, but whenever such data are not available they should be set at about 0.34
bar absolute pressure below useful pressure level and approximately the same quantity below
cut-in pressure.
2.4 Selection
Low-pressure safety pressostat:

Suited model: KP 1A
High-pressure safety pressostat:

Suited model: KP 5A
For more information see added catalogue on the disc or visit http://www.danfoss.com
3 Temperature switches
3.1 Thermostats
Thermostats are temperature-activated controllers that regulate the temperature level of a

refrigerated space by cycling the compressor on and off. They consist of a sensing element
and an electric contactor. Three types of temperature-sensing elements are commonly used in
refrigeration: fluid-filled bulbs connected to pressure bellows or diaphragm, bimetal strips or
compound bars, and electric resistances and semiconductors. The latter two are seldom used
in refrigeration, but as they are rather easy to install and regulate they promise well for the
future. Thermocouples as sensing elements in the thermostats are rarely used in refrigeration.
A problem with the pressure bellows type of element is they may lose their fluid charge. To
check this, compress the bellows by hand. If they move under finger pressure the fluid charge
has been completely or partially lost. The bellows assembly must be replaced, if accessible,
otherwise the thermostat must be changed.
Electric failure of the thermostats may be due to poor electrical connections (clean and tighten
them for correct operation) or to worn, pitted or corroded contact points (replace them). These
mechanical or electrical failures will cause the thermostat to remain either open or closed.
When it stays open the compressor will shut down even if the room temperature is above the
desired level. The use of a jumper lead across the terminals should immediately restart the
compressor.
When the thermostat remains closed the compressor will continue running even if the room
temperature is below the established level. Test by turning the control knob to a higher
temperature.
Faulty operation may also be traced to the incorrect installation of the sensing element,
usually the fluid bulb. If it is not securely attached to the evaporator surface and/or it is not
sufficiently clean and dry for good thermal contact, an excessive temperature difference is
necessary before the thermostat operates.
When the thermostat directly controls space temperature the bulb is fixed to the chamber wall
with a metallic bracket in one of the positions already described. If there is no thermal
insulation between the bulb and the bracket the equilibrium temperature of the bulb will be
higher than that of the room space.
157
3.2 Delivery thermostat
This safety device is used as there is the risk of increase in compressor delivery temperature
above 130C. This temperature is dangerous as the lubricating oil may break down,
notwithstanding high temperature decomposition, and form corrosive acids and sludge. There
are several causes for an excessively high delivery temperature.
The system is operating under an excessive compression ratio (absolute discharge
pressure divided by absolute suction pressure).
The suction pressure is very low because the constant pressure valve is not correctly
regulated. The lower the suction pressure the higher the discharge temperature, other
circumstances remaining constant. Again the two-stage compression systems offer an
advantage, as they avoid high discharge temperature.
Suction superheat is excessively high due to poor insulation of the suction pipe and/or
because it is too long. In an ammonia refrigerant system an automatic humidifier is an
advantage where the vaporization of partly vaporized liquid ammonia injected in the
suction line from the liquid receiver provides the necessary desuperheating of the
vapour.
The delivery superheat thermostat cycles-off the compressor if the discharge temperature is
too high. When in operation it avoids serious risk to the compressor, so it must be manually
reset.
4 Liquid level indicators.
4.1 Introduction
Liquid level can be indicated by visual indicators, electronic sensors, or a combination of the
two. Visual indicators include individual circular reflex level indicators (bullseyes) mounted
on a pipe column or stand-alone linear reflex glass assemblies (Figure 21.1). For operation at
temperatures below the frost point, transparent plastic frost shields covering the reflex
surfaces are necessary. Also, the pipe column must be insulated, especially when control
devices are attached. Electronic level sensors can continuously monitor the liquid level.
Digital or graphic displays of the liquid level can be locally or remotely monitored (Figure
21.2). Adequate isolation valves for the level indicators should be employed. High-
temperature glass tube indicators should incorporate stop check or excess-flow valves for
isolation and safety.
158
Fig. 21.1 Gage glass assembly for ammonia; Fig. 21.2 Electronic liquid level control
4.2 Selection
Model type: RT 280A
For more information about the liquid level control see added Danfoss catalogue or
http://www.danfoss.com
5 Regulating and throttling valves
5.1 Suction-pressure regulating valves
The suction pressure regulator is a downstream pressure regulator positioned in a compressor

suction line to respond to and limit compressor suction pressure. Typically, they are used to
prevent compressor motor overload due to high suction pressure related to warm start-up,
defrost termination, and intermittent high evaporator loading. The suction-pressure-regulating
valve is part of the compressor group.
5.2 Expansion valves
The expansion of the liquid refrigerant fluid from high to low pressure is achieved by passing
it through a device known as an expansion valve. Expansion valves have two functions:
metering the liquid refrigerant into the intercooler and low-pressure receiver and maintaining
a pressure differential between the high-and low-pressure sides of the refrigerating circuit.
There are several types of expansion valve, all automatic. In industrial plants it is advisable,
regardless of the type installed, to set up a hand operated controller in parallel with the main
159
expansion valve to avoid plant shut-down if the valve should break down. A solenoid valve is
mounted upstream of the expansion valve; both elements as a whole are placed between two
manually operated valves to isolate them for maintenance or repairs.
Thermostatic expansion valve
Thermostatic expansion valves are used in direct expansion systems. We will not go into
a full consideration of this type of valve.
Automatic expansion valve
The same remark can be made as for thermostatic expansion valves.
Low-pressure float valve
For multiple evaporator systems, low-side float valves are used to control the refrigerant
level in flooded evaporators.
The low-pressure float valve is used in industrial plants to maintain a constant liquid level
in the accumulator. The evaporator is therefore constantly filled to the desired level with
liquid refrigerant in all conditions of heat load and independently of evaporator
temperature and pressure. The vapour pressure in this flow controller and in the
accumulator is the evaporation pressure.
The low-pressure float valve can operate either continuously (its throttling action
modulates liquid flow into the evaporator in response to liquid level changes) or
intermittently (the valve is either fully open or fully closed in response to established
minimum and maximum liquid levels). A risk with this type of valve is that liquid may
pass through during the compressor off-cycle. This is avoided by fitting a solenoid valve
upstream of the float valve and wired in series with the compressor, shutting off the circuit
when the compressor is stopped. Another drawback is faulty liquid tightness, particularly
if the refrigeration circuit is not thoroughly cleaned in the mounting stage. In large-
capacity systems the float valve should be installed in such a way that a bypass line
equipped with a hand expansion valve permits the refrigerating plant to operate in the
event of float valve failure. Two hand-stop valves, placed one each side of the float valve,
allow its isolation for servicing without evacuating the large refrigerant charge from the
evaporator. A low-pressure float valve can be used in parallel with a thermostatic
expansion valve.
If the float valve allows too much liquid to pass through its control evaporating pressure
will be higher than normal and/or unevaporated liquid will return to the compressor,
causing pipe sweating and eventually knocking. Faulty operation may be due to the valve
seat not being properly cleaned, or to the ball float being punctured in which case it must
be replaced (there is the risk of high-pressure liquid being trapped inside the float, so
careful handling is required) or finally, the operating linkage may be jammed and must be
dismantled, cleaned and lubricated.
The float valve may not allow sufficient liquid to flow, the low-pressure circuit exhibiting
a lower than normal evaporating pressure and the cooling effect being noticeably reduced.
This malfunction may be due to a dirty inlet filter. The operating linkage may not be
160
moving freely and if so it must be dismantled, cleaned and lubricated. Finally the
accumulator chamber may be filled with gas, either because it is not properly insulated to
maintain constant temperature and pressure or because there is an obstruction in the vent
pipe.
High-pressure float valve
A high-pressure float control can be used to keep the condenser drained and to provide a
liquid seal between the high side and the low side. But, it can only be applied when the
refrigerating plant has only one evaporator. Therefore we will not go deeper into it.
5.3 Solenoid valves
Solenoid valves are widely used in refrigerating plants to control automatically the fluid
refrigerant flow (liquid line and evaporator outlet). Because gas defrosting is used solenoid
valves are fitted in the refrigeration circuit to reverse refrigerant flow.
A solenoid valve is simply an electrically operated valve consisting of an electromagnetic coil
which when energized draws a plunger which opens the valve port. Closing action is achieved
by gravity when the coil is de-energized. Solenoid valves controlling the refrigerant flow in
the liquid line feeding the evaporators are thermostatically activated.
Solenoids vary according to permitted pressure differences across the valve, drop in pressure
through the valve, the desired flow rate and the state of the circulating fluid. The valves can
be direct acting or pilot operated. Small solenoid valves are usually direct acting; they are
fitted in small-diameter pipes and their power demand is low, about 15 Watts.
Pilot-operated solenoid valves are used in large-diameter pipes as the pressure differences
across the valve provide the force to carry out the closing and opening actions.
Solenoid valves are simple and robust, and their functioning is generally reliable and accurate
as they usually operate in an on-off mode. This is also true when they work in a modulated
fashion (applying a changeable voltage to the electromagnetic coil of the valve).
They have some drawbacks. They are not gastight, particularly for large pipe sections. Their
most usual failure is coil breakdown, which is not foreseeable so it is impossible to detect
during maintenance operations. Unless they are vapour-tight condensation of water vapour on
their cold side may occur, particularly when they are installed in rooms with high humidity
and working in medium-temperature conditions. There is also some risk of causing a water
hammer effect in the liquid line.
Solenoid valves must be correctly installed. They must always be mounted in a vertical
position with the coil on top, unless they are especially. designed for horizontal installation.
They must also be mounted in line with the direction of flow, usually shown by an arrow on
the valve body.
Some operating faults may appear when a solenoid valve is at work. If the valve fails to open
the cause could be mechanical or electrical. Mechanical problems follow when the valve is
not mounted level on the pipe or when the plunger is stuck to the valve seat. Electrical failures
could be because of dirty or loose electrical connections, no power supply, and/or a damaged
coil. It is advisable to have several coils in stock for each type of solenoid valve. If the valve
does not fully close the valve seat may be either dirty or worn. The valve should be
dismantled and the valve seat and plunger cleaned or renewed, or a new valve should be
fitted. When the valve is noisy during operation it has obviously not been correctly installed
and is not level on the pipe; an alternative could be that the electrical connections are not
sound.
161
5.4 Selection
Expansion valves:
Suited model: AKVA 20
Solenoid valves:
See Danfoss catalogue
For more information see added catalogue or visit http://www.danfoss.com
6 Testing safety devices
As safety devices operate intermittently and operational conditions can gradually become
erratic, it is important to check them periodically for correct operation and to adjust them
when they malfunction. It is essential to check every control or safety element and all
equipment when starting up a newly installed plant. Pressure gauges, pressostats, thermostats,
valves, relief valves, compressors, fans, pumps, refrigerant pipes and liquid pipes must be
checked for proper operation on starting and whenever the refrigerating plant has been shut
down for a long period. It is advisable to be vigilant for some time as it is well-known that the
possibilities of breakdown are higher during start-up.
A maintenance programme for the operating plant that includes frequent automation checks is
necessary. Pressure gauges should be checked at regular intervals, normally once a month.
When the isolation manual valve is closed and the gauge connecting tube is slightly released
the pressure measured will be atmospheric. The pressure shown should be zero, and if it is not
the needle position must be readjusted. If this is not possible, note should be taken of the
adjustment to be done on the reading.
To check the high-pressure pressostat the refrigerating plant is started up without operating
the condenser cooling system (the air fans are not running) and the cutting pressure is checked
at which the pressostat cycles-off the plant. The pressostat action must be neat and at the same
time acoustic, and light signals should be activated. Cycling-on is not allowed until the
pressostat is manually reset.
The low-pressure pressostat is tested by stopping the refrigerant supply to the evaporator,
closing either a hand valve or a solenoid valve placed upstream of the expansion valve. The
pumping action of the compressor increases and the operating conditions are checked on the
suction pressure gauge. Manual resetting of this pressostat is optional.
Checking and regulation of both types of pressostat should be done at monthly intervals.
When the compressor is pump lubricated the oil pressostats cannot be checked by stopping
the oil pump. However, if the joints of the pressostat connecting pipes are slightly loosened
the pressure difference decreases until it becomes nil. The cutting action of the pressostat is
then tested, as is the operation of the time delay relay. This operation should be repeated once
every three months, verifying that the compressor does not cycle-on until the pressostat is
manually reset.
When the compressor is splash lubricated and protected against oil deficiency by an oil level
contactor, the oil level should be checked once a week when the compressor is stopped.
The delivery thermostat, controlling compressor discharge temperature, should be checked
and regulated once every three months.
162
7 Automation
7.1 Introduction
Automation of a refrigerating installation provides safe and profitable running in any

operating conditions without human intervention (integral automation) and at the same time
assures that the controlled parameters (temperature or relative humidity in the chamber, for
instance) do not differ from the fixed values and that they can be readjusted when they are not
within the established range.
Automation should satisfy two major objectives: it should provide a more accurate control
and reduce operating costs by minimizing the number of people employed in the plant. It also
serves another function of paramount importance - safety.
Control is always accurate as automatic equipment nowadays responds quickly to parameter
deviations, operates continuously and follows closely any equipment manoeuvre. This
accuracy is difficult to reach with manual control of the refrigerating operation. Sometimes
skilled operators have the advantage over automatic operation, as they are able to anticipate
some operating situations, such as cooling down the brine tunnel or blast freezing chamber
before they are loaded and start running. Moreover, the reduction in operating costs may be
small as the investment costs are usually much higher. The variety, complexity and costs are
such that a very strict maintenance programme must be undertaken by highly skilled technical
personnel with appropriate qualifications and experience. This increases maintenance costs
and offsets the benefits of a reduction in unskilled personnel. However, the advantages totally
justify the installation of automatic control.
Because we deal with a large and complex installations, semi-automation is recommended,
partly for economic reasons but also for safety as this installation must not be allowed to
operate without human supervision. The essential feature of a semi-automatic refrigerating
plant is that the choice of the operating periods is left to the operators' initiative. Once the
plant is in service, control is automatic as in totally automatic plants.
Automatic control must include the communication of information to allow supervision of the
plant's operation. All the sensing elements should be linked to indicator lights, measuring
devices, audiowarning devices, and so on, which are displayed in the machine room. They are
generally incorporated in a synoptic luminous panel. There must also be remote control of the
machinery in the machine room, either for manual control of the operation or for resetting the
control elements after operation failure. The automatic equipment should be verified at least
once a week with reliable apparatus.
As well as the refrigerating plant safety devices, the machine room must be provided with
ammonia leak detectors so that the alarm systems are activated and the plant shut down in
case of leakage. Extractor fans, water spray points and carbon dioxide fire extinguishers are
required in the event of ammonia leakage. These elements should be independent of the
general electric mains and be readily accessible outside the machine room.
The automatic sequence of machine start-up should be established with these requirements in
mind:
the compressor must not run if the condenser fans are not working;
the refrigerant liquid pump will not work until the compressor and evaporator fan are
running and the refrigerant fluid is circulating in any of the evaporators;
when the refrigerant fluid is not circulating in the evaporators, the various elements of
the refrigerating circuit should be stopped.
Although the decision will depend on the expertise of the maintenance personnel, the
accuracy of control can always be given second place after equipment robustness and long
service life. A stock of spare parts should be kept, giving priority to those that are the most
163
frequent causes of trouble. The list will be related to ease of supply. In some cases duplicates
for control apparatus should be stocked, but in general the following elements will suffice:
lubricating oil and refrigerant fluid; coils for each type of solenoid valve and a solenoid valve
of each type; sets of valves; sets of joints and driving belts; fan motors for air coolers; and
fuses, coils for electrical contactors, cut-outs, etc.
7.2 Selection
M&M Refrigeration has developed automation and control software (Engineered Control
System, ECS) that can serve for the installation we designed. The catalogue is added and
even so more information can be found on http://www.mmrefrigeration.com
8 Refrigerant circuit
The two main troubles encountered in the refrigerant fluid circuit (pipes, condensers,
evaporators, compressors, receivers and ancillaries) are refrigerant leakage and non-
condensable gas accumulation.
8.1 Refrigerant leakage
Ammonia leaks are detected by passing burning sulphur candles around the suspected joints.
The presence of ammonia will be indicated by a dense white smoke. Although ammonia is a
flammable refrigerant this method of detection does not represent any danger provided the
ammonia concentration is low. However, it must never be used when breathing has become
uncomfortable.
8.2 Non-condensable gases
Non-condensable gases accumulating in the high-pressure side of the refrigerant circuit will
cause a rise in delivery pressure, affecting compressor power consumption and wear.
The presence of these gases in the system is indicated by the standing head pressure shown on
the delivery gauge once the system is stabilized, that is when the condenser is in thermal
equilibrium with the environment. If this pressure is greater than that of the refrigerant vapour
that corresponds to the equilibrium temperature, then non-condensable gases are present in the
condenser. These must be removed from the system step by step by purging from the highest
point on the high-pressure side, slowly reducing the pressure and allowing stabilization by
short intervals after each reduction. Automatic purgers are installed for continuous non-
condensable gas removal.
164
1 Introduction
In a refrigeration system each phenomenon influences and is being influenced. For example
the heat transfer rate depends of the pressure, which in turn depends of pressure loss in the
ducting. In this chapter the pressure losses in the several pipes are calculated. Single-phase
flow, as well as two-phase flows occur.
2 Single-phase flow.
Here we give the calculation method to find the pressure drop for a single-phase flow.
Pressure drop caused by fluid friction in fully developed single-phase flows of all well-
behaved (Newtonian) fluids is described by the Darcy-Weisbach equation:
L u
2
p pipe = f (22.1)
D 2
0.25
64
f = 0.11 + (only for turbulent flows) (22.2)
D
H Re
if f >0.018: f = f
if f <0.018: f =0.85 f +0.0028
where
p = pressure drop, Pa
f = friction factor
L = length of pipe, m
D = internal diameter of pipe, m
= fluid density at mean temperature, kg/m3
V = average velocity, m/s
Valves and fittings cause pressure losses greater than those caused by the pipe alone. One
formulation expresses losses as
u2
p fitting = K (22.3)
2
where K = geometry- and size-dependent loss coefficient.
165
Table 22.1 K-factors in function of pipe diameter for several fittings
Nominal 90 90 45 Return Return Tee- Tee- Globe Gate Angle Swing

pipe standard Long- long- bend bend line branch valve valve valve check
diam, elbow radius radius standard long- valve
mm elbow elbow radius
25 0.43 0.41 0.22 0.43 0.43 0.26 1.0 13 - 4.8 2.1
32 0.41 0.37 0.22 0.41 0.38 0.25 0.95 12 - 3.7 2.1
40 0.40 0.35 0.21 0.40 0.35 0.23 0.90 10 - 3.0 2.1
50 0.38 0.30 0.20 0.38 0.30 0.20 0.84 9 0.34 2.5 2.1
65 0.35 0.28 0.19 0.35 0.27 0.18 0.79 8 0.27 2.3 2.1
80 0.34 0.25 0.18 0.34 0.25 0.17 0.76 7 0.22 2.2 2.1
100 0.31 0.22 0.18 0.31 0.22 0.15 0.70 6.5 0.16 2.1 2.1
150 0.29 0.18 0.17 0.29 0.18 0.12 0.62 6 0.10 2.1 2.1
200 0.27 0.16 0.17 0.27 0.15 0.10 0.58 5.7 0.08 2.1 2.1
250 0.25 0.14 0.16 0.25 0.14 0.09 0.53 5.7 0.06 2.1 2.1
300 0.24 0.13 0.16 0.24 0.13 0.08 0.50 5.7 0.05 2.1 2.1
Source: Engineering Data Book (Hydraulic Institute 1979)
Total pressure loss over the pipes and fittings:
p tot = p pipe + p fitting (22.4)
3 Two-phase flow.
Here we give the calculation method to find the pressure drop for a single-phase flow.
The total pressure drop of a fluid is due to the variation of potential energy of the fluid, kinetic
energy of the fluid and that due to friction on the walls of the flow channel. Thus, the total
pressure drop ptotal is the sum of the static pressure drop (elevation head) pstatic, the
momentum pressure drop (acceleration) pmom, and the frictional pressure drop pfrict:
p total = p static + p frict + p mom (22.5)
The static pressure drop for a homogeneous two-phase fluid is:
p static = H gHsen (22.6)
where H is calculated with formula (22.20).
For a horizontal tube, there is no change in static head, i.e. = 0 and H = 0 so pstatic = 0
while sin is equal to 1.0 for a vertical tube. The momentum pressure drop reflects the change
in kinetic energy of the flow and is for the present case given by:
p mom = m& total

2

(
1 x 2 ) +
x2
+
(
1 x2 ) +
x 2
(22.7)
L (1 ) G out L (1 ) G in
166
where m& total total is the total mass velocity of liquid plus vapour and x is the vapour quality.
The separated flow model considers the two phases to be artificially separated into two
streams, each flowing in its own pipe. The areas of the two pipes are proportional to the void
fraction . Numerous methods are available for predicting the void fraction. we use the
Steiner (1993) version of the drift flux model of Rouhani and Axelsson (1970) for horizontal
flows:
1
x x 1 x (1 x )[g ( L G )]0.25
= (1 + (1 x )) + + (22.8)
G G L 2
m& total L0.5
For vertical flows, the Rouhani and Axelsson (1970) expression can be used for void fractions
larger than 0.1:
1

gd i L2 4 x 1 x 1.18(1 x )[g ( L G )]0.25
1
1 + 0.2(1 x ) 2
x
= + + (22.9)
G m& total G L 2
m& total L0.5

The momentum pressure drop is calculable by input of the inlet and outlet vapour qualities
and void fractions.
For an evaporating flow, the kinetic energy of the outgoing flow is larger than that of the
incoming flow since the density of the vapour phase is less than that of the liquid. Hence, the
momentum pressure drop results in a lower pressure at the exit than at the inlet. Instead, for a
condensing flow the kinetic energy of the outgoing flow is smaller than that of the incoming
flow. Hence, the momentum pressure drop results in an increase in pressure at the exit than at
the inlet, i.e. a pressure recovery. For condensing flows, it is common to ignore the
momentum recovery as only some of it may actually be realized in the flow and ignoring it
provides some conservatism in the design.
The frictional pressure drop in two-phase flows is typically predicted using separated flow
models. The first of these analyses was performed by Lockhart and Martinelli (1949) and then
followed by many others. The basic equations for the separated flow model are not dependent
on the particular flow configuration adopted. It is assumed that the velocities of each phase
are constant, in any given cross-section, within the zone occupied by the phase. A common
method for in-tube flow is given below.
3.1 Friedel correlation
The correlation method of Friedel (1979) utilizes a two-phase multiplier:
p frict = p L 2fr (22.10)
where pL is calculated for the liquid-phase flow as
L 2 1
p L = 4 f L m& total (22.11)
di 2 L
167
The liquid friction factor L and liquid Reynolds number (and vapour friction factor G and
vapour Reynolds number with the vapour viscosity) are obtained from
0.079
f = (22.12)
Re 0.25
m& d
Re = total i (22.13)

using the liquid dynamic viscosity L. Its two-phase multiplier is
3.24 FH
2fr = E + (22.14)
FrH0.045We L0.035
The dimensionless factors FrH, E, F and H are as follows:
2
m& total
FrH = (22.15)
gd i H2
L fG
E = (1 x ) + x 2
2
(22.16)
G f L
F = x 0.78 (1 x )
0.224
(22.17)
0.91 0.19 0.7
G G
H = L 1 (22.18)
G L L
The liquid Weber WeL is defined as:
2
m& total di
We L = (22.19)
H
in which Friedel used the homogeneous density H based on vapour quality:
1
x 1 x
H = + (22.20)

G L
This method is typically recommended when the ratio of (L/G) is less than 1000 and is
applicable to vapour qualities from 0 x 1.
3.2 Lockhart-Martinelli
It is one of the oldest methods for horizontal piping but not so accurate. Despite having
vertical pipes, we use the model because the vertical ones form the minor part of the system
piping. The pressure drop related to friction is directly calculated by multiplying the single
phase pressure drop with a two phase coefficient: L2 and G2:
168
dp 2 dp 2 dp P P
F = L F = G F = WG G + WL L (22.21)
dx dx SL dx SG A A
2
dp u
F = f SG G SG (22.22)
dx SG 2
dp u2
F = f SL L SL (22.23)
dx SL 2
where
PG = part of the pipe in contact with vapour
PL,= part of the pipe in contact with the liquid
A = cross section area of the pipe
G = density of gas, kg/m3
L = density of liquid, kg/m3
f SG = friction factor of gas, dimensionless
f SL = friction factor of gas, dimensionless
The two coefficients, L and G, are related by the Lockhart-Martinelli parameter:
1
dp F 2
dx
SL f SL FL
X = = (22.24)
dp F f SG FG
dx
SG
Normally L and G are expressed like:
C 1
2L = 1 + + 2 (22.25)
X X
G = 1 + CX + X 2
2
(22.26)
with C one of the next values in table 22.2.
Table 22.2 C-values in function of the flow type in straight pipes.
Liquid Gas C
Turbulent Turbulent 20
Laminar Turbulent 12
Turbulent Laminar 10
Laminar Laminar 5
3.2.1 C-value for valves or fittings.
Chisholm (1965b) has examined pressure drops of two-phase flows and found that the
coefficient C2 in the equation for C is a function of relative radius R/D where R is the radius
of the bend and D is the pipe diameter. He proposed (1980):
169
2 .2
C2 = 1 +
k fo (
2+ R
D
) (22.27)
where kfo is the loss coefficient for all the flow assumed to be single-phase liquid.
Also following values can be considered for C2:
90 bends C2 = 1+35 D/L =1

180 bends C2 = 1+20 D/L =1
Tees C2 =1.75 =1
Gate valve C2 =1.5
Globe valve C2 =2.3
A most convenient expression (Chisholm and Sutherland 1969) for the coefficient C is
h h
0.5

0.5
h
0.5
C = + (C 2 ) fg g + f (22.28)
hg h f h g

(
where = 0.5 2 2 n 2 . )
For rough tubes (n = 0) = 1; for smooth tubes (n = 0.25) = 0.68.
3.3 Flow pattern depended model
To implement the map (figure 1), one first determines the Martinelli parameter X and FrG.
Using these two parameters on the top graph of figure 1, if their coordinates fall in the annular
flow regime, then the flow pattern is annular. If the coordinates of FrG and X fall in the lower
left zone of the top graph, then K is calculated. Using K and X in the middle graph, the flow
regime is identified as either stratified-wavy or as fully stratified. If the coordinates of FrG and
X fall in the right zone on the top graph, then T is calculated. Using T and X in the bottom
graph, the flow regime is identified as either bubbly flow or intermittent (plug or slug) flow.
1
G 2
FrG = u SG (22.29)
( L G )Dg
1
K = FrG Re SL2 (22.30)
1
dp 2
dz
T = L
(22.31)
g ( L G )

170
1
dp F 2
dx
SL
X = (22.32)
dp F
dx
SG
Fig. 22.1 Two-phase flow map of Taitel and Dukler (1976) for horizontal tubes.
In our case the two-phase flow only occurs in the evaporator and the piping that connects the
exit of the evaporator and the separator. Dependent of the vapour fraction stratified, stratified
wavy or annular flow appears.
171
3.3.1 The stratified flow pattern
Fig. 22.2 Stratified flow cross-section
The stratified flow pattern is characterized by the separation of the liquid and vapor phases by
a smooth interface. This flow pattern occurs at very low mass velocities when the Kelvin-
Helmholtz instability criterion is counter-balanced by the viscous forces. Figure 22.2 depicts
the cross-section of a stratified two-phase flow. The areas designated by A are cross-sectional
areas and S refers to the characteristic lengths (perimeter and interface). The internal diameter
is D. The subscripts refer to vapor (V), liquid (L) and the interface (i).
The geometrical relationships and flow characteristics of this flow pattern are as follows:
hL
wet = 2 arccos1 2 (22.33)
D
sen( wet )
L = wet (22.34)
2
D
PL = wet (22.35)
2

Pi = sen wet D (22.36)
2
PG = D PL (22.37)
uS
u= (22.38)

4 L A
DL = (22.39)
PL
4 G A
DG = (22.40)
PG + PL
172
uD
Re = (22.41)

u 2j
Wj = f j , j = L, G (22.42)
2
PG P
p frict = WG + WL L (22.43)
A A
3.3.2 The stratified-wavy flow pattern
Fig. 22.3 Three different stratified-wavy flow patterns of pure ammonia flowing at x =
0.20 and G = 26 kg/m2s, x = 0.20 and G = 60 kg/ m2s
and x = 80 and G = 41 kg/ m2s.
Fig. 22.4 Stratified-wavy flow cross-section
173
The stratified-wavy flow pattern is characterized by a wavy interface of the liquid as shown in
Figure 22.4. This flow pattern is a transitional one where waves exist but these are of a
reduced magnitude and are not able to reach the top of the tube. In a cross-sectional view, the
interface is curved. We use the same calculations and geometric parameters as for stratified
flow.
3.3.3 The annular flow pattern
The annular flow pattern is obtained when the liquid wets the entire tube periphery with the
vapor flowing at the center of the tube. Due to gravity, the film thickness is not uniform
around the periphery but is thicker at the bottom. The annular flow pattern has been
considered to be reached when the motion of the liquid flowing at the top of the tube.
Fig. 22.5 Annular flow pattern of pure ammonia flowing at x = 0.81

and G = 122 kg/m2s
Fig. 22.6 Annular flow cross-section.
174
4 Flow pattern map transition data
4.1 The stratified to stratified-wavy transition
The transition is obtained when the pressure and the shear work on a wave overcome the
viscous dissipation of the wave. Steiner formulated following expression in terms of void
fraction:
2 (1 )
1
3
G strat = 800 2 v g ( l v ) l cos v (22.44)
x (1 x )
where
= the void fraction, dimensionless
l = the fluid viscosity, Pa.s
x = vapor quality, dimensionless
where cos v = 1 for horizontal flow.
4.2 The stratified-wavy to annular transition
The transition is obtained when the waves wet the whole inside pipe surface. Zrcher et al.
proposed (22.45):
1
G2
( )
2
16 AV3 gD l v 2 G1 Fr
x 2 0.97 2
G wavy = 1 + (1 x ) + 50 75 exp
2 x 2 1 (2h 1)2 x(1 x )
2
25hl We L
l
where
hl = liquid height in the tube, m
D = pipe inside diameter, m
= the void fraction
G = mass velocity, kg/m2s
x = vapor quality, dimensionless
with Br = 3.0 Pa.
The Weber number We is the ratio of inertia to surface tension forces. The Froude number Fr
is a ratio of inertia to gravity forces. Both are expressed with the diameter of the tube as
characteristic length.
1 u l2
Fr = = (22.46)
We L gD u l l D gD l
2 2
q&
G1 = 48.24 (22.47)
q& krit
175
q&
G 2 = 9.65 + 1.053 (22.48)
q& krit
q crit = 0.131 v hevap [g ( l v ) ]
1
4 (22.49)
where
= surface tension [N/m]
q = heat flux, W/m2
5 Flow pattern map transition calculation
Table 22.3 lists the properties of ammonia necessary to find the vapour qualities in the
transition region between two flow patterns.
Table 22.3 Properties of saturated ammonia (T = -25C)
L, kg/m3 671,55
G, kg/m3 1,29
L, 10-4 Pa.s 2.28
G, 10-6 Pa.s 8.30
, 10-3 N/m 48.03
mtot, kg/m2s 48.76
q, W/m2 1213.35
hevap, kJ/kg 1342.66
Formula (22.8) and (22.44) give a vapour quality x = 0.05.
Formula (22.8) and (22.45) through (22.49) give a vapour quality x = 0.24.
6 Pressure loss calculations
6.1 Pressure loss in pipes, valves and fittings with single-phase flow
Table 22.3 lists the pressure drops for the several pipes that conduct a single-phase flow. The
liquid line between condenser and hp-receiver has the same dimensions and fittings as the
liquid line between the hp-receiver and the intercooler. It also conducts ammoniac in the
same thermodynamic state so it has the same pressure losses.
176
Table 22.3 Piping properties with their pressures losses
Suction Discharge Suction Discharge Liquid Liquid Liquid

line (low line (low line line (high line line overfeed
pressure) pressure) (high pressure) between between line
pressure) condenser intercooler
and hp- and liquid
receiver separator
Length,m 7 7 7 7 7 7 38.5
Di,mm 154.05 102.26 102.26 77.93 35.05 35.05 52.50
90 6 6 6 6 6 6 8
standard
elbows
k(90 0.29 0.31 0.31 0.34 0.46 0.46 0.38
standard
elbow)
stop 1 1 1 1 1 1 1
valves
k(stop 2.1 2.1 2.1 2.2 3.7 3.7 2.5
valve)
, mm 0.09 0.09 0.09 0.09 0.09 0.09 0.09
-6
,10 9.53 0.125 0,107 0.128 128.59 174.78 249.46
Pa.s
, kg/m3 1.30 3.30 4.45 8.39 588.24 628.93 671.14
m, kg/s 0.489 0.489 0.699 0.699 0.699 0.489 1.957
u, m/s 20.21 18.02 19.12 10.85 1.23 0.81 1.35
5
Re, 10 4.24 4.86 8.14 8.92 1.97 1.02 1.90
f, 10-2 1.81 2.00 1.94 2.06 2.56 2.26 2.35
p, Pa 1237.18 2845.72 4301.17 4843.35 5024.65 2327.87 14484.79
Before the liquid leaving the condenser reaches the pressure valve it has to flow through two
pipes: the pipe connecting the condenser with the high-pressure receiver and the pipe between
the high-pressure receiver and the intercooler, due to the fact that the valve is situated close to
the intercooler. The total pressure loss in these two pipes is p = 10049.30 Pa. The saturation
temperature related to the new pressure is 34.74C. To prevent partial evaporation we have to
subcool the liquid in the condenser. We subcool the liquid ammoniac 1C.
6.2 Pressure loss in horizontal pipes that conducts a two-phase flow
6.2.1 Pressure drop in the evaporator
The evaporator pipes conduct a two-phase flow. The evaporator not only conducts this flow,
by transferring heat it forces the vapor quality to increase. That is why we have to be more
cautious to calculate the pressure losses. We subdivide all the tubes of the evaporator in 20
equally sized pieces. Every part receives heat of the brine wherein it is immerged. So in every
part the vapour quality increases. This results in a pressure drop that changes throughout the
whole evaporator. The pressure and vapour quality change in turn influence the heat transfer.
We use the Friedel correlation to calculate the pressure drop. A more detailed description is
given in chapter 15. The pressure drop is 0.157 bar.
177
We also calculate the pressure drop via the flow pattern map method with the aim to compare
its result with the one calculated above.
First we predict the flow pattern with the map of Taitel and Dukler (see figure 22.1). We list
the ammoniac properties, necessary for the calculations in table 22.4.
Table 22.4 Properties of saturated ammoniac (T = -25C)
L, kg/m3 671,55
G, kg/m3 1,29
L, 10-4 Pa.s 2.28
G, 10-6 Pa.s 8.30
We calculate FrH (formula (22.29)) and X (formula (22.32)) to find the flow pattern at the end
of the evaporator (x = 0.25): FrH = 0.81, X = 0.16. These parameters form respectively the
ordinate and abscissa. The figure shows us that the flow regime is situated in the zone
between annular and stratified-wavy flow. But we already know that the transition appears
when the vapour quality x = 0.24. So annular flow only will occur in a small part of the
evaporator. We neglect this and assume stratified flow in the entire evaporator. We presume a
constant vapour quality in the first and in the second halve. In the first halve we take x =
0.125 and in the latter part x = 25. We do not take the turning bends into account. We just
want to estimate the overall pressure loss. Calculating the formulas (22.33) through (22.43)
give p1 = 4801.45 Pa and p2 = 11280.53 Pa. This gives a total pressure drop of p =
16081.98 Pa = 0.16 bar.
Both pressure drops calculated with two different methods mentioned above are almost the
same (0.157 0.16).
6.2.2 Pressure drop in the pipe between the evaporator and liquid separator
Here we use the flow pattern map method. We calculate FrH (22.29) and X (22.32) to find the
flow pattern (x = 0.25): FrH = 0.72, X = 0.15. Figure 1 predicts a stratified-wavy flow. Like
above we calculate the formulas (22.33) through (22.43) to find pfrict = 6853.10 Pa. This
pressure drop is due to friction in the pipes. We calculate the pressure drop due to fittings and
bends, enumerated in chapter 19 with formulas (22.21) through (22.28). For the 90 bends we
have D/L = 2/ and we compare the pressure losses in the angle valve with the losses in a
globe valve. Because we also miss the C2 -value for 45 bends we replace them by 90 bends.
The C2 values for bends, etc. are given for = 1 (rough pipes). We have half smooth pipes
so a safety factor to define the circulation pump is automatically incorporated.
p90 bends = 8519.24

p45 bends = 1248.51
pvalve = 1278.37
ptee = 327.64
The total pressure loss is ptotal = 18226.62 Pa.
178
Chapter 23: Thermal properties of tuna and

mackerel
.
1 Introduction
The thermal properties of the fish must be known to perform the various heat transfer
calculations involved in designing storage and refrigeration equipment and estimating process
times for refrigerating. Because the thermal properties of foods strongly depend on chemical
composition and temperature, and because many food items are available, it is nearly
impossible to find the thermal properties of foods and beverages for all possible conditions
and compositions. However, composition data for foods and beverages are readily available
from sources such as Holland et al. (1991) and USDA (1975). These data consist of the mass
fractions of the major components found in food items. The thermal properties of food items
can be predicted by using this composition data in conjunction with temperature-dependent
mathematical models of the thermal properties of the individual food constituents. Thermo-
physical properties of foods that are often required for heat transfer calculations include
density, specific heat, thermal conductivity and thermal diffusivity.
2 Thermal properties of food constituents
Constituents commonly found in food items include water, protein, fat, carbohydrate, fibre,
and ash. Choi and Okos (1986) developed mathematical models for predicting the thermal
properties of these food components as functions of temperature in the range of 40C to
150C (see Table 23.1); they also developed models for predicting the thermal properties of
water and ice (Table 23.2).
Table 23.1 Thermal properties of food components as function of temperature
Thermal property Food Component Thermal property model

Thermal conductivity, W/mK Protein K=1.7881*10-1+1.1958*10-3t-2.7178*10-6t2
Fat K=1.8071*10-1-2.7604*10-3t-1.7749*10-7t2
Carbohydrate K=2.0141*10-1+1.3874*10-3-4.3312*10-6t2
Fiber K=1.8331*10-1+1.2497*10-3t-3.1683*10-6t2
Ash K=3.2962*10-1+1.4011*10-3t-2.9069*10-6t2
Density, kg/m3 Protein =1.3299*103-5.1840*10-1t
Fat =9.2559*103-4.1757*10-1t
Carbohydrate =1.5991*103-3.1046*10-1t
Fiber =1.3115*103-3.6589*10-1t
Ash =2.4238*103-2.8063*10-1t
Specific heat, kJ/kgK Protein cp=2.0082+1.2089*10-3t-1.3129*10-6t2
Fat cp=1.9842+1.4733*10-3t-4.8008*10-6t2
Carbohydrate cp=1.5488+1.9625*10-3t-5.9399*10-6t2
Fiber cp=1.8459+1.8306*10-3t-4.6509*10-6t2
Ash cp=1.0926+1.8896*10-3t-3.6817*10-6t2
Soarce Choi and Okos (1986)
179
Table 23.2 Thermal properties of water and ice as function of temperature
Thermal property Thermal property model

Water Thermal conductivity, W/mK Kw=5.7109*10-1+1.7625*10-3t-6.7036*10-6t2
Density, kg/m3 w=9.9718*102+3.1439*10-1t-3.7574*10-3t2
Specific heat, kJ/kgK Cw=4.0817-5.3062*10-3t +9.9516*10-6t2
(for temperature range of 40C to 0C)
Specific heat, kJ/kgK 4.1762-9.0864*10-5t+5.4731*10-6t2
(for temperature range of 0C to 150C)
Ice Thermal conductivity, W/mK Kice=2.2196-6.2489*10-3t+1.0154*10-6t2
Density, kg/m3 ice=9.1689*102-6.0833*10-9t+9.5037*10-11t2
Specific heat, kJ/kgK Cice=2.0623+6.0769*10-3t
Source: Choi and Okos
3 Thermal properties of food
In general, the thermo-physical properties of a food are well behaved when the temperature of
the food is above its initial freezing point. However, below the initial freezing point, the
thermo-physical properties vary greatly due to the complex processes involved during
freezing. The initial freezing point of a food is somewhat lower than the freezing point of pure
water due to dissolved substances in the moisture in the food. At the initial freezing point, a
portion of the water within the food crystallizes and the remaining solution becomes more
concentrated. Thus, the freezing point of the unfrozen portion of the food is further reduced.
The temperature continues to decrease as the separation of ice crystals increases and the
concentration of the solutes in solution and depresses the freezing point further. Thus, the ice
and water fractions in the frozen food depend on temperature.
Because the thermo-physical properties of ice and water are quite different, the thermo-
physical properties of frozen foods vary dramatically with temperature. In addition, the
thermo-physical properties of the food above and below the freezing point are drastically
different.
3.1 Water Content
Because water is the predominant constituent in most foods, water content significantly
influences the thermo-physical properties of foods.
3.1.1 Initial freezing point
Foods and beverages do not freeze completely at a single temperature, but rather over a range
of temperatures. Thus, there is not a distinct freezing point for foods and beverages, but an
initial freezing point at which the crystallization process begins.
The initial freezing point of a food or beverage is important not only for determining the
proper storage conditions for the food item, but also for calculating thermo-physical
properties.
3.1.2 Ice fraction
To predict the thermo-physical properties of frozen foods, which depend strongly on the
fraction of ice within the food, the mass fraction of water that has crystallized must be
180
determined. Below the initial freezing point, the mass fraction of water that has crystallized in
a food item is a function of temperature. Tchigeov (1979) proposed an empirical relationship
to estimate the mass fraction of ice:
1.105 x wo
xice = (23.1)
0.8765
1+
ln(t f t + 1)
where
tf = initial freezing point, C
t = temperature of the fish, C
3.2 Density
Modelling the density of foods and beverages requires knowledge of the food porosity, as
well as the mass fraction and density of the food components. The density of foods and
beverages can be calculated accordingly:
1
= (23.2)
xi
i
where xi is the mass fraction of the food constituents, and i is the density of the food
constituents.
3.3 Specific heat
Specific heat is a measure of the energy required to change the temperature of a food item by
one degree. Therefore, the specific heat of foods or beverages can be used to calculate the
heat load imposed on the refrigeration equipment by the cooling or freezing of foods and
beverages. In unfrozen foods, specific heat becomes slightly lower as the temperature rises
from 0C to 20C. For frozen foods, there is a large decrease in specific heat as the
temperature decreases.
3.3.1 Unfrozen food
The specific heat of a food item, at temperatures above its initial freezing point, can be
obtained from the mass average of the specific heats of the food components. Thus, the
specific heat of an unfrozen food item cu may be determined as follows:
cu = c i x i (23.3)
where ci is the specific heat of the individual food components and xi is the mass fraction of
the food components.
181
Chen presents a simpler model for the specific heat of an unfrozen food item (1985). If
detailed composition data is not available, the following expression for the specific heat of an
unfrozen food item can be used:
cu = 4.19 2.30 x s 0.628 x s3 (23.4)
where cu is the specific heat of the unfrozen food item in kJ/kgK and xs is the mass fraction of
the solids in the food item.
x s = 1 x wo (23.5)
with xwo the initial water mass fraction.
3.3.2 Frozen food
Below the freezing point of the food item, the sensible heat due to temperature change and the
latent heat due to the fusion of water must be considered. Because latent heat is not released at
a constant temperature, but rather over a range of temperatures, an apparent specific heat must
be used to account for both the sensible and latent heat effects. Chens model can be used to
determine the specific heat and has the following form:
x s RT02
c a = 1.55 + 1.26 x s + (23.6)
M st 2
where
ca = apparent specific heat, kJ/kgK
xs = mass fraction of solids
R = universal gas constant
T0 = initial freezing point of water = 273.2 K
Ms = relative molecular mass of soluble solids in food item
T = food temperature, C
The relative molecular mass of the soluble solids within the food item may be estimated as
follows:
x s RT02
Ms = (23.7)
( x w0 xb ) L0 t f
where L0 = latent heat of fusion of water at 273.2 K = 333.6 kJ/kg
If the relative molecular mass of the soluble solids is unknown, Equation (23.7) may be used
to estimate the molecular mass. Substituting Equation (23.7) into Equation (23.6) yields
( x w 0 xb ) L0 t f
c a = 1.55 + 1.26 x s (23.8)
t2
182
3.4 Enthalpy
The change in enthalpy of a food item can be used to estimate the energy that must be added
or removed to effect a temperature change. Above the freezing point, enthalpy consists of
sensible energy, while below the freezing point, enthalpy consists of both
sensible and latent energy. Enthalpy may be obtained from the definition of constant-pressure
specific heat:
H
cp = ( )p (23.9)
T
where cp is constant pressure specific heat, H is enthalpy, and T is temperature. Mathematical

models for enthalpy may be obtained by integrating expressions of specific heat with respect
to temperature.
3.4.1 Unfrozen food
For food items that are at temperatures above their initial freezing point, enthalpy may be
obtained by integrating the corresponding expression for specific heat above the freezing
point. Thus, the enthalpy of an unfrozen food item H may be determined by integrating
equation (23.3) as follows:
H = H i xi = ci xi dT (23.10)
where Hi is the enthalpy of the individual food components and xi is the mass fraction of the
food components.
In the case of the method of Chen (1985), the enthalpy of an unfrozen food may be obtained
by integrating Equation (23.3):
H = H f + (t t f )(4.19 2.30 x s 0.628 x s3 ) (23.11)
The enthalpy at the initial freezing point Hf may be estimated by evaluating either Equation
(23.10) or (23.11) at the initial freezing temperature of the food as discussed in the following
section.
3.4.2 Frozen Foods
For food items below the initial freezing point, mathematical expressions for enthalpy may be
obtained by integrating the previously mentioned apparent specific heat models.
By integrating Equation (23.8) between a reference temperature tr (233.2 K) and the food
temperature T, Chen (1985) obtained the following expression for enthalpy below the initial
freezing point:
x s RT02
H = (t t r )(1.55 + 1.26 x s + ) (23.12)
M s tt r
183
Substituting equation (23.7) for the relative molecular mass of the soluble solids Ms simplifies
Chens method as follows:
( x wo xb ) L0 t f
H = (t t r )(1.55 + 1.26 x s + ) (23.13)
tt r
3.5 Thermal conductivity
Thermal conductivity relates the conduction heat transfer rate to the temperature gradient. The
thermal conductivity of a food depends on such factors as composition, structure, and
temperature.
Numerous researchers have proposed the use of parallel and perpendicular (or series) thermal
conductivity models based on analogies with electrical resistance (Murakami and Okos 1989).
The parallel model is the sum of the thermal conductivities of the food constituents multiplied
by their volume fractions:
k = xiv k i (23.14)
where xi is the volume fraction of constituent i. The volume fraction of constituent i can be
found from the following equation:
xi
i
x = v
i (23.15)
xi
( i )
The perpendicular model is the reciprocal of the sum of the volume fractions divided by their
thermal conductivities:
1
k= (23.16)
xiv
( ki
)
These two models have been found to predict the upper and lower bounds of the thermal
conductivity of most food items.
4 Calculations of the thermal properties of tuna and mackerel and their food
components
Table 23.3 lists the food components of tuna and mackerel. These are necessary to calculate
the other food properties.
184
Table 23.3 Mass fraction (10-2) of the food components of tuna and mackerel
Tuna Mackerel
xwo 68.72 64.20
x protein 23.96 19.25
x fat 5.53 14.54
xcarbohydrate 0 0
x fiber 0 0
x ash 1.81 2.00
Table 23.4 lists all the properties of tuna and mackerel. We calculate all these properties with
the formulas given above. We need these properties to calculate the freezing time, the
convection coefficient in the blast freezer, etc.
185
Table 23.4 Thermal properties of tuna, mackerel and their food components at entrance
(11C), final (-18C) and fully frozen temperature (-40C).
Temperature, C
11 -18 -40
tuna mackerel tuna mackerel tuna mackerel
xice , 10-2 0 0 57.93 54.13 61.25 57.23
xw , 10-2 68.72 68.72 10.78 10.07 7.46 6.97
kmoisture, 10-1 W/mK 5.90 5.37 4.90
kice, 10-1 W/mK 23.65 26.32
kprotein, 10-1 W/mK 1.92 1.56 1.27
kfat, 10-1 W/mK 1.50 2.30 2.91
kash, 10-1 W/mK 3.45 3.03 2.69
moisture, kg/m3 996.76 995.91 991.04
ice, kg/m3 919.24 922.12
protein, kg/m3 1324.20 1339.23 1350.64
fat, kg/m3 921.00 933.11 942.29
ash, kg/m3 2420.71 2428.85 2435.03
cmoisture, kJ/kgK 4.18 4.50 5.89
cice, kJ/ kgK 1.95 1.82
cprotein, kJ/kgK 2.02 1.99 1.96
cfat, kJ/kgK 2.00 1.96 1.92
cash, kJ/kgK 1.11 1.06 1.01
v
x moisture , 10-2 73.52 67.40 11.00 10.14 7.66 7.06
v
xice , 10-2 64.05 59.01 67.56 62.31
v -2 19.29 15.21 18.18 14.41 18.04 14.31
x protein , 10
x vfat , 10-2 6.40 15.21 6.02 15.62 5.96 15.49
v
x ash , 10-3 0.80 8.66 0.76 8.26 0.75 8.26
k(perpendicular 0.49 0.327 0.520 0.484 1.86 0.482
model), W/mK
k(parallel 0.38 0.454 1.62 1.51 0.48 1.74
model),W/mK
k, W/mK 0.43 0.391 1.07 0.998 1.17 1.11
cu , kJ/kgK 3.43 5.38
ca, kJ/kgK 3.28 3.28 2.22 2.26
H, kJ 358.75 350.65 56.03 56.69 0 0
186
Chapter 24: Heat transfer between air and tuna in the blast freezer
Chapter 24: Heat transfer between air and tuna in

the blast freezer
1 Introduction
In the blast freezer we hang the tuna at bars. To minimize space lost, the area between two
fishes is minimized to 3 cm approximately. Cool air flows around the tuna to decrease its
temperature. To calculate the freezing time, necessary to reach the end temperature we need
to know the convection coefficient. We discuss its calculation in this chapter.
2 Heat convection coefficient
To simplify the calculations we assume cylindrical shaped tuna. The heat transfer is not only
affected by the incoming air stream, though is also dependent of the mutual position of the
fish. To calculate the convection coefficient, we choose two extreme situations: (a) aligned
and (b) alternated. Afterwards we take the average to predict the heat transfer coefficient.
For the heat transfer correlations around cylinders, the Reynolds number is given by:
U m D
Re Dm = (24.1)

where
Um = the speed of the air in the minimum free space between the cylinders, m/s
D = the diameter of a cylinder, m
= the dynamic viscosity, Pa.s
= the specific mass, kg/m3
The surface temperature of the fish is necessary to determine ReDm because the properties in
formula (24.1) are temperature dependent. We predict a constant surface temperature. With
this temperature we calculate the film temperature. We take the fluid properties at this
temperature. With these properties and other data specified further, we calculate the
convective heat transfer coefficient. We insert the coefficient in formula (24.2) to find the
correct surface temperature.
Q=
[
m c p1 (Tin T f ) + c p 2 + c p 3 (T f Tend ) ] = hA(T Ts ) (24.2)

t
where
Q = average heat consumed by the fish, J/s
cp1 = specific heat above initial freezing point (Tf), J/kgK
cp2 = specific heat during freezing, J/kg
cp3 = specific heat below initial freezing point, J/kgK
h = convective heat transfer coefficient, J/m2K
A = heat transferring surface, m2
T = air temperature, K
187
Ts = surface temperature of the fish, K

t = freezing time, s
In the case of the aligned position, Um is given by:
ST
Um =U (24.3)
ST D
In the case of the alternated position, Um is given by (24.3) when 2(S D D ) > S T D , with
1
2 ST 2 2
S D = S L + . When 2(S D D ) < S T D , Um is given by formula (24.4).
2
ST
Um =U (24.4)
2(S D D )
Because of an extensive study, Zhukauskas was able to formulate a correlation to predict the
overall heat coefficient.
1
Pr 4
Nu = C Re n
Dm Pr 0.36
(24.5)
Prs
0.7 < Pr < 500

10 < Re Dm < 10
6

C and n from table 1

Propiedades at t film except for Pr , Prs
Pr at t , Pr at t
s s
Note: t is the air temperature, t s is the fish surface temperature
With Nu we find the convection coefficient h:
Nuk
h= (24.6)
l
Table 24.1 Coefficient C and n in function of ReDm and the fish mutual position
Interval of ReDm Aligned Alternated

C n C n
1000 - 2 105 ST/SL>0.7:0.27 0.63 ST/SL<2:0.35 (ST/SL)1/50.6
2 105 - 106 0.021 0.84 0.022 0.84
Because the correlation (24.5) is related to 20 rows, in most cases a correction factor, that
represents the deviation in amount of rows, has to be used. In our case the final amount of
rows is approximately 20. We neglect the correction factor.
188
2.1 Algorithm (1) to determine the heat convection coefficient
Step 1: calculate Um, (24.3) or (24.4)
Step 2: calculate ReDm and determine Pr, Prs and Pr, (24.1)
Step 3: find C and n, using table 24.1.
Step 4: calculate Nu, (24.5)
Step 5: determine h, (24.6)
3 Heat convection coefficient calculation.
Data: The tuna is hung in a blast freezer where the air temperature is -40C. The fish initial
temperature of 11C is reduced to -18C. The air velocity is about 3.55 m/s and the distance
between two fishes is approximately 3 cm for both mutual positions. The fish diameter is 0.20
m and the length is 0.50 m.
To determine the properties and dimensionless numbers, the fish surface temperature is
necessary. First, we predict a value close to the air temperature because we assume a higher
heat convection coefficient than conductivity coefficient. We start with Ts equal to the air
temperature and go through the calculations using the algorithm (1).
Table 24.2 Results of algorithm 1 for the aligned and altered arrangement of the fish.
Aligned Altered
Ts, C -40 -40
T , C -40 -40
U , m/s 3.55 3.55
D, m 0.2 0.2
ST, m 0.23 0.23
SD, m 0.325 0.230
SL, m 0.23 0.199
Um, m/s 9.07 9.07
Tf, K 233.00 233.00
(Tf), kg/m3 1.51 1.51
(Tf), 10-5 Pa.s 1.55 1.55
k(Tf), 10-2 W/mK 2.09 2.09
Re, 103 176.76 176.76
Pr (Tf), 10-1 7.27 7.27
Pr, 10-1 7.24 7.24
Prs (Ts), 10-1 7.27 7.27
ST/SL 1 1.15
C 0.27 0.36
n 0.63 0.6
Nu 486.34 486.34
h, W/m2K 50.80 47.17
189
Now we take the average of the two convective heat coefficients to predict the natural
circumstances:
50.80 + 47.17 W
h= = 49.99 2
2 m K
Experience thought us to expect a cooling time of 24 hours approximately. Now we can

predict more precise the fish surface temperature. Formula (24.7) gives the average heat
exchange. It says that the average heat exchange is the sum of three parts: a) heat extracted to
cool the fish to its initial freezing point, b) the latent heat of solidification and c) heat
extracted to freeze the fish until it reaches its final temperature. The three parts are divided by
the total freezing time.
Q=
[ ]
m c1 (T1 T f ) + hif + c3 (T f T2 )
(24.7)
24 * 3600
where
m = mass of one tuna, kg
c1 = specific heat of product above freezing, kJ/kgK
t1 = initial temperature of product above freezing, C
t2 = final temperature of product below freezing, C
tf = freezing temperature of product, C
hif = latent heat of fusion of product, kJ/kg
c3 = specific heat of product below freezing, kJ/kgK
We can also write the average heat exchange like formula (24.8). It relates the average heat
exchange to the heat convection between the cool air and the fish.
Q = hA(Ts T ) (24.8)
Temperature Ts can be derived from combining formulas (24.7) and (24.8):
Ts =
[
m c1 (T1 T f ) + hif + c3 (T f T2 ) 1 ]+ T (24.9)
24 3600 hA
In table 24.3 we list the new results of the algorithm (1).
190
Table 24.3 Results of algorithm (1) whit the correct fish surface temperature.
Aligned Altered
Ts, C -35.24 -35.24
T , C -40 -40
U , m/s 3.55 3.55
D, m 0.2 0.2
ST, m 0.23 0.23
SD, m 0.325 0.230
SL, m 0.23 0.199
Um, m/s 9.07 9.07
Tf, K 235.38 235.38
(Tf), kg/m3 1.49 1.49
(Tf), 10-5 Pa.s 1.55 1.55
k(Tf), 10-2 W/mK 2.11 2.11
Re, 103 174.39 174.39
Pr (Tf), 10-1 7.26 7.26
Pr(-40C), 10-1 7.24 7.24
Prs (Ts), 10-1 7.26 7.26
ST/SL 1 1.15
C 0.27 0.36
n 0.63 0.6
Nu 482.30 447.98
h, W/m2K 50.84 47.22
Now we take the average of the two convective heat coefficients to predict the natural
circumstances:
50.84 + 47.22 W
h= = 49.53 2
2 m K
As we see, the convection coefficients dont differ a lot. This value is further used to calculate
the freezing time of the tuna in the blast freezer. There we calculate a freezing time close to
24 hours. We decide that a recalculation of Ts according to formula (24.9) is redundant.
191
1 Introduction
Preservation of food is one of the most significant applications of refrigeration. Cooling and
freezing of food effectively reduces the activity of microorganisms and enzymes, thus
retarding deterioration. In addition, crystallization of water reduces the amount of liquid water
in food items and inhibits microbial growth.
In order for cooling and freezing operations to be cost-effective, refrigeration equipment
should fit the specific requirements of the particular cooling or freezing application. The
design of such refrigeration equipment requires estimation of the cooling and freezing times
of foods and beverages, as well as the corresponding refrigeration loads.
Numerous methods for predicting the cooling and freezing times of foods and beverages have
been proposed, including those based on numerical, analytical, and empirical analysis. The
designer is faced with the challenge of selecting an appropriate estimation method from the
many available methods. We review a selected procedure for estimating the freezing time of
fish.
2 Calculation methods
2.1 Thermodynamics of freezing
The cooling and freezing of food is a complex process. Prior to freezing, sensible heat must
be removed from the food to decrease its temperature from the initial temperature to the initial
freezing point of the food. This initial freezing point is somewhat lower than the freezing
point of pure water due to dissolved substances in the moisture within the food. At the initial
freezing point, a portion of the water within the food crystallizes and the remaining solution
becomes more concentrated. Thus, the freezing point of the unfrozen portion of the food is
further reduced. As the temperature continues to decrease, the formation of ice crystals
increases the concentration of the solutes in solution and depresses the freezing point further.
Thus, the ice and water fractions in the frozen food depend on temperature. The cooling and
freezing of foods and beverages can be described via the Fourier heat conduction equation:
T 1 T T T
= k + k + k (25.1)
c x x y y z z
where T is temperature, is time, is the density of the food, c is the specific heat of the food,
k is the thermal conductivity of the food and x, y, and z are the coordinate directions. For
regularly shaped food items with constant thermo-physical properties, uniform initial
conditions, constant external conditions and a prescribed surface temperature or a convection
boundary condition, exact analytical solutions for Equation (25.1) exist. However, in our case,
the fish is irregularly shaped with temperature-dependent thermo-physical properties.
Therefore, we cannot derive an exact analytical solution for the cooling and freezing times of
our fish. Accurate numerical estimations of the cooling and freezing times of the fish can be
obtained using appropriate finite element or finite difference computer programs. However,
192
the effort required to perform this task makes it impractical for our purpose. In addition, two-
dimensional and three-dimensional simulations require time-consuming data preparation and
significant computing time.
As mentioned at the beginning of this chapter, the freezing of foods is not an isothermal
process but rather occurs over a range of temperatures. In the following section, the basic
freezing time estimation method developed by Plank is discussed first, followed by a
discussion of those methods that modify Planks equation. The discussion then focuses on
those freezing time estimation methods in which the freezing time is calculated as the sum of
the pre-cooling, phase change, and sub-cooling times. The last section deals with freezing
time estimation methods for irregularly shaped food items.
2.2 Planks Equation
One of the most widely known simple methods for estimating freezing times of foods and
beverages is that developed by Plank (1913, 1941). In this method, convective heat transfer is
assumed to occur between the food item and the surrounding cooling medium. In addition, the
temperature of the food item is assumed to be at its initial freezing temperature and that this
temperature is constant throughout the freezing process. Furthermore, a constant thermal
conductivity for the frozen region is assumed. Planks freezing time estimation is as follows:
Lf PD RD 2
= + (25.2)
T f Tm h ks
where Lf is the volumetric latent heat of fusion, Tf is the initial freezing temperature of the
food, Tm is the freezing medium temperature, D is the thickness of the slab or the diameter of
the sphere or infinite cylinder, h is the convective heat-transfer coefficient, ks is the thermal
conductivity of the fully frozen food, and P and R are geometric factors. For the infinite slab,
P = 1/2 and R = 1/8. For a sphere, P = 1/6 and R = 1/24; and for an infinite cylinder, P =
and R = 1/16. Planks geometric factors indicate that an infinite slab of thickness D, an
infinite cylinder of diameter D and a sphere of diameter D, if exposed to the same conditions,
would have freezing times in the ratio of 6:3:2. Hence, a cylinder freezes in half the time of a
slab and a sphere freezes in one-third the time of a slab.
2.3 Modifications to Planks Equation
Various researchers have noted that Planks method does not accurately predict freezing times
of foods and beverages. This is due, in part, to the fact that Planks method assumes that
freezing of foods takes place at a constant temperature, and not over a range of temperatures
as is the case in actual food freezing processes. In addition, the thermal conductivity of the
frozen food is assumed to be constant; but in reality, the thermal conductivity varies greatly
during freezing. Another limitation of Planks equation is that it neglects the removal of
sensible heat above the freezing point. Furthermore, Planks method only applies to infinite
slabs, infinite cylinders, and spheres. Subsequently, researchers have developed improved
semi-analytical/empirical cooling and freezing time estimation methods that account for pre-
cooling and sub-cooling times, non-constant thermal properties, irregular geometries, and
phase change over a range of temperatures.
193
Cleland and Earle (1977, 1979a, 1979b) incorporated corrections to account for the removal
of sensible heat both above and below the initial freezing point of the food as well as
temperature variation during freezing. Regression equations were developed to estimate the
geometric parameters P and R for infinite slabs, infinite cylinders, spheres, and rectangular
bricks. In these regression equations, the effects of surface heat transfer, pre-cooling, and final
sub-cooling are accounted for by means of the Biot number, the Plank number, and the Stefan
number, respectively. In this section the Biot number is defined as
hD
Bi = (25.3)
k
where h is the convective heat-transfer coefficient, D is the characteristic dimension, and k is

the thermal conductivity. The characteristic dimension D is defined to be twice the shortest
distance from the thermal center of a food item to its surface. For an infinite slab, D is the
thickness. For an infinite cylinder or a sphere, D is the diameter. In general, the Plank number
is defined as follows:
C l (T f Tm )
Pk = (25.4)
H
where Cl is the volumetric specific heat of the unfrozen phase and H is the volumetric
enthalpy change of the food between Tf and the final food temperature. The Stefan number is
similarly defined as
C s (T f Tm )
Ste = (25.5)
H
where Cs is the volumetric specific heat of the frozen phase.
Hung and Thompson (1983) also improved upon Planks equation to develop an alternative
freezing time estimation method for infinite slabs. Their equation incorporates the volumetric
change in enthalpy H18 for the freezing process as well as a weighted average temperature
difference between the initial temperature of the food and the freezing medium temperature.
This weighted average temperature difference T is given as follows:
Cl C
(Ti T f ) 2 (T f Tc ) 2 s
T = (T f Tm ) + 2 2 (25.6)
H 18
where Tc is the final center temperature of the food and H18 is the enthalpy change of the
food between the initial temperature and the final center temperature, assumed to be 18C.
Empirical equations were developed to estimate P and R for infinite slabs as follows:
Ste
P = 0.7306 1.083Pk + Ste(15.40U 15.43 + 0.01329 ) (25.7)
Bi
R = 0.2079 0.2656U ( Ste) (25.8)
194
where U = T/(Tf Tm) (25.9)
In these expressions, Pk and Ste should be evaluated using the enthalpy change H18. The
freezing time prediction model is
H 18 PD RD 2
= + (25.10)
T h ks
Cleland and Earle (1984) applied a correction factor to the Hung and Thompson model
[Equation (25.10)] and improved the prediction accuracy of the model for final temperatures
other than 18C. The correction to Equation (25.10) is as follows:
H 18 PD RD 2 1.65Ste Tc Tm
= + 1 ln
T T
(25.11)
T h k s ks ref m
where Tref is 18C, Tc is the product final centre temperature and H18 is the volumetric
enthalpy change between the initial temperature Ti and 18C. The weighted average
temperature difference T, Pk, and Ste should be evaluated using H18.
Because of the fact that the equations to calculate the factors P and R only referred to the
infinite slab, Cleland and Earle (1982b) also introduced a geometric correction factor, called
the equivalent heat transfer dimensionality E to calculate the freezing times of irregularly
shaped food items. The freezing time of an irregularly shaped object shape, was related to the
freezing time of an infinite slab slab via the equivalent heat transfer dimensionality as
shape = slab /E (25.12)
Cleland et al. (1987a, 1987b) developed expressions for determining the equivalent heat
transfer dimensionality of infinite slabs, infinite and finite cylinders, rectangular bricks,
spheres, and two- and three-dimensional irregular shapes. Numerical methods were used to
calculate the freezing or thawing times for these various shapes. A non-linear regression
analysis of the resulting numerical data yielded the following form for the equivalent heat
transfer dimensionality:
E = G 1 + G 2 E1 + G 3 E 2 (25.13)
where
(
E1 = X 2.32 / 11.77 ) 1 + [1 X (2.32 / )] 0.73
1.77
(25.14)
12.50
1
1
(
E 2 = X 2.32 / 21.77 ) 1 + [1 X (2.32 / )] 0.73
1.77
(25.15)
22.50
2
2
x
X ( x) = (25.16)
Bi 1.34
+x
where the geometric constants G1, G2, and G3 are given in Table 25.1 and 1 and 2 are
given by:
195
second shortest dimension of food item

1 = (25.17)
shortest dimension of food item
longest dimension of food item
2 = (25.18)
shortest dimension of food item
For three-dimensional, irregularly shaped food items, E is
E = 1 .5
[
1 + 2 + 12 (1 + 2 ) + 22 (1 + 1 ) ( 1 2 )2 ]
0.4
(25.19)
1 2 (1 + 1 + 2 ) 15
The freezing time of the infinite slab is then calculated from the formulas provided by Hung
and Thompson, described above.
Table 25.1 Geometric constants for different shapes.
Shape G1 G2 G3
Infinite slab 1 0 0
Infinite cylinder 2 0 0
Sphere 3 0 0
Squat cylinder 1 2 0
Short cylinder 2 0 1
Infinite rod 1 1 0
Rectangular brick 1 1 1
Two-dimensional irregular shape 1 1 0
Source: Cleland et al. (1987a)
2.4 Algorithm for Freezing Time Estimation of the fish
Step 1: Determine the thermal properties of tuna.
Step 2: Determine the surface heat-transfer coefficient.
Step 3: Determine the characteristic dimension D and the dimensional ratios 1 and 2 .
Step 4: Using Equations (25.3) through (25.5), calculate the Biot number, the Plank number,
and the Stefan number.
Step 5: Using Equations (25.6) through (25.11), calculate the freezing time of an infinite slab.
Step 6: Using Equations (25.13) through (25.19), calculate the equivalent heat transfer
dimensionality for a finite cylinder (we assume a cylindrical form of the fish to facilitate these
calculations, nevertheless we have incorporated a factor that involves the specific form of the
fish to determine the heat transfer coefficient).
Step 7: Using Equation (25.12), calculate freezing time of the fish.

196
3 Freezing time calculations
3.1 Freezing time for tuna
Step 1: The thermal properties of tuna have been calculated in chapter 23.
Table 25.2 Properties of tuna at three different temperatures (-40C, -18C and 11C).
Property At 40C (fully At 18C (final At 11C (initial

frozen) temperature) temperature)
Density, kg/m3 1039.05 1038.05 1090.07
Enthalpy, kJ/kg 0 56.12 358.75
Specific heat, kJ/kgK 2.22 3.28 3.43
Thermal conductivity, 1.18 1.08 0.43
W/mK
Initial freezing point = 2.2C
Step 2: The surface heat-transfer coefficient has been calculated in chapter 24.
kJ
h = 49.53
kg
Step 3:
D = 0.20m, l fish = 0.5m

0.20
1 = = 0.33
0.15
0.50
2 = = 3.33
0.15
Step 4:
Bi = 9.26
Pk = 0.15
Ste = 0.260
Step 5:
T = 57.69C
P = 2.04
R = 2.62 10 2
U = 1.37
slab = 57.69h
197
Step 6:
E = 2.16
tuna = 26.65h
3.1.1 Solution
With an air velocity of 3.55 m/s the tuna would have to stay 26.65 hours in the blast freezer to
reach its final temperature. We see that this solution is approximately the same of the 24 hours
assumed at the beginning of our project.
3.2 Freezing time for mackerel
The thermal properties of mackerel have been calculated in chapter 23. We follow the same
steps as done above for tuna. The characteristic dimensions are: longest dimension = 0.30 m,
second shortest dimension = 0.05 m, shortest dimension = 0.03 m.
We find mackerel = 8.84 h
The freezing time of mackerel is approximately 9 hours instead of 24. If we now would
modify the calculation above, we would obtain a product load for mackerel of 606 kW. This
cooling load would exceed the evaporator, compressor and condenser capacities calculated
based on the cooling load of tuna. Because the selected compressors have the restraint of 330
kW maximum cooling capacity it is possible to calculate the time needed to freeze the
mackerel at maximum power.
Subtracting transmission load, internal load, air load and the 10% safety factor from the
compressors maximum cooling capacity brings us to a maximum product load of 248.335
kW.
As we calculated before the product load of mackerel is 202.215 kW. This corresponds with
17.471 109 J in 24 hours. Dividing the total energy to remove by the maximum capacity of
the compressors gives us the minimum time required to freeze the mackerel: 19.5 hours
198
Chapter 26: The freezing time of sardine
1 Introduction
Fish is frozen to achieve a longer conservation time and better food quality when consumed.
However we have to be precautious. When fish is cooled to quickly, physical (texture)
damage will occur. Slowly cooling on the contrary will decrease factory capacity. Therefore
we have to calculate an efficient cooling rate.
Factors influencing the freezing time are:
Heat of solidification according to the percentage of water to congeal. In the fish, 73.8
% is water and we must congeal 90 % of it.
Coefficient of thermal conductivity, k, that determines the speed at which the heat
penetrates in the layer already congealed.
Heat convection coefficient, h, between the product and the cooling medium (brine).
The thickness of the layer to congeal.
2 Freezing time calculation
We suppose a cylindrical geometry of the product to congeal. Formula (26.1) gives the time
necessary to reduce temperature until the final temperature is reached.
l 4k
t0 = D0 D0 + (26.1)
16kdT h
where
t0 = time necessary to freeze, s
l = heat of solidification, J/m3
k = Thermal conductivity, W/mK
h = heat convection coefficient between the brine and the fish, W/m2K
D0 = the smallest distance in the thickest section of the fish, m
dT = temperature difference between brine temperature and the average fish
temperature, C
2.1 Calculation of dT
Tin + Tout 8 + ( 12 )
Average fish temperature Tav : = = 2C
2 2
dT = Tbrine Tav = 20 (2) = 18C
199
2.2 Calculation of l:
In the fish, 73.8 % is water and we must congeal 90 % of that water:

l = 0.738 x 335 103 x 0.9 x 1000 J/m3 = 222.507 J/m3
2.3 Calculation of k
Because the fish doesnt have a fixed temperature, the texture and properties will change.
Therefore we average the properties at different temperatures. We calculate them using tables
and formulas of chapter 23 and we list them in Table 26.1.
Table 26.1 Heat conduction coefficients (10-1 W/mK) of the sardine its food components
in function of the temperature.
Temp,C
8.0 2.9 -7.1 -12.0
water 5.85 5.76 5.58 5.49
ice 22.69 23.09
protein 1.88 1.82 1.70 1.64
fat 1.59 1.73 2.00 2.14
ash 3.41 3.34 3.20 3.12
In table 26.2 we list the heat conduction coefficients according to the parallel and
perpendicular model. In the third row we calculate the average.
Table 26.2 Heat conduction coefficient (10-1 W/mK) of sardine in function of

temperature and calculation model.
Temp,C
8.0 2.9 -7.1 -12.0
k(parallel model) 0.497 0.490 1.502 1.616
k(perpendicular model) 0.393 0.390 0.565 0.585
k 0.445 0.440 1.034 1.101
Now we take the average of the ks (average of k(parallel model) and k(perpendicular model))
above:
8 2 .9 2 .9 ( 2 .2 ) ( 2.2 ) ( 7.1 ) ( 7.1 ) ( 12 )

k = 0.445 + 0.440 + 0.565 + 0.585
20 20 20 20
W
= 0.749
mK
200
2.4 Calculation of h
To find the heat convection coefficient between the fish and the brine we make two important
assumptions. First we assume that the brine flows perpendicular over the fish. We also
assume that the fish lays are mutually parallel positioned. Because the brine is forced to flow
around the fish we deal with forced convection. A general model is given by Zukauskas. The
Nusselt correlation is shown in formula (26.2).
p
Pr
Nu = c Re b Prb b (26.2)
Prs
where the index b indicates bulk and s indicates wall.
For a parallel arrangement next formulas are accurate.
0.25
Prb
Nu = 0.52c n Re 0.5
b Pr0.36
b
for Reb = 102-103 (26.3)
Prw
0.25
Prb
Nu = 0.27c n Re 0.63
b Pr 0.36
b
for Reb = 103-2 105 (26.4)
Prw
We need the fish skin temperature Ts to calculate the heat convection coefficient. This
temperature defines together with the brine temperature the film temperature. This one is
necessary to determine the physical properties of the brine to calculate the heat convection
coefficient. We start with the average fish temperature (Tav) during the freezing process (left
column of table 26.3 and 26.4). We notice that the Biot (hL/k) number is very large (Bi >
40), the internal resistance to heat transfer is much higher than the external resistance. We
may assume that the food surface temperature equals the temperature of the cooling medium
(right column of table 26.3 and 26.4).
Table 26.3 Properties of brine at two surface temperatures of sardine(271 and 253 K).
Ts, K
271 253
Tfilm, K 262 253
, kg/m3 1186 1190
, 10-3 Pa.s 4.80 7.00
k, 10-1 W/mK 4.20 4.10
cp, J/kgK 3300 3280
Pr 56.00 56.00
Prs 25.91 56.00
201
Table 26.4 Several steps and final solution when determining the heat convection
coefficient of the brine flowing around the sardine.
Ts, K
271 253
v, m/s 0.15 0.15
D0, m 0.04 0.04
Re 1111.88 765.00
Nu 115.72 61.26
h, W/m2K 1620.04 837.23
2.5 Calculation of t0
Formula (26.1) gives
0.738 * 0.9 * 335 *10 3 4 * 0.749

t 0 = 0.03 0.03 + = 1039.09s = 26.07 min
16 * 18 837.23
3 Freezing time of sardine
The sardine fillets will reach their final temperature (-12C) after a small half an hour in the
brine tank.
202
1 Gas Properties
1.1 Molecular Weight
Molecular weight: 17.03 g/mol
1.2 Solid phase
Melting point: -78 C

Latent heat of fusion (1.013 bar, at triple point): 331.37 kJ/kg
1.3 Liquid phase
Liquid density (1.013 bar at boiling point): 682 kg/m3

Liquid/gas equivalent (1.013 bar and 15 C (59 F)): 947 vol/vol
Boiling point (1.013 bar): -33.5 C
Latent heat of vaporization (1.013 bar at boiling point): 1371.2 kJ/kg
Vapour pressure (at 21 C or 70 F): 8.88 bar
1.4 Critical point
Critical temperature: 132.4 C

Critical pressure: 112.8 bar
1.5 Gaseous phase
Gas density (1.013 bar at boiling point): 0.86 kg/m3

Gas density (1.013 bar and 15 C (59 F)): 0.73 kg/m3
Compressibility Factor (Z) (1.013 bar and 15 C (59 F)): 0.9929
Specific gravity (air = 1) (1.013 bar and 21 C (70 F)): 0.597
Specific volume (1.013 bar and 21 C (70 F)): 1.411 m3/kg
Heat capacity at constant pressure (Cp) (1.013 bar and 15 C (59 F)): 0.037
kJ/(mol.K)
Heat capacity at constant volume (Cv) (1.013 bar and 15 C (59 F)): 0.028
kJ/(mol.K)
Ratio of specific heats (Gamma:Cp/Cv) (1.013 bar and 15 C (59 F)) : 1.309623
Viscosity (1.013 bar and 0 C (32 F)): 0.000098 Poise
Thermal conductivity (1.013 bar and 0 C (32 F)): 22.19 mW/(mK)
203
1.6 Miscellaneous
Solubility in water (1.013 bar and 0 C (32 F)): 862 vol/vol

Auto ignition temperature: 630 C
2 Material compatibility
Air Liquide has assembled data on the compatibility of gases with materials to assist in
evaluating which products to use for a gas system. Although the information has been
compiled from what Air Liquide believes are reliable sources (International Standards:
Compatibility of cylinder and valve materials with gas content; Part 1: ISO 11114-1 (Jul
1998), Part 2: ISO 11114-2 (Mar 2001)), it must be used with extreme caution. No raw data
such as this can cover all conditions of concentration, temperature, humidity, impurities and
aeration. It is therefore recommended that this table is used to choose possible materials and
then more extensive investigation and testing is carried out under the specific conditions of
use. The collected data mainly concern high-pressure applications at ambient temperature and
the safety aspect of material compatibility rather than the quality aspect.
Table 27.1 Material compatibility with ammonia
Material Material Compatibility

Metals
Aluminium Satisfactory
Brass Non recommended
Copper Non recommended
Ferretic Steel (e.g. Carbon Steels) Satisfactory
Stainless Steel Satisfactory
Plastics
Polytetrafluoroethylene (PTFE) Satisfactory
Polychlorotrifluoroethylene (PCTFE) Satisfactory
Vinylidene polyfluoride (PVDF) (KYNAR) Non recommended (*)
Polyamide (PA) (NYLON) Satisfactory
Polypropylene (PP) Satisfactory
Elastomers
Buthyl (isobutene - isoprene) rubber (IIR) Satisfactory
Nitrile rubber (NBR) Acceptable (**)
Chloroprene (CR) no data
Chlorofluorocarbons (FKM) (VITON) Non recommended (***)
Silicon (Q) Non recommended (***)
Ethylene - Propylene (EPDM) Satisfactory
Lubricants
Hydrocarbon based lubricant Non recommended (***)
Fluorocarbon based lubricant Satisfactory
Source: air liquide
(*) Notable acceleration of the process of ageing and significant loss of mass by extraction or
chemical reaction.
(**) but significant loss of mass by extraction or chemical reaction.
(***) Significant loss of mass by extraction or chemical reaction.
204
3 Health effects
Ammonia is not, strictly speaking, a poison and repeated exposure to it produces no additive
(chronic) effects on the human body. However, even in small concentrations in the air it can
be extremely irritating to the eyes, throat, and breathing passages.
Anhydrous ammonia primarily affects three areas of the body: eyes, lungs and skin
3.1 Eyes
Everything from mild irritation to destruction of the eye can occur depending on whether a
spray or gas is involved. Ammonia penetrates the eye more rapidly than other alkalis.
3.2 Lungs
In the lungs, liquid anhydrous ammonia causes destruction of delicate respiratory tissue.
Exposure to ammonia vapour may cause:

Convulsive coughing.
Difficult or painful breathing.
Pulmonary congestion.
Death.
3.3 Skin
Skin damage depends upon the length and concentration of exposure and can range from mild
irritation, to a darkened freeze-dry burn, to tissue destruction. Because liquid ammonia boils
at -28F, the expanding gas has the potential to freeze anything in its path of release,
including human flesh and organs.
Because water can absorb ammonia so readily, it is a factor that contributes to human toxicity.
Ammonia will keep spreading across contacted skin until the chemical is diluted by skin
moisture.
Alkalis effect tissue differently than acids, which tend to burn and seal off a wound. Alkalis,
such as ammonia cause liquidization of tissue and turn tissue into a sticky "goo" and mix with
this tissue, causing further damage. As a result, anhydrous ammonia burns keep spreading
until the chemical is diluted.
In addition to liquidization, super-cooled anhydrous ammonia spray causes a freeze dry effect
like frost bite when it hits the skin. The spray is also capable of freezing clothing to skin so
that if the clothing is removed incorrectly whole sections of skin can be torn off.
High concentrations in the air can also dissolve in the moisture of the skin or perspiration and
result in a corrosive action on the skin and mucous membranes.
205
4 First Aid
Decontaminate the victim as quickly as possible. Start with the eyes.

The whole body, or exposed area, must be flushed with generous
amounts of water; this includes the hair, ears, under the chin, and
armpits. Water sources such as showers, hoses, eye wash stations, or
stock tanks are acceptable.
Ensure trained personnel and adequate first aid supplies are readily
available.
4.1 Contact with the Eyes
Even if only a small amount of ammonia enters the eyes, irrigate the eyes with an abundance
of water for a minimum of 15 minutes. Continually and thoroughly flush the entire eye
surface and the inner lining of the eyelids. Eyes affected by ammonia close involuntarily, so
the eyelids must be held open so that water can flush the entire eye surface, as well as the
inner lining of the eyelid.
If there is no physician available, continue irrigation for an additional 15 minutes.
Do not wear contact lenses when handling anhydrous ammonia. If ammonia gets in the eyes,
the ammonia will get trapped under the lenses causing even more damage. They may also
prevent immediate flushing of the eye surface.
Serious eye injury should be treated by an ophthalmologist, but in an

emergency, wash with large quantities of water for 15 minutes or more as
quickly as possible. In fact, the only real hope for preventing permanent eye
injury lies in quick and generous washing.
One suggestion for those likely to be exposed is to carry a small, eight-ounce

squeezable squirt bottle filled with water, which can be used to get excess
ammonia out of the eyes until a larger water supply can be reached. This small amount of
water is not sufficient to remove all the ammonia. It is essential that the eyes be irrigated for a
minimum of 15 minutes as soon as possible.
Another emergency method is to duck the head in water and rapidly blink and move or rotate
the eyes about.
4.2 Contact with the Skin
It is essential that any ammonia spilled on the worker be removed immediately and that the
worker be moved to an uncontaminated area quickly.
Clothes that have been saturated by liquid ammonia may freeze to the skin. In any case, the
victim, still clothed, should get immediately under a shower, if available, or jump into a stock
tank, pond, or into any other source of water. Time is important! Remove clothes only after
206
they are thawed and they can be freely removed from frozen areas. If the clothing is removed
incorrectly, whole sections of skin can be torn off.
No salves, creams, ointments, or jellies should be applied to the skin during a 24-hour period
following the injury since this will prevent natural elimination of the ammonia from the skin.
After the 24-hour period, the medical treatment is the same for thermal burns. A physician
should view any second- or third-degree freeze burns of the skin.
4.3 Taken internally
This is what you should do if ammonia is ingested:

Call a physician.
If conscious, have the victim drink large amounts of water.
Do not induce vomiting if the victim is in shock, in extreme pain, or is unconscious.
If vomiting begins, place the victims face down with head lower than hips. This
prevents vomit from entering the lungs and prevents severe injury.
4.4 Inhalation
In all inhalation exposures, severe or minimal:

Take the exposed workers at once to a clean, uncontaminated area.
Watch workers exposed to low concentrations for a short period of time. They will
usually require no treatment and can be released.
For severe exposure to higher concentrations:

Call a physician.
Administer oxygen by an individual who is trained and authorized to do so by a
physician. This will help relieve pain and symptoms of lack of oxygen.
Begin artificial respiration immediately if the patient is not breathing.
Keep victim warm (but not hot) and rested until transported to the hospital.
5 Summary
In any accident involving contact with ammonia with the eyes or skin:
Immediately flush the affected area with large quantities of clean water.
Place the injured person into a container of clean water or under an emergency
shower.
Provide the injured worker with first aid treatment and call a physician at once in the
case of extreme exposure. Give the physician a complete account of the incident.
Seconds count, wash the ammonia away with water immediately.
207
Ammonia is an economical choice for industrial systems. Although ammonia has superior
thermodynamic properties, it is considered toxic at low concentration levels of 35 to 50
mg/kg. Large quantities of ammonia should not be vented to enclosed areas near open flames
or heavy sparks. Ammonia at 16 to 25% by volume burns and can explode in air in the
presence of an open flame. The importance of ammonia piping is sometimes minimized when
the main emphasis is on selecting major equipment pieces. Mains should be sized carefully to
provide low-pressure drop and avoid capacity or power penalties caused by inadequate piping.
Rusting pipes and vessels in older systems containing ammonia can create a safety hazard.
Oblique X-ray photographs of welded pipe joints and ultrasonic inspection of vessels may be
used to disclose defects. Only vendor-certified parts for pipe, valving, and pressure-containing
components according to designated assembly drawings should be used to reduce hazards.
Cold liquid refrigerant should not be confined between closed valves in a pipe where the
liquid can warm and expand to burst piping components. Rapid multiple pulsations of
ammonia liquid in piping components (e.g., those developed by cavitation forces or hydraulic
hammering from compressor pulsations with massive slugs of liquid carryover to the
compressor) must be avoided to prevent equipment and piping damage and injury to
personnel. Hydraulic shock, also known as hydraulic hammering, can be particularly
hazardous. In stream and water systems, it is called water hammer. Symptoms in an operating
system are loud clattering or banging or pipes moving suddenly and erratically. The hazard
arises out of great, potentially destructive forces that are suddenly generated inside the piping
as a result of hydraulic shocks (Lloyko 1992; Shelton and Jacobi 1997a, 1997b) The
fundamental flow-phenomenon descriptions of the transients that can cause these shocks are
vapour-propelled, liquid slugs, and condensation-induced shock. The phenomenon typically
arises when high-pressure warm vapour and cold liquid mix in the same pipe or other
refrigerating system element. Parts of the system where this is possible most often are within,
and in the piping around, air-cooling units before, during, and after hot-gas defrosting and in
parts of gas-pressure liquid circulating systems, but other parts of the system can also be
vulnerable (Glennon and Cole 1998).
The conditions that are most conducive to the development of hydraulic transients are
pressures lower than approximately 138 kPa, with saturated or subcooled liquid present,
which are then exposed to high-pressure vapours at pressures equal to or greater than those
commonly used for hot-gas defrosting (586 kPa). Standing liquid in a horizontal pipe has
been shown to be able to be excited sufficiently to cause hydraulic shocks when exposed to
high-pressure gas flowing over the top of it. In situations where high-pressure gas will be
introduced regularly, it is necessary that steps be taken as part of the operating protocol to rid
the pipe or other system element of liquid before the introduction of the hot gas.
To that end, all valves in horizontal lines should be installed with their stems horizontal, and
air-cooling units should be subjected to a complete pumpdown before the introduction of hot
gas for defrosting. Failures occur most often at the pipe or header end-closure. Such sections
of pipe, where unavoidable, should be kept very short, less than 24 pipe diameters in length.
Most service problems are caused by inadequate precautions during design, construction, and
installation (IIAR Standard 2, ASHRAE Standard 15). Ammonia is a powerful solvent that
208
removes dirt, scale, sand, or moisture remaining in the pipes, valves, and fittings during
installation. These substances are swept along with the suction gas to the compressor, where
they are a menace to the bearings, screws, and lubricant. Most compressors are equipped with
suction strainers and/or additional disposable strainer liners for the large quantity of debris
that can be present at initial start-up. Moving parts are often scored when a compressor is run
for the first time. Damage starts with minor scratches, which increase progressively until they
seriously affect the operation of the compressor or render it inoperative.
A system that has been carefully and properly installed with no foreign matter or liquid
entering the compressor will operate satisfactorily for a long time. As piping is installed, it
should be power rotary wire brushed and blown out with compressed air. The piping system
should be blown out again with compressed air or nitrogen before evacuation and charging.
See ASHRAE Standard 15 for system piping test pressure.
209
1 Introduction
Fishery products may undergo undesirable changes in flavour, odour, appearance, and texture
during frozen storage. These changes are attributable to dehydration (moisture loss) of the
fish, oxidation of the oils or pigments, and enzyme activity in the flesh. The rate at which
these changes occur depends on (1) the composition of the species of fish, (2) the level and
constancy of storage room temperature and humidity, and (3) the protection afforded the
product through the use of suitable packaging materials and glazing compounds.
2 Composition
The composition of a particular species of fish affects its frozen storage life considerably. Fish
with high oil content, such as tuna and mackerel have a comparatively short frozen storage
life because of the development of rancidity as a result of the oxidation of the oils and
pigments in the flesh. The development of rancidity is less pronounced in fish with a low oil
content. Therefore, lean fish such as haddock and cod, if handled properly, can be kept in
frozen storage for many months without serious loss of quality.
3 Temperature
The quality loss of frozen fish in storage depends primarily on temperature and duration of
storage. Fish stored at 29C has a shelf life of more than a year. In Canada, the Department
of Fisheries recommends a storage temperature of 26C or lower. Storage above 23C,
even for a short period, results in rapid loss of quality. Time-temperature tolerance studies
show that frozen seafoods have memory; that is, each time they are subjected to high
temperatures or poor handling practices, the loss in quality is recorded. When the product is
finally thawed, the total effect of each exposure to mistreatment is reflected in the quality of
the product at the consumer level. Continuous storage at temperatures lower than 26C
reduces oxidation, dehydration, and enzymatic changes, resulting in longer product shelf life.
From the time they are frozen until they reach the consumer, frozen seafoods should be kept
at temperatures as close to 26C as possible. The shelf life of frozen fish products stored at
different temperatures is given in Table 29.1. Note the increase in shelf life at the lowest
temperatures.
For many years, it was thought too costly to operate refrigerated warehouses at temperatures
lower than 23C. However, improvements in the design and operation of refrigeration
equipment have made such temperatures economically possible. The production of surimi by
West Coast-based factory ships has resulted in the construction of ultra-cold rooms for its
storage. Japanese standards call for this product to be kept at 30C.
210
Table 29.1 Shelf life (months) of tuna and mackerel in function of their storage
temperature (C)
Product Temperature, C Shelf Life, Months

Whole blue fin tuna -12 4
-18 to 20 8
-29 12
Packaged mackerel fillets -9 2
-18 3
-23 3 to 5
4 Humidity
A high relative humidity in the cold storage room tends to reduce the evaporation of moisture
from the product. The relative humidity of air in the refrigerated room is directly affected by
the temperature difference between room cooling coils and room temperature. A large
temperature difference results in decreased relative humidity and an accelerated rate of
moisture withdrawal from the frozen product. A small temperature difference between the air
and evaporator cooling coils results in high relative humidity and reduced moisture loss from
the product.
5 Packaging and glazing
Adequate packaging of fishery products is important in preventing product dehydration and

consequent quality loss. Individual fish, whether frozen in the round or dressed, cannot
usually be suitably packaged; therefore, they must be protected by a glazing compound. A
glaze acts as a protective coating against the two main causes of deterioration during storage:
dehydration and oxidation. It protects against dehydration by preventing moisture from
leaving the product and against oxidation by mechanically preventing air contact with the
product. It may also minimize these changes chemically
with an antioxidant.
6 Maximum Storage life
Maximum storage life of fishery products can be obtained by employing the following
procedures:
Select only high-quality fish for freezing.
Use moisture-vapour-resistant packaging materials and fit package tightly around
product, or use a modified atmosphere and oxygen-barrier package.
Freeze fish immediately after processing or packaging.
Glaze frozen fish prior to packaging.
Glaze round, unpackaged fish prior to cold storage.
Put fish in frozen storage immediately after freezing and glazing, if required.
Store frozen fish at 26C or lower.
Renew glaze on round, unpackaged fish as required during frozen storage.
211
Chapter 30: Average weather conditions in Santander
Chapter 30: Average weather conditions in

Santander
Fig.30.1 Average weather conditions in Santander
212
Reference list
Reference list
Chapman, A.J. (1984). Transmisin de calor. Madrid, Grefol.
Gmez, N. (2003). Instalacin frigorfica para congelacin de pescado por inmersin en

salmuera. Proyecto fin de carrera, Santander, Escuela tcnica superio de Ingenieros
industriales y de telecommunicacin, Universidad de Cantabria.
Koelet, P.C. (1997). Fro industrial: Fundamentos, diseo y aplicaciones. Madrid, Igrara.
Welty, J.R., Wicks C.E., Wilson R.E., Rorrer G. (2001). Fundamentals of Momentum, Heat,
and Mass Transfer. 4th edition, United Sates of America.
De Paepe, M. Thermische installaties (cursusnotas). Gent, Faculteit Toegepaste

Wetenschappen, Universiteit Gent, 196 p.
Vandevelde, P. Warmte- en verbrandingstechniek deel 1 (cursusnotas). Gent, Faculteit

Toegepaste Wetenschappen, Universiteit Gent.
ASHRAE. 1999, 2000, 2001 and 2002 ASHRAE HandbookCD. Atlanta.
Rapin, P.J. (1986). Instalaciones Frigorificas tomo 2. Madrid, marcombo boixareu editores.
Stegmann, R. Practical guide to low temperature refrigeration, ASHRAE Journal 42 (1)

(2000), 4250.
Kilicarslan, A. An experimental investigation of a different type vapor compression cascade

refrigeration system, Applied Thermal Engineering 24 (2004), 26112626
Zrcher, O., Favrat, D., Thome, Thome, J.R. Evaporation of refrigerants in a horizontal tube:
an improved flow pattern dependent heat transfer model compared to ammonia data.
International Journal of Heat and Mass Transfer 45 (2002), 303-317.
Thome, J.R. Void fractions in two-phase flows (new textbook being written).
Thome, J.R. Two-Phase Pressure Drops (Wolverline Databook III).
Thome, J.R. Two-Phase Flow Patterns (Wolverline Databook III).
Zrcher, O., Favrat, D., Thome, Thome, J.R. Development of a diabatic two-phase flow
pattern map for horizontal flow boiling. International Journal of Heat and Mass Transfer 45
(2002), 291-301.
Zrcher, O., Favrat, D., Thome, Thome, J.R. Evaporation of refrigerants in a horizontal tube:
an improved pattern dependent heat transfer model compared to ammonia data. International
Journal of Heat and Mass Transfer 45 (2002), 303-317.
213
Reference list
Ohadi, M.M., Li, S.S., Radermacher, R., Dessiatoun, S. (1996). Critical review of available
correlations for two-phase flow heat transfer of ammonia. Int J. Refrig. Vol 19, No F, 272-284
http://www.fao.org/documents/show_cdr.asp?url_file=/DOCREP/003/V3630E/V3630E08.ht
m
http://www.fao.org/DOCREP/004/T0098E/T0098E00.htm#TOC
http://www.airliquide.com
http://www.osha.gov/
http://www.bbc.co.uk/weather/world/city_guides/results.shtml?tt=TT004230
http://www.et.dtu.dk/coolpack.
http://en.wikipedia.org/wiki/Atmospheric_pressure
214

Design Ref 2

Uploaded by

Copyright:

Available Formats

Design Ref 2

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Design Ref 2

Uploaded by

Copyright:

Available Formats

University of Cantabria

Refrigerating Plant Design

Tutor: Prof. Dr. Ir. Michel De Paepe

Thesis is presented to obtain the Diploma of the Second Cycle of Electromechanical

Study year 2005-2006

Refrigerating Plant Design

Tutor: Prof. Dr. Ir. Michel De Paepe

Thesis is presented to obtain the Diploma of the Second Cycle of Electromechanical

Study year 2005-2006

Santander, june 2006

Prof. Dr. Ir. Michel De Paepe De Muynck Wouter

Wouter De Muynck and Davy Poelman

Thesis is presented to obtain the Diploma of the Second Cycle of Electromechanical

Study year 2005-2006

Tutor: Prof. Dr. Ir. Michel De Paepe

Department of Flow, heat and combustion mechanics

Key words: thermodynamic cycles, heat transfer, two-phase flow, refrigeration

Supervisor(s): Delfin Sili & Valentin Castaera

Chapter 2: Thermodynamic cycle selection

Part 2 :The blast freezing chamber........................................................................................... 40

Chapter 3: The floor construction

Chapter 4: The floor ventilation

Chapter 5:Refrigerating space calculation

Chapter 6:Air circulation in the blast freezer

Chapter 7: Refrigerating load

Chapter 9:Evaporator defrosting

Part 3:The brine tunnel............................................................................................................. 89

Chapter 10: The brine tank

Chapter 11:Brine selection

Chapter 12:The agitators

Chapter 13: The conveyer belts

Chapter 15: The evaporator

Chapter 16: The brine basin

Part 4 ...................................................................................................................................... 123

Chapter 17: Screw compressors

Chapter 18: Condensers

Chapter 19: Piping

Chapter 20: Auxiliary circuit instruments

Chapter 21: Refrigeration circuit controls

Chapter 22: Pressure drops

Chapter 23: Thermal properties of tuna and mackerel

Chapter 25: Freezing time for tuna and mackerel

Chapter 26: The freezing time of Sardine

Chapter 27: Ammonia (R-717)

Chapter 28: Safety considerations

Chapter 29: Storage of frozen fish

Chapter 30: Average weather condition in Santander

Reference list.. 213

3 General description of the brine tank

2 lifting conveyer belt (lifts

4 General description of the blast freezer

Internal dimensions: Length: 18 m

Fig. 1.3 Blast freezing room arrangement.

To achieve an uninterrupted vapour retarder/insulation envelope a total encapsulation

Fig. 1.4 Total exterior vapour retarder system

Fig. 1.5 Air circulation in the blast freezer room.

Chapter 2: Thermodynamic cycle selection

1 Introduction: Multistage Vapour Compression.

We choose the condenser temperature 35C. Average summertime temperature of Santander

Each refrigeration system consists of a compressor, a condenser, an expansion valve, and an