ORIQPl'%-76-001 C2

MarineRefrigeration
Same
TipsonDesign,
tnstallatian,
andServicing

OregonStateUniversity
ExtensionMarineAdvisoryProgram
A LandGrant/ SeaGrantCooperative
Special Report 488
Reprinted June 1983

50ANCOPY
Ogiy

CIRCIjLATtII8
COI'7
SoaGrantDepository
 3. Thermo King Service School Training Manual on Basic
Refrigeration. A limited supply is free and avail-
able in this area by contacting: Don Bachelder or
Bill Pascoe, Standard Parts and Equipment Company,
P. 0. Box 42294, 5251 S.E. McLoughlin Boulevard,
Portland, Oregon 97202 -800-452-8i07!.

information in these notes is organized according to the

following outline:

~ae
i. ENGINEERING DESIGN

A. Refrigeration System Design

1. Sizing of the System-

A Rule-of-Thumb
Horsepower vs Refrigeration Tons
Insulation

2. Configuration and Controls

Oil Return
P-Traps
Oil Separators
Sensors and Controls

3. Sizing and Selection of Components

a. Refrigerant Selection 10
R-12/R-22/R-502
Changing Refrigerants
b. Sizing and Selection of Hardware--- ----
Compressor
Electric Power
Expansion Va3.ve
Evaporator
Filters/Driers
Receivers
Condenser
Piping

B. Brine Spray System Design------

1. Pumps js
Sizing and Selection
16
2. Chillers
Closed Chillers
Open Coil Chillers
Operating Temperatures
 ~ae

3. Brine Preparation, Circulation, and Spray----- ls

Brine Freezing Points
De foamers
Spray Pipe Configuration
Spray Heads
23
C. Installation of Equipment
Materials
Piping
Expansion Valve
Condenser

26
I I. MAINTENANCE, TROUBLESHOOTING,AND REPAIR
26
A. System Shutdown
26
B. Refrigerant Shortage
About Troubleshooting
Leak Detection
Making Repairs
Adding Refrigerant

C. Component Servicing 30
Comp re s so r
Lubrication
Head Pressure
Expansion Valve
Condenser
What To Do After It Fails
Filters/Driers
Receiver
Chiller

INDEX 36

Brand names are used in this publication for illustrative

purposes only; their mention does not in any way constitute
endorsement of these products.
 I. En ineerin Desi n

A. Refri eration S stem Desi n

1. Sizing of the System

Wnen designing a system, several questions must be answered.

--What areas will the vessel fish in; Water temperature
will influence the condenser performance.
--What type of fishing will be done? For example, for a
tuna freezing system, the size will depend upon whether
the boat is engagedin bait fishing, purse seining, or
jig fishing. Should the system be sized for the maximum
possible catch rate, or should some average such as a
ton a day! be anticipated?

--What power source is available to operate the refrigera-

tion? A choice of electric or hydraulic motor, auxiliary
engine, or direct drive off the main engine will influence
the methods of control discussed in the next section!.
--Is the hold insulated? How much heat leakage into the
hold must be assumed through the deck, hull, engine room
bul khead!?

--What is the tonnage capacity of the boat? How much will

frozen fi:sh, already in the hold, help refrigerate warmer
fish dumped on top?

--How continuously will the refrigeration equipment be op-

erating?

For tuna, the system should be capable of freezing the fish

in 24 hours. If a ton of fish were caught every 24 hours, the
amount of heat which must be removed to cool and freeze the fish
from 'F to about 18 'F is equivalent to aboutone refrigeration
ton. {But a "one ton" compressor will not be adequate for
the job, as will be explained below.!
This characteristic is known as the "flywheel effect." Someboats
havemadeuse of "ice builders" to achievethe sameeffect. During
the night, when no warm fish are loading the system, ice frozen on
coils can provide a certain amountof residual cooling during the
next day's fishing, whenthe refrigeration capacity of the equip-
ment may be inadequate.
A*
A refrigeration ton is defined as the amount of heat that must be
removed from a ton of 3Z' fresh water, in order to turn it into ice
within 24 hours. It is equivalent to 12,000 BTU/hr.
 The following rule-of-thumb for sizing a refrigeration sys-
tem is applied to California boats; it should serve equally well
for the smaller Oregon albacore trollers:

Provide 30 horsepower of refrigeration capacity for 100

tons of hold capacity t'or 10 horsepower for a 30-ton boat,
or 6 horsepower for a 20-ton boat!.

This rule-of-thumb is based on the following assumptions:

a. a "reasonable" amount of insulation,
b. about 85' condenser water temperature,
c. designed for "peak" day, and
d. no overloading t'for example, one-fifth of the hold capacity
dumped into warm brine the first day!.
How does "horsepower" relate to "refrigeration tons?"
"Refrigeration tons" describes a rate of removing heat, say,
from the fish hold or brine. It is the capacity of the refrig-
eration system. "Horsepower" describes the amount of work you
need to put into the compressor to remove that heat. The capa-
city will change with different conditions of evaporator and con-
denser temperature, compressor speed, and type of refrigerant.
But if the latter three conditions--condenser temperature, com-
pressor speed, and type of refrigerant--are held more or less
constant as they usually are!, the capacity will vary a lot
with evaporator temperature, Figure l shows this variatton under
some typical conditions for a common three cylinder compressor.

z' 12
0

10
0

uJ

R4

U FI GURL 1

0
40 -30 -20 -l 0 0 10 20 30 40
EVAPORATOR TE MP. ' F

jlI
Compressor running at l,7SO RPM, using R- 22, condenser temperature
at 108 'F, liquid is subcooled li' after leaving condenser and
before entering t.-xpansion valve.
As the temperature of the evaporating refrigerant gets lower
and therefore, as the suction pressure goes down!, the system
is able to deliver less and less capacity.

Horsepower required by the compressor also decreases with

evaporator temperature, but not as much. Figure 2 shows this
variation pictured with the results of Figure l. At an evap-
orator temperature of around 40 'F, the "horsepower" and "tons"
of this system are about the same, and the terms can be used
interchangeably. However, as temperature decreases, the tons
are a lot lower, but the ower must still be designed into the
system to drive it when t e temperatures go back up. This is
why "horsepower" is a better description of a system than "tons."
A "7-1/2 ton compressor" is rated at a certain suction tempera-
ture often +40 'F!, a given speed,* and with a given refrigerant.
As an example, assume you were initially operating with warm
brine, and warm fish are dumped into the hold. The system may be
operating at 40 'F suction temperature, using about 13 horsepower
to get 13 tons of refrigeration. As the temperature decreases
to, say, l0 'F suction temperature, you' re now using the 13
horsepower available in the motor to get only 6-1/2 tons of re-
frigeration. Seiners with freezing coils frequently use a 5 F
suction temperature and, therefore, don't get much work out of
their compressors. In refrigerated seawater sprayers, 2S 'F

l4

IJJ

~2

Cl
z l0

z
0

z6
0

FIGURE 2

0
-40 -30 -20 -I0 0I 0 20 30 40
EVAPORATOR TEMP. F

A 10-ton unit at 1,200 RPMmight put out 14 tons at 1,750 RPM.

suction temperature is common; so a higher capacity is possible
with the same !horsepower driving motor.

Heat leakage into the fish hold depends on insu1ation,

seawater and air temperature, frequency of opening the hatch,
and hold temperature. Tuna boats with~pray brine systems fre-
quently use 18'-20' brine temperature.* It is necessary to have
a water-tight fish hold and shaft alley and ventilation around
the ribs. But it's not always necessary to have insulation all
around, especially on wooden boats Figure 3!. Sometimes it is
simpler and less expensive just to increase the refrigeration
capacity to take care of a higher amount of heat leakage. Many
of the San Pedro purse seiners with 75-120 ton hold capacities
have no insulation.

H ULL
LANKING

FRAME L
AIR SPACE

Much of the discussion in this section has been directed

to brine freezing of tuna. Many of these systems, with some
modification, would also be capable of chilling shrimp. For
example,
a system
consistingof a threecylinder"RA"Tecumseh*
*
compressor and a small W. E. Stone chiller would probably be
adequate for chilling 15,000 lb/day of shrimp to 34 'F.
2. Configuration and Controls

A basic refrigeration system appears in Figure 4. High

pressure refrigerant gas coming from the compressor condenses
into liquid in the condenser and collects in the receiver.
At the expansion valve, the high pressure liquid changes into
a low pressure liquid which evaporates as it flows through the
This is why, when changing from a freezer to a chiller, it may
be necessary to re-size the condenser. The suction pressure
would go up, the compressor would do more work, thus the condenser
must be increased in size so it can remove more heat.

In southern California, the practice is to maintain this until

close to delivery time, when temperature is increased to 26-27'.
This is said to increase the weight of the fish.
A*A
Theuse of productnamesanywhere
in these notes doesnot imply
endorsement of those products by Oregon State University.
coils, picking up heat from warmfish or brine. Whenrefrigerant
arrives at the control bulb, it is heat-laden gas. The dia-
gramimplies that the condenseris insta11ed higher than the
receiver, which, in turn, is installed higher than the com-
pressor. Although this is the usual configuration, the ele-
vation is not important, and any arrangementwould work. The
thermostatic control bulb should be located in the engine room
as shown. This is to make it accessible and to keep copper
materials out of the fish hold. The piping leaving the coils A!
is shownto loop ~u before going through the bulkhead, then,
after leaving the control bulb, down B! into the compressor.
This loop is important for proper return of oil, as explained
below,

EVAPORATOR

Any compressor will pump some lubricating oil into the

system. Proper return of the oil is an important considera-
tion in piping design. Oil is easily carried through the system
when liquid refrigerant is present, especially with R-12, which
mixes with oil like oil with gasoline--you can't tell them
apart. However, when the refrigerant is evaporated to a vapor
and the lubricating oil is left behind, piping should be si2'.ed
so that the Speed of the gas flow is high enough to carry the
oil back to the compressor. This is important. Suppose the
design calls for L-L/8 inch pipe and you happen to use a 2-inch
pipe because that's what is available. It will only fill up
with oil until the cross-sectional area is back down to that of
a L-l/8 inch pipe, and tge gas velocity is high enough to carry
the remaining oil along.

Try to avoid collecting points where oil can accumulate

in the piping circuit. One common collecting point might be
somewhere in the middle of the evaporator coils. If, for
example, an expansion valve were undersized, or if a drier were
partially clogged, or if you were short of refrigerant, chances
are that the refrigerant liquid would be completely boiled away
before it reached the end of the coil point "C" of Figure 4,
instead of near the control bulb!. The result for relatively
large diameter coils is that it is very difficult for the oil
to be forced out past point A! and back into the compressor B!.
For similar reasons, it is generally recommended that for coil
banks, series piping should be used instead of parallel runs,
which have lower gas velocities and sometimes "dry" sections
which collect oil. The oil may accumulate in the coils to the
extent that evaporator temperatures are thrown way off. An-
other possibility is that the compressor could be robbed of
adequate lubrication, resulting in failure.
On multiple refrigerant circuits, it's important to drop
each return line down into the main suction header, rather than
push the gas u~ into the main header Figure 5!. This will pre-
vent slugs of oil from clogging individual circuits, leading to
uneven distribution of refrigerant and perhaps to erratic be-
havior of the expansion valve. This configuration might make it
necessary for oil separated out near the control bulb to be
pushedu~ to the compressor
througha return line illustrated
by D! xn Figure S. If the return line is properly sized, the
gas velocity will usually be sufficient to carry the oil up
the vertical rise to the compressor. However, if this is more
than 10 or 20 feet, it may be necessary to use a "P-trap" and
double suction riser, which serves as a gas velocity increaser

FIG
RETURN HEADER

Note that too small a pipe is also bad, because excessive pressure
drops will affect system performance.
to carry oil upwardFigure6!. During periodsof low refriger-
ant flow, the P-trap fills with oil, forcing refrigerant gas up
the right-handriser E!. Because all the gasgoesup oneriser,
the velocity is higher andthe remainingoil droplets are carried
along. Later, whenthe refrigerant gasflow returns to a high
volume,the pressureat point F! increasesto a valuesufficient
to blow the oil out of the P-trap and into the compressor,after
whichgas is returned along both risers at a high velocity. Note
that both legs of the doublesuction riser haveloops at the top
wherethey join the mainreturn line. This preventsoil from
draining into oneof the legs during a period of low flow. In
general,oneP-trap is neededfor every20 feet of rise. If the
rise is only a few feet, a P-trap is not necessary.

F IGUR.

Oil return can also be a problem on headered coil systems.

The recommendedprocedure is to have the refrigerant inlet at
the bottom, thus making it like a flooded, or "wet expansion"
WX!system. In general, the inlet header is smaller than the
outlet header. As seen in Figure 7, the oil return problems
are apparent. Another type of evaporator, the dry expansion
DX! system, feeds from the top; the refrigerant goes downhill.
The oil return problems will be similar.
How about using an oil separator to take care of oil return
problems? These are not recommended.Often, an oil separator
is a crutch to compensate for improper system design. Unless
it is w~aoversized, someoil will leak past; the ultimate
methodof returning oil must be in the sizing and configuration
 DESIGN 2.
INTERNAL VIEWS OF
HEADE
GNS

F IG

of piping. Besides, if the float sticks while you' re at sea,

you have real problems. Standard separators used in Freon
systems are ~tin comparedto those used in ammoniasystems.
If you must use an oil separator, take a look at the manufac-
turer's size recommendation, then go one or two sizes larger.

The general philosophy of system control should be to

makethings as automatic but as simple as possible. You shouldn' t
have to continually adjust hand valves in order to control tem-
peratures.

Some controls are for emergency situations. If there is a

force-feed oil system on the compressor, by all means have an
oil pressure safety switch. If you were to lose refrigerant,
causing oil to collect in the evaporator coils, you could de-
stroy a compressor rather quickly. Other sensors might be
installed to shut down in emergencies. Some examples are sen-
sors to detect high condenser water temperature caused, for
example,by seaweedplugging up a condenserintake!, or high head
pressure.

Sensors that detect refrigerant suction pressure or temper-

ature, or brine and hold temperaturecan also control by cycling
the compressor on and off. This is obviously easiest with an
electric motor drive--the sensor activates the switch. With a
hydraulic system, the sensor could activate a solenoid valve to
bypasshydraulic oil.* With the direct-drjge auxiliary engine
system, this type of control is difficult.
iVote that in this case, a solenoid valve on the hydraulics alone
is sufficient--it's not necessary to have one also on the liquid
refrigerant line.
AAFor shut-off of an auxiliary engine, put a solenoid
valv fuel line.
 One example of a temperature control system is to put a
sensor in the brine pump discharge line, set it for maybe 31'
in a Refrigerated Seawater RSW! system. This would switch
the compressor on and off while allowing the seawater or brine!
pump to keep going. Even if your system is direct-drive and
difficult to cycle on and off, it's a good. idea to monitor brine
pump discharge. Often, during a slow fishing period when the
temperatures start going down, the operator forgets about
things--perhaps causing a chiller to freeze-up. It's easy to
install a thermostat that will ring a bell in the pilot house
when the temperature gets down to, say, 30 or 31 'F. It's an
inexpensive item $20-$25! and you can set it to any temperature
just like a room thermostat. Simply install a stainless steel
well on the brine pump discharge line avoid copper and cupro-
nickel in contact with brine or refrigerated seawater!; the
thermostat-sensing bulb goes in the we11. In the event a closed
chiller does freeze-up, some systems have a valve in the com-
pressor discharge line to allow hot gas to bypass the condenser,
and thaw out the chiller.

Other types of controllers regulate temperatures in a more

continuous way. For example, head pressure can be regulated by
a condenser water flow regulator--an increased water flow will
maintain a constant condenser temperature and head pressure!
when the refrigeration load increases. Brine temperatures are
sometimes controlled, on hydraulically-driven systems, by regu-
lating hydraulic fluid flow--less oil will cause the compressor
to turn slower. Evaporator temperatures can be regulated in
several ways. One common system has an evaporator pressure regu-
lator also called a back pressure regulator! in the line from
the evaporator to the compressor. This valve senses upstream
evaporator! pressure and works to control to some pre-set value--
constant evaporator pressure means constant evaporator tempera-
ture.

A disadvantage of the back pressure regulator is that at

low refrigeration demand, it will shut way down while trying
to maintain a constant pre-set! back pressure. When it does
this, the downstream suction! pressure decreases to such an
extent that it goes into a vacuum--a bad situation from the
standpoint of leakage. A better alternative for controlling
evaporator pressure/temperature is a variable capacity compressor.
An example is the three cylinder Carrier which senses suction pres-
sure and unloads down to two or one cylinders when the capacity
required is low.
 10

3. Sizing and Selection of Components

a. Refrigerant Selection

Three refrigerants commonly used for marine refrigeration

systems are R-12, R-22, and R-S02. Selection of the proper one
depends on the operating temperatures and pressures and on the
maximum compressor efficiency point this occurs at high suction
pressures and densities!. At the evaporator design temperature,
you also want to make sure that the compressor suction is not
running at a vacuum, since this would tend to suck in air and
moisture if a leak occurred.

For brine freezing or refrigerated seawater in the fishing

industry, we can just about rule out R-SOZ, which has an appli-
cation for low temperature work -20 to -SO 'F!. It is the most
expensive of the three about $1.65/lb wholesale in 1976! and
at common brine freezing or RSW conditions, its operating pres-
sure and discharge temperature are unnecessarily high.

Some characteristics of the remaining two:

--R-12 will dissolve in oil and so it carries oil through

the system well. High-side pressure is on the order
of 120 psig.* Wholesale cost is around $.80/lb 976!.
--R-22 will not completely dissolve in oil but is an
acceptable refrigerant for proper oil circulation.
High-side pressure is on the order of 200 psig ~
Wholesale cost is around. $1.20/lb 976!.
The best refrigerant to use is R-ZZ. Because R-22 has a much
higher heat-carrying capacity than R-12, you don't have to pump
as much of it. Thus, it is possible to run a compressor so
much slower about one-third slower! to get the same refrigeration
capacity. At a slower running speed,the equipmentlasts longer.*
By using R-22 instead of R-12, it's sometimes possible to buy a
one size smaller compressor, which could save you $400-$500. A
compressor will be more efficient with R- 22 than with R-12 be-
cause the suction pressure will be higher. And although R-22
costs a bit more, because of the lower pumping rate, you can
go one size smaller on all pipes and valves, thus more than
saving any increased refrigerant expense. More on this in the
following section, "Sizing and Selection of Hardware" !.
How about cnanging an R-12 system to R-22? First of all,
you must have an open-type belt-driven or direct-drive! com-
psig = pounds per square inch gauge.
**
For Carrier compIessors, try to run them under 1,200 RPM for
a trouble-free long life.
pressor on.which you can vary the speed. Changing to R-22 in a
hermetically-sealed, electric motor-driven compressor just won't
work. It will try to pump too much capacity, draw too many amps,
overheat, and shut off. The capacity will increase about one-third;
so, to get the samecapacity with R-22 as you got with R-12, de-
crease the compressor speed about one-third from 1,200 RPMto
SOURPM, for example!. Heat transfer characteristics in the evap-
orator will be about the same with either refrigerant, so it' s
not necessary to make any changes there. But the expansion valve
will have to be changed. For example, a three ton R-12 valve would
run about four and one-half tons of R-22 at the same pressure, mak-
ing it too large for the job .
b. Sizing and Selection of Hardware
Balanced design is important. It is necessary to match the
compressor, condenser, evaporator, expansion valve, and other
components.

Compressor
We've already looked at recommendedsystem and compressor
sizing Section 1, "Sizing of the System"!: a boat with 30 tons
of hold capacity would need a compressor of about 10 horsepower,
which, rated at 40 'F suction temperature, would be close to 10
refrigeration tons. Rememberthat compressor speed, refrigerant,
and operating pressures strongly affect its capacity. If you
happen to have a compressor that's too big for the job, just slow
it down. For example, a compressor putting out lS tons at 1,200
RPMwould put out five tons at around 400-S00 RPM. A manufac-
turer will generally give a minimum allowable speed of 600-700
RPM. This is their recommended minimum for adequate lubrication
on compressors with farce-feed oil systems. Usually it's pos-
sible to go below this to, say, 400-500 RPM!without having
any problems.
A previous section " Configuration and Controls" ! discussed.
evaporator temperature control, and how this might influence the
selection of a compressor. For RSW refrigerated seawater! sys-
tems, temperature control is more critical than in brine spray/
freezer systems. There are two ways to go. The cheaper way is
to buy a conventional compressor plus some intricate valves
"gadgets"! such as back pressure regulators' The better way
although more expensive! is to buy a variable capacity compres-
sor* which will unload. cylinders to maintain a constant suction
temperature.

A variable capacity compressor has another advantage. Fre-

quently, when starting-up a system with warm brine, the high ini-
tial horsepower required Figure 2! will overload or stall a
Carrier is one company making compressors with this capability.
 12

motor if it's not sized quite right. With a conventional compres-

sor, it's possible to overcome this by putting a throttle flow
restricting device! in the suction line to limit capacity. But
a variable capacity compressor is generally designed to start-up
on only one cylinder--the others don't start working until oil
pressure in the compressor is sufficiently high.
Electric Power

If one is using electric power to drive pumps and compres-

sors, a rule-of-thumb is to choose a generator size that will
put out one kw per horsepower needed. For example, if one in-
stallation had a 10 horsepower compressor, a three horsepower
brine pump, one horsepower condenser pump, plus six horsepower
for other convenience items, a 20 kw generator would be needed.
Expansion Valve

Choosing the proper size for an expansion valve is a matter

of experience. One doesn't necessarily choose a 10-ton valve
with a 10-ton compressor. Frequently, it depends on the length
of pipe used in the evaporator coil. As an example, for an open
coil spray-brine freezing system having L50-200 feet of one inch
pipe submerged in brine, one would probably choose around a
three ton valve

In general, expansion valves are adjusted to 7' superheat.

"Superheat" is the amount measured in d,egrees! that the refrig-
erant gas heats up after all the liquid has been evaporated.
Because the pressureecerough the evaporator that is, the coils,
chiller, or freezer plates! is more or less constant, boiling of f
of the liquid refrigerant will take place at a temperature which
is also more or less constant. Say, as in the example in Figure 8,
this temperature is 20 F. As the refrigerant flows through the
evaporator, it picks up heat and boils away at the constant evap-
orator suction! pressure, until it gets toward the end of the

EX PANSION

EVAPORATOR

F I GURE

27 F
 13

evaporatorwhere
theliquidportiondisappears, leavingonlygas
at 20'F. Beyondthis point,asmore heatflowsinto therefrig-
erant,its temperature
will beginto rise. It is this temperature
rise that is called "superheat." Sevendegreessuperheat
means
thatbythetimeit getsta thecontrol
bulb,it will have
warmed
up 7', as in Figure 8.
Theexpansion valvewill sense
superheat in theexample,
27'F!. If it falls below,thevalvewill closeslightly, caus-
ingtheevaporator temperature to risebackupto 27'. If the
temperatureincreases,the valvewill openandlet in morere-
frigerant,causingtheevaporator temperature
to coolbackdown
to 27'. Thus,expansion valvessensetemperaturedifference or
hp*h.hd'h~dd
amountanticipatedfor conditionsof start-upwith warm brine,
highersuperheatsare sometimesadjusted. This is not common,
however,
andin general,wetry for the lowestsuperheat pos-
sible without flooding the compressar.
Theperformance of theexpansion valveis shown bythefrost
pointoccurring aroundthecontrolbulb,or thermostat bulb.
Whenseveraldifferent refrigeration circuits or return lines
aredumpedbackinto a common suctionline asin Figure5!, it' s
possible
to seewhichcircuitsarenotrunning properlyand,
therefore,whichexpansion valvesmayneedadjustment! by look-
ing at the frost line for eachcircuit.
Mostvalves, suchas Alco, will modulatethe capacityar
tonnage
avera range
of 25-304.Forexample,
a threetonvalve
will vary fromalmosttwo ta four tons in capacity.
Besure to buy goodquality expansion valves, since they
areveryimportant.It's alsoa good
ideato buythe typeof
valve suchas that made by Alco! that allowsyouto interchange
the valvestems,making it easyto change to a largeror smaller
size withouttakingthe valveout of the line. Forexample, if
a three ton valvewerefoundto be too small, it's better to put
in an orifice capableof makingit a four ton valve, rather than
trying to adjustthe smalleroneto the limit of its capacity.
Finally, buy a valve that canbe silver-solderedin. Stayaway
fromthreadedconnections--incold lines, waterwill follow the
threadsback--alternatefreezingandthawingwill loosenthe
nut.

Evaporator
All comments
related to evaporatarsappearin a later sec-
tion entitled "Chillers."!
Thisfrost point frequentlyextendstwoto five feet beyond
the
bulb, which is o.k.
 14

Filters/Driers

An undersized filter or drier can initiate premature evapo-

ration of refrigerant. Be sure these are sized to take the
maximumexpected flow of refrigerant. In fact, it's generally
a gooagidea to insta11 a drier that is a size or two larger than
the manufacturer' s recommendation; thus, it can be used to dry
out a system after a repair. More on this in a later section
entitled "Maintenance, Troubleshooting, and Repair."!
Receiver

Receivers can be purchased in a range ofsizes. It is also

fairly easyto makea receiver out of a Freon*bottle. Try to
run with the receiver no more than about one-quarter full of
liquid. If a leak occurs, it will show up rather quickly. Also,
if most of the spare refrigerant were stored in the receiver,
it might all be lost before the leak is detected. Store spare
refrigerant in a bottle elsewhere.

Condenser

Several types of condensers are available. Air-cooled units

are not too satisfactory because of the corrosive atmosphere eat-
ing up the fins. Water-cooled types consist of keel-coolers,
double pipe heat exchangers, and shell-and-tube heat exchangers.
Generally, try to operate the condenser with a S to 6 'F tempera-
ture rise in the seawater flowing through.

Piping

Pipe sizing has already been discussed in relation to the

need for high gas velocities to produce an adequate oil return.
Piping sizes relate, therefore, to a choice of refrigerants.
We know that to produce a given amount of refrigeration, less
R-22 is needed than R-12. Therefore, although R-22 costs a
little more, it's possible to use smaller line sizes than with
R-12 and save money. As an example, instead of a 1-1/8 inch
suction line needed for R-12, we could drop it one size to
7/8-inch to get the same high flow velocity necessary to return
the oil. Another example: a 100-ton boat using 1-3/8 inch li-
quid line for Refrigerant 12 would require a 1 1/8 inch liquid
1ine for Refrigerant 22.*
It is also important that pressure drop in the evaporator
is minimized. A common design might be for a coil bank to be

Note that Freon" is a trade name of the Dupont Company. Pro-

bably, any liquid refrigerant bottle would do.
It's interesting to note that a 1-3/8 inch valve costs $25 mcrre
than a l-l/8 inch valve, and the same different.ial might be true
as the size reduces further.
 1S

200 feet long, consisting of one inch pipe. Using R-22, you
might have no appreciable pressure drop. However, if R-L2 were
put into the same lines and the same capacity were needed, the
pressure drop would be excessive. Therefore, we would have to
either decrease the length of the lines to 125 feet or so! or
increase the diameter to operate without' excessive pressure drop.

B. Brine S ra S stem Desi n

l. Pumps

When choosing a pump for seawater or brine, keep in mind

the following characteristics:

a. It should be of the open-impeller type. Closed or semi-

enclosed impellers will clog up. Although screens are
generally not necessary, a metal shield with I/2-inch
punched or drilled holes is often used to keep out any
trash that might clog the pump. Open impellers are Like
"garbage disposals"--they' re able to grind up almost every-
thing and send it through. One problem with this is
getting trash through the spray heads.!
b. The pump should be made out of some material that can
withstand strong brines. Electrolysis is not generally
a problem with pumps in brine, because the temperatures
are too low.

c. The seal should consist of a stuffing box rather than a

mechanical seal.' The latter will not stand up; servic-
ing a bad mechanical seal at sea is difficult and wastes
precious time.

d. The capacity of pump needed for spray brine systems is on

, the order of four gallons-per-minute per ton of fish
aboard. It usually works out that this flow is needed
to get the proper heat transfer on the coils.! It is
possible to have too large a pump--brine pumps used
around. here Oregon! are often oversized because they
also serve to circulate seawater through crab tanks.
One consequence is erosion on the baffles of a closed
chiller obviously not a problem on open coil chillers!.
Another consequence is that excess power means heat into
the system. An oversized pump controlled by bypassing
the circulation adds a lot of heat. One horsepower is
equivalent to about 2,500 BTU/hr, requiring almost another
one-quarter ton of refrigeration.
Note that Canadian laboratories recommend "all iron" open-impeller
pumps for the RSW systems used on board salmon packers.
 16

e. The head pressure of the pump should be on the order of

35-45 feet of seawater. Translated into pressure, this
is approximately 15-20 psi on a pressure gauge. Thirty-
three feet of seawater is equivalent to one atmosphere of
pressure, or 14.7 psi.!

It's nice to have pumps driven with hydraulics because

you can control the speed. Foaming, which frequently
occurs under high rates of spray, could be reduced when
such a high rate is not needed. A later section con-
tains some notes on anti-foaming additives.! There is
one problem with hydraulic "sump pumps" in the bilge.
If the hydraulic lines need changing, the results are
messy.

g. Some of the pump manufacturers who turn out satisfac-

tory pumps for this type of job are: Pacific, Jacuzzi,
and Gould.

2. Chillers

There are two basic types of chillers in use: closed chillers

and open coil chillers. For reasons that will be discussea, open coil
chillers are generally recommended.

Closed Chillers

This consists of a box or cylinder containing horizontal

refrigerant pipes and vertical half-baffles to force the brine
or seawater! up and down over the horizontal pipes. At aver-
age operating conditions, brine temperatures will drop 6 to 7 'F.
This could increase to 8 or 9 'F for a high refrigeration load,
or decrease to 2 to 3 'F if the brine were cooled to 18 'F and
the load decreased.

One problem with closed chillers is that they are suscept-

ible to plugging up with dirt and fish parts. They will also
freeze-up if the brine strength drops, if the brine flow is re-
stricted, or if the evaporator coil temperature drops too low.
For example, you' re in danger of freezing-up a closed chiller
if you try to cool seawater to less than 34 'F.! It is a good
idea to have a thermometer on the discharge of the brine pump
after the brine has cooled. Common types of stainless steel
thermometers for heat exchangers are made by Weston and Texas
Gauge. They are dial-type, have adapters for pipe plugs, and
can be oriented in. any direction. One problem is that they are
sensitive to vibration.!

Another problem with closed chillers is that an excessively

high brine flow rate or insufficient spacing between baffle
plates can lead to erosion. A Bell and Gossett chiller that was
 17

ars a o originally used eight inch spacing on the

a er chan ed to three inch spacing,
but the brine was overly restricted
res and erosion too p ace. To
remedy the situation, it was necessary to note ea

incn pipe ~ This bypassed some of e

force it back and forth over the refri 'g crated coilsy an e-
creased efficiency, but the flushing heLped to eep is sc
and debris from piling up in dead corners.

Open Coil Chillers

0 en coil systems similar to that diagrammed. in Figure 9

ed. The sketch shows a wooden boat with a fiber-
the hold. The vent atc
a trip, might be opened while
ile the
the oa
boat ra in
is r. port to allow pro er
air circulation etween
betwee the ribs. The waterproo
alley also has a fiber
has 'g lass coating. Lnsta ll the evaporator coils
elow the to of t eh h s a ft a lley cover- then place
'l b dl . For spray brine
zin s stems in southern Cali fornia waters, use L2 feet o f one

expe
O
B
ed LO t" bt ld e"u b
120 feet. An acceptable design for the evaporator coi s is
use iron pipe on four in.ch center rs with standard elbows. T is
is probabLy the cheapest wa
way to go. 1 f you were to use we
w lded e
return bends, you could space the coils closer, r but ui this is mo re

AEI EY,
E SIDE
SSOVER
 Electrolysis is not really a problem, since the brine
temperatures
are generallyquite low. Also not a problemis
freeze-upon the coils due to loss of brine strength or drop
in evaporatorcoil temperature. It is important to.note, how-
ever, that if freeze-updoesoccur in the hold, the brine wash-
ing over ice will not cool as muchas brine washingover re-
frigerated coils--it might get downto about 40 'F.
Onepossible difficulty with installing an opencoil chiller
occurs when the hold has a relatively flat bottom if, for ex-
ample,the engineroomwereaft!, allowingno sump.But the
advantagesover closed.chillers are many--youcan seewhat's
goingon, freeze-upsand.plug-upsare not a problem. In ad-
dition, experiencehas shownthat for boats with a hold capa-
city of about35 tons or less, open-pipesubmerged
coil chillers
are far more economical than closed chillers.
Operating Temperatures
Assuming
you knowthe freezing temperatureof your brine,
makesure that the operating temperature of the brine flowing
out of the chiller is kept 5-7 F abovethe freezing point.
Also, in cupronickel chillers, don't let the refrigerant suc-
tion temperaturebe any colder than 10' belowthe brine freez-
ing temperature.Ona steel chiller you cango a little lower
because the heat transfer is not as good.
To monitor these temperatures, good gauges are important,
especially for measuring
suctionpressure. A cheapsuction
gaugewhichmightbe off by S psi could throwyour calculations
off, causinga freeze-up of the chiller. A headpressuregauge
is not as important as that for low pressure suction!. Most
of thesepressuregaugeshavethe correspondingsaturation
boiling! temperatures
of both R-1?andR-22printed on the face.
3. Brine Preparation, Circulation, and Spray
Brine Preparation
Sprayseawatersystemsare presentlybeing installed for
the purposesof holdingshrimp. It appearsthat they mightbe
held on the boat longer than the present two- or three-day limit;
also, RSW
systems
appearto eliminatelabor, savetime, andget
aroundthe problemof ice scarcity. But somequestionsremain
to be answered:

What temperatures are necessary for adequate preser-

vat ion?
How long can the shrimp be held?
 19

How much warm shrimp can be added and over what period
of time?
What is the effect of salt in seawater or brine! on
the product?

Some experiments with additives have taken place. National

Marine Fisheries Service in Seattle has looked at CO~ in sea-
water as a preservative to enhance peelability, combat salt
uptake, and maintain color. Shrimp in the Gulf of Mexico are
generally frozen in a glucose-salt brine. And obviously, if
you want to operate with seawater or brine! at 33 'F or lower,
salt is a necessary additive to prevent chiller freeze-ups.
Seawater freezes at 27~ 'F, and as stated in the previous sec-
tion, this should be 5-7 'F below the temperature of the sea-
water leaving the chiller.
Salt brine has been most commonly used for freezing on
bait boats of San Pedro and on some albacore boats in this area.
Practice in San Pedro is to run brine temperatures at 18 or 19'
while on the trip. The brine should be kept at a freezing point
of about 10 'F below the operating temperature of the brine--
so for this case, freezing point should be about 8-10 'F. This
corresponds to a total salt weight of about 175 and a reading of
oS salometer degrees!. You need this 10' leeway or "margin of
error" because the brine will always dilute as you add wet fish.
It's difficult to mix in more salt wl~enthe brine is cold, but
it dissolves easily when it is warm. Make sure you don't just
dump the salt into a pile in the bilge and forget about it, be-
cause it may not dissoLve and may clog things up. Whenyou add
salt to the brine, check its salinity. If you notice that it
does not rise, find out where the excess salt is going and why
it is not dissolving. A good practice is [o calculate how much
salt is needed, add that much and no more.
Defoamers represent another frequently used. additive.
Foaming in the brine results occasionally from impurities in
the "kiln-dried, rock salt" but more commonly from fat and other
material from the fish. It does not relate to the salinity
of. the brine. A major problem with foam is that it may cause
Loss of prime in the circulating pump. There are several types
of defoamers. One brand is put out by How Corning. Make sure
that the defoamer used is FDA-approved. It should be premixed
in a bucket of hot or warm brine, then poured into the hold.
If it is dumped in directly, it might not mix properly. Some-
times the spraying characteristics can affect foaming--more
However, sometimes it is necessary to add makeup salt on a long trip,
The experience of one local fisherman was to add 4 bags 00 lbs!
o f s alt on a 15- day trip.
A helpful bulletin on this subject is "Preparation of Salt Brines
for the Fishing Industry," OSU Sea Grant Publication 822. It in-
cludes techniques and tables .
sprayholes and, therefore, lowervelocities might reducefoam-
ing. Foams
can also be knockeddown"mechanically"--asystem
observed in a California cat food plant used a brine spray sup-
plied fromoverhead
sprinklers to breakup foambubbles.
Circulation and Spray
A typical Oregonfishing vessel might have300to 400
gallons of brine in the systern. This requiresa tight tank in
the hold andno leaks fiberglass lining makesa goodseal!.
The intake screenat the brine pumpshouldbe accompaniedby
a large openarea. For example,for a 25-tonhold, provide
an openareaof 12 squarefeet--a cylindrical screenshould
be six to ten feet long.
PVCplastic pipe is quite suitable for brine pumping.For
the sprayingsystem,the pipe andsprinklerheadsshouldbe laid
out as if the tank were a lawn and you were designing a full-
coverageirrigation system. This mightmean providingsome
kind of swing-outor detachableline on the undersideof the
hatch cover. Figure 10 is a sampleconfiguration. Try to put

F I GVRE l0

thepipein seriesasfar aspossibleto preventdeadends,

accumulationof debris, andclogging. Byavoidingparallel lines
in thepipecircuit, it is possibleto flush the system
by open-
ingjust oneendandflushing eitherondeckor intoa sock
or nylonstockingplaced overtheendto catchscalesandde-
bris; Onetypical installation at the hatchcombingappears
in FigureIl. Another scheme Figure12!usesa rubber elbow
that allows the hatchline to be swungout of the way; it has
a plug at the end to allow easy flushing.
 ETACHABLE
LENGTH OF
F I GURE 11
PIP E

FIGHT!; 12

Figure 13 shows an elbow and pipe plug with a horizontal

slot cut into the middle. This produces a spray over the area
covered by the hatch combing.
Still another configuration Figure 14! employs a gate
valve at the end of the pipe which is easily reached to control
the spray strength might be important if you have an oversized
pump!. One trick recently seen was to run a boot or a fire

ONTAL

GURZ 13
hosetype arrangement
from the gate valve all the wayback to
the bilge,' this serves to control foamingwhenexcessbrine
is being pumped through the system.
A typical spray systemdesignuses a 2$ x 2 pumpabout
150gal/min! anda 2-inch diameterPVCpipe with 12 or 13 holes
having a diameterof 7/16-inch. Theholes, which could be in
either the top or bottomof the pipe, needto be small enough
to spray out with somevelocity, although this spray action
does create a lot of foam. You will have to raise or lower the
pipe to get the best sprayingpattern andadjust baffles and
holes as appropriate. Thereare two ideas frequently usedfor
deflectors. One is to use engine block soft plugs {around two
inches in diameter! fastened to the ceiling of the hold with a
screw or nail. This is placed directly over the hole in the
top of the brine spraypipe Figure 1S!. Thesecondidea is
shownin Figure 16. A tablespoonis bent at right anglesand
fastened to the PVCspray pipe with tape or a hose clamp. The
difficulty with tablespoonsis that they frequently bendout of
shape, and it is hard to get at them to adjusts

TABLESPOON
CLAMPED

F IGURE 15
Pvc PIPE FIGURE 16
C. Installation of E ui ment

Materials

Copperandbrine don't mix. Copperions are badbecause

theycanbe toxic andpreliminaryfindingsshowtheycanlead
to discoloration in cannedtuna meat, This mayevenresult from
cupronickelchillers, especially if they're on deckin warm
climates. Coatedcopper coils in brine tanks are bad because
anycoatingtendsto comeoff. Coils madeof galvanizedor
black iron are no problem with brine and fish. In fact, you
mightas well makethemout of blackiron versusgalvanized.
First of all, black iron is cheaper,andwhenyou weld, the
galvanizing is destroyedanyway. It is possible to use
"galvalloy" to protect the weld, but it's still the mostvul-
nerable part of the metal.!

Piping
Figure 17 showsa through-bulkhead
piping designdiscussed
earlier--steel pipe in the fish hold, copper in the engine room.
The riser in the fish hold is frequently installed as a matter of
convenience--the engine room is often too crowded. to go in very
low. Oil return is no problem as long as the copper pipe for
that particular circuit drops downinto the top of the suction
line before it rises back up into the compressor. Make sure
 24

that the control bulb for the expansion valve is located on

the horizontal! copper line for good heat transfer and rapid
response.

Whenpiping up the system, the main thing is cleanliness.

Keep things closed as much as possible--a commonproblem is sand
from sand blasting. As a precaution, install a good suction line
filter. When first starting-up the system, this filter will
keep any dirt out of the compressor. No manufacturer will take
back a failed compressor containing foreign matter.! If a sys-
tem is clean when installed and kept closed, it should operate
like a home refrigerator and last indefinitely.
Copper pipe and fittings should be joined by silver solder'
Regular solder 0-S05 lead! is too soft and does not have suf-
ficient strength. Steel pipe and fittings should be joined
by welding. Occasionally, copper must be joined to steel as
in the through-bulkhead fitting shownin Figure 17. For this
case, it's commonto bore out, for example, a one inch pipe to
1-1/8 inches or a 3/4-inch pipe to 7/S-inch, insert the copper
pipe or tubing and join with 454 silver solder. The pipe or
nipple is welded to the bulkhead and to the steel evaporator
coils. If stainless steel instead of black iron! were used for
the through-bulkhead fitting, it would be appropriate to use
Sil-Fos. Try to avoid using threaded unions to join pipe
sections in refrigerant lines. Especially avoid threaded con-
nections between two different kinds of metal. The expansion
and contraction resulting from temperature changes will cause
the union or threaded joint to loosen and leak.
Also stay away from flared fittings if you can. Sometimes,
they' re hard to avoid--gaugesfrequently connect in with a flareg
fitting. But there are a lot of breaks in these 1/4-inch lines.
One problem is that they' re usually tapped in at or near the
compressor--a region of high vibration. A solution is to get
awayfrom the compressor. For example,on the high-side, go all
the way back to the receiver to tee in a high pressure gauge or
switch.

Sil-Fos is a brazing alloy composedof a phosphor-bronze alloy

mixed with about 155 silver. Its melting point of 1,250 'F is
higher than soft solder. FromAir Conditionin andRefri eration
S stem Servicin , by Joe Ammons; ic erson o xns,
1*
A tip on installation--use a few drops of oil grease, butter,
whatever is available!--on the flare block to prevent galling of
the tube, and around the nut to prevent twisting the tube while
tightening or loosening.
 Alwaysfit a flexible section betweenthe compressorand
the refrigerant lines carrying vapor to and from the compressor.
Make sure that the refrigerant line is rigidly clamped to the
boat structure, just beyond the point where the flexible fitting
fastens on.

Expansion Valve
Buy an expansionvalve with soldered connections--in and.out.
Install it in the engine room, not in the hold. This is to avoid
having.copper materials in the fish hold, and to makeit easier
to service in case of a problem.

Condenser

Condensers are designed to be "counter-flow," that is, sea-

water entering at the bottom and flowing out the top will be
flowing at a direction opposite to the refrigerant which enters
as a gas at the top and leavesas a liquid at the bottom Figure 3.8!.
Usually the refrigerant pipes makefour passes. If the sea-
water exit and inlet are installed in opposite ways that is,
in the top and out the bottom!, the performanceof the condenser
will be reduced. A similar problem occurs when chillers are
not installed right-side up.

REFRIGERANT

REF R IGERANT
FIGURE l8
 26

II. MAINTENANCE, TROUBLESHOOTING,AND REPAIR

A. S stem Shutdown

Whenyou' re through using the sys tem for a while, it's a good
idea to close the king valve just downstream of the receiver!
and pump the system partially down. That is, pump most of the
refrigerant liquid into the receiver. You don't have to pump
it all back.! The reason for doing this is to prevent a slug
of Txquid entering the compressor crankcase on start-up, causing
oil to be blown out.

Once the system has been pumpeddown, chances are the com-
pressor valves will leak a little and eventually pressurize the
suction side. This is o.k., since pressure is better than va-
cuum in the event of system leaks. However, if the shutdown
is to be for an extended period of time, you should also close
the service valves to isolate the low side as much as possible.
Without oil sloshing against the compressor shaft seal, it' ll
tend to dry out and leak. If the service valves weren't shut off,
you might then find yourself out of refrigerant when you go to
start-up at the beginning of the season.
How do you check whether the system is performing well at
the beginning of the season? Oneway is to have a curve of tem-
perature versus time, madewhenthe systemwas working well--then
comparethe new curve with that of the original system. It
might be only necessary to see how manyhours it takes to get
down to the proper temperature. For this, you would need to have
a thermometer on the brine pump discharge. If you have a chiller,
it's easy to look at the temperature difference across the chil-
ler--for example, S' might indicate it is working properly. For
an open coil chiller, obviously this is difficult.
B. Refri erant Shorta e

A shortage of refrigerant often showsup as a shortage of

oil. As discussed in a previous section {"Configuration and
Controls" !, inadequate refrigerant will cause the BP boil-off
point! to locate partway through the evaporator coils. Be-
cause liquid refrigerant in this case will not flow all the way
throughthe coils, compressor
oil will not be returnee. Instead,
it collects in the coils--and a shortage is indicated.
But be careful about troubleshooting. There are other rea-
sons that an oil shortage might show up: a slowed-down compressor,
a heavy.load of warmfish dumped
into a systemhaving an under-
sized expansion
valve, improperpipe sizing for someflow
conditions, a plugged or restricted expansionvalve or drier.
A goodsystemdesign will allow the BP to go to the evaporator
exit no matter what the operating engine speed and load.
 27

This brings us to an important general rule for troubleshooting:

Troubleshooting should consist of 955 diagnosing
and thinking, without any wrenches or screwdrivers
in your hand~ Generally, a problem will have just one
cause--think it out before making a move.

For the case of oil shortage, this means that you shouldn' t--
add oil, tear down a compressor, change a drier, adjust an ex-
pansion valve--before thinking the situation over. Check for
an oil leak. If there isn't one, and if the compressor is
short, the oil m~u
t be in the systemsomewhere--probably
in the
evaporator coils. The cause might be a shortage of refriger-
ant. Check the sight glasses. Frequently, the receiver will
have sight glasses in the side, showing whether or not the level
of liquid is adequate. If it doesn' t, there are two ways to
determine the position of the liquid level in the receiver. One
way is to briefly turn off the water to the condenser while
the system is in operation. Hot gas will heat up the non-liquid
area of the receiver, which you can feel with your hand; below
the liquid level, the walls of the receiver will be cold. The
second way of determining liquid level is to pour boiling water
on the receiver. Again, the metal below the level of the liquid
will be cold."*
Check the liquid-line sight glass. If bubbles are seen, it' s
usually a good. ind.ication of a shortage of refrigerant; and
this means a leak somewhere.*** If you' re at sea, the best
solution is usually to add some refrigerant and worry about the
leak when you get home. Adding more refrigerant--whether you' re
using R-12 or R-ZZ--will return the oil from the coils.
Leaks can mean losing refrigerant or "gaining" air and mois-
ture if the low side were to be in a vacuum!. Loss of refrig-
erant can be. detected in several ways. With R-12 systems, watch
for any signs of oil accumulating on a joint or fitting. This
is a sign of a leak--enough refrigerant gas has leaked out to
One refrigeration serviceman removed 27 gallons of oil from
coils on a SO-ton boat.
**Howmuchliquid should be in. the receiver? Xou actually need
only just enough to cover the outlet. The receiver should be
no more than one- fourth full.
It's difficult to identify a refrigerant shortage by merely
observing the frost line on the coils. Often, this frost line
is some distance downstream of the actual BP, due to cold gas
creating frost.
 28

leave behind a residue of oil on the outside surface. Look

for leaks around gauge and pressure switch lines which are
connected. directly to the compressor. Cracks in lines or flares
can often be spotted and capped until the end of the trip.

Leak detectors are of generally two types:

l. Electronic type: These cost about $120, are quite sen-

sitive, but have filters which, when dirty, can prevent
the instrument from functioning.

2. Halide torch: These have a copper reaction plate, and

when the user has had a little practice, are capable
of detecting leaks of 1/2-ounce per year a leak so
small, it would hardly raise a bubble on a soap film!.
One recommended make is the Prestolite leak detector.

The electronic type is quite fragile, easily broken, and

therefore not recommended for use on fishing vessels. In fact,
the halide torch shouln't be kept aboard a boat either. Usually
by the time you go to use it, the copper reaction plate is
corroded and it won't work.

Another leak problem is to have air leaking into the system.

You can detect this by looking at tables for equi~i rium temper-
ature/pressure conditions appropriate for different gases. [or
example,with R-12, 70 'F will correspondto about 70 psig,"
60 'F corresponds to about 60 psig actually 57.8 psig! . If
the pressures are higher than these, it could indicate that
there is air in the system. Be careful about diagnosing this,
however--bad gauges are the problem more often than air in the
system,j To purge the system of air, run the condenser water
well after the refrigeration has been shut off. This will con-
dense all possible liquid. Both R-12 and R-22 vapors are much
heavier than air--the Freon/air mixture above the liquid will
separate, with the air going to the highest point. This point is
where you can open a valve for example, a service valve on the
compressor high-side! and let the air out.
To make a repair on the system, such as fixing a leak or
changing an expansion valve, keep the following steps in mind:
1. Get ever thin ready beforehand--tools, equipment, parts.
2. Pump t e system down to l or 2 psig.
3. Break the connection with this small amount of refrigerant
pressure on the system.
4. Make the repair as quickly as possible.

This won't occur on systems using R-22, which doesn't combine

with oil in the same way.
**
psig = pounds per square inch gauge, the amount of pressure
ab o v8 amb ien.t .
 Kith 1 or 2 psig in the piping, refrigerant vapor comes out
at all times; nothinggets in. Of c~ourse,
youcouldevacuate
the system before making a repair, but it's almost never necessary.
flow about welding on pipes containing Freon? It's true
that Freon plus high temperature leads to phosgene, a poisonous
gas. But at I psig, there won' t be enough phosgene around. to
do any harm. Be careful about welding on a pressurized pipe,
however. Often the heat will build up the pressure to such
a level that gas will bubble out through the weld. In this case,
you' ll have to go to the other end of the pipe and vent it.
During some repairs, the system does have to be left open
for quite a while, and air and moisture do get into the piping.
One textbook solution to this is to f irs t evacuate with a two-
stage vacuum pump, then purge with dry nitrogen followed by an-
other evacuation with the vacuum pump. As a practical matter,
few people have either a two-stage vacuum pump plus gauges! on
hand or the time t'several days to a week! to go through this pro-
cess. A better solution is to get rid of the moisture with drier
elements. Put in an oversized drier-- instead of a $20 drier, put
in a $~0 drier--one or two element changes will take all the mois-
ture out, and you' re done. Once the moisture is out, tXe air can
be purged as described earlier.
How about adding more refrigerant to the system? Adding
refrigerant is just like transferring fuel t'Figure 19!. The
filling bottle, usually disposable, commonly holds 2S pounds
of refrigerant. The filling valve is opened and the main king!

EXPANSION
LVE L COILS

F I GU RF. 19
 30

valve used for pumping down! is turned or it could remain

right-side up, filling the system with refrigerant gas!. As
you' re filling, watch the sight glass; you should see a. steady
flow of bubbles. If you have a system that holds, for example,
50-75 pounds of refrigerant and you' re pouring in a slug from a
25-pound bottle, it's probably o.k, to put the entire ZS pounds
in if the receiver is of an adequate size. But as discussed
previously, spare refrigerant should be kept in the spare
bottle and not in the system.

C. Com onent Servicin

Compressor

One of the difficulties with refrigeration systems on

fishing vessels in this area is that they may be used only about
60 days out of the year. Such long periods of down-time are
hard on systems. In the compressor, watch out for seals going
bad as a results--especially mechanical seals made out of neo-
prene and carbon face.*
Adequate lubrication obviously is very important. Keep
in mind, once again, that a system which is low on refrigerant
can run the risk of losing proper lubrication for the compressor--
movementof the refrigerant is necessary to keep the oiL circulat-
ing. Most compressorshave sight glasses showingthe level of
oil; if you can see the oil anywhere in the sight glass, it' s
probably o.k. On a splash lubrication system, the oil needs to
be only deep enough so the dippers hit. Oil pressure gauges are
frequently used to measurethe health of a pressurized lubrica-
tion system. The thing to watch out for is that these gauges
don't stay accurate very long.
It's easy to blame one component, such as the compressor,
for a problem that might be caused by something else in the sys-
tem. For example, a head pressure that is too high might have
any one of several causes: a restriction in the condenserwater
flow, an increase in the cooling water temperature, air in the
system, or even a bad high-side pressure gauge check to see
that it goes to zero when the pressure is reLeased!.
Whatshouldthe headpressurebe?* For R-12,pressure
should be an ntne range of 10S-125 psig, with discharge tempera-
tures around 140-160 'F. For R-Z2, pressures should be in the
range of 200-215psig, discharge temperaturesin the area of
200 'F. Temperature of the hot gas line from the compressor
This is a frequent problem in automotive air conditioning
systems, too.
Headpressure switches are frequently used to shut downwhenthe
head pressure gets too high, Theseswitches ver seldomfail.
Generally, adjusting or changinga pressure switc or t ermostat!
is a crutch, not a cure.
 3L

will usually be too hot to allow you to hold your hand against
it. However, the 1 ine should be below 212 'F. Check this by
splashing water or spitting on it. If the water sizzles, causing
steam, the line is over 212' and you have compressor problems.
A possible cause is that the discharge valve isn't closing tight-
ly. Therefore, gas would be sucked back in, causing too much to
be compressed during the compression cyc1e, resulting in. discharge
gas which is too hot.

It's also possible to check for bad valves by slowly clos-

ing off the suction line. The compressor should be able to
pump a good vacuum fairly rapidly--thelow side gauge should go
down to around 22-28 inches of vacuum. Following this, stop
the machine--the vacuum should hold. If it comes right back up,
there is a compressor problem--probably broken suction or dis-
charge valves.

Try to maintain even head pressures, because these affect

the rating of the expansion valve. For example, if you bought
a 5-ton valve for R-12 and ran it at 75 psi instead of 120, you' d
have a 2-1/2 ton valve. This is because the rating of the valve
p* I ~d *I .1
uneven head pressures xs to install a water regulator to control
water flow rate through the condenser.

For a similar reason, one occasionally sees refrigerant li-

quid slugging back to the compressor, especially with R-12.
One cause of this is a sudden increase in head pressure pos-
sibly due to operating in warmer temperature water!. If the
head pressure were to jump from, say, 120 to 180 psi, the
pressure drop across the open expansion valve is suddenly higher,
causing a greater volume of refrigerant to be pushed through the
system. Thus, for a short time, the capacity of the system is
increased; all of the refrigerant is not evaporated in the coils;
and some flows past the control bulb into the compressor.

Expansion Valve

There shouldn't be any servicing at all done on expansion

valves. First of all, buy goodvalves.* * Then, oncethe ex-
pansion valve is set and ad>usted, walk away from it and leave
A gauge which measures both pressure and vacuum will almost always
show units of "psi" for pressure above ambient, and "inches of
mercury" for "vacuums" below ambient. A true, absoLute vacuum
if it were attainable! would show up on the gauge as 29.92 inches.
*0
Broken suction or discharge valves in the compressor can frequently
cause loss of oil as well. Oil can be blown out of the compressor
within 15 minutes if one of these failures occurs.
*A*
One fisherman that had a habit of buying surplus expansion valves
for a dollar apiece ended up losing 100 tons of fish in two trips.
 32

it alone. If, for some reason, the valve doesn't hold its ad-
>ustment, it should go back to the factory. There is no reason
to take the cap off. If you occasionally detect inadequate flow
through the valve, the difficulty is usually elsewhere. This is
especially true of a valve that has been working for six months
or more. There might be dirt or moisture causing a problem, but
it ' s not the valve ' s f ault. Very f ew valves f ail.
As stated previously, expansion valves sense temperature
differen e, that is, the amount of superheat. So in order to
check that the valve is operating properly, it's necessary to
use two temperature sensors or a temperature sensor and a
pressure gauge!.* Expansion valves are usually set for about
7' superheat. If frost is seen on the compressor,it meansthat
the expaiis'oil salve is either set wrong or it's bad."" The genera1
solution to this problem is to buy a new valve. If you are using
a Sporlen or Alco valve, you can merely change the seat or the
inside of the valve.

Condenser

Discussion above has shown the importance of good water flow

through the condenser. Take the access plate off periodically
and clean out debris and fouling. For a shell-and-tube, or
double pipe, do this during lay-up or off-season. For a keel
cooler, you can take care of it when you haul out.
What happens if a condenser fails and water runs into the
refrigerant circuit7 This is a real messthat's going to take
a considerable amount of time to fix up. The following describes
the general procedure:
l. Assumingit is the condenserthat has failed, removeit,
clean and repair it, or throw it away and install a new
one.
2. Removethe compressor, overhaul and clean it, and put it.
back.
3. It is difficult to clean the sludge out of the receiver.
The best and probably cheapest solution is to throw it away
and install a new one. At this point, the compressor,
condenser, and receiver are clean and ready to go.
4. Removethe expansion valve and blow the refrigerant
lines clean with dry nitrogen. Then, with lines dis-
connected from the coils or "chiller," or "evaporator"!
blow the bulk of the water out of the coils with nitro-
gen. You~renowready to work on the final drying of
the coils. For this a "system cleaner" is needed.
Note that if you are attemptingto measurethe temperatureof the
gasby measuring
the outsidepipe wall temperature,remember
that
the wall will be warmer than the gas. For a copper pipe, the wall
will be about I/2 'F warmer; through a steel wall pipe, the drop
will be around 6-7 'F.
*It is possibleto havethe compressor
frost over for a short period
of time for example,if the brine pumpshuts off! and this doesn't
indicate a bad expansion valve. But it shouldn't happen as a
permanent thing.
 33

5. A system cleaner is essentially a distiller. One type

commonly used is made by York, is about 10 inches in
diameter and 4 feet high, and contains screening and
about 6 pounds of chemicals--either silica gel or ac-
tivated. alumina. Hook it up between the coiX and com-
pressor according to the diagram in Figure 20.
6. If the refrigerant used was R-ZZ, change to R-12 while
using the system cleaner. The reason for this is that
water will float on R-12 better than on R-22. Thus,
removal of water from the coils and separation in the
cleaner will be more efficient with R-12.
7. It takes sometimes 16 to 24 hours of running to clean
out and separate all the water in the system. Sight
glasses on the side of the cleaner indicate water or
refrigerant levels. Heat applied to the cleaner will
cause refrigerant to vaporize and flow to the com-
pressor. Chemicals silica gel or activated alumina!
must be changed frequently. The procedure is long and
difficult, but it is better than the alternative, which
is to run a vacuum pump for a few weeks, a technique
which still may not completely dry the system.

IQUID R-I2 R-I2 GAS TO

sOR

PLIED
OSE:BUD
R-I2
ED FROM
H~O.
LvE
ER

FIGURE 20

Occasionally, R-ll is used for this kind of job because it is

slightly cheaper. However, the advantage of R-12 is that the
refrigerant is recycled and, in many cases, left in the system
when the cleaning is comple ted.
Filters/Driers

Use a sight glass with a moisture indicator to continually

show the condition of the drier. Install it downstream of the
drier so that dirt or darkened refrigerant won't cloud the in-
dicator on initial start-up. The moisture indicator is actually
a piece of litmus paper--usually green whenit's dry, yellow
in the presenceof moisture. Theseindicators are 1004de-
pendable. If the oneon your systemturns color, or if you
openup the piping to do,work, changeboth the drier andthe
indicator. Because indicators are dependable and because a
refrigeration systemshouldbe completelysealed, you should
h~ h g d
Occasionally a filter or drier maybecomeclogged with
dirt. Anyrestriction in a liquid refrigerant line will be
accompanied
by a temperaturedrop. Obviously, a built-in, in-
tentional restriction is the expansion valve. However, when
the restriction is a drier or filter that is plugged or perhaps
undersized, the device acts somewhatlike the expansionvalve--
the pressuredrops over this restriction and someof the re-
frigerant is evaporated,thus decreasingthe temperature. You
can detect this by grabbing the tubing on either side. It
mightbe, for example,80 on the upstream
side and70' on the
downstream side. The result of this pre-evaporation is to
rob the last evaporator coils of refrigerant, causing a. de-
crease in capacity and an accumulationof oil in these last
coils.

Receiver

Occasionallythe flow into the receiver appearsto be

intermittent, as seenthroughthe top sight glass. That is,
it appearsto "rain" refrigerant for a while; it then stops
andyou can't seeanythingcomingin; thenonceagain, it
commences to "rain." This actually represents slugging of
liquid fromthe condenser into the receiver. Thesituation
is not particularly harmfulto the system;it is caused by
heat from the engine roomflowing into the line betweenthe
condenserand the receiver Figure 21!. For example,if the
seawatercooling the condenseris 60' and the engineroomis
90', the liquid leavingthe condensermaygassify and"slug"
into the receiver. Dueto this gas, a "false pressure" of
around1 psi is createdin the condenser.A solution to this
is to install an equalizer line /2- or S/8-inch coppertubing!
as diagrammed in Figure 21.
 T
SSES

Chiller

The biggest problem with chillers is clogging with scales

or salt. The resUlt might be a decrease in flow rate, heat
transfer rate, performance, and perhaps a freezing-up if the
salt is not being properly dissolved.
 36

INDEX

29-30 Horsepower 2-4

Adding Refrigerant
28,29 Hot Gas Bypass 9
Air in the System
Anti foaming 16i 19-20, 21-22 Hydraulic Drives 8, 16

Backpressure Regulator Icebuilders

Evaporator Pressure Insulation
Regulator! 9, 11
Brine Preparation 18-19 King Valve 5, 26, 29

2-4, 10-11 Leak Detectors 28

Capacity, Refrigeration
16-18 Leaks, Refrigerant 27-28
Chiller s
Open Coil 17-18
16-17, 26, 35 Moisture Indicator 34
Closed
Ccmpressor Multiple Circuits 6, 13
Performance 30-31
Sizing 1-4, 11-12 Oil Return 5-8, 23, 26-27, 31
10-11 . Oil Separators 7-8
Speed
Valves 26
P-Trap 6-7
Variable Capacity 9, ll
Condenser 4, 14, 25, 32 Piping
Condenser Failure 32-33 Brine Spray 20-22
Control Bulb 4-5, 12-13, 23-24 Refrigerant 4-7, 10, 14-15, 23-25
Controls 8-9, 11 Pumrxiown 26
Copper, Use of 5, 9, 23, 24 Pumps 15-16, 22
Purging the Systemof Air 28
Deflectors for Spray 22
Double Suction Riser 6-7 R-12
Drier 5, 14, 29, 34 5, 10-11, 14-15, 27-28, 30-31, 33
Dry Expansion DX! System 7 R-22 10-11, 14-15, 27-28, 30-31
Drying a System After It Has R-502 10
Been Opened 29, 32-33 Receivers 14, 27, 34-35
Refr igerant
Electric Power Generator 12 Adding 29-30
Equalizer Tubefor Condenser 34-35 Changingfrom one to Another 10-11
Evaporator Pressure Regulator Selection 10
Backpressure Regulator! 9, ll Shortage 26
ExpansionValve ll, 12-13, 25, 31-32 Wholesale Prices 10
Refrigeration SystemDiagram 5
Filters 14, 24, 34 Refrigeration Ton 1-4
Flared Fittings 24 Repair Procedurefor Piping 28-29
Flexible Sections 25
Flywheel Effect 1 Seals 15, 26, 30
Foaming 16, 19-20, 21-22 Shutdown 26
Frostpoint 13, 32 Sightglass, Refrigerant Line 27, 30
Sil-Fos 24
Gauges 18, 28, 30, 31 Sizing
Compressor 1-4, 11-12
Halide Torch 28 Condenser 4, 14
Head Pressure 30-31 Electric Power Generator 12
Headers 7 ExpansionValve 12-13, 31
Filters/Driers 14
Pumps 15-16, 22
Nr'iI
I0I iL >EnGRr',N1'
DIPt>ITOIIY
PEL' LIB''HY Rl]~lGING Rt:CI.'VE0
';,'1jTP.,'<~
URI, NPRr" AG,'JI, G.P1I'U! NAtlDNAL
SEA
Il&gflyggy>Tppy
5ARRAGANStT1;
R I02882 OATE,
AUG1