Related Commercial Resources

CHAPTER 45

FACTORY DEHYDRATING, CHARGING, AND TESTING

Dehydration (Moisture Removal) ................................................................................................ 45.1
Moisture Measurement ................................................................................................................. 45.3
Charging ...................................................................................................................................... 45.4
Testing for Leaks .......................................................................................................................... 45.4
Performance Testing .................................................................................................................... 45.5

P ROPER dehydration, charging, and testing of packaged refrig-

eration systems and components (compressors, evaporators,
and condensing coils) help ensure proper performance and extend
of 10 to 15 ppm on bulk shipments. Controls at the factory are
needed to ensure these moisture levels in the oils and refrigerant are
maintained.
the life of refrigeration systems. This chapter covers the methods Newer insulating materials in hermetic motors retain much less
used to perform these functions. It does not address criteria such as moisture compared to the old rag paper and cotton-insulated
allowable moisture content, refrigerant quantity, and performance, motors. However, tests by several manufacturers have shown that
which are specific to each machine. the stator, with its insulation, is still the major source of moisture in
compressors.
DEHYDRATION (MOISTURE REMOVAL)
Dehydration by Heat, Vacuum, or Dry Air
Factory dehydration may be feasible only for certain sizes of
equipment. On large equipment, which is open to the atmosphere Heat may be applied by placing components in an oven or by
when connected in the field, factory treatment is usually limited to using infrared heaters. Oven temperatures of 180 to 340°F are usu-
purge and backfill, with an inert holding charge of nitrogen. In most ally maintained. The oven temperature should be selected carefully
instances, this equipment is stored for short periods only, so this to prevent damage to the synthetics used and to avoid breakdown of
method suffices until total system evacuation and charging can be any residual run-in oil that may be present in compressors. Air in the
done at the time of installation. oven must be maintained at low humidity. When dehydrating by
Excess moisture in refrigeration systems may lead to freeze-up heat alone, the time and escape area are critical; therefore, the size
of the capillary tube or expansion valve. It also has a negative effect of parts that can be economically dehydrated by this method is
on thermal stability of certain refrigeration oils [e.g., polyol ester restricted.
(POE)]. Contaminants can cause valve breakage, motor burnout, The vacuum method reduces the boiling point of water below
and bearing and seal failure. Chapter 6 has more information on the ambient temperature. The moisture then changes to vapor,
moisture and other contaminants in refrigerant systems. which is pumped out by the vacuum pump. Table 3 in Chapter 6 of
Except for freeze-up, these effects are not normally detected by the 2005 ASHRAE Handbook—Fundamentals shows the relation-
a standard factory test. Therefore, it is important to use a dehydra- ship of temperature and pressure for water at saturation.
tion technique that yields a safe moisture level without adding for- Vacuum is classified according to the following absolute pres-
eign elements or solvents. In conjunction with dehydration, an sure ranges:
accurate method of moisture measurement must be established. Low Vacuum 29.92 to 1.0 in. Hg
Many factors, such as the size of the unit, its application, and type of Medium Vacuum 1.0 in. Hg to 1 µm Hg
refrigerant, determine acceptable moisture content. Table 1 shows
moisture limits recommended by various manufacturers for partic- High Vacuum 1 to 10–3 µm Hg
ular refrigeration system components. Very High Vacuum 10–3 to 10–6 µm Hg
Ultrahigh Vacuum 10–6 µm Hg and below
Sources of Moisture
The degree of vacuum achieved and the time required to obtain
Moisture in refrigerant systems can be (1) retained on the sur- the specified moisture level are a function of the (1) type and size of
faces of metals; (2) produced by combustion of a gas flame; (3) con- vacuum pump used, (2) internal volume of the component or sys-
tained in liquid fluxes, oil, and refrigerant; (4) absorbed in the tem, (3) size and composition of water-holding materials in the sys-
hermetic motor insulating materials; (5) derived from the factory tem, (4) initial amount of moisture in the volume, (5) piping and
ambient at the point of unit assembly; and (6) provided by free fitting sizes, (6) shape of the gas passages, and (7) external temper-
water. Moisture contained in the refrigerant has no effect on dehy- atures maintained. The pumping rate of the vacuum pump is critical
dration of the component or unit at the factory. However, because only if the unit is not evacuated through a conductance-limiting ori-
the refrigerant is added after dehydration, it must be considered in fice such as a purge valve. Excessive moisture content, such as a
determining the overall moisture content of the completed unit. pocket of puddled water, takes a long time to remove because of the
Moisture in oil may or may not be removed during dehydration, volume expansion to vapor.
depending on when the oil is added to the component or system. Vacuum measurements should be taken directly at the equipment
Bulk mineral oils, as received, have 20 to 30 ppm of moisture. (or as close to it as possible) rather than at the vacuum pump. Small
Synthetic POE lubricants have 50 to 85 ppm; they are highly hygro- tubing diameters or long tubing runs between the pump and the
scopic, so they must be handled appropriately to prevent moisture equipment should be avoided because line/orifice pressure drops
contamination. Refrigerants have an accepted commercial tolerance reduce the actual evacuation level at the equipment.
If dry air or nitrogen is drawn or blown through the equipment
The preparation of this chapter is assigned to TC 8.1, Positive Displace- for dehydration, it removes moisture by becoming totally or par-
ment Compressors. tially saturated. In systems with several passages or blind passages,
flow may not be sufficient to dehydrate. The flow rate should obtain

45.1
Copyright © 2006, ASHRAE
45.2 2006 ASHRAE Handbook—Refrigeration

Table 1 Typical Factory Dehydration and Moisture-Measuring Methods for Refrigeration Systems
Component Dehydration Method Moisture Audit Moisture Limit
Coils and tubing 250°F oven, –70°F dry-air sweep Dew point recorder 10 mg
Evaporator coils
Small –70°F dew point dry-air sweep, 240 s P2O5 25 mg
Large –70°F dew point dry-air sweep, 240 s P2O5 65 mg
Evaporators/condensers 300°F oven, 1 h, dry-air sweep Cold trap 200 mg
Dry-air sweep Nesbitt tube 8.5 mg/ft2 surf. area
Condensing unit (0.25 to 7.5 ton) Purchase dry P2O5 25 to 85 mg
Dry-air sweep Nesbitt tube 8.5 mg/ft2 surf. area
Air-conditioning unit Evacuate to 240 µm Hg P2O5 35 ppm
3 h winding heat, 0.5 h vacuum Refrigerant moisture check 25 ppm
Refrigerator 250°F oven, dc winding heat, vacuum Cold trap 200 mg
Freezer –70°F dew point dry-air ambient, −40°F dew point air sweep P2O5 10 ppm
Compressors
dc Winding Heat
0.5 h dc winding heat 350°F, 0.25 h vacuum/repeat Cold trap 200 mg
dc winding heat 190°F, 0.5 h vacuum Cold trap 1200 mg
2 to 60 ton semihermetic dc winding heat, 30 min, evacuation, N2 charge Cold trap 1000 to 3500 mg
Oven Heat
250°F oven, 4 h vacuum Cold trap 180 mg
250°F oven, 5.5 h at –60°F dew point air Cold trap 200 mg
0.5 to 12 ton hermetic 300°F oven 4 h, –70°F dew point air 3.5 min Cold trap 150 to 400 mg
50 to 100 ton Oven at 270°F, 4 h evacuate to 1000 µm Hg Cold trap 750 mg
1.5 to 5 ton hermetic 340°F oven, –100°F dew point dry air, 1.5 h Cold trap 100 to 500 mg
2 to 40 ton semihermetic 250°F oven, –100°F dew point dry air, 3.5 h Cold trap 100 to 1100 mg
5 to 150 ton open 175°F oven, evacuate to 1 mm Hg Cold trap 400 to 2700 mg
Scroll 2 to 10 ton hermetic 300°F oven 4 h, 50 s evacuation and 10 s –70°F Cold trap 300 to 475 mg
dew point air charge/repeat 7 times
Hot Dry Air, N2
3 to 5 ton Dry air at 275°F, 3 h Cold trap 250 mg
7.5 to 15 ton Dry air at 275°F, 0.5 h vacuum Cold trap 750 mg
20 to 40 ton Dry N2 sweep at 275°F, 3.5 h evacuate to 200 µm Hg Cold trap 750 mg
Dry N2 Flush
Reciprocating, semihermetic N2 run, dry N2 flush, N2 charge — —
Screw, hermetic/semihermetic R-22 run, dry N2 flush, N2 charge — —
Screw, open N2 run, dry N2 flush, N2 charge — —
Evacuation Only
Screw, open, 50 to 1500 ton Evacuate <1500 µm Hg, N2 charge — —
Refrigerants As purchased Electronic analyzer Typically 10 ppm
Lubricants
Mineral oil As purchased Karl Fischer method 25 to 35 ppm
As purchased and evacuation Hygrometer 10 ppm
Synthetic polyol ester As purchased Karl Fischer method 50 to 85 ppm

optimum moisture removal, and its success depends on the overall system or component. The dry air used should have a dew point
system design and temperature. between –40 and –100°F. Heat sources are the same as those men-
tioned previously. Heat can be combined with a vacuum to acceler-
ate the process. The heat and dry-air method is effective with open,
Combination Methods hermetic, and semihermetic compressors. The heating temperature
Each of the following methods can be effective if controlled care- should be selected carefully to prevent damage to compressor parts
fully, but a combination of methods is preferred because of the or breakdown of any residual oil that may be present.
shorter drying time and more uniform dryness of the treated system. Advantages and limitations of the various methods depend
Heat and Vacuum Method. Heat drives deeply sorbed moisture greatly on the system or component design and the results expected.
to the surfaces of materials and removes it from walls; the vacuum Goddard (1945) considers double evacuation with an air sweep
lowers the boiling point, making the pumping rate more effective. between vacuum applications the most effective method, whereas
The heat source can be an oven, infrared lamps, or an ac or dc cur- Larsen and Elliot (1953) believe the dry-air method, if controlled
rent circulating through the internal motor windings of semiher- carefully, is just as effective as the vacuum method and much less
metic and hermetic compressors. Combinations of vacuum, heat, expensive, although it incorporates a 1.5 h evacuation after the hot-
and then vacuum again can also be used. air purge. Tests by manufacturers show that a 280°F oven bake for
Heat and Dry-Air Method. Heat drives moisture from the 1.5 h, followed by a 20 min evacuation, effectively dehydrates com-
materials. The dry air picks up this moisture and removes it from the pressors that use newer insulating materials.
Factory Dehydrating, Charging, and Testing 45.3

MOISTURE MEASUREMENT saturated vapor pressure of water at a constant temperature) for each
fluid must be determined. For many fluids, this constant must be cor-
Measuring the correct moisture level in a dehydrated system or rected for the operating temperature at the sensor.
part is important but not always easy. Table 1 lists measuring meth- Christensen Moisture Detector. The Christensen moisture de-
ods used by various manufacturers, and others are described in the tector is used for a quick check of uncharged components or units on
literature. Few standards are available, however, and acceptable the production line. In this method, dry air is blown first through the
moisture limits vary by manufacturer. dehydrated part and then over a measured amount of calcium sulfate
Cold-Trap Method. This common method of determining (CaSO4). The temperature of the CaSO4 rises in proportion to the
residual moisture monitors the production dehydration system to quantity of water it absorbs, and desired limits can be set and mon-
ensure that it produces equipment that meets the required moisture itored. One manufacturer reports that coils were checked in 10 s
specifications. An equipment sample is selected after completion with this method. Moisture limits for this detector are 2 to 60 mg.
of the dehydration process, placed in an oven, and heated at 150 to Corrections must be made for variations in desiccant grain size, the
275°F (depending on the limitations of the sample) for 4 to 6 h. quantity of air passed through the desiccant, and the difference in in-
During this time, a vacuum is drawn through a cold-trap bottle strument and component temperatures.
immersed in an acetone and dry-ice solution (or an equivalent), Karl Fischer Method. In systems containing refrigerant and
which is generally held at about –100°F. Vacuum levels are oil, moisture may be determined by (1) measurement of the dielec-
between 10 and 100 µm Hg, with lower levels preferred. Important tric strength or (2) the Karl Fischer method (Reed 1954). In this
factors are leaktightness of the vacuum system and cleanliness and method, a sample is condensed and cooled in a mixture of chloro-
dryness of the cold-trap bottle. form, methyl alcohol, and Karl Fischer reagent. The refrigerant is
Vacuum Leakback. Measuring the rate of vacuum leakback is then allowed to evaporate as the solution warms to room tempera-
another means of checking components or systems to ensure that no ture. When the refrigerant has evaporated, the remaining solution is
water vapor is present. This method is used primarily in conjunction titrated immediately to a dead stop electrometric end point, and the
with a unit or system evacuation that removes the noncondensables amount of moisture is determined. This method requires a 15 g
before final charging. This test allows a check of each unit, but too sample of refrigerant and takes about 20 min. Multiple checks are
rapid a pressure build-up may signify a leak, as well as incomplete run to confirm results. This method is generally considered inaccu-
dehydration. The time factor may be critical in this method and must rate below 15 ppm; however, it can be used for checking complete
be examined carefully. Blair and Calhoun (1946) show that a small systems because this method does not require that oil be boiled off
surface area in connection with a relatively large volume of water the refrigerant. Reed points out that additives in the oil, if any, must
may only build up vapor pressure slowly. This method also does not be checked to ensure that they do not interfere with the reactions of
give the actual condition of the charged system. the method. The Karl Fischer method may also be used for deter-
Dew Point. When dry air is used, a reasonably satisfactory check mining moisture in oil alone (ASTM Standard D117; Morton and
for dryness is a dew-point reading of the air as it leaves the part Fuchs 1960; Reed 1954).
being dried. If airflow is relatively slow, there should be a marked An alternative method is available. A 5 to 10 g refrigerant sam-
difference in dew point between air entering and leaving the part, ple is injected directly into Karl Fischer reagents at a constant flow
followed by a decrease in dew point of the leaving air until it even- rate using a pressure-reducing device such as a capillary tube. After
tually equals the dew point of the entering air. As is the case with all the refrigerant is completely passed through the reagent, the mois-
systems and methods described in this chapter, acceptable values ture content is determined by automatic titration of a dead stop
depend on the size, usage, and moisture limits desired. Different electrometric end point. This alternative method takes about 1 h to
manufacturers use different limits. perform and is typically considered to be accurate to 5 ppm.
Gravimetric Method. In this method, described by ASHRAE Electrolytic Water Analyzer. Taylor (1956) describes an elec-
Standard 35, a controlled amount of refrigerant is passed through a trolytic water analyzer designed specifically to analyze moisture
train of flasks containing phosphorous pentoxide (P2O5), and the levels in a continuous process, as well as in discrete samples. The
weight increase of the chemical (caused by the addition of moisture) device passes the refrigerant sample, in vapor form, through a sen-
is measured. Although this method is satisfactory when the refrig- sitive element consisting of a phosphoric acid film surrounding two
erant is pure, any oil contamination produces inaccurate results. platinum electrodes; the acid film absorbs moisture. When a dc volt-
This method must be used only in a laboratory or under carefully age is applied across the electrodes, water absorbed in the film is
controlled conditions. Also, it is time-consuming and cannot be electrolyzed into hydrogen and oxygen, and the resulting dc current,
used when production quantities are high. Furthermore, the method in accordance with Faraday’s first law of electrolysis, flows in pro-
is not effective in systems containing only small charges of refrig- portion to the weight of the products electrolyzed. Liquids and
erant because it requires 200 to 300 g of refrigerant for accurate vapor may be analyzed because the device has an internal vaporizer.
results. If it is used on systems where withdrawal of any amount of This device handles the popular halocarbon refrigerants, but sam-
refrigerant changes the performance, recharging is required. ples must be free of oils and other contaminants. In tests on desic-
Aluminum Oxide Hygrometer. This sensor consists of an alu- cants, this method is quick and accurate with R-22.
minum strip that is anodized by a special process to provide a porous Sight-Glass Indicator. In fully charged halocarbon systems, a
oxide layer. A very thin coating of gold is evaporated over this struc- sight-glass indicator can be used in the refrigerant lines. This device
ture. The aluminum base and gold layer form two electrodes that consists of a colored chemical button, visible through the sight
essentially form an aluminum oxide capacitor. glass, that indicates excessive moisture by a change in color. This
In the sensor, water vapor passes through the gold layer and method requires that the system be run for a reasonable length of
comes to equilibrium on the pore walls of the aluminum oxide in time to allow moisture to circulate over the button. This method
direct relation to the vapor pressure of water in the ambient surround- compares moisture only qualitatively to a fixed standard. Sight-
ing the sensor. The number of water molecules absorbed in the oxide glass indicators have been used on factory-dehydrated split systems
structure determines the sensor’s electrical impedance, which mod- to ensure that they are dry after field installation and charging, and
ulates an electrical current output that is directly proportional to the are commonly used in conjunction with filter driers to monitor
water vapor pressure. This device is suitable for both gases and liq- moisture in operating systems.
uids over a temperature range of 158 to –166°F and a pressure range Special Considerations. Although all methods described in this
of about 10 µm Hg to 5000 psig. The Henry’s Law constant (satu- section can effectively measure moisture, their use in the factory
ration parts per million by mass of water for the fluid divided by the requires certain precautions. Operators must be trained in the use of
45.4 2006 ASHRAE Handbook—Refrigeration

the equipment or, if the analysis is made in the laboratory, the proper the specified positive pressure and submerged in a well-lighted tank
method of securing samples must be understood. Sample flasks filled with clean water. It may take a few minutes for development
must be dry and free of contaminants; lines must be clean, dry, and of a small bubbles trace to visualize a small leak. Note that bubbles
properly purged. Procedures for weighing the sample, time during can develop on the surface as a result of outgassing, and develop-
the cycle, and location of the sample part should be clearly defined ment of a trace is a key factor. This method of leak testing is not as
and followed carefully. Checks and calibrations of the equipment sensitive as the mass spectrometer or electronic leak detection meth-
must be made on a regular basis if consistent readings are to be ods, but is suitable for high-volume production.
obtained. Soap Bubble Leak Detection. High-rate leaks from a pressur-
ized system can be found by applying a soapy liquid solution to the
CHARGING suspected leak areas. Bubbles that form in the solution indicate
refrigerant leakage.
The accuracy required when charging refrigerant or oil into a unit
depends on the size and application of the unit. Charging equipment Fluorescent Leak Detection. This system involves infusing a
must also be adapted to the particular conditions of the plant; equip- small quantity of a fluorescent additive into the oil/refrigerant
ment may be manual or automatic. Standard charging is used where charge of an operating system. Leakage is observed as a yellow-
extreme accuracy is not necessary or the production rate is not high. green glow under an ultraviolet (UV) lamp. This method is suitable
Fully automatic charging boards check the vacuum in the units, for halocarbon systems. Because the additive is in the oil, thorough
evacuate the charging line, and meter the desired amount of oil and cleanup is needed after the leak is fixed to avoid a false positive
refrigerant into the system. These devices are accurate and suitable caused by leftover oil residue. It may also be a problem to identify
for high production. fluorescent glow in daylight.
Refrigerant and oil must be handled carefully during charging; Pressure Testing. The test article is sealed off under pressure or
the place and time of oil and refrigerant charging greatly affect the vacuum, and any decrease or rise in pressure noted over time indi-
life of a system. To avoid unnecessary complications (foaming, oil cates leakage. Dry nitrogen is often used as the medium for pressure
slugging, improper oil distribution, etc.), the unit should be charged testing. The limitations of this method are the time required to con-
with oil before the refrigerant. Charging with refrigerant should duct the test, the lack of sensitivity, and the inability to determine the
avoid liquid slugging during initial start-up; the best way to do this location of any leak that may exist.
is to charge the unit at the high-pressure side. Refrigerant lines must Electronic Leak Testing. The electronic leak detector consists of
be dry and clean, and all charging lines must be kept free of moisture a probe that draws air over a platinum diode, the positive ion emis-
and noncondensable gases. Also, new containers must be connected sion of which is greatly increased in the presence of a halogen gas.
with proper purging devices. Carelessness in observing these pre- This increased emission is translated into a visible or audible signal.
cautions may lead to excess moisture and noncondensables in the Electronic leak testing shares with halide torches the disadvantages
refrigeration system. that every suspect area must be explored and that contamination
Oil storage and charging systems should be designed and main- makes the instrument less sensitive; however, it does have some
tained to avoid contamination and direct contact between oil and air. advantages: mainly, increased sensitivity. With a well-maintained
Regular checks for moisture or contamination must be made at the detector, it is possible to identify leakage at a rate of 10–3 mm3/s
charging station to ensure that oil and refrigerant delivered to the (standard), which is roughly equivalent to the loss of 1 oz of refrig-
unit meet specifications. Compressors charged with oil for storage erant in 100 years. The instrument also can be desensitized to the
or shipment must be charged with dry nitrogen. Compressors with- point that leaks below a predetermined rate are not found. Some
out oil may be charged with dry air. models have an automatic compensating feature to accomplish this.
The problem of contamination is more critical with improved
TESTING FOR LEAKS sensitivity, so the unit under test is placed in a chamber slightly pres-
surized with outside air, which keeps contaminants out of the pro-
Extended warranties and critical refrigerant charges add to the duction area and carries contaminating gas from leaky units. An
importance of proper leak detection before charging. audible signal allows the probe operator to concentrate on probing,
The U.S. Environmental Protection Agency (EPA) established without having to watch a flame or dial. Equipment maintenance
an allowable leakage rate for certain refrigerants (e.g., no more than presents a problem because the sensitivity of the probe must be
0.1 oz per year of R-22 at 150 psig). A system that has 4 to 6 oz checked at short intervals. Any exposure to a large amount of refrig-
of refrigerant and a 5 year warranty must have virtually no leak, erant causes loss of probe sensitivity. A rough check (e.g., air under-
whereas in a system that has 10 to 20 lb of refrigerant, the loss of water testing) is frequently used to find large leaks prior to use of the
1 oz of refrigerant in 1 year would not have much affect on system electronic device.
performance. Any leak on the low-pressure side of a system oper- Mass Spectrometer. In this method, the unit to be tested is evac-
ating below atmospheric pressure is dangerous regardless of the uated and then surrounded by a helium-and-air mixture. The vac-
size of the refrigerant charge. uum is sampled through a mass spectrometer; any trace of helium
Before any leak testing is done, the component or system should indicates one or more leaks. Many equipment manufacturers use the
be strength tested at a pressure considerably higher than the leak test mass spectrometer leak detection method because of its high sensi-
pressure. This test ensures safety when the unit is being tested under tivities: mass spectrometers can detect leaks of 10–7 mm3/s. Test
pressure in an exposed condition. Applicable design test pressures levels for production equipment are typically set near 10–2 mm3/s.
for high- and low-side components have been established by Under- This method is normally used to measure the total leakage rate from
writers Laboratories (UL), the American Society of Mechanical all joints simultaneously. The main limitation for this method is that
Engineers (ASME), the American National Standards Institute the costs for test equipment and consumables are higher than for
(ANSI), and ASHRAE. Units or components using composition other leak detection methods.
gaskets as joint seals should have the final leak test after dehydra- The required concentration of helium depends on the maximum
tion. Retorquing bolts after dehydration helps to reduce leaks past leak permissible, the configuration of the system under test, the time
gaskets. the system can be left in the helium atmosphere, and the vacuum
level in the system; the lower the vacuum level, the higher the
Leak Detection Methods helium readings. The longer a unit is exposed to the helium atmo-
Water Submersion Testing. A water submersion test is a sphere, the lower the concentration necessary to maintain the
method of leak and strength testing. The test article is pressurized to required sensitivity. If, because of the shape of the test unit, a leak is
Factory Dehydrating, Charging, and Testing 45.5

distant from the point of sampling, a good vacuum must be drawn, Compressor Testing
and sufficient time must be allowed for traces of helium to appear on
The two prime considerations in compressor testing are power
the mass spectrometer.
and capacity. Secondary considerations are leakback rate, low-volt-
As with other methods described in this chapter, the best testing age starting, noise, and vibration.
procedure in using the spectrometer is to locate and characterize cal-
Testing Without Refrigerant. A number of tests measure com-
ibrated leaks at extreme points of the test unit and then to adjust
pressor power and capacity before the unit is exposed to refrigerant.
exposure time and helium concentration to allow cost-effective test-
In cases where excessive power is caused by friction of running gear,
ing. One manufacturer reportedly found leaks of 0.05 oz of refrig-
low-voltage tests spot defective units early in assembly. In these
erant per year by using a 10% concentration of helium and exposing
tests, voltage is increased from a low or zero value to the value that
the tested system for 10 min.
causes the compressor to break away, and this value is compared
The sensitivity of the mass spectrometer method can be limited
with an established standard. When valve plates are accessible, per-
by the characteristics of the tested system. Because only the total
formance can be tested by using an air pump for leakback tests. Air
leakage rate is found, it is impossible to tell whether a leakage rate
at fixed pressure is put through the unit to determine the flow rate at
of, for example, 1 oz per year is caused by one fairly large leak or
which valves open properly. The air pressure exerted against the
several small leaks. If the desired sensitivity rejects units outside the
closing side of the valve indicates its efficiency. This method is effec-
sensitivity range of tests listed earlier in this chapter, it is necessary
tive only when the valves are reasonably tight, and is difficult to use
to use a helium probe to locate leaks. In this method, the compo-
on valves that must be run in before seating properly.
nent or system to be probed is fully evacuated to clear it of helium;
then, while the system is connected to the mass spectrometer, a fine Extreme care should be taken when a compressor is used to
jet of helium is sprayed over each joint or suspect area. With large pump air because the combination of oil, air, and high temperatures
systems, a waiting period is necessary because some time is caused by compression can result in a diesel effect or an explosion.
required for the helium to pass from the leak point to the mass spec- In a common test using the compressor as an air pump, the dis-
trometer. To save time, isolated areas (e.g., return bends on one end charge airflow is measured through a flowmeter, orifice, or other
of a coil) may be hooded and sprayed with helium to determine flow-measuring device. When the volumetric efficiency of the com-
whether the leak is in the region. pressor with refrigerant is known, the flow rate that can be expected
with air at a given pressure may be calculated. Because this test adi-
Special Considerations abatically compresses the air, the discharge pressure must be low to
prevent overheating of discharge lines and oil oxidation if the test
There are two general categories of leak detection: those that lasts longer than a few minutes. (The temperature of adiabatic com-
allow a leak check before refrigerant is introduced into the system, pression is 280°F at 35 psig, but 540°F at 125 psig.) When the com-
and those that require refrigerant. Methods that do not use refriger- pressor is run long enough to stabilize temperatures, both power and
ant have the advantage that heat applied to repair a joint has no flow can be compared with established limits. Discharge tempera-
harmful effects. On units containing refrigerant, the refrigerant must ture readings and speed measurements aid in analyzing defective
be removed and the unit vented before any welding, brazing, or sol- units. If a considerable amount of air is discharged or trapped, the air
dering is attempted. This practice avoids refrigerant breakdown and used in the test must be dry enough to prevent condensation from
pressure build-up, which would prevent the successful completion causing rust or corrosion on the discharge side.
of a sound joint. Another method of determining compressor performance re-
All leak-testing equipment must be calibrated frequently to quires the compressor to pump from a free air inlet into a fixed vol-
ensure maximum sensitivity. The electronic leak detector and the ume. The time required to reach a given pressure is compared
mass spectrometer are usually calibrated with equipment furnished against a maximum standard acceptable value. The pressure used in
by the manufacturer. Mass spectrometers are usually checked using this test is approximately 125 psig, so that a reasonable time spread
a flask containing helium. A glass orifice in the flask allows helium can be obtained. The time needed for measuring the capacity of the
to escape at a known rate; the operator calibrates the spectrometer compressor must be sufficient for accurate readings but short
by comparing the measured escape rate with the standard. enough to prevent overheating. Power readings can be recorded at
The effectiveness of the detection system can best be checked any time in the cycle. By shutting off the compressor, the leakback
with calibrated leaks made of glass, which can be bought commer- rate can be measured as an additional check. In addition to the
cially. These leaks can be built into a test unit and sent through the pump-up and leakback tests noted above, a vacuum test should also
normal leak detection cycles to evaluate the detection method’s be performed.
effectiveness. Ensure that the test leak site does not become closed; The vacuum test should be performed by closing off the suction
the leakage rate of the test leak must be determined before and after side with the discharge open to the atmosphere. The normal vacuum
each system audit. obtained under these conditions is 1 to 1.5 psia. Abrupt closing of
From a manufacturing standpoint, use of any leak detection the suction side also allows the oil to serve as a check on the priming
method should be secondary to leak prevention. Improper brazing capabilities of the pump because of the suppression of the oil and
and welding techniques, unclean parts, untested sealing compounds attempt to deaerate. This test also checks for porosity and leaking
or improper fluxes and brazing materials, and poor workmanship gaskets. To establish reasonable pump-up times, leakback rates, and
result in leaks that occur in transit or later. Careful control and anal- suctions, a large number of production units must be tested to deter-
ysis of each joint or leak point make it possible to concentrate tests mine the range of production variation.
on areas where leaks are most likely to occur. If operators must scan In any capacity test using air, only clean, dry air should be used
hundreds of joints on each unit, the probability of finding all leaks in order to prevent compressor contamination.
is rather small, whereas concentration on a few suspect areas
Observing performance while testing compressors of known
reduces field failures considerably.
capacity and power best establishes the acceptance test limits
described. Take precautions to prevent oil that has been used repeat-
PERFORMANCE TESTING edly to lubricate many compressors from becoming acidic or con-
Because there are many types and designs of refrigeration sys- taminated.
tems, this section only presents specific information on reciprocat- Testing with Refrigerant. Calorimeter and flow meter testing
ing compressor testing and covers some important aspects of methods for rating positive-displacement compressors are described
performance testing of other components and complete systems. in ASHRAE Standard 23. This type of testing is typically conducted
45.6 2006 ASHRAE Handbook—Refrigeration

on an audit basis. If the purpose of the testing is not an accurate deter- The primary function of the factory performance test is to ensure
mination of the unit’s capacity and efficiency, alternative methods that a unit is constructed and assembled properly. Therefore, all
can be used, such as testing on vapor or desuperheating stands. The equipment must be compared to a standard unit, which should be
vapor stand requires an expansion device (TXV) and a heat typical of the unit used to pass the Air-Conditioning and Refrigera-
exchanger (or condenser) large enough to handle the heat equivalent tion Institute (ARI) and Association of Home Appliance Manufac-
to the motor power. The gas compressed by the compressor is cooled turers (AHAM) certification programs for compressors and other
until its enthalpy is the same as that at suction conditions. It is then units. ARI and AHAM provide rating standards with applicable
adiabatically expanded back to the suction state. This method elim- maximum and minimum tolerances. Several ASHRAE and Interna-
inates the need for an evaporator and uses a smaller heat exchanger tional Organization for Standardization (ISO) standards specify
(condenser). On small-capacity compressors, a piece of tubing that applicable rating tests.
connects discharge to suction and has a hand expansion valve can be Normal causes of malfunction in a complete refrigeration system
used effectively. The measure of performance is usually the relation- are overcharging, undercharging, presence of noncondensable gases
ship of suction and discharge pressures to power. When a water- in the system, blocked capillaries or tubes, and low compressor effi-
cooled heat exchanger (condenser) is used, the discharge pressure is ciency. To determine the validity and sensitivity of any test proce-
usually known, and the water temperature rise and flow are used as dure, it is best to use a unit with known characteristics and then
capacity indicators. Operation of the desuperheating stand is similar, establish limits for deviations from the test standard. If the estab-
but in addition to a condenser and TXV, it also requires a hot-gas lished limits for charging are ±1 oz of refrigerant, for example, the
bypass valve (HGBV). Liquid refrigerant from a condenser and hot test unit is charged first with the correct amount of refrigerant and
discharge gas are mixed by the HGBV to provide adequate suction then with 1 oz more and 1 oz less. If this procedure does not estab-
pressure and temperature to the compressor: the HGBV controls suc- lish clearly defined limits, it cannot be considered satisfactory and
tion pressure and the TXV, acting as a quench valve, controls super- new values must be established. This same procedure should be fol-
heating. Note that higher range and stability during operation are lowed regarding all variables that influence performance and cause
achieved by using a desuperheating stand instead of a vapor stand. deviations from established limits. All equipment must be main-
As a further refinement, flow-measuring devices can be installed tained carefully and calibrated if tests are to have any significance.
in the refrigerant lines. This system is charge-sensitive if predeter- Gages must be checked at regular intervals and protected from
mined discharge and suction pressures and temperatures are to be vibration. Capillary test lines must be kept clean and free of contam-
obtained. This is satisfactory when all units have the same capacity ination. Power leads must be kept in good repair to eliminate high-
and one test point is acceptable, because the charge desired can be resistance connection, and electrical meters must be calibrated and
determined with little experimentation. When various sizes are to be protected to yield consistent data.
tested, however, or more than one test point is desired, a liquid In plants where component testing and manufacturing control
receiver after the condenser can be used for full-liquid expansion. have been so well managed that the average unit performs satisfac-
The refrigerant must be free of contamination, inert gases, and torily, units are tested only long enough to find major flaws. Sample
moisture; the tubing and all other components should be clean and lot testing is sufficient to ensure product reliability. This approach is
sealed when they are not in use. In the case of hermetic and semi- sound and economical because complete testing taxes power and
hermetic systems, a motor burnout on the test stand makes it imper- plant capacity and is not necessary.
ative not to use the stand until it has been thoroughly flushed and is When the evaporator load is static (e.g., for refrigerators or freez-
absolutely acid-free. In all tests, oil migration must be observed ers), time, temperature, and power measurements are used to mea-
carefully, and the oil must be returned to the crankcase. sure performance. Performance is determined by the time elapsed
The length of a compressor performance test depends on various between start and first compressor shutoff or by the average on-and-
factors. Stable conditions are required for accuracy. If oil pump or off period during a predetermined number of cycles in a controlled
oil charging problems are inherent, the compressor should be run or known ambient temperature. Also, concurrent suction and dis-
long enough to ensure that all defects are detected. charge temperatures in connection with power readings are used to
establish conformity to standards. On units where the necessary
Testing Complete Systems connections are available, pressure readings may be taken. Such
In a factory, testing of any system may be done at a controlled readings are usually possible only on units where refrigerant loss is
ambient temperature or at an existing shop ambient temperature. In not critical because some loss is caused by gages.
both cases, tests must be run carefully, and any necessary correc- Units with complicated control circuits usually undergo an oper-
tions must be made. Because measuring air temperature and flow is ational test to ensure that controls function within design specifica-
difficult, production-line tests are usually more reliable when sec- tions and operate in the proper sequence.
ondary conditions are used as capacity indicators. Measurements of
water temperature and flow, power, cycle time, refrigerant pres- Testing of Components
sures, and refrigerant temperatures are reliable capacity indicators. Component testing must be based on a thorough understanding
When testing self-contained air conditioners, for example, a fixed of the use and purpose of the component. Pressure switches may be
load may be applied to the evaporator using any air source and either calibrated and adjusted with air in a bench test and need not be
a controlled ambient or shop ambient temperature. As long as the checked again if there is no danger of blocked passages or pulldown
load is relatively constant, its absolute value is not important. For tripout during the operation of the switch. However, if the switch is
water-cooled units, in which water flow can be absolutely controlled, brazed into the final assembly, precautions are needed to prevent
capacity is best measured by the heat rejected from the condenser. blocking the switch capillary.
Suction and discharge pressures can be measured for the analysis. Capillaries for refrigeration systems are checked by air testing.
Suction and discharge pressures and temperatures can be used as When the capillary limits are known, it is relatively easy to establish
an indirect measure of capacity in units with air-cooled condensers. a flow rate and pressure drop test for eliminating crimped or
As long as the load is relatively constant, the absolute value is not improperly sized tubing. When several capillaries are used in a dis-
important. Air distribution, velocity, or temperature over the test tributor, a series of water manometers check for unbalanced flow
unit’s coil must be kept constant during the test, and the perfor- and can find damaged or incorrectly sized tubes.
mance of the test unit must then be correlated with the performance In plants with good manufacturing control, only sample testing
of a standard unit. Power measurements supplement the suction and of evaporators and condensers is necessary. Close control of coils
discharge parameter readings. during manufacture leads to the detection of improper expansion,
Factory Dehydrating, Charging, and Testing 45.7

poor bonding, split fins, or uneven spacing. Proper inspection elim- ASHRAE. 1992. Method of testing desiccants for refrigerant drying. ANSI/
inates the need for costly test equipment. In testing the sample, ASHRAE Standard 35-1992.
either a complete evaporator or condenser or a section of the heat ASTM. 2002. Standard guide for sampling, test methods, and specifications
transfer surface is tested. Because liquid-to-liquid is the most easily for electrical insulating oils of petroleum origin. Standard D117-02.
American Society for Testing and Materials, West Conshohocken, PA.
and accurately measurable method of heat transfer, a tube or coil can
Blair, H.A. and J. Calhoun. 1946. Evacuation and dehydration of field instal-
be tested by flowing water through it while it is immersed in a bath lations. Refrigerating Engineering (August):125.
of water. The temperature of the bath is kept constant, and the capac- Goddard, M.B. 1945. Moisture in Freon refrigerating systems. Refrigerating
ity is calculated by measuring the coil flow rate and the temperature Engineering (September):215.
differential between water entering and leaving the coil. Larsen, L.W. and J. Elliot. 1953. Factory methods for dehydrating refriger-
ation compressors. Refrigerating Engineering (December):1325.
REFERENCES Morton, J.D. and L.K. Fuchs. 1960. Determination of moisture in fluorocar-
ASHRAE. 2005. Methods of testing for rating positive displacement refrig- bons. ASHRAE Transactions 66:434.
erant compressors and condensing units. ANSI/ASHRAE Standard Reed, F.T. 1954. Moisture determination in refrigerant oil solutions by the
23-2005. Karl Fischer method. Refrigerating Engineering (July):65.
ASHRAE. 2004. Designation and safety classification of refrigerants. Taylor, E.S. 1956. New instrument for moisture analysis of “Freon” fluori-
ANSI/ASHRAE Standard 34-2004. nated hydrocarbons. Refrigerating Engineering (July):41.

Related Commercial Resources