1、7.1CHAPTER 7CONTROL OF MOISTURE AND OTHER CONTAMINANTS IN REFRIGERANT SYSTEMSMoisture 7.1Other Contaminants 7.6System Cleanup Procedure After Hermetic Motor Burnout . 7.8Contaminant Control During Retrofit. 7.9Chiller Decontamination. 7.10MOISTUREOISTURE (water) is an important and universal contami
2、nantM in refrigeration systems. The amount of moisture in a refrig-erant system must be kept below an allowable maximum for satis-factory operation, efficiency, and longevity. Moisture must beremoved from components during manufacture, assembly, and ser-vice to minimize the amount of moisture in the
3、 completed system.Any moisture that enters during installation or servicing should beremoved promptly.Sources of MoistureMoisture in a refrigerant system results from Inadequate equipment drying in factories and service operationsIntroduction during installation or service operations in the fieldLea
4、ks, resulting in entrance of moisture-laden airLeakage of water-cooled heat exchangersOxidation of some hydrocarbon lubricants that produce moistureWet lubricant, refrigerant, or desiccantMoisture entering a nonhermetic refrigerant system throughhoses and sealsDrying equipment in the factory is disc
5、ussed in Chapter 5. Properinstallation and service procedures as given in ASHRAE Standard147 minimize the second, third, and fourth sources. Lubricants arediscussed in Chapter 12. If purchased refrigerants and lubricantsmeet specifications and are properly handled, the moisture contentgenerally rema
6、ins satisfactory. See the section on Electrical Insula-tion under Compatibility of Materials in Chapter 6 and the sectionon Motor Burnouts in this chapter.Effects of MoistureExcess moisture in a refrigerating system can cause one or all ofthe following undesirable effects:Ice formation in expansion
7、valves, capillary tubes, or evaporatorsCorrosion of metalsCopper platingChemical damage to motor insulation in hermetic compressors orother system materialsHydrolysis of lubricants and other materialsSludge formationIce or solid hydrate separates from refrigerants if the water con-centration is high
8、 enough and the temperature low enough. Solidhydrate, a complex molecule of refrigerant and water, can form attemperatures higher than those required to separate ice. Liquid waterforms at temperatures above those required to separate ice or solidhydrate. Ice forms during refrigerant evaporation when
9、 the relativesaturation of vapor reaches 100% at temperatures of 32F or below.The separation of water as ice or liquid also is related to the sol-ubility of water in a refrigerant. This solubility varies for differentrefrigerants and with temperature (Table 1). Various investigatorshave obtained dif
10、ferent results on water solubility in R-134a andR-123. The data presented here are the best available. The greaterthe solubility of water in a refrigerant, the less the possibility that iceor liquid water will separate in a refrigerating system. The solubilityof water in ammonia, carbon dioxide, and
11、 sulfur dioxide is so highthat ice or liquid water separation does not occur.The concentration of water by mass at equilibrium is greater inthe gas phase than in the liquid phase of R-12 (Elsey and FlowersThe preparation of this chapter is assigned to TC 3.3, Refrigerant Contami-nant Control.Table 1
12、 Solubility of Water in Liquid Phase ofCertain Refrigerants, ppm (by weight)Temp., F R-11 R-12 R-13 R-22R-113R-114R-123R-134aR-410AR-502160 460 700 4100 460 450 2600 4200 1780150 400 560 3600 400 380 2300 3600 1580140 340 440 3150 344 320 2000 3200 7200 1400130 290 350 2750 290 270 1800 2800 5800 12
13、20120 240 270 2400 240 220 1600 2400 4600 1080110 200 210 2100 200 180 1400 2000 3600 930100 168 165 1800 168 148 1200 1800 2800 81090 140 128 1580 140 120 1000 1500 2200 69080 113 98 1350 113 95 900 1300 1700 58070 90 76 1140 90 74 770 1100 1300 49060 70 58 44 970 70 57 660 880 950 40050 55 44 830
14、55 44 560 730 700 33540 44 32 26 690 44 33 470 600 510 27830 34 23.3 573 34 25 400 490 370 22520 26 16.6 14 472 26 18 330 390 270 18010 20 11.8 384 20 13 270 320 190 1460 15 8.3 7 308 15 10 220 250 130 11510 11 5.7 244 11 7 180 200 88 9020 8 3.8 3 195 8 5 140 150 59 6930 6 2.5 152 6 3 110 120 39 534
15、0 4 1.7 1 120 2 90 89 25 4050 3 1.1 91 1.5 70 66 3060 2 0.7 68 1 53 49 2270 1 0.4 50 0.6 40 35 1680 0.8 0.3 37 0.4 30 25 1190 0.5 0.1 27 0.2 22 18 8100 0.3 0.1 19 0.1 16 12 5Data on R-134a adapted from Thrasher et al. (1993) and Allied-Signal Corporation.Data on R-123 adapted from Thrasher et al. (1
16、993) and E.I. DuPont de Nemours it also varies with tem-perature. Table 2 shows the distribution ratios of water in the vaporphase to water in the liquid phase for common refrigerants. It canbe used to calculate the equilibrium water concentration of the liq-uid-phase refrigerant if the gas phase co
17、ncentration is known, andvice versa.Freezing at expansion valves or capillary tubes can occur whenexcessive moisture is present in a refrigerating system. Formationof ice or hydrate in evaporators can partially insulate the evaporatorand reduce efficiency or cause system failure. Excess moisture can
18、cause corrosion and enhance copper plating (Walker et al. 1962).Other factors affecting copper plating are discussed in Chapter 6.The moisture required for freeze-up is a function of the amountof refrigerant vapor formed during expansion and the distribution ofwater between the liquid and gas phases
19、 downstream of the expan-sion device. For example, in an R-12 system with a 110F liquidtemperature and a 20F evaporator temperature, refrigerant afterexpansion is 41.3% vapor and 58.7% liquid (by mass). The percent-age of vapor formed is determined by(1)wherehL(liquid)= saturated liquid enthalpy for
20、 refrigerant at liquid temperaturehL(evap)= saturated liquid enthalpy for refrigerant at evaporatingtemperaturehfg(evap)= latent heat of vaporization of refrigerant at evaporatingtemperatureTable 1 lists the saturated water content of the R-12 liquid phaseat 20F as 3.8 ppm. Table 2 is used to determ
21、ine the saturated vaporphase water content as3.8 ppm 15.3 = 58 ppmWhen the vapor contains more than the saturation quantity(100% rh), free water will be present as a third phase. If the temper-ature is below 32F, ice will form. Using the saturated moisture val-ues and the liquid-vapor ratios, the cr
22、itical water content of thecirculating refrigerant can be calculated asMaintaining moisture levels below critical value keeps free waterfrom the low side of the system.The previous analysis can be applied to all refrigerants and appli-cations. An R-22 system with 110F liquid and 20F evaporatingtempe
23、ratures reaches saturation when the moisture circulating is139 ppm. Note that this value is less than the liquid solubility,195 ppm at 20F.Excess moisture causes paper or polyester motor insulation tobecome brittle, which can cause premature motor failure. However,not all motor insulations are affec
24、ted adversely by moisture. Theamount of water in a refrigerant system must be small enough to avoidice separation, corrosion, and insulation breakdown.Polyol ester lubricants (POEs), which are used largely with hy-drofluorocarbons (HFCs), absorb substantially more moisture thando mineral oils, and d
25、o so very rapidly on exposure to the atmo-sphere. Once present, the moisture is difficult to remove. Hydrolysisof POEs can lead to formation of acids and alcohols that, in turn, cannegatively affect system durability and performance (Griffith 1993).Thus, POEs should not be exposed to ambient air exc
26、ept for verybrief periods required for compressor installation. Also, adequatedriers are particularly important elements for equipment containingPOEs.Exact experimental data on the maximum permissible moisturelevel in refrigerant systems are not known because so many factorsare involved.Drying Metho
27、dsEquipment in the field is dried by decontamination, evacuation,and driers. Before opening equipment for service, refrigerant mustbe isolated or recovered into an external storage container (seeChapter 9). After installation or service, noncondensable gases (air)should be removed with a vacuum pump
28、 connected preferably toboth suction and discharge service ports. The absolute pressureshould be reduced to 1 mm of mercury or less, which is below thevapor pressure of water at ambient temperature. External or internalheat may be required to vaporize water in the system. Take care notto overheat th
29、e equipment. Even with these procedures, smallamounts of moisture trapped under a lubricant film, adsorbed by themotor windings, or located far from the vacuum pump are difficultto remove. Evacuation will not remove any significant amount ofwater from polyol ester lubricants used in HFC systems. For
30、 thisreason, it is best to drain the lubricant from the system before dehy-dration, to reduce the dehydration time. A new lubricant chargeshould be installed after dehydration is complete. Properly disposeof all lubricants removed from the system, per local regulations.Table 2 Distribution of Water
31、Between Vapor and Liquid Phases of Certain RefrigerantsTemp., FWater in Vapor/Water in Liquid, mass %/mass %R-12 R-22 R-123 R-134A R-404A R-407C R-410A R-507A20 15.3 0 13.1 20 11.9 30 11.2 40 9.9 0.546 0.978 0.829 0.493 0.520 0.61150 9.0 0.566 0.965 0.844 0.509 0.517 0.65760 8.2 0.586 0.952 0.859 0.
32、525 0.514 0.70370 7.5 0.606 6.20 0.939 0.874 0.541 0.511 0.74980 6.3 0.626 5.40 0.926 0.889 0.557 0.508 0.79590 6.1 0.646 4.84 0.913 0.904 0.573 0.505 0.841100 5.5 0.666 4.62 0.900 0.919 0.589 0.502 0.887110 0.686 4.57 0.887 0.934 0.605 0.499 0.933120 0.706 4.52 0.874 0.949 0.621 0.496 0.979Data ada
33、pted from Gbur the color changesat a low enough level to be safe. Manufacturers instructions must befollowed because the color change point is also affected by liquid-line temperature and the refrigerant used.Moisture MeasurementTechniques for measuring the amount of moisture in a compres-sor, or in
34、 an entire system, are discussed in Chapter 8. The followingmethods are used to measure the moisture content of various halo-carbon refrigerants. The moisture content to be measured is gener-ally in the parts-per-million range, and the procedures requirespecial laboratory equipment and techniques.Th
35、e Karl Fischer method is suitable for measuring the moisturecontent of a refrigerant, even if it contains mineral oil. Although dif-ferent firms have slightly different ways of performing this test andget somewhat varying results, the method remains the commonindustry practice for determining moistu
36、re content in refrigerants.The refrigerant sample is bubbled through predried methyl alcoholin a special sealed glass flask; any water present remains with thealcohol. In volumetric titration, Karl Fischer reagent is added, andthe solution is immediately titrated to a “dead stop” electrometricend po
37、int. The reagent reacts with any moisture present so that theamount of water in the sample can be calculated from a previous cal-ibration of the Karl Fischer reagent. In coulometric titration (AHRI Standard 700C), water is titratedwith iodine that is generated electrochemically. The instrument mea-s
38、ures the quantity of electric charge used to produce the iodine andtitrate the water and calculates the amount of water present.These titration methods, considered among the most accurate,are also suitable for measuring the moisture content of unused lubri-cant or other liquids. Special instruments
39、designed for this particu-lar analysis are available from laboratory supply companies.Haagen-Smit et al. (1970) describe improvements in the equipmentand technique that significantly reduce analysis time.The gravimetric method for measuring moisture content ofrefrigerants is described in ASHRAE Stan
40、dards 35 and 63.1. It isnot widely used in the industry. In this method, a measuredamount of refrigerant vapor is passed through two tubes in series,each containing phosphorous pentoxide (P2O5). Moisture presentin the refrigerant reacts chemically with the P2O5and appears asan increase in mass in th
41、e first tube. The second tube is used as atare. This method is satisfactory when the refrigerant is pure, butthe presence of lubricant produces inaccurate results, because thelubricant is weighed as moisture. Approximately 200 g of refriger-ant is required for accurate results. Because the refrigera
42、nt mustpass slowly through the tube, analysis requires many hours.DeGeiso and Stalzer (1969) discuss the electrolytic moistureanalyzer, which is suitable for high-purity refrigerants. Other elec-tronic hygrometers are available that sense moisture by the adsorp-tion of water on an anodized aluminum
43、strip with a gold foil overlay(Dunne and Clancy 1984). Calibration is critical to obtain maximumaccuracy. These hygrometers give a continuous moisture readingand respond rapidly enough to monitor changes. Data showing dry-down rates can be gathered with these instruments (Cohen 1994).Brisken (1955)
44、used this method in a study of moisture migration inhermetic equipment.Thrasher et al. (1993) used nuclear magnetic resonance spectros-copy to determine the moisture solubilities in R-134a and R-123.Another method, infrared spectroscopy, is used for moisture analy-sis, but requires a large sample fo
45、r precise results and is subject tointerference if lubricant is present in the refrigerant.DesiccantsDesiccants used in refrigeration systems adsorb or react chemi-cally with the moisture contained in a liquid or gaseous refrigerant/lubricant mixture. Solid desiccants, used widely as dehydratingagen
46、ts in refrigerant systems, remove moisture from both new andfield-installed equipment. The desiccant is contained in a devicecalled a drier (also spelled dryer) or filter-drier and can be installedin either the liquid or the suction line of a refrigeration system.Desiccants must remove most of the m
47、oisture and not react unfa-vorably with any other materials in the system. Activated alumina,silica gel, and molecular sieves are the most widely used desiccantsacceptable for refrigerant drying. Water is physically adsorbed onthe internal surfaces of these highly porous desiccant materials.Activate
48、d alumina and silica gel have a wide range of pore sizes,which are large enough to adsorb refrigerant, lubricant, additives,and water molecules. Pore sizes of molecular sieves, however, areuniform, with an aperture of approximately 0.3 nm for a type 3Amolecular sieve or 0.4 nm for a type 4A molecula
49、r sieve. The uni-form openings exclude lubricant molecules from the adsorptionsurfaces. Molecular sieves can be selected to exclude refrigerantmolecules, as well. This property gives the molecular sieve the ad-vantage of increasing water capacity and improving chemical com-patibility between refrigerant and desiccant (Cohen 1993, 1994;Cohen and Blackwell 1995). The drier or desiccant manufacturercan provide information about which
