1、80 2016 ASHRAEThis paper is based on findings resulting from ASHRAE Research Project RP-1641.ABSTRACTThe presence of unsaturated fluorocarbon contaminantsintherefrigerantsusedinHVAC AHRI2013). In addition to R-40, which may cause deadly explosions,unsaturated halogenated contaminants were also found
2、 in coun-terfeitrefrigerantgasthatwasusedinplaceofR-134a(McCamp-bell Analytical 2013). Unsaturated fluorocarbons, also known asEffects of Halogenated UnsaturatedContaminants on the Reliability ofHVAC Calm and Hourahan 2010, 2011). Their presence in therefrigerants used in HVAC Leck et al. 2013;Majur
3、in et al. 2014). Therefore, it may not be appropriate to clas-sify all unsaturated compounds as unstable and blanket themunder the same restrictive limit.ASHRAE RP-1641 (Rohatgi and Clark 2015) aimed atdetermining the effects of halogenated unsaturated contaminantson the stability of refrigerant/lub
4、ricant systems and recommend-ing a concentration limit specific to the unsaturated contaminantbelow which the refrigerant/lubricant system is thermally stable.EXPERIMENTAL METHODSBased on a literature search and review, three refrigerant/lubricant mixtures with their corresponding contaminantswere s
5、elected for stability study in sealed-tube tests accordingto the test matrix shown in Table 1.Sealed Tube PreparationThe lubricants tested were 3MAF polyoester (POE) andwhite naphthenic mineral oil, which are representative ofcommercially used POE and mineral oil lubricants. The mois-ture content of
6、 the lubricants was adjusted to less than 50 ppmby drying with gentle heat (65C 149F) under vacuum. Thecopper, aluminum, and steel catalyst coupons (copper CDA110 or C11000, steel 1010, and aluminum 1100) wereprepared by punching 20 32mm(3/4 1/8 1/16 in.)coupons from thin sheets. The coupons were he
7、ld together byaluminum wire such that the steel and copper were separatedby the aluminum. These prepared coupon sandwiches werethoroughly cleaned and kept dry prior to use. The test tubeswerecleanedbyrinsingfirstwithdeionizedwater,thenbytworinses with methanol and one rinse with toluene. They weredr
8、iedat175C(347F)andkeptdryindesiccatorspriortouse.The metal coupons were first placed in the tube, which wasthen necked down to a size through which a standard cannulacould fit. Next1goflubricant was added accurately with asyringe and cannula. The tube was evacuated to 30 microns.With the tube submer
9、ged in liquid nitrogen, 1 g of refrigerantwas charged through condensation from a calibrated gas-handling system. The tube was again evacuated and thecondensible gas contaminant was charged through condensa-Table 1. Test MatrixRefrigerant/Lubricant ContaminantContaminant Concentration,ppmTemperature
10、/Time,C/days (F/days)R-134a/POEControl (no contaminant) 0175/14(347/14)R-1130 (CAS # 75-35-4)10005000R-1130(E) (CAS # 156-60-5)10005000R-1131 (CAS#460-16-2)100100030,000R-1234yf (CAS#754-12-1) 30,000R-1234yf/POEControl (no contaminant) 0R-1225ye(Z) (CAS # 5528-43-8)500030,000R-1233xf (CAS # 2730-62-
11、3)500030,000R-1243zf (CAS # 158664-13-2)500030,000R-123/Mineral oilControl (no contaminant) 0R-1122 (CAS # 359-10-4) 30,000R-1123 (CAS # 359-11-5) 30,000R-1131 (CAS # 460-16-2) 30,000Published in ASHRAE Transactions, Volume 122, Part 2 82 ASHRAE Transactionstionfromthecalibratedgas-handlingsystem.Fi
12、nally,thetubeneck was sealed and annealed. Control tubes without contam-inant were also prepared for comparison. Triplicate sealedtubes were prepared for each test condition listed in Table 1.The sealed tubes were placed in drilled holes in large alumi-num blocks, which were heated in air-circulatin
13、g ovens.Sample AnalysesAfteraging,thetubeswerevisuallyinspectedforchangesincolorofthelubricantorinthemetalsurfaces,signsofpartic-ulate, precipitate, metal corrosion, and copper plating. Thelubricant in the sealed tubes was analyzed for total acidnumber (TAN) and by ion chromatography (IC) to determi
14、nerefrigerant and/or lubricant decomposition. The lubricant wasalsoanalyzedbyinductivelycoupledplasma(ICP)tomeasurethe concentrations of dissolved metals.TAN: The TAN was determined for the lubricant accord-ing to a modified ASTM D664 (ASTM 2011). The methodwas modified to accommodate the small 1 mL
15、 sample size byreducing the alcoholic potassium hydroxide (KOH) titrantconcentration from 0.1 to 0.01 normal. This yielded sufficientsensitivity to determine acid numbers down to 0.10 mg KOH/g with a standard deviation of 0.05 mg KOH/g.IC: In the determination of anion concentrations by IC,about1gof
16、thelubricant sample was added to a preweighedcup containing 30 mL deionized water. The water/lubricantmixture was stirred continuously for 24 hours to allow forextraction of halide ions and acid anions from the lubricant.ThewaterextractwasthenanalyzedbyIC.Theconcentrationsof halide ions (such as flu
17、oride and chloride), organic anions(such as formate, acetate, butyrate, pentanoate, and hexa-noate), and inorganic anions (such as nitrate, sulfate, carbon-ate, and bicarbonate) were obtained by calibrating the ionchromatograph with standard solutions so the peak areas wereproportional to the anion
18、concentrations.Concentrations of Metals: Spectrochemical analysis byICP was performed according to ASTM-D5185 (ASTM2013) to determine the elemental metal concentrations (inparts per million) in the lubricant.EXPERIMENTAL RESULTSVisual InspectionThevisualchangesfortheR-134a/POEsystemareshowninTable2a
19、ndFigures1through3.Inthissystem,althoughthepresence of R-1234yf as a contaminant did not change thevisual appearance of the refrigerant/lubricant after aging, thepresence of R-1130 and R-1130(E) resulted in significantchanges, especially at the higher concentrations. The othersystems with their corr
20、esponding contaminants did not showany significant visual change from the controls.Table 2. Visuals of R-134a/POE System After Aging at 175C (347F) for 14 DaysRefrigerant/Lubricantwith ContaminantContaminantConcentration, ppmLiquid MetalsR-134a/POE(control)0 Clear, color unchanged, no deposit All me
21、tals unchangedR-134a/POE withR-11301000Faint cloudiness, color darker than unaged,no depositSteel with light to medium copper plating,copper and aluminum unchanged5000Light amount of tan floc, color darker thanunaged, faint white deposit on tube wall andin bottom of tubeSteel with light to medium co
22、pper plating,copper and aluminum unchanged30,000Color dark brown, much darker thanunaged, with heavy tan floc; whitish-tandeposit on tube walls and in bottom of tubeSteel dark brown with medium amount ofcopper plating, copper dull with brownstains, aluminum unchangedR-134a/POE withR-1130(E)1000Faint
23、 cloudiness, color darker than unaged,no depositSteel with medium to heavy copper plating,copper and aluminum unchanged5000Faint cloudiness, color darker than unaged,no depositSteel dark brown with medium copperplating, copper slightly dull, aluminumunchangedR-134a/POE withR-1131100Extremely faint c
24、loudiness, color slightlydarker than unaged, extremely faint whitedeposit in bottom of tubeAll metals unchanged1000Extremely faint cloudiness, color slightlydarker than unaged, no depositAll metals unchanged30,000 Clear, color darker than unaged, no deposit All metals unchangedR-134a/POE withR-1234y
25、f30,000 Clear, color unchanged, no deposit All metals unchangedPublished in ASHRAE Transactions, Volume 122, Part 2 ASHRAE Transactions 83Total Acid Number (TAN)ThechangesinTANs,obtainedbysubtractingtheTANofthe control from the TAN of the sample containing contami-nant, are shown in Table 3. In the
26、R-123/mineral oil system,the presence of contaminant did not result in significantchangeinTANafteraging,evenatahighconcentration.IntheR-1234yf/POEsystem,althoughtherewasnoincreaseinTANwhen the concentration of contaminant was 5000 ppm, therewas a small increase in TAN when the concentration ofcontam
27、inant was 30,000 ppm. In the R-134a/POE system,showninFigure4,thepresenceofthecontaminantR-1131didnotcausesignificantincreasesinTANafteragingevenathighconcentration, while the presence of R-1234yf at 30,000 ppmcausedsmallincreasesinTAN.GreaterincreasesinTANwererecorded with R-1130 and R-1130(E), esp
28、ecially at concen-trations of 5000 ppm or higher. At 1000 ppm the increases inTAN were small.Ion Chromatography Results, Total Organic Acid(TOA) AnionsICwasconductedtomeasuretheconcentrationsofhalideions (fluoride and chloride), organic acid anions (such asformate, acetate, butyrate, pentanoate, and
29、 hexanoate), andinorganicanions(suchasnitrate,sulfate,carbonate,andbicar-bonate). Fluoride and chloride are indicative of refrigerantand/or contaminant decomposition, while total organic acid(TOA) anion concentrations along with TAN are indicative oflubricant decomposition. The TOA represents the su
30、m of theconcentrations of the organic anions detected by IC as shownin Table 4. The changes in TOA shown in Table 3 wereobtained by subtracting the TOA of the control from the TOAof the sample containing contaminant.Figure 1 Sealed tubes containing R-134a/POE with R-1234yf contaminant.Figure 2 Seale
31、d tubes containing R-134a/POE with R-1130 contaminant.Figure 3 Sealed tubes containing R-134a/POE with R-1130(E) contaminant.Published in ASHRAE Transactions, Volume 122, Part 2 84 ASHRAE TransactionsThere was no chloride ion detected in any of the sealedtubes after aging. Fluoride ions were present
32、 in the tubescontaining R-1234yf, indicating of a small amount of refrig-erant decomposition. In the R-1234yf/POE system, althoughthere was no increase in TOA when the concentration ofcontaminants was 5000 ppm, there were increases in TOAwhen the concentration of contaminants was 30,000 ppmespeciall
33、y with R-1225ye(Z) and R-1243zf. In the R-134a/POE system shown in Figure 5, the presence of R-1234yf didnotcauseasignificantincreaseinTOAafteragingevenathighconcentration, while the presence of R-1131 at 30,000 ppmcaused a small increase in TOA. With R-1130 and R-1130(E)there was no increase in TOA
34、 when the concentration ofcontaminant was 1000 ppm and there were small increases inTOA when the concentration of contaminant was 5000 ppm.With R-1130 at 30,000 ppm, the TOA after aging wasextremely high (greater than 13,000 ppm).Concentrations of MetalsAs shown in Table 5, the only significant conc
35、entrationsof metals detected after aging were in the R-134a/POE systemwith R-1130 at 30,000 ppm.Table 3. Summary of Changes in Total Acid Numbers (TANs) andTotal Organic Acid (TOA) Anion ConcentrationsRefrigerant/Lubricantwith ContaminantContaminant Concentration,ppmChanges in TAN,mg KOH/gChanges in
36、 TOA,ppmR-134a/POEwith R-11301000 0.05 875000 0.42 9130,000 6.27 13045R-134a/POEwith R-1130(E)1000 0.05 1435000 0.38 158R-134a/POEwith R-1131100 0.07 421000 0.05 4130,000 0.04 296R-134a/POEwith R-1234yf30,000 0.35 9R-1234yf/POEwith R-1225ye(Z)5000 0.10 41730,000 0.16 755R-1234yf/POEwith R-1233xf5000
37、 0.14 36430,000 0.29 172R-1234yf/POEwith R-1243zf5000 0.23 32130,000 0.12 405R-123/mineral oilwith R-112230,000 0.01 43R-123/mineral oilwith R-112330,000 0 20R-123/mineral oilwith R-113130,000 0 1Figure 4 Changes in TAN of R-134a/POE system.Published in ASHRAE Transactions, Volume 122, Part 2 ASHRAE
38、 Transactions 85DISCUSSIONTo allow for easy comparison between the differentcontaminants tested and the controls, numerical values wereassigned to the visual observations as shown in Table 6. Thetotal value of the visuals for each test condition was obtainedby adding the values from the liquid phase
39、 (color, cloudiness,particulate, deposit, and film) and the metals (aluminum,copper, steel, and copper plating). The changes were calcu-latedbysubtractingthetotalvalueofthecontrolfromthetotalvalueofthesamplecontainingcontaminant.Negativechangesindicatethatthesystemcontainingcontaminantshowedbetterre
40、sults than the controls.To establish acceptability, pass/fail criteria were chosenasshowninTable7.Itshouldbeemphasizedthatthesecriteriaare highly subjective and selected mainly to provide ways ofdetermining the significant effects of the contaminants andrecommending a concentration limit specific to
41、 the contami-nant below which the refrigerant/lubricant system is stable.Different applications would require different acceptancecriteria, and the ones chosen here may be either too lenient ortoo stringent for a particular application.Theresultsofthepass/failevaluationofthecontaminantsare shown in
42、Table 8. Based on these results, it could beconcluded that the contaminants in the R-134a/POE systemR-1130, R-1130(E), and R-1131 at a concentration of1000 ppm passed the criteria of Table 7 across the board. Atthis concentration they showed changes in visuals, TAN,TOA,andconcentrationsofmetalscompa
43、rabletothecontrols.Contaminant R-1234yf was not tested at 1000 ppm; however,at30,000ppm,itpassedthetestsbasedonvisualscores,TOA,and concentration of metals, but failed the TAN test becauseits TAN after aging was 0.35 mg KOH/g greater than thecontrol. Using linear interpolation between the control(co
44、ncentration = 0, TAN = 0) and the test data (concentration= 30,000 ppm, TAN = 0.35 mg KOH/g) the increase in TANat1000ppmwouldbe0.01mgKOH/g,thusmakingR-1234yfacceptable at 1000 ppm in the R-134a/POE system based onthecriteriaofTable7.ThecontaminantsintheR-1234yf/POE(R-1225ye(Z), R-1233xf, and R-1243
45、zf) passed the criteria ofTable 4. Ion Chromatography ResultsRefrigerant/Lubricantwith ContaminantAnion Concentration, ppmFluoride Propanoate Chloride Formate Pentanoate 2-Ethyl Hexanoate TOAUnaged POE 0 0 0 0 0 0 0Unaged mineral oil 0 0 0 0 0 0 0R-134a/POE(control)0 0 0 20 0 230 250R-134a/POEwith R
46、-1130 at 30,000 ppm0 21 0 765 12509 13295R-134a/POEwith R-1130(E) at 5000 ppm0 0 0 157 0 251 408R-134a/POEwith R-1131 at 30,000 ppm0 0 0 109 0 437 546R-134a/POEwith R-1234yf at 30,000 ppm44 0 0 68 0 191 259R-1234yf/POE(control)42 0 0 110 0 464 574R-1234yf/POEwith R-1225ye(Z) at 30,000 ppm112 11 0 26
47、4 0 1054 1329R-1234yf/POEwith R-1233xf at 30,000 ppm51 5 0 165 0 576 746R-1234yf/POEwith R-1243zf at 30,000 ppm69 14 0 179 0 786 979R-123/mineral oil(control)00000 9191R-123/mineral oilwith R-1122 at 30,000 ppm0 0 0 0 0 134 134R-123/mineral oilwith R-1123 at 30,000 ppm00000 7171R-123/mineral oilwith
48、 R-1131 at 30,000 ppm00000 9090Published in ASHRAE Transactions, Volume 122, Part 2 86 ASHRAE TransactionsTable 8 across the board at a concentration of 5000 ppm, andthe contaminants in the R-123/mineral oil (R-1122, R-1123,and R-1131) passed at a concentration of 30,000 ppm.CONCLUSIONS AND RECOMMENDATIONSThe results from this study have shown that the 40 ppmlimit set by AHRI Standard 700 is too restrictive for manyunsaturated halogenated contaminants found in new and
