US3912617A - Blended refrigeration oil composition - Google Patents

Abstract

A wide boiling range (e.g. for a given VGC) refrigeration oil, having good chemical and thermal stability and good miscibility with fluorinated hydrocarbon refrigerants, comprises a blend of from 50-75 percent of a hydrorefined naphthenic oil component and 50-25 volume percent of a wax free paraffinic component (such as a hydrogenated polyolefin oil).

Description

United States Patent Mills et al.

BLENDED REFRIGERATION OIL COMPOSITION Inventors: Ivor W. Mills, Media; John J.

Melchiore, Wallingford, both of Pa.

Sun Oil Company of Pennsylvania, Philadelphia, Pa.

Filed: Feb. 14, 1974 Appl. No.: 442,552

Assignee:

US. Cl 208/14; 208/19 ClOg 41/00 Fieldof Search 208/14, 19, 264

References Cited UNITED STATES PATENTS 10/1944 Reid .f. 208/19 Oct. 14, 1975 3,576,898 4/1971 Blake et a1 260/6839 3,676,521 7/1972 Steams et a1. 260/683.15 D 3,715,302 2/1973 Mills et a1. 208/14 3,791,959 2/1974 Mills et al. 208/19 Primary Examiner Herbert Levine Attorney, Agent, or FirmGeorge L. Church; Donald R. Johnson; Bisson: Barry A.

[5 7] ABSTRACT 8 Claims, N0 Drawings BLENDED REFRIGERATION OIL COMPOSITION CROSS REFERENCE TO RELATED APPLICATIONS This application is related to application Ser. No. 175,775 of Ivor W. Mills and Glenn R. Dimeler which refers to and was copending with our application Ser. No. 63,303 filed 8-12-70 entitled Refrigeration Oil Composition Having Wide Boiling Range, now US. Pat. No. 3,715,302 which was issued Feb. 6, 1973, and was copending with our application Ser. No. 200,185, filed Nov. 18, 1971, now US. Pat. No. 3,791,959 issued Feb. 12, 1974.

The following patents and applications, all assigned to The Sun Oil Company (as is the present application), are related to the disclosure of the present application in that they disclose methods of obtaining hydrorefined naphthenic oils, wax-free paraffmic components and methods of reducing the basic nitrogen content of component oils which can be used to make the blended refrigeration oil composition of the present invention. Other methods for obtaining hydrorefined naphthenic and wax-free paraffmic fluids, useful as such blending components, are well known to the art.

The disclosure of said Ser. No. 63,303 and of Ser. No. 175,775 is hereby incorporated in the present application.

BACKGROUND AND SUMMARY OF THE INVENTION High temperature compressors operating with prior art naphthenic refrigeration oils develop coke deposits in the throttle valves. Apparently, these coke deposits are initiated by small particles of Mylar or other synthetic materials which are suspended or solubilized in the oil but deposit on the throttle valves because the oil vaporizes on hot spots. The deposits initiate decomposition of the refrigeration oil and cause it to coke.

A possible solution to this coking problem is to make a refrigeration oil with a high boiling point. One means of obtaining a higher boiling point is to use a paraffinic rather than a naphthenic oil. Refrigeration oils have been marketed which apparently consist of a paraffinic oil which has been acid treated. These oils usually contain some aromatics and although high boiling, have a narrow boiling range. In general, such paraffinic refrigeration oils lack the chemical stability of naphthenic refrigeration oils and/or they have a high floc point and poor miscibility with refrigerants (e.g. R-l2).

The present invention relates to a refrigeration oil which has good chemical stability and also good thermal stability and which has a wider boiling range for a given VGC than does a prior art paraffinic refrigera- 2 has good miscibility with conventional refrigerants, has good sealed tube stability and permits the operation of compressors at higher operating temperatures (e.g. a coil temperature above 125C.) than do present refrigeration oils.

A wide boiling range (eg for a given VGC) refrigeration oil of the present invention, having good chemical and thermal stability and good miscibility with fluorinated hyerocarbon refrigerants, comprises a blend of from 50-75 volume percent of hydrorefined naphthenic oil component and 50-25 volume percent of a wax-free paraffinic component. Preferably, the hydrorefined naphthenic oil component (which can be a blend of two or more hydrorefined oils) has an SUS viscosity at 100F. in the range of -750 SUS and contains in the range of 15-50 wt aromatic hydrocarbons (e.g. 20-35 wt%). The paraffinic component is chosen so that the resulting naphthenic-paraffinic blend has a viscosity at F. in the range of 100-500 SUS (preferably -300 SUS, typically -250 SUS) and has a maximum natural (i.e. doesnot contain a pour point depresant) floc point of 35F. Preferably the blend contains less than 10 ppm (more preferably less than 5) of basic nitrogen and contains at least 10 wt (more preferred in the range of 15-35 weight aromatics. The paraffmic component can contain or consist of hydrorefined paraffmic oil having a low wax content. The hydrorefined naphthenic component can be a raffinate from solvent extraction (either before or after hydrorefining) as with furfural, to reduce the aromatic content. The blends of the present invention have a greater aromatic content, and have a generally lower natural floe point (for a given viscosity at 100F.) than do the transmission oil blends of US. Pat. No. 3,450,636 to Rausch, which issued June 17, 1969.

The paraffinic component can be a wax-free, hydrogenated polyolefin oil or a high viscosity index, hydrocracked lube or a mixture of such components.

The preferred polyolefin oils are polymers or copolymers of CF olefin which have a pour point no greater than 3 5F., and preferably below 50F. The hydrogenation can be from 50 to 100% of saturation and, preferably, is to a bromine number no greater than 10, more preferably less than 5. Preferred polyolefins include ethylene-propylene copolymer, polypropylene, polybutene (especially polysobutene), and poly loctene).

The polyolefin (e.g., polybutene in the C -C range) can be partially or fully hydrogenated. Preferred polyolefins include those described in US. Pat. No. 3,576,898 to Blake et al., issued Apr. 27, 1971, and in the following United States patent applications (the disclosure of which is incorporated herein by reference):

Serial No. Patent No. Filing Date Inventors 28,942 now 4-15-70 Hirschler-Driscoll abandoned 52,300 3,775,503 7-6-70 Driscoll-Duling-Gates-Warren 52,301 3,778,487 7-6-70 Driscoll-Duling-Gates 78,191 3,676,521 10-5-70 Stearns-Duling-Gates 133.637 4-13-71 Duling-Glazier 135,295 4-19-71 Driscoll-Haseltine 144,165 3,715,313 5-17-71 Haseltine-Driscoll 162,896 7-15-71 Duling-Gates tion oil. The novel oil also retains a floc point no greater than 35F. (more preferred less than 40F),

The hydrocracked lube oils which can be employed as the parafi'mic component are preferably solvent ex- TABLE I Properties of Hydrocracked Oils* Aniline Viscosity ASTM Gravity Wt. Point (SUS, 100F) VI API Aromatics F.

I I03 34.2 12 220 200 107 33.3 I l 235 500 107 31.5 13 250 All oils dcwaxed to u 0F. pour point by chilling in a solvent and stabilizing by extraction with furfural.

The high V.I. hydrocracked lube component is obtained by hydrocracking a high viscosity distillate or dewaxed distillate from a paraffinic crude (such as Lagomedio) and typically has a VI in the range of 90405. The hydrocracked lubes can be stabilized by extraction with aromatic selective solvents, such as furfural or phenol. One process for preparing such a high V.l. lube oil comprises fractionating the stock material (such as an atmospheric residuum from Lagomedio crude) into three fractions, boiling at (a) from 720855F., (b) 855 to 980F. and (c) the residuum or a fraction boiling at from 986 1070F., solvent extracting fraction (b) with a solvent having preferential solubility for aromatics such as furfural, recombining the three fractions, and hydrocracking the combined fractions at from 720 to 800F. using a hydrogen partial pressure of from 2,000 to 3,000 psi, and a sulfided nickel-tungsten catalyst supported on silica-alumina and containing a minor amount of a fluoride (e.g. Gulf GC-6). The higher boiling fraction is deasphalted if required.

The novel oil can be made by blending a low basic nitrogen content paraffinic oil (having a viscosity at 100F. of about 500 SUS) with a low basic nitrogen content hydrorefined naphthenic oil (e.g. having an SUS viscosity at 100F. of about 150). A preferred process for insuring a low basic nitrogen content (e.g. having less than ppm, typically 1 ppm or less) in the naphthenic and paraffinic components and/or in the blend comprises acid or acid-clay contacting the blend and/or one or more of the component oils. For example, the Lewis acid contacting (preferably with neutralization) of the aforementioned applications Ser. Nos. 622,398 (now U.S. Pat. No. 3,462,358); 652,026 (now U.S. Pat. No. 3,502,567); 657,438; 850,717 now abandoned and 850,779 now U.S. Pat. No. 3,586,752. A preferred method of reducing the basic nitrogen in a blend of a hydrorefined naphthenic oil and a dewaxed parafflnic oil or in one or both of these component oils is to contact the oil with an adsorbent comprising an acid-activated adsorbent clay, more preferably an admixture or combination of an acid-activated adsorbent clay and a fullers earth bleaching clay such as attapulgite. Useful adsorbent admixtures and process conditions for such contacting are disclosed in U.S. Pat. No. 3,369,999, for the decolorizing of waxes.

FURTHER DESCRIPTION By naphthenic distillate, we refer to a distillate fraction (or a mildly acid treated distillate fraction, or a solvent raffinate fraction or an acid-treated raffinate) usually from vacuum distillation, of a crude which is classifled as naphthenic (including relatively naphthenic" by the viscosity-gravity constant (VGC) classification method. Preferably, such crudes are Grade A (waxfree), typically Gulf Coastal, and include, for example, Refugio, Mirando, and Black Bayou. The lower VGC oils can be obtained from mid-continental crudes; however, dewaxing-may be necessary (as by extraction or isomerization) to insure, for 50-70 SUS transformer oils, that the final blended oil has a pour point of less than 60F. Such fractions will have a VGC in the range of 0.820 to 0.899 and, usually a viscosity in the range of 30200 SUS at F. (typically, 40-70). In some cases the crude (and distillate) can have a VGC as high as 0.94 (such crudes are characterized as mildly aromatic). Deep furfural extraction (e.g., about 50% yield) of a high VGC Grade A crude can be used to produce a wax-free, lower VGC fraction (e.g. 0.83 VGC) which can be used in low floc point (or pour point) blends.

One important test of refrigeration oil quality is the floc point, which is a measure of the tendency of wax separation from the oil under conditions which simulate actual operation in a refrigeration system. The second important test, which can be correllated with actual use conditions, is the 14-day sealed tube stability test. These tests are further described in U.S. Pat. No. 3,715,302.

In the oil of the present invention, it is important that either the blended oil or all of its components be as free of basic nitrogen as is practical. Preferred means of reducing basic nitrogen being contacting with acid (e.g. H 50 and neutralization, or by contacting with an acid-activated clay (preferably a mixture of acidactivated clay and attapulgite). Such contacting to reduce the basic nitrogen in the oil insures good 14 day sealed tube stability (e.g., maximum of 2.5% R-22, more preferably 1.5% maximum, typically less than 1.0%).

The acid or acid clay treatment of the paraffmic component also causes improved results in the Falex failure load test, which is a test to determine lubricating quality of a refrigeration oil.

One low pour point, low basic nitrogen content oil, which can be used as a naphthenic component in the blends of the present invention is a high VGC oil produced by refining a blend of heavy catalytic gas oil (I-ICGO) and catalytic slurry oil (CS0). The blend is in about yield porportion (from cracking of a given feed stock, such as virgin gas oil). The heavy gas oil is the distillate product (from the cracked gas oil) which boils mainly above about 600F. and can have an end point at about thermal cracking temperature at about 5 mm Hg absolute pressure. The slurry oil can be the recycle from any catalytic cracking process, for example, see Ser. No. 730,999.

The blend of HCGO and CS0 (or a distillate fraction of the blend, e.g., 500850F.) is refined by furfural extraction, followed by naphtha washing of the extract to obtain a product containing at least 99 wt aromatics and having a pour point below 0F. (in contrast to the 40-70F. pour point of the original blend).

The 99+% aromatic product (which is useful as a heat exchange medium or as a seal swell agent) is then processed by one of two routes. Preferably, the product is mildly catalytically hydrocracked (e.g., no more than 20% conversion to products boiling below 600F.) to

obtain some ring opening and some ring hydrogenation. This product can be distilled to obtain a high boiling franction, useful in blends of the present invention.

Alternatively, instead of a hydrocracking step (which can require a prior hydrodesulfurization step), the 99+% aromatic product of the naphtha wash (after furfural stripping) can be hydrorefined, (as with sulfided NiMo catalyst) under hydrogenation conditions which decrease the ultraviolet absorptivity at 260 milimicrons by at least to about 50% to produce a low basic nitrogen, high VGC oil, the high boiling distillate fractions of which can be used as a naphthenic blending component in the present invention. This hydrorefined product can be subjected to mild catalytic cracking to further adjust the VGC.

An alternative to the hydrorefming under hydrogenation conditions (which reduce the 260 UVA) is to conduct the hydrorefining at about the autofining point (which maintains aromaticity).

For example, a (642F. initial to 827 end point, 95% point at 800F.) distillate fraction of such a blend was extracted with 250 Vol.% furfural at about 80F. and the extract, in solution with the furfural, was washed twice with 100% naphtha at about 80F., then the furfural was flash evaporated to produce a 99.6 wt aromatic oil having a refractive index of about 1.68. This oil was especially useful a heat transfer medium or as a seal swell agent for automatic transmission fluids or in hydraulic oils. The oil could alos be further refined by hydrocracking and/or hydrorefining, described above, to produce a high VGC, low basic nitrogen, low pour point oil which could be used a component of the blends of the present invention. This oil and its process of manufacture is the invention of lvor W. Mills and will be claimed in a later filed application.

lLLUSTRATlVE EXAMPLES In the following examples, SUS viscosity is at ]00F. and parts are by volume, unless otherwise indicated.

EXAMPLE l I A hydrorefined naphthenic oil component having a viscosity of 150 SUS was obtained by blending 100 SUS and 500 SUS hydrorefined naphthenic oils. Each of these oils was obtained by severe hydrorefming (as defined in US Pat. No. 3,462,358) of naphthenic acidfree naphthenic distillate. The hydrorefming was at 625F,, 1200 psig of 80% hydrogen, 0.25 LHSV, in the presence of sulfided NiMo oxide catalyst. The 150 SUS blended composition was contacted at about 240F. with a mixture, per barrel of oil, of 10 pounds of acidactivated clay and 10 pounds of attapulgite. The resulting clay contacted 150 SUS hydrorefined naphthenic oil contained less than 1 ppm of basic nitrogen.

A duosol solvent refined and methylethyl ketone dewaxed (to a pour point of 0F) paraffinic lube distillate, having a SUS viscosity of 500, was contacted at 240F. with a mixture, per barrel of oil, of 10 pounds of acidactivated clay and 10 pounds of attapulgite. The resulting 500 SUS paraffinic component contained less than 1 ppm of basic nitrogen.

A blended refrigeration oil, of the present invention, was obtained by blending 68 parts by volume of the 150 SUS hydrorefmed naphthenic oil component and 32 parts of the 500 SUS paraffinic component. The properties of the blended refrigeration oil are reported in Table l, along with a typical range of properties of a preferred blended refrigeration oil composition.

Table 2 hereof reports the distillation ranges in F., obtained under vacuum and corrected to at l atmosphere, of the hydrorefined naphthenic component, the refrigeration oil blend and the dewaxed paraffinic component. Note the wide boiling range of the blended refrigeration oil, compared with the boiling range of each component. In one preferred embodiment, at least volume percent of the naphthenic component will distill below 800F., corrected to 1 atmosphere, and at least 80 volume percent of the paraffinic component will distill above 800F., corrected to one atmosphere, the latter requirement being the most important. More preferred, at least 50% of the paraffinic component should boil above 900F.

EXAMPLE 2 A blended refrigeration oil was obtained as in Example 1,,except that the blend contained 63 parts of the 'l50 SUS hydrorefmed naphthenic oil component and 37 parts of the 500 SUS dewaxed paraffinic component. This blended refrigeration oil had the same properties as those reported in Table l for the oil of Example 1 except that the viscosity at F. was 235, the pour point was 25F., the aniline point was 199.8F. and the weight percent (gel) aromatics was 27.5.

The performance of the blended refrigeration oil of this example was evaluated after 2,000 hours of operation with R- l 2 in a compressor at a boiling temperature of about C. The condition of the Mylar, the valves, the copper plating on the shafts and of the other operating parts after the 2,000 hours was significantly better than the condition of similar parts from a compressor run for 2,000 hours at 160C. using a conventional naphthenic refrigerant (which failed before 2,000 hours).

In another series of use tests, the blended refrigeration oil of this example was compared with a low basic nitrogen content dewaxed paraffinic refrigeration oil and with a refrigeration oil comprising 60% alkylated benzene and 40% mineral oil. The tests were run for 3 months at 160C. winding temperature. The naphthenic-paraffmic blend of the present invention exhibited better wear and life properties than did the other two oils.

This was a severe test of thermal stability since these compressors normally operate within the l30l40C. temperature range. Generally, prior art naphthenic refrigeration oils are used in compressors where the temperature is in the l00-l 10C. range.

Coking of compressor valves is not a new phenomenon in refrigeration systems. Knowledgeable refrigeration engineers believe that the long life of a refrigeration compressor is ultimately ended by carbonized deposits on the valve leaves. This causes improper seating of the valves, makes the compressor inefficient, and if allowed to continue, would eventually cause a compressor failure. In some cases, the deposit on the discharge valve leaf is quite soft and can be wiped off with a finger; however, such a deposit can cause severe reduction in efficiency.

Even modern compressors which use carefully selected materials can be destroyed by carbonization of valve leaves by operating them at high discharge temperatures. For example, deposits have been found on the valves (e.g., suction leaf and valve stop) from a refrigerator compressor operated'for one week at a discharge cavity temperature of about 405F.

While high discharge temperatures are not a design objective, they are an unavoidable result of design economies. Such changes have been the trend, whether they take the form of smaller size and heavier loading of compressor, thinner cabinet walls or smaller condenser sizes. These trends have reached a stage where further economies involve the danger of carbonized valves from the high discharge temperature. This temperature therefore, becomes critically important for new system design and for field reliability.

Two reasons have generally been offered in the art to explain carbonized valve deposits. Traditionally, deposits have been attributed to the breakdown of oil, particularly to the chemical reactions which occur between the oil and the refrigerant in the presence of metals. The maximum reaction occurs at the discharge valve which is the hottest spot in a compressor.

More recently, emphasis has been shifted to the thermal degradation of motor insulation. (ASHRAE, Guide and Data Book, Systems 1970 Chapter 28, Refrigerant Systems Chemistry, pp. 415-416). According to this theory, the insulating materials used in the motor decompose at elevated temperatures and release the decomposition products into the oil. The oil which circulates through through the system, carries these decomposition products to the discharge valve where they react or decompose further to form hard, tenacious deposits on the valve leaves.

Both of these theories fail to account for observed deposits. For example, in a compressor made in 1954, and which had to be discarded in 1970, the motor insulation theory offers a plausible explanation. The insulation system was polyvinyl formal, paper and cotton lead wire and tie cord, which had a nominal Class A rating. Under occasional heavy loading in the field, it is possible that some decomposition products were indeed released to the circulating oil. On the other hand, the used motor was in excellent condition and showed no clear evidence of high temperature operation. The magnet wire was still flexible and had good dielectric strength. The phase insulating paper could be folded repeatedly without cracking, except in the outermost layer. The cotton tie cord had a breaking strength of l2-l 5 lbs, which is over 10,000 psi based on the nominal cross-section, The excellent condition of the insulating materials, of course, does not preclude high temperature excursions of short duration. However, in case of the compressor which was operated for one week at 405F. discharge temperature, the motor insulation theory does not remain valid. This compressor was specially built with polyimide film and polyimide wire insulation. The lead wires had tetrafluoroethylene insulation and only the tie cords were polyethylene terephthalate. The test was run under controlled conditions with winding temperatures at 250260F. It is most unlikely that the Polyimides, which are considered to be extremely stable materials, would decompose at these temperatures within the seven days of operation. Therefore, the use of an insulation of high thermal rating did not prevent the carbonization of valves.

Similar to the hypothesis regarding motor insulation, the theory of chemical reactions between oil and Refrigerant 12 would be a plausible explanation for the 1954 model field unit. After all, at that time there were no quantitative laboratory methods for evaluating the chemical reactivity of oils, and it is quite possible that the oil used in that particular compressor may have been much more reactive than those available today.

8 Such high reactivity could eventually cause carbonization.

Again, in the case of motor insulation theory, this explanation is unsatisfactory when the nature of the deposits in the 405F. discharge temperature case is considered. The compressor in this case had the least reactive naphthenic oil on the market, sold under the trade name Suniso 3G". Assuming that the conversion of R12 to R22 in a sealed tube test at C. for 14 days in presence of valve steel is a reasonable measure of the chemical reactivity of an oil, the particular oil in that compressor shows a typical value of 0.1-0.2 percent conversion. This value is sometimes hereinafter referred to as a Laboratory Reactivity Index. The use of an oil with a very low Laboratory reactivity lndex did not prevent the formation of carbonized deposits at the discharge valve. The possibility of contaminants also had to be ruled out, since all precautions were taken in the selection of the components, in the cleanliness of the system, and in the dehydration and evacuation of the unit.

A particularly interesting aspect of carbonization of valves with the low reactivity oil was that, except in the vicinity of the discharge valve, there were no other visual signs of oil breakdown. The oil in the sump remained clear with negligible color change. The bearings did not wear, copper plating was negligible, and the general appearance of the compressor parts was excellent. Even in the field unit of 1954, the compressor showed only a faint blush of copper plating at spots, the cellulosic motor insulation showed no signs of charring at any place, the bearings were totally unaffected, and again, except for the immediate vicinity of the discharge valve seat, the remainder of the valve and valve plate were clean and shiny. The oil in the compressor had an ASTM color 2 and appeared clear and clean.

Much has-been written during the past decade regarding the chemical reactions of oils and refrigerants in the presence of metals and contaminants. Laboratory tests, which give consistent results between different laboratories, have been developed. Attempts have been made to interpret these laboratory data in terms of the reactions at the discharge valve of a compressor, with the hope that such interpretation would be useful in predicting the life of an operating compressor. See, for example H. O. Spauschus and G. C. Doderer: Chemical Reactions of R22, ASHRAE Transactions, Vol. 71, Part I, 1965, p. 162. The underlying assumption has been that the laboratory data, in terms of the reactivity index, correlates with system performance, at least as far as the failure mechanism at the discharge valve is concerned. In other words, the chemical reactivity has been equated with the coking of discharge valves.

As-"a result of this work, oil suppliers have directed their development efforts in the past decade to improving'the laboratory reactivity index of refrigeration oils, At least one commercial oil has a reactivity index of 0.2, which is a four fold improvement over the product marketed a few years ago. The old product itself was quite satisfactory, and had about half the reactivity index as white oil with lubricity additives.

Therefore, carbonization cannot necessarily be prevented by the use of oils with a low laboratory reactivity index. It, therefore, becomes apparent that this carbonization is the direct result of high discharge temperature. The term thermal stability" will be used to define the tendency of an oil to resist-coking at these elevated temperatures.

EXAMPLE 3 Two runs have been made which provide an insight into the critical properties of a thermally stable oil, and

illustrate the advantages of the present invention,

These runs were done on complete systems, i.e.: on domestic U.S. refrigerator-freezers The critical cavity temperature was monitored by a thermocouple mounted inside the compressor next to the discharge valve and the results are based on this discharge cavity temperature. The commercial refrigerator freezer systems had to be modified in order to obtain the high cavity temperatures. Shortening of the auxiliary oil cooler condenser and an adjustment of the overload protector were usually enough to operate the systems at high cavity temperatures. In a typical coking experiment, the refrigerator-freezer was operated in a 110F. ambient continuously for 7 days with no door opening. The compressor was then torn down and the valves examined for deposits.

The first experiment used a commercial 200 SUS paraffinic oil (Oil B, Table 3) marketed under the trade name Suniso 21. This oil is not recommended for refrigeration application by the supplier because of its poor chemical stability. This particular sample had a laboratory reactivity index of 6.0. In a sealed tube test, the contents became black and gummy, the steel corroded, and a fair amount of acid was formed.

In the coking experiment" however, this oil formed a negligible amount of coke. The valve leaves had darkened with a thin varnish-like coating on the entire valve leaf. This varnish was quite different from the carbon build-up, in that it did not appear to interfere with the seating of the valve.

Aside from the chemical reactivity index, Oil B differed from Oil A in two ways. One was its structure. Oil B was paraffinic, whereas Oil A. was naphthenic. The other was its distillation range. The entire distillation curve of Oil B was about lOF. higher than that of Oil A.

The second run was conducted to check which of these two differences were most important for thermal stability. Oil C is a refrigeration grade, 500 SUS naphthenic oil, marketed as Suniso GS. Its reactivity index is approximately 1.0. Its distillation range is approximately the same as Oil B. The valve leaves with this oil were also free of interfering deposits when run in a coking experiment. Some varnish had formed, but there was no buildup of deposits at any place.

It appears that the boiling range of an oil is the key to thermal stability. The structure per se, whether naphthenic or paraffinic is incidental and a low laboratory reactivity index is not essential for good thermal stability.

It must also be recognized that good chemical stability, the appropriate viscosity grade, good viscosity-temperature relations, good solubility characteristics with refrigerants and good lubricating properties are among the many time tested requirements of a refrigeration oil. Thermal stability is simply an additional requirement, to the preceding oil properties.

Both Oil B and Oil C. lack some of the traditional re quirements. The better thermal stability of Oil C is believed to be due to the increased viscosity. The higher viscosity causes difficulty in cold start-up and also causes greater frictional losses. In the case of Oil B, the extreme reactivity was a major concern. Moreover,

10 paraffinic oils have poorer solubility relations with refrigerants, require expensive dewaxing to meet the low floc requirements and are believed to have inadequate lubricating properties. The traditional naphthenic oils do not have these drawbacks.

The present invention provides an oil which is basically naphthenic, yet has a high boiling range without significant increase in viscosity. The properties of one such oil (Oil D) are shown in Table 3.

The valves of a compressor remained clean after 7 days at 405F. using Oil D. The entire valve area remained clean, no varnish or carbon build up occurred. In addition to thermal stability, this oil satisfies all the other requirements of a refrigeration oil. The slightly higher viscosity creates no problems under cold startup and the frictional losses are not noticeably different with this oil than the presently used commercial Oi] A.

The boundary lubricating properties were found to be as good as the present oils in laboratory tests in the refrigerant environment, and confirmed by numerous system tests. The higher floc point of Oil D poses no problems. This was confirmed by operating a system with R22 and with R502 (a well known isotropic mixture of 48.8% chlorodifluoromethane and 51.2% chloropentafluoroethane) at 45F. evaporator. temperature. No congealing of the oil phase was detected and the oil return to the compressor was quite satisfactory.

Long term 4000 hour life tests, at 400F. temperature in the discharge cavity, were also conducted to prove the performance of Oil D. Under these conditions, some deposits were formed, but the nature and the extent of the deposits qualitatively were considered to be tolerable. In contrast, the traditional oils, which show some build up at 405F. in 7. days, would completely clog the value ports and stop the compressor in about 3 weeks.

EXAMPLE 4 Surprisingly, Oil B, which had a very high laboratory reactivity index, still had excellent thermal stability. However, even if there were no interfering deposits on the valves, the varnish build-up and the dark appearance of the valve leaves were not desirable. Also, after a few days in air, the valve leaves showed rust, which was not the case with Oil D. Two experiments were performed to determine the effect of Laboratory Reactivity Index. Oil B was refined to obtain an index of about 0.1. A less refined version of Oil D, having an index of y A about 4.0 was also procured. This oil was made as in Example 1 except that the paraffinic component was not clay contacted. In both cases, the well refined version showed a clean valve leaf whereas the less refined versions showed a darkened appearance and some formation of a varnish like deposit. The valve leaves with the less refined oils also rusted after a few days in air and in some cases, the varnish could be peeled off as flakes exposing a rusted steel surface underneath. The

1 1 formed, the R22 formation would be higher. This was not always the case. In some experiments when deposits were noted, the R22 formed was only 0.2-0.3 percent and in others it was 0.7-1.0 percent.

However, the physical appearance of the valve leaves after teardown offered a clue. It was repeatedly noted that when the high boiling oils were used, the valve leaves were considerably wetter, i.e., covered with a layer of oil. With the present commercial oils, the valve leaves appeared dry. The formation of deposits probably depends on whether the oil refrigerant mixture impinges on a dry or on a wet surface. If the surface is wet, a liquid phase reaction similar to that in a sealed tube test would occur between the oil and refrigerant. If the oil mist comes in contact with a dry valve surface, not only would the reaction occur in a different manner, but the reaction products would tend to accumulate on the valves. The high boiling oils probably keep the valve leaf wet and permit any reaction products to circulate freely in the system.

The new Oil'D discussed earlier was formulated from general use, i.e., for refrigeration down to 40F. and for air conditioning. For specialty applications, where thermal stability may be the prime requirement, the correlation between thermal stability and the boiling range can be put to further advantage. This is done by extending the boiling range even higher than in the case of Oil D. An experimental oil was made, (Oil E) with a boiling range which extended to a final boiling point of over I 150F. The temperature at the discharge cavity was 430F440F, and the duration of the test was seven days. The steel had darkened, but no interfering deposits were formed.

This temperature level at the discharge cavity already I strains the other component of the system, and further extension of the thermal stability correlation seems more of academic interest than of any practical significance. Nevertheless, some experiments were made with Oil F, which was a specially formulated synthetic paraffinic white oil whose boiling range was far higher than the available mineral oils. After considerable difficulty, one refrigerator freezer unit was operated for an extended period, at a temperature of 475F. in the discharge cavity. After 820 hours and with the system still operating satisfactorily, the test was stopped.

Tear down analysis showed no interfering deposits nor'va rnish on the valve leaves. Other components also showed the effects of a high temperature, such as the embrittlement of the motor insulation and discoloration of some of the steelsurfaces.

It should be emphasized that the correlation of thermal stability of an oil with deposit formation has been found largely with mineral hydrocarbon oils. The synthetic oil was also a branched chain paraffinic hydrocarbon type. For other synthetic non-hydrocarbon type oilsor even for alkylated benzene type oils the present thermal stability correlation may or may not hold. Other characteristicsthermal cracking or chemical reactions with the various materials in the system may well be the controlling factors.

The trend towards higher operating temperatures of refrigeration compressors is limited by the danger of coking of discharge valves. The phenomenon of coking is found is be the direct result of oil breakdown rather than an indirect result of the thermal degradation .of motor insulation. However, existing laboratory tests cannot be used to predict the carbonizing tendency of a refrigeration oil. Some oils which show a very low re- 12 activity in laboratory tests may indeed form coke and others which show poor stability in laboratory tests may have excellent non-coking characteristics. The term thermal stability" as distinguished from chemical stability" is suggested to define the non-coking characteristic of an oil. I

Thermal stability is found to be related to the boiling range of the refrigeration oil. The higher the boiling range, the higher is thermal stability.

For satisfactory performance at high operating temperatures, both thermal stability and chemical stability are found to be necessary. Refrigeration oils, totally of mineral origin, which combine these characteristics without sacrificing the other traditional requirements have been developed. One commercial oil of this nature has been extensively tested and found to be suitable for low temperatures of down to 4O"F. as well as for air conditioning applications such that it can be used across the board as an improved refrigeration oil.

For specialty purposes where the thermal stability may be the prime requirement, the correlation of the boiling range has been utilized to an extent that other components in the system are strained and coking is not the weak link. This has been achieved with mineral oils without resorting to any synthetic materials.

Synthetic materials do have an advantage because they can be tailored to provide a higher boiling range than is possible with mineral oils, without a corresponding increase in viscosity. Thus, specially formulated synthetic oils of a paraffinic hydrocarbon type can be used without danger of coking at incredibly high discharge temperatures. This has been demonstrated by operating a refrigerator-freezer system at 475F. in the discharge cavity for over one month.

In the discussion above, coil temperature has been loosely used as a synonym for the more technically correct term winding temperature".

The blends of the present invention can in some instances advantageously contain such additives as a metal deactivator, an antifoam, an antiwear agent, an antioxidant or mixture of antioxidants, (such as zinc dialkyl dithiophosphate and/or ditertiarybutyl paracresol.

TABLE I TYPICAL PROPERTIES OF BLENDED REFRIGERATION OIL Ex. I Blended Refriger- Typical Test Method* ation Oil Range Viscosity, SUS/100F. D2161 227 210-240 Viscosity, SUS/2 10F. D2l6l 46 45-52 Flash, COC, F. D92 355 340 min Fire. COC, F. D92 390 Pour, F. D97 -35 25 max Viscosity-gravity Constant Calculated 0.847 Color D1500 L 1.0 2.0 max Gravity, API D287 26.0 2527 Total Acid No., mgKOH/g D974 0.00 0.05 max Dielectric Strength, KV D877 30 25 min Inorganic Chloride & Sulfatcs D878 None Basic Nitrogen, ppm I less than 5 Free Sulfur D989 None Corrosive Sulfur; Class D1275 No. I Total Sulfur, 7r D129 0.04 Aniline Point, F. D6I I I Refractive Index Dl747 1.4945

Aromatics. Gel. "/1 30 Fine, F. 47 35 max Power Factor/25C., Initial D924 0.0001

TABLE l-continued TYPICAL PROPERTIES OF BLENDED REFRIGERATION OIL ASTM Test Designations **Floc and scaled tubc tests are with R-l2 refrigerant and are dcserihed in this specification. The basic nitrogen test is described in Serial No. 850.779.

TABLE 2 The dislillalions were made at reduced pressure and corrected (or extrapolated) to one atmosphere.

TABLE 3 Typical Properties of Refrigeration Oils Oil Oil Oil Oil* Oil Oil A B C D E F Viscosity, SUS 100] I55 205 500 230 I90 220 Viscosity Index 91 55 58 27 I40 APl Gravity 22.8 31.3 24 26 23.5 38.7 Color ASTM 0.5 1.0 1.0 1.0 0.5 0.5 Molecular Wt. 300 394 370 340 330 640 Composition CA 71 I4 5 7 l 1 12.5 CN '71 43 31 46 40 42 0 CP "/1 43 64 47 49 45.5 I00 Gel Aro matics 71 34 I5 22 27.5 30 0 Flash Point F 340 395 420 375 350 470 Pour Point F 0 40 30 60 Floc Point "F 68 40 -50 45 40 No wax 14 day scaled tube Reactivity. "/1 R22 formed 0. 6.0 1.0 0.6 0. 0

TABLE 3-continued Typical Properties of Refrigeration Oils Oil Oil Oll Oil* Oil Oil A B C D E F Boiling Range Initial BP "F 580 740 640 580 580 800 50% F 720 840 810 770 750 920 Final BP "F 840 950 950 990 l 150 I200 blend of high boiling paraffinic oil & naphthcnic oil according to present invention The invention claimed is:

l. A composition, useful as a refrigeration oil containing no more than 10 ppm basic nitrogen, comprising a blend of from 50-75 volume percent of a hydrorefined naphthenic oil component and from 50-25% of a hydrogenated C C. polybutene oil, said blend con taining in the range of 15-35 weight aromatic hydrocarbons and having an SUS viscosity at F. in the range of l25300 SUS and a natural floc point no higher than 35F. in CCl F refrigerant.

2. The composition of claim 1 and having a basic nitrogen content no greater than 5 ppm and wherein said blend has an SUS viscosity at 100F in the range of 210-240, an SUS viscosity at 210F in the range of 45-52, a flash point of at least 340F, a pour point no higher than 25F, a D1500 color no greater than 2.0, an API gravity in the range of 2527, a total acid number no greater than 0.05 mg KOl-l/g, and a minimum dielectric strength of 25 KV, and wherein no greater than 1.0 weight percent CHClF 'is determined after 14 days of scaled tube stability testing at 347F with CCI F refrigerant in the presence of a steel strip.

3. The composition of claim 1 wherein no greater than 2.5 Wt. CHCIF is determined after l4-days of sealed tube stability testing at 347F. with CCI F refrigerant in the presence of a steel strip.

4. The composition of claim 1 and having better thermal stability than a naphthenic refrigeration oil of about the same viscosity at 100F.

5. The composition of claim 4 and which, compared with said naphthenic refrigeration oil, produces significantly less coke deposit on throttle valves when used for about 2,000 hours in a compressor which is operated at an average coil temperature in the range of l40l70F.

6. The composition of claim 5 wherein the basic nitrogen content is less than 5 ppm and wherein no greater than 1.0 Wt of CHCIF is determined after l4-days of sealed tube stability testing at 347F. with CCI F refrigerant in the presence of a steel strip.

7. A composition according to claim 1 and wherein said polybutene has a bromine number less than 5;

8. A composition according to claim 1 and having a viscosity at 100F. in the range of -250 SUS.

Claims (8)

1. A COMPOSITION, USEFUL AS A REFRIGERATION OIL CONTAINING NO MORE THAN 10 PPM BASIC NITROGEN, COMPRISING A BLEND OF FROM 50-75 VOLUME PERCENT OF A HYDROREFINED NAPHTHENIC OIL COMPONENT AND FROM 50-25% OF A HYDROGENATED C-16-C40 POLYBUTENE OIL, SAID BLEND CONTAINING IN THE RANGE OF 15-35 WEIGHT % AROMATIC HYDROCARBONS AND HAVING AN SUS VISCOSITY AT 100*F. IN THE RANGE OF 125-300 SUS AND A NATURAL FLOC POINT NO HIGHER THAN -35*F. IN CC12F2 REFRIGERANT.

2. The composition of claim 1 and having a basic nitrogen content no greater than 5 ppm and wherein said blend has an SUS viscosity at 100*F in the range of 210-240, an SUS viscosity at 210*F in the range of 45-52, a flash point of at least 340*F, a pour point no higher than -25*F, a D1500 color no greater than 2.0, an API gravity in the range of 25-27, a total acid number no greater than 0.05 mg KOH/g, and a minimum dielectric strength of 25 KV, and wherein no greater than 1.0 weight percent CHClF2 is determined after 14 days of sealed tube stability testing at 347*F with CCl2F2 refrigerant in the presence of a steel strip.

3. The composition of claim 1 wherein no greater than 2.5 Wt. CHClF2 is determined after 14-days of sealed tube stability testing at 347*F. with CCl2F2 refrigerant in the presence of a steel strip.

4. The composition of claim 1 and having better thermal stability than a naphthenic refrigeration oil of about the same viscosity at 100*F.

5. The composition of claim 4 and which, compared with said naphthenic refrigeration oil, produces significantly less coke deposit on throttle valves when used for about 2,000 hours in a compressor which is operated at an average coil temperature in the range of 140*-170*F.

6. The composition of claim 5 wherein the basic nitrogen content is less than 5 ppm and wherein no greater than 1.0 Wt % of CHClF2 is determined after 14-days of sealed tube stability testing at 347*F. with CCl2F2 refrigerant in the presence of a steel strip.

7. A composition according to claim 1 and wherein said polybutene has a bromine number less than 5.

8. A composition according to claim 1 and having a viscosity at 100*F. in the range of 150-250 SUS.