A Reliable and Practical Accelerated Test Method for Predicting the

Aug 10, 1994 - Navy Technology Center for Safety andSurvivability, U.S. Naval Research Laboratory, ... long term storage stabilities of aviation turbi...
3 downloads 0 Views 582KB Size
Energy & Fuels 1995,9, 183-187

183

A Reliable and Practical Accelerated Test Method for Predicting the LongTerm Storage Stabilities of Aviation Turbine Fuels Based on Hydroperoxide Formation Seetar G. Pande and Bruce H. Black Go-Centers, Inc., Ft. Washington, Maryland 20744

Dennis R. Hardy* Navy Technology Center for Safety and Survivability, U.S. Naval Research Laboratory, Code 6181,Washington, D.C. 20375-5342 Received August 10, 1994. Revised Manuscript Received October 24, 1994@

A reliable method is needed for predicting the long term storage stabilities of the military aviation turbine fuel reserves as well as for evaluating the effectiveness of antioxidants approved for these fuels. This paper addresses the successful development of such a method. The proposed test method entails measuring the hydroperoxide concentration formed after stressing at 100 "C and 345 kPa air overpressure for 24 h with the option to extend to 48 h should this be necessary. Application of the proposed test method to worldwide current production fuels attests to the sensitivity of the method in differentiating fuels with differing storage stabilities. Introduction Like other hydrocarbon-based fuels, aviation turbine fuels are susceptible to autoxidation during long term storage. The reaction products of primary concern in the autoxidation of aviation turbine fuels are hydroperoxides: these products have been found to be detrimental t o the elastomers in aircraft fuel systems resulting in fuel pump f a i l ~ r e s . l - ~The formation of other autoxidation products during storage, such as soluble gums and filterable insolubles, occurs to a small extent with current aviation turbine fuels and does not pose a storage problem as occurs with diesel fuels. This significant difference in the storage stability problems between diesel fuels and aviation turbine fuels defined the need for separate methodshest conditions for predicting their long term storage stabilities. Thus, whereas an accelerated method for assessing the storage stability of distillate fuels via gravimetry was recently developed,5 the same could not be used for aviation turbine fuels. To meet the US Navy's need for ensuring the storage stability of their strategic reserves of fuels, a reliable and practical accelerated test method for predicting the long term storage stabilities of aviation turbine fuels Abstract published in Advance ACS Abstracts, December 1,1994. (1)Smith, M. Aviation Fuels; Foulis, G . T., & Co, Ltd Henley-onThames, England, 1970; Chapter 51. (2) Hazlett, R. N.; Hall, J. M.; Nowack, C. J.; Craig, L. Hydroperoxide Formation in Jet Fuels. In Proceedings ofthe Conference on Long Term Storage Stabilities of Liquid Fuels, Tel Aviv, Israel, Dec 1983; The Israel Institute of Petroleum and Energy: Tel Aviv, Israel, 1983; pp B132-B148. (3)Determination of the Hydroperoxide Potential of Jet Fuels; CRC Report No. 559, Coordinating Research Council, Atlanta, GA, April 1988. (4) Shertzer, R. H. Aircraft Systems Fleet SupporVOrganic Peroxides in JP-5 Investigation. Final Report NAPC-LR-78-20, Naval Air Propulsion Center, Trenton, NJ, 27 September 1978. (5) Standard Test Method for Assessing Distillate Fuel Storage Stability by Oxygen Overpressure, ASTM D 5304-92. American Society for Testing and Materials, Annual Book of ASTM Standards, Philadelphia, PA, 1992. (a)Ibid. footnote 5. @

was developed. The ability to predict the peroxidation potential of aviation turbine fuels in as short a time as possible is particularly important from (a) the procurement aspect, especially in the case of future fuels;2 and (b) in the case of fuel reserves on an ongoing basis, since antioxidants are depleted during long term storage. In addition, a method is needed for evaluating antioxidants specified for aviation turbine fuels (MIL-T-5624N specification) based on their effectiveness. The proposed test method entails measuring the hydroperoxide concentration formed after stressing a t 100 "C and 345 kPa air overpressure for 24 h with the option to extend to 48 h should this be necessary. The maximum peroxide number allowed by the military specification, MIL-T-5624N, is 8.0 ppm (lmequivkg of sample). The 100 "CY24 h stress test conditions simulate approximately 9 months storage a t ambient conditions and meets the Navy's protocol for periodic testing on a 6 monthly basis. The proposed method, which culminated through the joint efforts of three laboratories6involved establishing a stress temperature/duration and overpressure of air/ oxygen that would be reliable and practical in predicting long-term storage at ambient conditions. The reliability in using an accelerated stress test, Le., elevated temperatures including 100 "C, to simulate long-term ambient storage has been verified by Fodor et al.'v8 and confirmed by Hardy and Black.g The overpressure of aidoxygen was employed to obviate the depletion of oxygen that occurs at elevated temperatures. The need for adequate oxygenation of fuels on accelerated aging (6) Navy Position Paper. Storage Stability Assurance of US Navy Aircraft Strategic Fuel Reserves. Prepared at the Navy Technology Center for Safety and Survivability, Naval Research Laboratory, Washington DC, July 1994. (7) Fodor, G. E.; Naegeli, D. W.; Kohl, K. B. Peroxide Formation in Jet Fuels. Energy Fuels 1988,2, 729-734. (8) Fodor, G. E.; Naegeli, D. W.; Kohl, K. B. Peroxide Formation in Jet Fuels. Prepr. Pap.-&. Chem. SOC.,Diu. Fuel Chem. 1990,35(4), 1267-1276.

0887-0624/95/2509-0183$09.00/0 0 1995 American Chemical Society

P a n d e et al.

184 Energy & Fuels, Vol. 9, No. 1, 1995

fuel no. NO.90-22 NO.90-26 NO.91-7

NO.90-23 NO.91-4 NO.91-33

Table 1. JP-5 Fuels Used in the Overpressure Studies at NRL description Blending stock doped with 22.6 mg/L 2,g-di-tert butyl 4-methylphenol WWS:" hydrocracked, hydrotreated; additives: antiicing inhibitor antioxidant corrosion inhibitor JP-5 quality Jet A,b but without additives; the refiner reported this fuel to be representative of a fuel batch produced by hydrotreatment of straight run kerosene distilled from a refinery slate consisting of 76% sour Arabian crudes and 24% mixed sweet domestic crudes WWS:" hydrofined (1400 psig); additives: antioxidant corrosion inhibitor additive free hydrotreated; additive free

+

+

+

"Worldwide survey fuel: additional information on these fuels is given in Table 3. bMet all the specifications of a JP-5, with the exception of additives.

has been demonstrated by Fodor et aL7 as well as by Black et al.lOJ1For example, Black et aZ.l0J1found that the hydroperoxide concentrations formed in capped versus vented bottle tests for five fuels stressed a t 65 "C112 weeks, and at atmospheric pressure were significantly lower (27-74-fold) in the capped versus the vented bottles for three of the test fuels. In addition, for bottle storage tests at 43 "C, Fodor et aZ.12found that oxygen depleted in the liquid phase is apparently replenished by oxygen in the vapor phase. The above results focus on the importance of adequate oxygenation in accelerated stress tests, particularly for those fuels that are susceptible to peroxidation. Furthermore, spurious implications can ensue as a consequence of oxygen depletion in capped-bottle tests.2 For example, the results obtained in the extended Coordinating Research Council study, CRC 1,2are likely due to oxygen depletion1' and not to a change in mechanism at higher temperatures as was originally proposed.2 Because of the importance of adequate oxygenation in accelerated stress tests, it was necessary to determine the magnitude of airloxygen overpressure that would be ample from the peroxidation aspect, in reliably simulating aging at ambient conditions, as well as from the safety considerations. The purpose of this paper is twofold: (1)investigate the effect of airloxygen overpressure with respect to establishing the overpressure that should be employed in the test method being developed; and (2) demonstrate the usefulness of the proposed test method in differentiating the peroxide potential of current production fuels and, hence, in predicting the long-term storage stability of aviation turbine fuels.

(b) refinery processing (e.g., hydrocracking, hydrotreating); and (c) formulation (e.g., derived from blending stocks (No. 9022),and usdnonuse of additives including antioxidants). Such differences are well-known to impact on storage stability. A brief description of t h e test fuels is given in Table 1. (b) Four overpressures: 103,345,a n d 690 kPa air, and 241 k P a p u r e oxygen. The three a i r overpressure values may also be expressed as 21,69,and 138 kPa, oxygen partial pressures, respectively. (c) Stress conditions: 100 "C over a 96 h stress period with hydroperoxide analyses at 24 h intervals. The 100 "C stress temperature was selected over 120 "C for practical reasons, which include t h e following: a steam b a t h could be used, should an oven not be available; for safety considerations, e.g., pertinent to fuels with lower flash points; and, it is an easier temperature to calibrate. Method. The basic method is similar to that described in ASTM D53045except for t h e stress conditions employed-it is these critical factors that must be established. Briefly, the test fuel (100mL) is contained in an amber borosilicate 125 m L bottle, t h e opening of which is covered with perforated aluminum foil. The fuel is stressed in an appropriate pressure vessel at 100 "C for a stress duration of 96 h a n d at the test overpressure of air or oxygen being examined. Hydroperoxide analyses were performed at 24 h intervals, in duplicate, using t h e potentiometric method described by Morris et aZ.13 A typical pressure vessel for conducting t h e accelerated stress tests has been described in previous stability studies.14 This particular reactor, termed a low pressure reactor (LPR), was found15 to effectively analyze reaction rates in a more controlled environment than previous bottle tests (ASTM D4625) a n d was more cost effective than t h e reactor used by Fodor et aL7s8J2 The LPR has t h e capacity to hold u p to 22 (125mL) bottles a n d is commercially available from a number of vendors.5a

Results Experimental Section Test Matrix. The test matrix comprised t h e following: (a) Six JP-5 fuels of diverse origins, which include differences in (a)crudes (i.e., were obtained from different refineries) (9) Hardy, D. R.; Black, B. H. Navy Aircraft Mobility Fuels R&D Program: Status of NRL Contributions for Second, Third, and Fourth Quarter of FY92: Ser 6180/644.2;Naval Research Laboratory, Washington, DC, October 1992. (10)Black, B. H.; Hardy, D. R.; Beal, E. J. Accelerated Peroxide Formation in Jet Fuel Using Conventional and Oxygen Overpressure Methods. Prepr. Pap.-Am. Chem. SOC.,Diu. Fuel Chem. 1990, 35(4), 1277-1285. (11)Black, B. H.; Hardy, D. R.; Beal, E. J. Accelerated Peroxide Formation in Jet Fuel at 65 'C in Capped and Vented Bottles. Energy Fuels 1991,2, 281-282. (12)Fodor, G. E.; Naegeli, D. W. Development of a Method to Determine the Autoxidation of Turbine Fuels. Final Report BFLRF No. 280 (AD A260578), prepared by Belvoir Fuels and Lubricant Research Facility, Southwest Research Institute, San Antonio, TX, June 1987.

As shown in Table 2, which ranks the test fuels in an apparent increasing order of storage stability, the effect of airloxygen overpressure on hydroperoxide formation appears to be dependent both on the fuel and its stress duration. For example, the two fuels that peroxidized highly with increasing stress duration (No. 90-22 and No. 90-26 a t >48 h) appear to exhibit an increase in hydroperoxide formation with an increase in air over(13)Morris, R. E.; Black, B. H.; Mitchell, C. s.An Improved Method for Measuring Hydroperoxides in Fuel Using Potentiometry and Alternatives to Chlorofluorocarbon Solvents. Fuel Sci. Technol. Int. 1994,12 (7&8),1003-1018. (14)Hardy, D. R.; Hazlett, R. N.; Beal, E. J.;Burnett, J. C. Assessing Distillate Fuel Storage Stabilityby Oxygen Overpressure. Energy Fuels 1989,3, 20-24. (15)Turner, L. M.; Nowack, C. J. Comparison of Accelerated Predictive Techniques to Measure the Oxidative Tendencies of Aviation Fuels. Proc. 4th Int. Conf. Stability Handling Liq. Fuels, Orlando, FL, NOU19-22, 1991 1991, 231.

Long-Term Storage Stabilities of Aviation Fuels Table 2. Effect of Air/Oxygen Overpressures on Fuel Peroxidation at 100 "C with Time

fuelno. No. 90-22

No. 91-7

No. 90-23

No. 90-26

No. 91-4

No. 91-33

calcb equiv stress storage timea time (h) (approx) 24 9months 48 1.5yrs 72 2.2 yrs 96 3 yi-s 24 9months 48 1.5yrs 72 2.2yrs 96 3 yrs 24 9months 48 1.5yrs 72 2.2yrs 96 3yrs 24 9months 48 1.5yrs 72 2.2~1-s 96 3yrs 24 9 months 48 1.5yi-s 72 2.2~1-s 96 3yrs 24 9months 48 1.5~1-s 72 2.2~1-s 96 3yrs

hydroperoxide concentration in ppm at specified overpressures: airloxygen air Pure 103 345 690 oxygen kPa kPa kPa 241kPa 1.27 2.91 4.02 4.20 4.44 10.02 10.67 14.76 10.88 28.93 44.97 76.41 23.55 204.81 575.01 1387.8 2.34 5.13 5.39 4.27 8.35 12.47 14.91 11.23 20.14 19.72 24.28 21.50 30.89 34.34 36.38 41.47 4.96 4.95 5.71 5.84 9.11 11.64 13.32 14.62 18.28 19.52 22.58 21.20 27.79 30.29 38.66 37.35 1.18 1.41 1.85 2.24 3.81 4.28 5.21 8.26 5.10 8.51 16.90 18.96 10.33 16.35 47.97 220.0 0.66 1.78 1.75 2.65 3.71 4.17 6.38 5.69 5.40 6.69 8.06 8.04 7.68 7.99 11.06 10.80 1.49 1.33 3.19 2.20 0.83 0.92 3.17 2.34 2.38 0.53 3.18 1.84 1.84 0.70 2.87 1.80

a At time zero, the hydroperoxide concentration was below the level of detection in all cases. Calculation is based on a n approximate doubling of the reaction rate for every 10 "Cincrease in temperature as derived from Arrhenius equation. Thus, at 100 "C, the rate is approximately 256 times that at 20 "C.

pressure. In contrast, for the remaining test fuels, with the exception of the most stable fuel (No. 91-33), hydroperoxide formation rate appears to increase more so, due to the increase in stress duration than to the increase in overpressure. For the most stable fuel, No. 91-33, the same increases in both overpressure and stress duration had little effect-its hydroperoxide formation was significantly low (1-3 ppm). Repeatability. In general, hydroperoxide determinations were performed in duplicate. The difference between these two measurements was usually less than f 0.3%. However, if the difference between duplicate analyses exceeded f1.5%, a third determination was performed. The accelerated stress tests were single determinations. Nevertheless, for fuels that showed no significant change in hydroperoxide concentration with increasing oxygen concentration for the same stress duration up to 96 h, the repeatability at the four levels of overpressure examined was approximately f8%. Such fuels include all fuels in Table 2 except No. 9022 and No.90-26.

Discussion Effect of Air/Oxygen Overpressure. In cases where an overpressure effect appears to be operative, viz., the two fuels that peroxidized highly (No. 90-22 and No. 90-26) at >48 h, the significant increase in hydroperoxide formation at 241 W a oxygen versus 690 kPa air suggests that the apparent increase is attributable to the oxygen concentration rather than to the

Energy & Fuels, Vol. 9, No. 1, 1995 186

overpressure magnitude. In experiments on the accelerated aging of middle distillate fuel, which were designed specifically to differentiate between the effects of pressure and oxygen concentration, Hardy et al.14 found concentration to be the operative factor. The increase in insolubles was ascribed to an equilibrium shift and not to an increase in reaction rate. Fodor et ~ 1 also . found ~ that the partial pressure of oxygen had no effect on the rate of peroxide formation for model jet fuels stressed a t 80 "C/790 and 1140 kPa of oxygen and 100 "C/240 and 790 kPa of oxygen. It is important to point out that because of the high oxygen concentration levels employed by Fodor et u Z . , ~ oxygen is likely not a limiting reagent. Consequently, an oxygen concentration effect was not operative. However, the earlier literature16 states that a t oxygen pressures above 50-100 mm (6.56-13.1 E a ) autoxidation is independent of oxygen pressure even though the actual oxygen concentration is still very low. In those studies, the systems examined were pure compounds. In this study, the significant differences in peroxidation for similar increasing levels of aidoxygen concentration focus on a fuel dependence factor. These results emphasize the differences in composition among fuels and the different system examined in the earlier literature16 (a single component system versus fuel-a complex system). Nevertheless, for all six fuels examined at a maximum stress duration of 48 h, the effect of pressure on hydroperoxide formation does not appear to be significantly different a t 345 and 690 kPa air overpressures. Also, for the same stress duration (224 h), a t 103 kPa air overpressure (21 kPa partial pressure oxygen), hydroperoxide values of the fuels appear to be generally lower than at higher levels of oxygen (241 W a overpressure). The scatter in the data for fuel No. 91-33 reflects hydroperoxide concentration to be the net result of the formation and destruction of a relatively small concentration of hydroperoxides. Thus, the hydroperoxide concentration for the very stable fuel No. 91-33 can be best considered as basically unchanged with time and oxygen concentration. Furthermore, from a criterion viewpoint, the values are below the maximum specification limit of 8.0 ppm and, therefore, will not pose a problem in passing such a fuel. The overall results suggest that (a) 103 kPa air may not be adequate and (b) 345 kPa air overpressure has a similar effect to 690 kPa. Thus, for safety reasons, 345 kPa air overpressure was selected to obviate the depletion of oxygen due to accelerated rates of peroxidation a t higher temperatures. The overall results of temperature, pressure, and stress duration indicate that test conditions of 100 "C a t 345 kPa air overpressure for 24 h would be adequate for realistically predicting aviation turbine fuel storage stability at ambient conditions for approximately 9 months. The pressure studies also indicate that the stress test could be extended t o 48 h (1.5years ambient storage) should the results of the 24 h test deem this necessary, e.g., marginally stable fuels. In addition, the data shown in Table 2 focus on the potential instability of certain fuels, on long term storage, and hence the need for periodic assurance testing. (16)Walling, C. Free Radicals in Solution; J. Wiley & Sons, Inc: New York, 1957;p 421.

Pande et al.

186 Energy &Fuels, Vol. 9, No. 1, 1995 fuel no. NO. 90-24 NO.90-27 NO.90-14 NO.90-23 NO.90-7 NO.90-26 NO.90-5 NO.90-13 NO.90-16 NO.90-6 NO.90-12 NO.90-11 NO.90-9 NO.90-8

Table 3. Description of Worldwide Survey Fuels sourceltype of crude and treatment Alaskan North slope (85%) San Joaquin Valley (15%);HF AO: Du Pont AO-37; CI: IPC 4445 FSII: 0 . 1 8 % ~ DIEGME; AO: AITEC 4733; CI: NALCO 5403 Kuwait Export HC; HT FSII: 0.18%~DIEGME; AO: STIA 26/24; CI: NALCO 5403 Saudi Arabian (LighWAlgerian Mild HT same as No. 90-24 same as No. 90-24 Alaskan North slope; NGL Injection, secondary and tertiary AO: IPC 4650; CI: IPC 4445 recoveryic HC; SD, and hydrodesulfurization same as No. 90-27 same as No. 90-27 FSII:0.15-0.2%~DIEGME; AO: Nalco 5275; CI: Nalco 5403 West Texas Intermediate; sweet 0.2-0.4 w t Ti sulfur; SD and Merox treatment same as No. 90-14 same as No. 90-14 West Texas Sour 41% Prudhoe Bay 43% South Texas AO: Ethyl 733 (80%active) Mix 13% Can0 Limon 3%; SD; mild HT same as No. 90-5 same as No. 90-5 Venezuelan Furrial, Colombian Can0 Limon, and Mexican FSII: 0.16%~methacarbatol; AO: Ethyl 733; CI: DCI4A Olmeca; caustic wash Venezuelan Lagomedio, Mesa Ecuadorian Oriente, same as No. 90-12 Alaskan North slope HawkinsKabinddSouth Louisiana; SD AO: name not given same as No. 90-9 same as No. 90-9 additives a

+

+

+

+

aAO, antioxidant; CI, corrosion inhibitor; FSII: fuel system icing inhibitor. 'Refinery process techniques: HC, hydrocracking; HF, hydrofined; HT, hydrotreatment; SD, straight distillation. "rude production techniques used other than normal pumping or natural pressure.

Use of the Accelerated Test Method To Determine the Long-TermStorage Stabilities of Worldwide Fuels. The proposed accelerated test method (100 "C/345 kPa air for 24 h) was used to determine the long term storage stabilities of a total of 14 worldwide finished JP-5 fuels. The fuels employed in this study were field samples and were obtained from Navy depots. A description of the additives' content and the source/ type of crude, are given in Table 3. The accelerated stress duration of 24 h was extended to 96 h with hydroperoxide analyses at 24 h intervals. The results, given in Table 4, list the fuels generally in an increasing order of storage stability. The order is based primarily on the 24 h test period and secondarily on the 48 h and longer stress duration. The results of the 24 h test period appear to be adequate in differentiating most of the fuels. Three levels of storage stability are identified, viz., low, moderate, and high stabilities (see Table 4). As described in Table 4, this categorization is based on the stress time taken to exceed the 8.0 ppm maximum hydroperoxide concentration allowed in the military specification, MIL-T-5624N. The apparent exponential increases in hydroperoxides for No. 90-27 at 72 h and above, and for No. 90-26 a t 96 h, confirm the need for periodic testing, particularly so, for fuels of marginal-poor storage stability ('8.0 ppm hydroperoxides). The overall results also indicate that fuels of high storage stability appear not to deteriorate or deteriorate extremely slowly with aging. For example, after 96 h stress duration, which is equivalent to approximately 3 years ambient storage, four of the eight high storage stability fuels continue t o meet the 8.0 ppm hydroperoxides specification; the remaining four fuels exhibited no change in their initial low hydroperoxide content ( e1.5 ppm). Again, the scatter in the hydroperoxide concentration with increase in stress duration for many of the fuels that exhibited high stability may also be explained as likely due to the net result of formation and destruction of a relatively small

Table 4. Application of the Accelerated Test Method to Worldwide Fuels (Test Conditions: 100 "C at 345 Wa Air Overpressure) hydroperoxide concn, ppm at fuel no. low stability;" NO. 90-24b NO. 90-27' moderate Stability NO.90-14' NO. 90-23 NO. 90-7 NO.90-26 high stabilitf NO.90-5b NO. 90-13 NO.90-16 NO.90-6' NO.90-12' NO. 90-11' NO.90-gb NO. 90-8'

24h

48h

72h

96 h

15.1 8.1

23.7 29.2

42.4 523.0

40.6 1437.2

7.2 4.8 4.0 1.8

15.5 10.0

39.3 17.9 11.3 16.6

33.6 28.9 14.6 241.3

3.6 2.1 2.7 1.9 1.5 1.4 0.7 0.4

2.3 5.0 3.8 2.7

4.4

5.3 6.4 5.0 4.2 0.9

8.0

4.3

1.3 1.0

0.3 0.9

3.1

4.8 7.4 1.0 1.3 1.a 0.5

1.0

0.9 0.6

Exceeded specification limit of 8.0 ppm maximum hydroperoxide concentration at 24 h stress duration. Accelerated storage stability test performed by Naval Air Warfare Center. Exceeded 8.0 ppm specification limit at 72 h or greater. Did not exceed 8.0 ppm specification limit at 24-96 h stress duration.

hydroperoxide concentration, as mentioned earlier for another stable fuel, No. 91-33 (Table 2). The above results attest to the sensitivity of the proposed accelerated test conditions in identifying aviation turbine fuels of differing storage stabilities. The reliability of the test conditions employed for simulating long-term storage at ambient conditions has been addressed (see Introduction for reports on the validation of elevated temperatures, and Discussion for the air overpressure selected). Furthermore, the method is considered practical based on the relatively short stress duration period: 24 h a t 100 "C versus 3 weeks at 65 "C as was recommended in the last CRC r e p ~ r t .Also, ~ the method uses a pressure vessel that is commercially available. Thus, the proposed test method offers a

Energy & Fuels, Vol. 9, No. 1, 1995 187

Long-Term Storage Stabilities of Aviation Fuels

reliable and practical means of predicting the long term storage stabilities of fuels. Applicability of the test method should be considered to include all commercial turbine fuels with and without additives, should the need arise for such determinations. In addition, the method offers a much needed means of evaluating antioxidants specified for aviation turbine fuels (MILT-5624N) based on their effectiveness.

Conclusions A rapid and practical method was developed for reliably predicting the long term storage stabilities of aviation turbine fuels. Usefulness of the method extends to the procurement of fuels (e.g., for strategic reserves and future fuels) and to the qualification of antioxidants for military specification, MIL-T-5624N. The method is based on the collective results of extensive rigorous studies performed by three laboratories.

The proposed accelerated method entails conducting the stress test in a pressure vessel at 100 "C,345 kPa overpressure air, for 24 h with the option t o extend to 48 h if deemed necessary. The method is considered reliable for it employs: a stress temperature that earlier studies validate to be predictive of long-term ambient storage for aviation turbine fuels; and an overpressure of air that is ample in preventing oxygen depletion at this elevated temperature. Application of the test method to worldwide current production fuels indicates it to be adequate in differentiating fuels with differing storage stabilities. Studies that extended the stress test duration to 96 h confirm the Navy's current protocol for periodic testing on a 6 monthly basis.

Acknowledgment. The authors thank the Naval Air Warfare Center for permission to present their data included in Table 4. EF940160U