APTD-156- OPTIMUM WORKING FLUIDS FOR AUTOMOTIVE RANKINE ENGINES VOLUME I - EXECUTIVE SUMMARY U.S. ENVIRONMENTAL PROTECTION AGENCY Office of Air and Water Programs Office of Mobile Source Air Pollution Control Advanced Automotive Power Systems Development Division Ann Arbor, Michigan 48105 ------- APTD-1563 Prepared By D. R. Miller, H. R. Null, Q. E. Thompson Monsanto Research Corporation 800 North Lindbergh Blvd. St. Louis, Missouri 63166 Contract No. 68-04-0030 EPA Project Officer: K. F. Barber Prepared For U.S. ENVIRONMENTAL PROTECTION AGENCY Office of Air and Water Programs Office of Mobile Source Air Pollution Control Advanced Automotive Power Systems Development Division Ann Arbor, Michigan 48105 June 1973 ------- The APTD (Air Pollution Technical Data) series of reports is issued by the Office of Air Quality Planning and Standards, Office of Air and Water Programs, Environmental Protection Agency, to report technical data of interest to a limited number of readers. Copies of APTD reports are available free of charge to Federal employees, current contractors and grantees, arid non-profit organizations - as supplies permit - from the Air Pollution technical Information Center, Environmental Protection Agency, Research Triangle Park, North Carolina 27711 or may be obtained, for a nominal cost, from the National Technical Information Service, 5285 Port Royal Road, Springfield, Virginia 22151. This report was furnished to the U.S. Environmental Protection Agency by Monsanto Research Corporation in fulfillment of Contract No. 68-04-0030 and has been reviewed and approved for publication by the Environmental Protection Agency. Approval does not signify that the contents necessarily reflect the views and policies of the agency. The material presented in this report may be based on an extrapolation of the "State-of-the-art." Each assumption must be carefully analyzed by the reader to assure that it is acceptable for his purpose. Results and conclusions should be viewed correspondingly. Mention of trade names or commercial products does not constitute endorsement or recommendation for use. Publication No. APTD - 1563 ii ------- ACKNOWLEDGEMENT The authors acknowledge with gratitude the helpful counsel of Messrs. K. Barber, S. Luchter and F. P. Hutchins, all of the Environmental Protection Agency's Advanced Automotive Power Systems Development Division. Many persons within Monsanto Research Corp- oration and Monsanto Company contributed substantially to the research program reported here. These included R. C. Binning, J. A. Conover, L. L. Fellinger, R. L. Green, R. L. Koch, R. J. Larson, G. J. Levinskas, R. A. Luebke, N. F. May, J. T. Miller, L. Parts, A. C. Pauls, P. F. Pellegrin, J. V. Pustinger, W. R. Richard, A. D. Snyder, W. N. Trump, W. M. Underwood, J. A. Webster, E. P. Wheeler and F. J. Winslow. Their manifold inputs are gratefully recognized, as also are those of Professor P. A. S. Smith of the Department of Chemistry, University of Michigan, who served as a consultant. Also recognized are the contributions made by D. B. Wigmore and R. E. Niggemann of Sundstrand Aviation, Division of Sundstrand Corporation, both in the Research Program and in the Engine Optimization Studies performed for this report. iii ------- CONTENTS Volume I. Executive Summary Page Introduction 1 Summary 2 Conclusions 27 Recommendations 30 Volume II. Technical Section Volume III. Technical Section - Appendices Volume IV. Engine Optimization iv ------- INTRODUCTION This is the final report of a research program performed by Monsanto Research Corporation for the Environmental Protection Agency, Advanced Automotive Power Systems Development Division (AAPSDD) under Contract 68-04-0030. The objective of the work was to determine the best working fluids for Rankine cycle automotive power plants. The Rankine Engine has been identified by EPA as a potential alternative to the internal combustion engine for the automotive application in the event the internal combustion engine cannot satisfy the 19.75/76 Federal Emission Standards. Development work has been proceeding on the Rankine engine, as well as other alter- native power plants, under the direction of AAPSDD since initiation of the Clean Air Act Amendments of 1970. In a Rankine cycle power plant, the working fluid, which is repeatedly vaporized, expanded and recondensed, plays a key role in determining the competitiveness of this system. The thermo- dyriamic, chemical and physical properties of the fluid have a strong influence on the power plant efficiency, materials of con- struction, component weight and size, the degree of hazard, if any, and to some extent the power plant cost. Therefore, to make certain that the Rankine engine could be evaluated at its greatest potential, this study program was initiated to search out the best fluids currently available for the automotive Rankine engine appli- cation. To accomplish this objective, specific guidelines were established for the fluid screening process. These guidelines, contained within the contract work statement, covered all significant fluid char- acteristics: availability, compatibility, thermal stability, cycle efficiency, cold start ability, applicability in principal engine types, owner cost, toxicity, flammability and environmental hazard. Quantitative targets set for these characteristics were intended to approximate the upper limit of what could be achieved with presently known materials and current technology. The results indeed reveal that the criteria were well chosen, as only two possible fluids from over 100 candidates survived the screening process, and these did not totally satisfy all of .the stated requirements. In reviewing the results it should be kept in mind that less stringent or different selection criteria could very well have led to the selection of different candidates. The volume of reportable material resulting from this program has dictated the physical subdivision of the detailed technical matter into three separate volumes (II through IV) with this Volume I serving as an executive introduction and summary. ------- SUMMARY A translation of system and working fluid contractual goals into fluid screening criteria was the first step toward meeting con- tract objectives. Contractual goals are listed in Table 1, while the derived screening criteria are given in Table 2. A detailed discussion of the derivation of the criteria is given in Volume II. It may be noted that the screening criteria are mainly of the go/ no-gO) variety useful in the early elimination of unsuitable materials. Additional requirements on cost, hazards, compatibility and system operability were treated on an ad hoc basis involving only a few of the more .promising candidate fluids. On the basis of predicted thermal stability and melting/boiling points, 112 candidate fluid constituents were identified and, where possible, procured. Included in the candidate list (Table 3) for comparison are several organic fluids previously used or suggested as Rankine working fluids. Except for water, all identified candidates are organic chemicals. Although inorganic candidates were sought and a list of 32 possibilities was similarly compiled, none was judged to be compatible with low-cost metals at the 712°F minimum peak fluid temperature needed to meet cycle efficiency requirements. Binary and ternary mixtures of several of the 112 constituent candidates were also identified with the expectation that such mixtures offer wider latitude in matching diverse fluid re- quirements . Identified fluids were next subjected to screening tests under the first five criteria. Criteria 1 through 3 required mainly literature work. A special steel ampoule thermal exposure test was employed to screen fluids for 720°F thermal stability under Criterion 4. Criterion 5, designed to distinguish between fluids requiring "reasonable" and "too-much" regeneration, drew on published thermophysical properties or, when these data were not available, on recognized property prediction procedures to com- pute the I-factors. Candidate screening results under Criteria 1 through. 5 are summarized in Table 3- Here it is seen that the majority of initial candidates failed to survive 200 hours at 712°P in the Criterion 4 steel ampoule test. The few which did are listed in Table 4 with their corresponding capsule survival times, I-factors (at normal boiling points), and Criterion 3 (normal boiling point) results. Brackets in the I-Factor column delineate the fluids satisfactory per se (innermost bracket) from those suitable only as components of a mixed (or complex) working fluid (outermost bracket) and from those satisfactory as neither, all based upon the tolerable regenerator size, Criterion 5- ------- Table 1. PROPULSION SYSTEM AND WORKING FLUID GOALS Propulsion System Characteristic Type Expander Approx. Net2 Full Power Max. Weight Max. Volume Regenerator Size Min. WF6 Condensing Temp. Max. Low Ambient Temp. Min. High Ambient Temp. Max. Starting Time (idle at -20°F. Max. Starting Time (65% power) @ 60°F, Average Annual Usage Life Expectancy Working Fluid Characteristic Presently available? Min. Carnot Cycle Efficiency Min. Ideal Cycle Efficiency Max. Cost5 to Owner, 5 yr. period Compatible with Low Cost Materials? Compatible Lubricant Needed? Health Hazard Fire and Explosion Hazard Environmental Hazard Goal1 1-stage recip. and 1-stage turbine 1453 HP 1600 Ib 35 cu ft Nil or small 220°F -40°F 125°P 25 sec *J5 sec 350 hr 3500 hr Yes Goal1 30% $100 Yes Yes - recip. Maybe - turbine Nil Nil Nil NOTES: 1. Prime sources: 1971 Vehicle Design Goals (EPA) as amplified in consultation with EPA 2. Gross expander shaft less feed pump shaft 3- Sundstrand calculation based on performance requirements of 1971 Vehicle Design Goals (EPA) 4. Prime source: contract work statement 5- Initial factory fill plus materials only for any subsequent refills 6. "Working Fluid" ------- Table. 2. WORKING FLUID SCREENING CRITERIA 1. Fluid component(s) known and available at reasonable cost. 2. Flow point1 below: a) 20°C (68°F) for mixture component b) -29°C (-20°F) for final working fluid 3. Normal boiling point 65 to 120°C (150 to 250°F) 4. Stable in mild steel ampoule test for at least 200 hours at 720°F. 5. I-Factor2 (at normal dew point) between: a) 0.65 arid °° for mixture component b) 0.75 and 1.5 for final working fluid 6. Reference ideal cycle efficiency at least 30% (UAk < ») 7. Equivalent real cycle efficiency - report only - (UAk = 125 Btu/HP-hr-F) 8. Isentropic expansion enthalpy drop no greater than 200 Btu Ib (turbine expander only, equivalent real cycle) 9. Isentropic expansion density ratio no greater than 25 (reciprocating expander only, equivalent real cycle). NOTES: 1. Lowest temperature at which fluid will freely flow (melting po.int in case of pure compound). 2. Parameter related to slope of saturated vapor curve on T-S diagram (1 = vertical) ------- Table 3. SUMMARY OF SCREENING TEST RESULTS, CRITERIA 1 to 5 « Result, Defined Organic Component Criterion No. FSN Name I 1 3 1 5 1 benzotrifluoride 0 0 0 X X 2 o-bromobenzotrifluoride 3 3 m-bromobenzotrifluoride 0 0 X X 4 p-bromobenzotrifluoride 0 0 X X 5 2-bromo-l,4-difluorobenzene 00X6 6 o-bromofluorobenzene 0 0 X X 7 m-bromofluorobenzene 0 0 X X 8 bromopentafluorobenzene 0 0 X X X 9 p-bromofluorobenzene 0 0 X X 10 p-bromophenyl trifluoromethyl ether 0 0 X X 11 o-chlorobenzotrifluoride 0 0 X X 12 p-chlorobenzotrifluoride 0 0 X X 13 m-chlorobenzotrifluoride 0 0 X X X 14 o-chlorofluorobenzene 0 0 X X 0 15 m-chlorofluorobenzene 0 0 X X 0 16 p-chlorofluorobenzene 0 0 X X 0 17 chloropentafluorobenzene 0 0 0 X X 18 1,2-dichlorohexafluorocyclobutane 0 0 X X X 19 2,3-dichlorooctafluorobutane 18 20 1,2-dichlorooctafluorocyclohexene-1 20 21 1,2-dichlorotetrafluorocyclobutene-l 0 0 0 X 22 o-difluorobenzene 00000 23 p-difluorobenzene 00000 24 m-difluorobenzene 00000 25 2,4-difluorotoluene 0 0 0 " 0 X 26 2,5-difluorotoluene 0 0 0 0 X 27 2,6-dimethylfluorobenzene 0 0 X 0 X 28 2,3-dimethylfluorobenzene 0 0 X 0 X 29 3,4-dimethyIfluorobenzene X 30 l,3-di(trifluoromethyl)benzene 0 0 0 0 X 31 l,4-di(trifluoromethyl)benzene 0 0 0 X X 0 means pass X means fail 42 means failed by implication by FSN-42 ------- Table 3. SUMMARY OF SCREENING TEST RESULTS. CRITERIA 1 to 5 (Cont'd.) Defined Organic Component FSN Name 32 o-fluoroanlsole 33 m-fluoroanlsole 34 p-fluoroanisole 35 fluorobenzene 36 p-fluorobenzotrifluoride 37 m-fluorobenzotrifluoride 38 o-fluorobenzotrifluoride, 39 fluoroether No. 3 40 2-fluoro-3-methylpyridine 41 p-fluorophenyl trlfluoromethyl ether 42 2-fluoropyridine 43 o-fluorotoluene 44 p-f luorotoluene 4 5 m-f luorotoluene 46 hexafluorobenzene 47 methylpentafluorobenzene 48 octafluorotoluene 49 pentafluorobenzene 50 pentafluoropropanol-1 51 perfluoro(methylcyclohexane) 52 perfluoroalkane-70 53 perfluorotributylamine 54 perfluoro-n-hexane 55 perfluoro-2-butyltetrahydrofuran 56 perfluoro(dimethylcyclohexanes) 57 phenyltrlfluoromethyl ether 58 phenyltrlfluoromethyl sulflde 59 1,2,4,5-tetrafluorobenzene 60 1,2,3,5-tetrafluorobenzene 61 1,2,3,4-tetrafluorobenzene 62 2,3,5j6-tetrafluorotoluene Result, Criterion No. I 2 1 4 5_ 0 0 X X X 0 0 X X X 0 0 X X 00000 37 0 0 0 X X 0 X 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 X 0 0 0 0 0 0 0 0 0 X 42 X X X X 0 X 0 0 X 0 X X X X X 0 X X 0 X 0 0 0 X X 0 0 X X X 0 0 0 0 0 X 0 0 0 0 0 0 0 OX X 0 0 X X 00000 0000 0 X X 0 means pass X means fail 42 means failed by implication by FSN-42 ------- Table 3. SUMMARY OF SCREENING TEST RESULTS. CRITERIA 1 to 5 (Cont'd.) Defined Organic Component FSN Name 63 trifluoroacetophenone 64 1,2,4-trifluorobenzene 65 1,3>5-trifluorobenzene 66 2,2,2-trifluoroethanol 67 tris(trifluoromethyl)-s-triazine 68 resorcinol bis(trifluoromethyl) ether 69 3-chloropyridine 70 3-methylpyridine 71 pyrazine 72 pyrlmldine 73 toluene 7*1 methanol 75 1-propanol 76 benzene 77 monochlorobenzene 78 pyridine 79 tetrachloroethylene 80 trichloroethylene 81 trichloroethane 82 2,2,3,3-tetrafluoropropanol 83 perfluorosuccinic anhydride 84 perfluoroglutaric anhydride 85 undecafluoropropoxy-2-propanol-l 86 2,4,6-trifluoropyrimidine 87 hexafluoroacetone trihydrate 88 hexamethyldisilazane 89 perfluoroether P-ID (Allied) 90 water 91 acetic acid 92 morpholine 93 ethanol 0 means pass X means fail 2 means failed by implication by FSN-42 Result, Criterion No. 1 0 0 0 0 0 X 0 0 0 0 0 0 0 0 0 0 0 0 0 X 0 X 0 2 0 0 0 0 0 0 0 X X 0 0 0 0 0 0 0 0 0 0 0 3 X 0 0 0 0 X X 0 0 0 0 0 X 0 X 0 0 0 0 4_ X 0 0 X 0 X 0 0 X X 0 X 0 X X 80 X X X 5. 0 0 X 0 0 0 0 0 0 0 0 0 0 0 X 0 Corrosive 0 0 0 0 0 0 0 0 0 0 0 0 X X 0 0 X 0 X 0 0 X X X X X 0 X 0 ------- Table 3. SUMMARY OP SCREENING TEST RESULTS. CRITERIA 1 to 5 (Cont'd.) Defined Organic Component FSN Name 94 neopentanol 95 ethylene glycol monomethyl ether 96 anisole 97 thiazole 98 N-methylmorphollne 99 thlophene 100 1,1,1,2,1,1,1-heptafluoro-2,3,3-tri- chlorobutane 101 perfluoroheptene-1 102 2,3,5,6-tetrafluorophenol 103 pentafluorophenol 101 N-nonafluorobutyloctafluoromorphollne 105 2-methylpyridine 106 1-methylpyrldine 107 pentafluoropyridlne 108 2-chlorothiophene 109 2,6-dimethylpyridine 110 2-methylpyrazine 111 hexamethyldisiloxane 112 trlmethyl borate Result, Criterion No. 0 0 0 0 0 I 5_ 0 X 0 X X 0 0 X X 0 0 0 X X X 0 0 0 X X 000X0 0 0 0 X 0 0 0 X 0 X X 0 X X 0 0 0 X 0 0 X 0 X 00X00 0 0 0 0 X X X 0 0 0 0 X X 0 0 0 0 X 0 0 0 X Defined Mixtures MFSN Components 1 FSN-73 and PSN-78 2 FSN-71 and PSN-90 3 PSN-90 and FSN-93 1 FSN-78 and FSN-91 5 FSN-75, FSN-76 and FSN-90 6 FSN-76, PSN-90 and FSN-93 7 FSN-73 and FSN-75 8 FSN-75 and FSN-76 000 0 0 0 X X 000X0 0 0 0 X X 0 0 0 X 0 0 0 X 0 0 0 X 0 0 0 X 0 means pass X means fail 142 means failed by implication by FSN-12 ------- Table 3. SUMMARY OF SCREENING TEST RESULTS. CRITERIA 1 to 5 (Cont'd.) Defined Mixtures MFSN 9 10 11 12 13- 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 FSN-73 FSN-74 FSN-75 FSN-74, FSN-73 FSN-71 FSN-78 FSN-23 FSN-30 FSN-61 FSN-23 FSN-66 FSN-70 FSN-71 FSN-90 FSN-90 FSN-90 FSN-90 FSN-46 FSN-46 FSN-70, FSN-90, FSN-70, Components and FSN-74 and and FSN-76 FSN-90 FSN-78 and FSN-90 and and and and and and and and and and and and and and and and FSN-76 FSN-90 FSN-90 FSN-78 FSN-78 FSN-78 FSN-96 FSN-90 ("Fluorinol") FSN-90 FSN-90 FSN-105 FSN-106 FSN-109 FSN-110 FSN-107 FSN-49 FSN-90 and FSN-105 FSN-105 and FSN-106 FSN-90 and FSN-106 Result , Criterion 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 2 0 0 0 0 0 X 0 0 0 0 0 0 X 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 X 0 0 0 0 0 0 0 0 X 0 0 0 0 0 0 n X X X X 0 X 0 X X X X 0 X 0 0 0 X 0 0 0 0 0 * No. 5 0 0 0 X X 0 0 0 0 0 0 0 0 0 means pass X means fail 42 means failed by implication by FSN-42 ------- Table 4. THERMALLY STABLE COMPONENTS AND I-PACTORS PSN Name 56 perfluoro(dimethylcyclohexanes) 89 perfluoroether P-ID (Allied) 111 hexamethyldisiloxane 67 trls(trifluoromethyl)-s-trlazlne 51 perfluoro(methylcyclohexane) 30 1,3-di(trifluoromethyl)benzene 28 2,3-dimethylfluorobenzene 27 2,6-dimethylfluorobenzene 48 octafluorotoluene 105 2-methylpyridlne 25 2,4-difluorotoluene 26 2,5-difluorotoluene 73 toluene 46 hexafluorobenzene 1|9 pentaf luorobenzene 59 lf2,1,5-tetrafluorobenzene 60 1,2,3,5-tetrafluorobenzene 107 pentafluoropyridine 109 2,6-dimethylpyridine 64 1,2,4-trifluorobenzene 65 1,3,5-trlfluorobenzene 35 fluorobenzene 22 o-difluorobenzene 23 p-difluorobenzene 24 m-difluorobenzene 70 3-methylpyrldlne 106 4-methylpyridine 78 pyridine 76 benzene 90 water tt 0 = pass, X = fail Stable Hrs . at 720°P >336 >336 312 >312 >336 >336 >288 >336 >336 >288 >336 >336 >336 >336 >336 >336 >336 >336 >288 >336 >336 >336 >336 >336 >336 288 240 >336 >336 NBPt I- Result^ Factor Grit. 3 0.27 0 0.28 X 0.35 0 0.36 0 0.40 0 0.42 0 0.50 X 0.50 X 0.55 0 0.56 X 0.60 0 0.60 0.66 0 0 0.68 0 0.71 o 0.74 0.75 0.75 0.75 0.76 0.76 0.77 0.78 0.79 0.81 0.82 0.82 0.87 0.89 2.81 0 0 0 X 0 0 0 0 0 0 X X 0 0 0 10 ------- Chemicals in the lower part of Table 4, i.e. the primary auto- motive candidates, all fall in several chemical categories with the characteristics shown in Table 5. Table 5- ADVANCED FLUID COMPONENT CLASSIFICATIONS I. Aromatics (benzene, toluene, pyridine and methylpyridines) A. inexpensive (articles of present commerce) B. flammable C. water solubility 1. Benzene, toluene - insoluble 2. Pyridines - soluble in all proportions D. recognized toxicity II. Polyfluorinated aromatics (fluorinated benzene, toluene, pyridine and methylpyridines) A. inherently expensive, several in limited production B. Flammability 1. fluorobenzene, difluorobenzenes, trifluoro- benzenes - flammable in normal air (by flash point test) 2. tetra-, penta-, hexa-fluorobenzenes, penta- fluoropyridine, octafluorotoluene - non- flammable in normal air (by flash point test) C. water insoluble D. toxicity not well established, thought to be of same order as parent aromatics III. Water 11 ------- Of the simple fluids within the innermost I-factor.bracket in Table 4, not all meet the Criterion 2 requirement for final working fluids (flow point below -29°C). Those which do are listed in Table 6 with their melting points (same as flow point) and an indication of flammability in normal air. Table 6. SIMPLE FLUID CONDIDATES Melting Flam. FSN Ft. (°C) Test* 60 1,2,3,5-tetrafluorobenzene -48 0 107 pentafluoropyridine -44 0 64 1,2,4-trifluorobenzene -32 X 35 fluorobenzene -42 X 22 o-difluorobenzene -34 X 24 m-difluorobenzene -78 X 78 pyridine -42 X •flashpoint in normal air; 0 = pass, X = fail When a requirement for non-flammability is added to Criteria 1 to 5, then, the only simple candidates to emerge are: l,2,3»5-tetrafluorobenzene, and pentafluoropyridine While test amounts of these materials are available, their prices are so high as to suggest that even high-volume production would not lower them enough to meet owner cost requirements of $100/5 y supply. This added impetus to a search for mixed (or complex) fluids to take advantage not only of the wider choice of physical parameters so afforded, but also of the possibility of lowering the cost below that of the most expensive component. Water is an especially attractive component of such a mixture since it can be expected to reduce flammability and toxicity as well as cost. Some 31 mixtures were defined at various stages of the search (Table 3). Most of these were eliminated from consideration when they failed the Criterion 4 test, or were found to contain com- ponents which had failed. Two more (MFSN's 1 and 13) are quite flammable and were eliminated on that basis. The remaining group is constituted as in Table 7. 12 ------- Table 7. COMPLEX FLUID CANDIDATES 1. hexafluorobenzene with a. pentafluorobenzene (MFSN-28) b. pentafluoropyridine (MFSN-27) 2. water with a. pyridine (MFSN-15) b. 2-methylpyridine (MFSN-23) c. 3-methylpyridine (MFSN-21) d. 4-methylpyridine (MFSN-24) e. binary combinations of b, c and d (MFSN-29, 30, 3D Each of the three methylpyridines violates either the I-factor or boiling point criterion. Retention of these as candidate components is justified primarily on the grounds that they are the only water-soluble organics (other than pyridine itself) found to have reasonable thermal stability. Predictive computer routines were used to enable cycle evaluations of the advanced candidates of Tables 5, 6 and 7. Their functions are listed in Table 8. These routines were used extensively to analyze the behavior of advanced candidates in the "Reference Ideal Cycle", defined as follows: Reference Ideal Cycle - An ideal Rankine cycle (clockwise rectangle on a pressure-entropy diagram) operating between a top (bulk fluid) temperature of 712°F and a bottom tem- perature of 220°F (saturated liquid at condenser outlet) and with the peak fluid pressure chosen as the lesser of: a. that just preventing expansion penetration into the 2-phase region; b. 1000 psia 13 ------- Table 8. COMPUTER ROUTINES EMPLOYED 1. Estimation of pure compound a. critical constants (P, T, V) b. perfect vapor heat capacity coefficients c. vapor pressures d. liquid densities e. enthalpies of vaporization 2. Calculation of liquid activity coefficients from vapor pressure data on particular complex fluids. 3. Calculation of vapor phase properties by a special general- ization of the Redlich-Kwong equation of state to cover com- plex as well as simple vapor mixtures. 4. Prediction of principle PVT/thermodynamic functions of either simple or complex fluids, using as input, data found in the literature, or generated as in 1 to 3 above, and having as an option (when treating simple fluids) the use of the Hirschfelder, Buehler, McGee and Sutton formulation. 5- Calculation of specified ideal or real Rankine cycles (only with the Hirschfelder et al. formulation). 6. Plotting of temperature-entropy diagrams for any fluid processed in i». ------- Reference ideal cycles calculated for many of the Table 5-7 candidates are compiled in Table 9. The first column is a test of Criterion 6, which requires an ideal cycle efficiency of at least 30% at any level of regenerator effectiveness. All straight organic fluids pass, whereas the pyridine-water mixtures fail. While water itself appears disqualified, it is capable of higher efficiencies by raising its upper operating temperature. Other entries in the table demonstrate compliance with certain interim (non-criterion) goals; namely, 1. condenser pressure from 5 to 50 psia 2. ideal expansion density ratio less than about 40 (to serve in a single expansion reciprocator) 3. ideal expansion enthalpy drop less than about 200 Btu/pound (to serve in a single-stage impulse turbine) 4. I-factor in the approximate range of 0.65 to » com- ponents) or 0.75 to 1.5 (final fluids) The ideal Rankine cycle with a fully effective regenerator is only of limited applicability in evaluating working fluids, especially as concerns cycle efficiency. A 100% effective re- generator is by definition infinite in size. Since the space available for an automotive regenerator is severely restricted in the engine compartment, and since the cycle efficiency is highly sensitive to extent of regeneration (as well as to ex- pander efficiency), a more realistic "Equivalent Real Cycle" was defined as follows: Equivalent Real Cycle - A real Rankine cycle which matches the Reference Ideal Cycle at the two extreme corners (ex- pander and pump inlets), has engine and pump efficiencies of 75% each, has pressure losses in the regenerator, vapor generator and condenser representative of those predicted by system contractors (in January 1972), and has an adiabatic pump and expander. The "size" of the regenerator may be limited by restricting its q/At (log mean) ratio (or equiv- alently its UAk product) to a "reasonable value". Equivalent Real Cycle calculations employing many of the same simple fluids recorded in Table 9 are compiled in Table 10. Com- parison of the entries with the reciprocating/turbine suitability Criteria 8 and 9 reveals all fluoroaromatics and benzene suitable for either engine type and all aromatics but benzene suited only for turbine use. 15 ------- Table 9. SUMMARY OF REFERENCE IDEAL CYCLE CALCULATIONS ON THERMALLY STABLE CANDIDATES Aromatic3 Benzene Toluene Pyridine 3-Methylpyrldine Fluoro Aromatlcs Fluorobenzene m-Dlfluorobenzene p-Dlfluorobenzene 1.2,3,5-Tetrafluoro benzene 1,2,4,5-Tetrafluoro benzene Pentafluorobenzene Octafluorotoluene Pentafluoropyrldine Hexafluorobenzene Water Pyridine/Water Solutions 6056 30$ 60% 30% 60% Pyridine 2-Methylpyridine 3-Methylpyrldine 30% 60% f4-Methylpyridine Ideal Cycle Efflc.E* 31.8 31.7 31.5 32.1 31-9 31.6 31-7 30.6 30.8 30.3 30.2 30.3 31.1 20.5 27.4 28.3 27.9 27-9 27-3 27.7 27.6 28.6 Condenser Pressure (psia) 29.4 12.3 10.5 4.2 25.9 28.5 23.6 28.2 22.8 26.6 11.9 28.2 30.5 17.3 21.3 19.6 22.8 21.6 19-3 15-9 13.9 14.3 Expansion Density Ratio 40.3 145-9 145.1 338.1 42.4 5.8 Expansion Enthalpy Drop (Btu/lb) 81.3 79-3 93-9 97.3 13-0 226.0 I-Pactor at 220°F 0.85 0-75 0.98 0.89 46.6 43-7 54.6 70.1 56.9 59-1 0.65 0.70 0.71 0.76 54.7 15-2 86.9 41.6 40.3 44.8 35.3 29-7 38.2 35.2 0.77 0.68 0.55 0.75 0.54 3.34 36.3 47.1 33.8 49.4 39.1 65-4 40.1 76.4 154.9 125.9 160.8 107.3 167-1 112.3 165-5 114.8 1.66 1.33 1.48 0.89 1.75 1.21 1.73 1.20 * Full regeneration on all simple fluids: full regeneration or regeneration only to the dew line, whichever occurs first, on all solutions. ** Concentrations shown are mol percents. 16 ------- Table 10. SUMMARY OF EQUIVALENT REAL CYCLE CALCULATIONS ON SIMPLE THERMALLY STABLE FLUIDS Aromatics benzene toluene pyrldine 3-methylpyridlne Fluoro Aromatics f luorobenzene m-dif luorobenzene p-d if luorobenzene 1 j 2, 3 > 5- tetraf luorobenzene pentaf luoropyridine hexaf luorobenzene Notes: 1. Regenerator Q/AT(log Real Cycle Effic.( 22.8 22.7 23-7 24.8 22.1 21.9 22.3 21.2 20.9 20.7 mean) Expansion Density %1} Ratio2 15-1 46.4 13-5 83.6 17.0 16.2 19-5 15.7 15.4 15-2 = UAk = 125 Enthalpy Drop,( Btu/lb) 57.4 56.5 67.3 70.2 49-5 40.2 41.8 30.4 27-0 24.8 Btu/HP-hr-°F 2. Isentropic, from engine inlet to exhaust enthalpy 3. Condenser pressures Table 9 and 220 °F I-factors same as in 17 ------- A searching examination of the complex fluid candidates listed in Table 7 was carried out to further limit the number of final can- didates. This included experimental studies of flammability, toxicity and freeze (flow) point, in addition to the development of criteria data discussed above. The first choice was obvious. Of the nonflammable fluids, only the highly fluorinated benzenes stand a reasonable chance of approaching the cost goal in large-scale production. Hexa- fluorobenzene is too high melting to be used alone, but mixtures with tri-, tetra- and pentafluorobenzenes possess the requisite low flow points. Such mixtures actually result from the fluor- ination of benzene, so processing economies could be expected. As a particular model representing this class, the first final Rankine candidate fluid, designated RC-1, was fixed as follows: RC-1 Concentration Constituent Mol Wt Mol '% Wt % pentafluorobenzene 168.1 60.0 57-5 hexafluorobenzene 186.1 40.0 42.5 The only obvious way to avoid using the expensive fluoro compounds and still secure some measure of fire hazard protection is to employ an organic-water solution. Only the pyridines, with their water solubility, permit this approach. Mixtures involving two or more of the three methylpyridines and pyridine itself, all con- templated in this category, offer some advantage in processing economics because of the simultaneous production of all four in practice. In consideration of freeze (flow) point depression, cycle utility, and flash/fire points, it appeared that a mixture of 2-methylpyridine in water was the best binary choice, so as the second final candidate the representative model defined as follows was chosen: RC-2 Concentration Constituent Mol Wt Mol % water 18.0 65.0 2-methylpyridine 93-1 35-0 18 ------- Thorough experimental and computational characterization of the two final candidates followed their definition. Physical property measurements included flow points, vapor pressures, liquid densities3 specific heats, viscosities and thermal conductivities. Selected values appear in Table 11. Computations using these and pure com- ponent literature data then provided such derived properties as vapor viscosities and thermal conductivities, ideal vapor heat capacity coefficients, and liquid activity coefficients. Thermodynamic tables and T-S diagrams were next prepared, and the two principal Rankine cycles were calculated, as summarized in Table 12. It is significant to compare the cycle efficiencies of the two fluids. While RC-2 fails the 30$ ideal cycle efficiency stricture, it actually shows a slightly higher real cycle effi- ciency than does RC-1. The main reason is that the fixed "size" regenerator is more effective in recovering the lesser superheat of the RC-1 fluid. This supports the original use of the I-factor screening criterion. Application-oriented test work was also undertaken to further explore candidate utility. Included were stability-compatibility tests in dynamic pumping loops and steel ampoules, animal acute toxicity experiments, flammability tests, lubrication studies, and manufacturing cost estimations. Highlights are listed in Table 13- Fluid RC-2 was found to be incompatible with SAE 4130 steel boiler tubing at 720°F. Failure of the steel to passivate was evidenced by formation of hydrogen gas and magnetic iron oxide. An attempt to prepassivate the steel with water/steam reduced, but did not stop, the water-iron reaction. An idea of the inhalation toxicity hazards of the two final can- didates was garnered from comparisons of the rat LC50 results with (a) expected inhalation doses in a hypothetical "worst case" Rankine automobile accident involving driver entrapment and fluid release, and (b) known inhalation accidents involving common vol- atile fluids. The inhalation hazards of both candidates are con- sidered comparable to those of gasoline. Products of flame de- composition of each fluid were identified. Attempts to find lubricants with all characteristics desired in the reciprocating engine application were largely unsuccessful. Desired characteristics were (a) stable, compatible, and a liquid at 712°F; (b) incompletely soluble in the working fluid at ambient temperature; and (c) solidification point (in suspension in working fluid) below -20°F. All lubricant candidates known to be stable above 700°F were completely miscible in RC-1 and most were highly soluble in RC-2. A number of highly silylated benzenes were syn- thesized which had minimal solubility in RC-2 and thermal stability comparable to the RC-2. These compounds suffered from inherent expense and too-high solidification points. 19 ------- Table 11. SELECTED PROPERTIES OF FINAL CANDIDATES Final Candidate Property Average molecular weight Flow point, °F Normal boiling point, °F Condensing pressure @ 220°F, psia Heat of vaporization 6 NBPt, Btu/lb I-Factor at 220°F Liquid density at 77°F, g/ml Specific heat at 77°F, Btu/lb RC-1 175.3 172 30 79-1 0.72 1.549 0.290 RC-2 200 21 378 1.38 0.985 0.668 20 ------- Table 12. FINAL CANDIDATE CYCLE CALCULATION SUMMARY Final Candidate: Cycle1 : Cycle Efficiency Per 100 Net Cycle HP Fluid rate, Ib/hr Eng. exhaust, cfm Pump rate, gpm Pump HP Heat flows, kBtu/hr Vapor gen.., kBtu/hr Regen., kBtu/hr Conden. , kBtu/hr Engine, % Ef f ic . 2 Pressure ratio Density ratio Isentropic 3 Exhaust quality, % Enthalpy drop, Btu/lb Pump, % Effic.2 Pressure ratio Regenerator, % Effect. UAK, Btu/HP-hr-°F Log mean temp, diff., °F Notes: 1. "Reference ideal" earlier 2. Input specified 3. From engine inlet RC-1 Ideal Real 30.0 21.0 7470 11500 259 3^0 12.2 18.8 7.34 16.8 - 848 1220 391 692 409 95.0 100 75 36.6 29.8 44.6 37.9 44.6 15.4 176 190 36.6 25-8 100 75 36.6 42.1 95-3 .91.8 12.0 12. 52 RC-2 Ideal Real 28.8 22.2 1760 2520 233 310 3.93 5-66 2.43 5-04 883 1150 64.8 213 629 890 100 75 43.6- 35.4 35 -.6 32.9 35.6 12.1 108 118 149 106 100 75 43.6 50.1 100 98.4 12. 52 32.6 55-6 0 23.0 and "equivalent real" as defined condition to exhaust enthalpy 21 ------- Table 13- PERFORMANCE PARAMETERS OF FINAL CANDIDATES Candidate Stability. Compatibility 1000-hour dynamic loop test Failure mode Toxic Hazards Inhalation 4-hr rat LC50, ppm 30-min rat LC50, ppm Oral, rat, LD50, mg/kg Dermal, rabbit, LD50, mg/kg Irritation score Rabbit skin (8 max.) Rabbit eye (110 max.) Flammability Flash point, modified COC, °F Fire point, modified COC, °F Auto ignition temp., °F Hot compartment ignition rating (benzene3!*!) Cost (to owner, 5 yr supply), est.$ RC-1 Pass RC-2 Fail Decomposition 16,000 40,0002 12,000 >8,000 3 10 8,000* - 8101 2501 5.6' 711 e e e 0 112 130 145 1,060 8 <100 Notes: Experiments conducted with a 40/60 mol % mixture of 2-methylpyridine/water Experiment conducted with a 50/50 mol % mixture of penta-/hexafluorobenzene by Imperial Smelting Corp- oration 22 ------- A low cost synthetic fluid - biphenyl/terphenyl eutectic - was evaluated as a miscible lubricant in friction-wear experiments involving cast iron rubbing against various typical mating materials. This material has promise as a miscible lubricant in either RC-1 or RC-2. No petroleum-based lubricants can be ex- pected to survive the high temperatures demanded in this appli- cation. Fairly detailed economic studies were carried out to project the ultimate cost to the driving public of the RC-1 fluid. Present manufacturing techniques are not suited to high volume low-cost production. A considerable technological investment would be necessary to bring the cost to the level indicated in Table 13- The RC-2 fluid is inherently cheaper and is currently abundant. Ancillary experimental studies were carried out to probe the ultimate thermal stability of "Fluorinol 85" with and without added refrigeration oil. The upper temperature limit of these fluids was placed near 650°F. Another study of a class of semi- organic heat stable fluids called "carboranes" revealed a flammability deficiency. The two final candidate fluids were subjected to a Rankine Cycle Engine Optimization Analysis where the size, weight, and efficiency of an engine for each of these fluids was determined. Fluid RC-1 was used to optimize an engine with a turbine expander and fluid RC-2 was used to optimize an engine with a reciprocating expander. An optimum automotive engine was determined by using a criterion that weighed engine size and weight against engine overall efficien (miles per gallon) over a time-weighted average of five operating load points representing country, suburban, and city driving con- ditions . The Sundstrand Rankine Cycle Design Optimization Program used is a comprehensive analytical design procedure in which detailed design point and off-design calculations are made, and components meeting cycle requirements are calculated. In the design procedure a design point cycle is first calculated, from which engine com- ponents are designed. The performance of the system using these components is then evaluated by means of off-design cycle calcula- tions. The overall system and its performance is then evaluated against an optimizing criterion, the results of which are used to vary the inputs to the design point cycle analysis. A new design point cycle is then calculated, components designed, and off-design performance evaluated. This procedure is repeated until an optimum system, based on the optimizing criterion, is found. 23 ------- The first step in defining an optimum engine is to define the load points for which the engine is to be optimized, and to define the parameters of the optimizing criterion. For the purposes of this study only steady state load points were to be used, and the optimizing criterion was a trade-off between system efficiency and system weight. The optimizing criterion was based on the fact that high-efficiency systems tend to be high-weight systems, and reduced-weight systems result in lower efficiencies. As system weight is directly related to system volume, system size is also being optimized by this criteria. The load points used in the analysis consisted of one design point load and four off-design load conditions. These load points were derived from the Federal driving cycle and EPA's prototype vehicle performance specification. The load points selected were: Speed Power % Time Design Point maximum power acceleration 70 mph 108 HP 0.5 Off-design 1 steady speed 60 mph 32 HP 16.5 Off-design 2 steady speed 30 mph 8.4 HP 33-0 Off-design 3 acceleration 25-1 mph 9-92 HP 28.0 Off-design 4 acceleration l.?4 mph 1.143 HP 22.0 The power is the power output from the transmission to the drive shaft, the engine power required being determined by the mechanical losses, accessory requirements and the cycle calculations. These load points were also used to calculate the fuel load required for a 200 mile range. The optimizing criterion was constructed using miles-per-gallon as representative of engine efficiency, weighed against system weight in pounds. In the optimizing process a weighting factor was varied to place more emphasis on system efficiency or system weight to see how this affected the system configuration. The optimum con- ditions found in each calculation are listed in Tables 14 and 15. " The turbine engine was optimized as a supercritical engine assuming a turbine inlet temperature of 712°F (required for a 200°F con- v denser in order to meet the EPA specifications of 42% carnot cycle efficiency), and a 425 psia turbine inlet pressure. This pressure was selected to keep pump work to a minimum since the fluid prop- erties did not indicate any significant benefit in increased adiabatic head or reduced regenerator size by going to a higher pressure. The turbine tip speed was limited to 1200 ft/sec to allow economical materials, and the condenser assumed was the AiResearch design for the EPA Thermo Electron engine. 24 ------- Table 14. PRINCIPAL RESULTS_OF__ENGINE OPTIMIZATION CALCULATIONS Engine Type Working Fluid Weighting Factor, Effic. (Vs Weight) Overall Miles/Gallon Weights, Pounds Expander and Gear Box Condenser and Pan Regenerator Burner and Vapor Generator Transmission and Drive Train All Other Components and Fuel Total System Volumes, Cubic Feet Expander and Gear Box Condenser and Fan Regenerator Burner and Vapor Generator Transmission Unoccupied Space Total System Turb. RC-1 0.5 173 195 380 355 322 1571 Turb. RC-1 0.97 178 172 234 368 355 319 1626 Recip. RC-2 0.5 13.33 13.81 18.89* 225 142 36 281 355 322 1361 0.23 7-55 4.13 9.58 1.50 15.^1 38.3 0.28 7.71 5-01 8.83 1.50 15.57 38.9 2.25 5-56 0.82 9.79 1.50 13-3 33-2 Based on 3% of stroke clearance volume and square wave operation of valves porting 20$ of piston area 25 ------- Table 15. OPTIMIZED DESIGN POINT PARAMETERS Engine Type Working Fluid Weighting Factor, Ef f Ic . (Vs Weight) Condensing Temperature, °F Regenerator Press. Drop/Press. In, % Regenerator Effectiveness, % Condenser Effectiveness, % Condenser Air Power, kW Expander Speed Tip Speed, ft/sec Pitch Diam. , in. RPM (3 in. stroke) Admission Arc/Total Circum. , % Intake Angle, Degrees Flue Gas Press. Split (Vap. Gen/ Total) % Turb. RC-1 0.5 179 1.27 93 81 5-48 677 4.96 - 7.7-1 - 80.4 Turb. RC-1 0.97 179 1.27 95 81 4.93 677 6.00 - 100. - 89.4 Recip RC-2 0.5 242 1.43 87 80 5.0 1900 54 72.6 26 ------- The optimization analysis for the turbine engine was performed over six values of the optimization weighting factor (which can vary from 0 to 1; the higher the value, the greater the emphasis on high efficiency). The result of this analysis showed that there was no appreciable variation in the weight of the engine or in miles per gallon over the.range of the weighting factor. It was also seen that there was little variation in the design point condenser temperature, which varied from 179°P to 200°F. It appeared from an analysis of the results that this situation was the result of the restriction of condenser size to one fitting the frontal region of a normal engine compartment. The size restriction on the condenser so severely restricts the condenser size for the power level required that essentially the same size condenser results with all values of the optimization weighting factor. As a result, the condenser influence on the rest of the power system is such that there is very little variation in the optimum engine weight and efficiency throughout the range of the optimizing weighting factor. The reciprocating engine was analyzed as a subcritical engine using RC-2. The design assumed an expander inlet temperature of 712°F and a 700 psia inlet pressure for the same reasons as in the turbine engine, and to keep cylinder bore diameter a reasonable^ size for an automotive engine. The reciprocator was a single acting, single expansion engine whose speed was limited to 2000 rpm to keep valve train forces within reasonable limits. The condenser core was that of the AiResearch design developed for the Thermo Electron engine. As can be seen by the miles per gallon, the reciprocating engine shows very encouraging results. However, it must be taken into consideration that the simplified reciprocator model used assumed a three percent of stroke clearance volume and "square wave" valve operation that had flow areas equal to one-fifth the piston area. These rather optimistic assumptions result in low valve losses and high volumetric efficiency which may be difficult to attain in an actual engine. The result of these assumptions is a high efficiency expander which results in the high miles per gallon.figures. It should be borne in mind when comparing these figures to that of the turbine, that the turbine expander model used was a well developed turbine analysis that gave realistic results based on current state-of-the-art technology. CONCLUSIONS 1. No organic or aqueous organic fluids were found to completely satisfy all the requirements established in the contract for Rankine cycle automotive power plant working fluids. 27 ------- Fluid RC-1 comes nearest to satisfying the automotive organic working fluid requirements and goals set by AAPSDD. It meets all of the established performance criteria for both turbine and reciprocating engines, including efficiencies, temperature, pressure and density limitations, and regenerator size. Additionally, it is considered fire-safe and unlikely to cause serious injury on accidental human exposure (subject to verification in extended toxicity testing recommended as a follow on). The fluid is quite stable at temperatures to at least 720°F in low-cost system materials, and it is expected to last for at least the 5-year span of initial vehicle owner- ship. It possesses a distinct advantage over water in not freezing at temperatures down to -4o°F. A low-cost, soluble lubricant is available if needed. The main contractual deficiency of the fluid is its projected cost of no less than $112/over its expected five year life, as compared to a $100 contract limit. To arrive at the $112 cost it was necessary to assume (a) successful advancement of the state of fluor- ination technology beyond that now practiced, and (b) the existence of a 50 million pound per year benzene fluorination plant. Current prices of the higher fluorinated benzenes are many times the projected price. Turning to cycle practicality, the comparatively high flow rate (Table 12) of the RC-1 fluid per net cycle horsepower (related to its higher molecular weight) may be considered a disadvantage despite the lack of a contractual limit. All in all, this fluid is worthy of further development. Fluid RC-2 is a back-up candidate for automotive use, either in a turbine or reciprocating engine. It satisfies all per- formance criteria except the 30% minimum ideal cycle efficiency. Even so, the fluid is capable of high efficiency usage in a real cycle with a compact regenerator. Although the fluid is flammable, it is less so than common fuels. In burning it releases considerably less energy on account of its water con- tent, and its products of combustion are not hazardous to humans. The fluid itself is unlikely to cause serious injury. It is marginally stable at ?20°F in steel. SAE 4130 steel is not a compatible material for boiler tube construction with RC-2. The fluid acts chemically to corrode this steel, forming F6304 and hydrogen. A more suitable steel alloy needs to be found before this fluid can be considered a contender with RC-1. Assuming a satisfactory steel were found, the RC-2 fluid would cost the vehicle owner about $100 in the first five years of vehicle life. This presumes three system refills. Tradi- tional lubricants cannot be used with this fluid above ?00°F, but low cost synthetics offer promise when mutual solubility is permissible. Synthetic silane lubricants may be considered if immiscibility is mandatory, but further development is needed. Fluidity above -40°F is another feature of importance. The prime ingredient of the fluid is commercially available at 28 ------- low cost now, but in lesser quantity than would be commanded by a million cars-per-year market. RC-2 is worthy of further development only in the unlikely event that candidate RC-1 must be abandoned. ^. Many organic fluids possess the state and thermodynamic prop- erties needed to attain engine performance goals. 5. Very few fluids possess the requisite thermochemical stability in steel at 712°F (and higher). Those that do fall in these categories: Aromatics Fluoroaromatics Aromatic tertiary amines Perfluoroaliphatics Water 6. Flammability eliminates the aromatics from contention, and places the aromatic tertiary amines in a contingency category. Since the aromatic tertiary amines (or pyridines) are quite flammable and water soluble, their safe usage in automobiles is contingent upon a suitable reduction in flammability effected by mixing with water. 7. An excessive demand for regenerator space to maintain tol- erable cycle efficiency rules out the perfluoroaliphatics. This is clearly shown by the I-factor criterion and the supporting cycle calculations. 8. Additional restrictions on vapor pressure, melting point (or "flow" point), cost, and toxicity further narrow the choice to mixed fluorobenzenes and aqueous pyridines, from which categories the final candidates RC-1 and 2 were chosen. 9- The search rationale adopted at the outset was basically sound, although some adjustments were necessary. Particu- larly valuable features were: a. an efficient thermal stability screening program b. development and effective use of the I-factor con- cept for thermodynamic screening - including con- firmation via I-factor - efficiency correlations c. maximal use of computation throughout, but especially in these areas: 29 ------- prediction of thermodynamics of non-ideal/ ideal mixtures I-factor and physical property predictions ideal and real cycle calculations d. the concept of progressive advancement of candidates, exposing each to increasingly rigorous examination. 10. As shown by the optimization studies, both the RC-1 and RC-2 final candidate fluids are appropriate fluids for automotive- size engines and are conducive to good efficiencies for automotive application. RECOMMENDATIONS Relative to fluid RC-1, it is recommended that 1. it be considered an acceptable organic fluid for a prototype engine development program. 2. static thermochemical stability tests be performed to learn how far above 720°F the fluid might be driven 3. providing prototype experience is favorable, the fluid be subjected to a carefully planned set of exposure tests in the areas of chronic inhalation chronic dermal contact environmental fate as necessary to qualify the fluid for widespread use 4. encouragement be given to prospective manufacturers to develop the necessary low-cost processes for large scale production. Relative to fluid RC-2 and assuming abandonment of RC-1, it is recommended that: 5. additional work be undertaken to find a suitable steel and/or steel passivation technique to avoid the iron- water reaction at high temperature 6. if appropriate, additional synthesis work be done to finalize an immiscible silane lubricant 30 ------- 7. a fire hazard simulation be performed - preferably in- volving a mock-up burner-vapor generator and a simulated high pressure fluid leak 8. the first three recommendations under RC-1 be followed 9- additional engine optimization studies be carried out using computer models of components which more closely represent that for automotive type advanced state-of-the- art designs. These models would give a better descrip- tion of component weight, size, effectiveness, and efficiency. 10. further engine optimization studies be performed that optimize more parameters than were done for this pro- gram. A more detailed optimizing criteria would also be beneficial in determining an optimum engine. 31 ------- TECHNICAL REPORT DAYA (Please rrcJ IniLn.ctiuns o;i //(..• ri true beJ 1. REPORT NO. APTD - 1563 3. RECIPIENT'S ACCESSION NO. 4. TITLE AND SUBTITLE Optimum Working Fluids for Automotive Rankine Engines :Volume I- Executive Summary 5. REPORT DATE 6. PERFORMING ORGANIZATION CODE 7. AUTHORlS) D.R." Miller, H. R. Null, Q. E Thompson 8. PERFORMING ORGANIZATION REPORT NO. 9. PERFORMING ORGANIZATION NAME AND ADDRESS Monsanto Research Corporation, SOONorth Lindbergh Blvd St. Louis, Missouri 63166 and Sundstrand Aviation Rockford, Illinois 10. PROGRAM ELEMENT NO. 11. CONTRACT/GRANT NO. 68 - 04 - 0030 12. SPONSORING AGENCY NAME AND ADDRESS U. S. ENVIRONMENTAL PROTECTION AGENCY Office of Mobile Source Air Pollution Control Advanced Automotive Power Systems Development Division Ann Arbor, Michigan. *~48105 13. TYPE OF REPORT AND PERIOD COVERED 14. SPONSORING AGENCY CODE 15. SUPPLEMENTARY NOTES 16. ABSTRACT The objective was to determine the best working fluids for Rankine cycle automotive ppwer plants. This study program was initiated to search out the best fluids currently available for the automotive Rankine engine application. Specific guidelines were established for the fluid screening process. These guidelines covered all significant fluid characteristics: availability, compatibility, thermal stability, cycle efficiency cold start ability, applicability in principal engine types, owner cost, toxicity, flammability and environmental hazard. This Volume i serves as an executive introductidjr and summary. The best Rankine candidate fluid, designated RC-1, was fixed as follows.: pentafluorobenzene 60 Mol%, hexaflurorbenzene 40 Mol%. The best binary choice, RC-2 was fixed as follows: water 65 Mol%, 2-methylpyridine 35Mol%. Fluid RC-1 was used to optimize an engine with a turbine expander and fluid RC-2 was used to optimize an engine with a reciprocating expander. The optimization studies indicate that both the RC--1 and RC-2 final candidate fluids are appropriate fluids for automotive-size engine and are conducive to good efficiencies for autonotive application. Additional conclu- sions and recommendations are given. The results of the study are summarized and presented in fifteen tables. 17. KEY WORDS AND DOCUMENT ANALYSIS DESCRIPTORS b.lOENTIFIERS/OPEN ENDED TERMS C. COSATI 1'iclU/GfoUp Air Pollution Rankine cycle Engines Design Turbines Reciprocating Engines Working Fluid Mixed Fluorobenzenes Aqueous Pyridines 13 B 21 G 21 E 14 B 7 C 13. DISTRIBUTION STATEMENT Release Unlimited 19. SECURITY CLASS (This Report) Unclassified 21. NO. OF HAGtS 36 20. SECURITY CLASS (This page) «-s Unclassified 22. PRICE EPA Form 2220-1 ,-V-73) 32 ------- |