ORNL-TM-.3777 HEAT TRANSFER SALT FOR HIGH TEMPERATURE STEAM GENERATION E. G. Bohlmann MASTER PISTRIBUTIDH OF TWiT ncmuecyy 13 g naiye ) O, OAK RIDGE NATI OPERATED BY UNION CARBIDE CORPORATION e FOR THE U.S. ATOMIC ENERGY COMMISSION This report was prepared as an account of work sponsored by the United States Government. Neither the United States nor the United States Atomic Energy Commission, nor any of their employees, nor any of their contractors, subcontractors, or their employees, makes any warranty, express or implied, or assumes any legal liability or responsibility for the accuracy, completeness or usefulness of any information, apparatus, product or process disclosed, or represents that its use would not infringe privately owned rights. ORNL-TM-3777 Contract Na. W-7405-eng-26 Reactor Chemistry Division HEAT TRANSFER SALT FOR HIGH TEMPERATURE STEAM GENERATION E. G. Bohlmann DECEMBER 1972 — NOTICE This report was prepared as an account of work sponsored by the United States Government, Neither the United States nor the United States Atomic Energy Commission, nor any of their employees, nor any of their contractors, subcontractors, of their employees, makes any warranty, express or implied, or assumes any legai fiability or responsibility for the accuracy, com- pleteness or usefulness of any information, apparatus, product or process disclosed, or represents that its use would not infringe privately owned rights, OAK RIDGE NATIONAL LABORATORY Oak Ridge, Tennessee 37830 operated by UNION CARBIDE CORPORATION for the U.S. ATOMIC ENERGY COMMISSION g P T L PARTARIERARY U UL Ler e L TS YT i‘?»fi*!fi%x»u;ww GF Thid vLuUaBRT i BIUME fih‘ [ \e/*a 4 : ABSTRACT AND SUMMARY. . INTRODUCTION., . . . . . COMPOSITION AND STABILITY CORROSION . . . . - . . PHYSICAL PROPERTIES . . Freezing Point . . Viscosity. . « « + & Electrical Resistivity DenSity. . . . ° . . o © + Thermal Conductivity . @ * Heat Transfer Coefficient. Thermal Expansion. . Specific and Latent Heat Efficiency . . . . - SAFETY PRECAUTIONS. . . Principal Hazards. . Combustion Hazards . APPENDIX A. . . « « . . APPENDIX B, . . . « . & REFERENCES, . . « « « o 0 @ ¢ o e iii CONTENTS 16 23 23 23 23 26 26 26 28 28 28 31 31 31 35 37 39 HEAT TRANSFER SALT FOR HIGH TEMPERATURE STEAM GENERATION E. G. Bohlmann ABSTRACT AND SUMMARY A literature review and canvass of vendors reveals no experience and sparse, short term data bearing on the possible use of heat transfer salt (HTS - 407% NaNO:, 7% NaNOi, 53% KNOj3) at temperatures > 1000°F, Stability data found indicate that in a steam generator loop running with a maximum temperature of 1100°F about 0.5% of the NaNO, would decompose per day according to the overall reaction 5NaNO, -+ 3NaNO3 + Naz0 + Ny . In the size loop envisioned for a Molten Salt Breeder Reactor (MSBR) the nitrogen produced would amount to 3400 ft®/d. Austenitic stainless steels are recommended materials of construction above 850°F and short term test data suggest high nickel alloys like Incoloy, and Inconel would also be suitable. I would infer that Hastelloy-N would be resistant also. A tertiary HTS steam generator loop would provide an effective barrier to transfer of tritium formed in an MSBR to the steam generator. The tertiary loop is required to eliminate the possibility of a violent moderator graphite - HIS reaction. How- ever, projected use in an MSBR at temperatures to v 1100°F would require a substantial development program directed at stability and/or composition re-adjustment and verification of materials compatibility. Thus, the presence of appreciable Na;0 and/or NaOH, depending on steam leakage, at such temperatures might appreciably alter the corrosion picture. An industrial installation for drying NaOH was operated one and one half years with a maximum temperature of 980°F without difficulties due to corrosion or stability. The salt loop was constructed of ASTM 106 Grade A steel. INTRODUCTION It is estimated that a molten salt breeder reactor will produce NZSOO c; T per day. Transfer of appreciable percentages of this 3T to the steam penerator system must be prevented to avoild resultant contamination and containment problems. A substantial barrier to such transfer would be provided by the use of heat transfer salt - HTS (Na-K, nitrite-nitrate mixture) in the reactor steam generator circuit. The HTS would be expected to block the tritium transfer in several ways: 1. Conversion of hydrogen to water by the reactions: (Cf Appendix A). a. NaNO3; + Hp =+ NaNO2 + H20 AF = =44 keal b. Ho + 1/2 0,* = H,0 AF = -46 keal * From NaNO; I NaNOz + 1/2 0» 2. HTS systems are compatible with steam (used as inert cover gas in industrial applicationsl) go the water formed in 1 would not be corrosive to the steam generator tubing, thereby eliminating that possible alternate tritium transport mechanism, 3. The tritiated water could be stripped from the HTS as required. 4. Protective oxides, which inhibit tritium diffusion, are formed on alloys in contact with HTS.2 Originally, consideration was given to the substitution of the HTS for the fluoroborate secondary coolant. However, the probability of a 3,4 between the HTS and the graphite moderator (see violent interaction section, Safety Precautions) in the event of contact during an accident eliminated this possibility. Such a violent interaction has been con- firmed by an investigation carried out by Bamberger and Ross (Cf. Appendix B). Use of HTS as the secondary coolant would also necessitate extensive demonstration of compatibility with the very high radiation intensities in the primary heat exchanger. An alternate possibility is the use of HTS in a separate, tertiary steam generator loop; 'LiF-BeF», or 92 NaBF,-8NaF, would be used in the secondary loop. Increased cost would make this approach untenable were it not for other positive considerations. These include: 1. The low, 288°F, melting point of HTS makes possible significant simplification of the steam system; cheaper alloys would also be useable. 2, Steam leaks into HTS will not enhance corrosion as it would in 92NaBF 4,~8NaF or other fluoride coolants. Assuming all the water leaking into the fluoroborate participates, n 5 moles/h is equivalent to 1 mpy (uniform) for the entire secondary loop of an MSBR. 3. Use of "LiF-BeF, in the secondary loop would eliminate poisoning of the fuel by leakage of secondary salt into the fuel; minor reverse leakage would also be more tolerable since the secondary loop could be included in the reactor cell. A substantially cheaper alternate might be NaF-ZrF, which has a reasonably low cross section and which probably is removable by the presently considered fuel processing methods. Robertson5 has estimated the added cost of a third HTS loop, taking into account savings due to the above and other considerations. He con- cluded, "Addition of the (L;B and HTS) loops would increase the power production costs by 0.2 - 0.3 mills/kwhr, making the total cost about 5.5 mills/kwhr." The added capital cost amounted to 10 to 13 million dollars (206 total) depending on whether Incoloy 800 could be substituted for Hastelloy N in the secondary loop. While this estimated increased cost was not favorable to the third loop approach to tritium control, this survey of the ''state of the art" as regards use of HTS under MSBR conditions was also conducted. At first blush it appeared that the widespread industrial experience would minimize the development required for such adaptation; unfortunately the survey showed such optimism was not well founded. COMPOSTTION AND STABILITY Heat transfer salt was developed by duPont in the thirties6 for use in the chemical and petroleum process industries, notably the Houdry catalytic cracking and refining units. The usual weight 7 composition is 40NaNOz, 7NaNO3, 53 KNOj3; which melts at 288°F; the melting point is not greatly affected by appreciable alterations as shown in Table 1. Table 1¥ - The Freezing Points of NaNO2-NaNO3;-KNOj Mixtures % NaNO; % NaNOs % KNO3 Freezing by wt. by wt. by wt. Point °F 100 0 0 540 50 0 50 282 4 3 53 284 42 3 55 287 40 0 60 289 40 7 53 288 ("HTS'") 38.5 11 50.5 293 38 6 56 292 35 7 58 311 34 13 53 305 34 3 63 328 30 20 50 305 30 10 60 312 30 0 70 333 100 0 586 0 100 633 %* Table 11, page 375, from Ref. 1. 1 Apparently the technology grew ''like Topsy," as regards both composition and materials of construction, out of many years experience with salt baths used in heat treating and cleaning metals. Presumably because of this and proprietary considerations, there is a paucity of quantitative data on stability and corrosion in the open and vendor literature; this in spite of widespread use in substantial quantities — one Houdry unit used n 10% 1b. Low cost (still 15¢/1b) and tractability under "'rcugh and ready" handling also may have contributed. At the time of its introduction the low, 288°F, melting point was a major advantage since it was below the temperature at which steam was available in most plants. The need for steam jackets and tracers, design for complete gravity drainage, and other freeze-melt problems, however, gave the competitive nod to organics like "Dowtherm" for temperatures below 700°F. A patented European development available under license in the U.S. 7,8 eliminates these problems by "salt dilution."” by American Hydrotherm Corp., This involves adding sufficient water to the salt at startup—and starting at 500°F on cooldowmto keep the salt in solution at ambient (¢ 50°F) temperature. The water is gradually flashed off at 200 - 500°F while circulating through the system; this eliminates the severe thermal shock of starting circulation at 500°F in a cold system or need for provision of an alternate preheat capability. Ten pilot plant and thirty-two plant size systems of this kind have been in operation since 1961 or are now under construction. American Hydrotherm claims a 207% saving on capital costs for the use of salt dilution instead of dry salt,7 with no resultant materials or handling problems. A point by point comparison by American Hydrothermg’lo of "Dowtherm A" with HTS using salt dilution indicates substantially lower costs for the latter. The dilution process also eliminates problems which may be caused by increases in melting point due to oxidation of nitrite to nitrate or formation of relatively insoluble Na;CO; from pickup of 002.8 J. Alexander, Jr., and S. G. Hindinll have published detailed studies of the freezing points of the system sodium and potassium nitrites and nitrates and the effects of impurities on them. Commercially used measures for countering these effects also discussed included: 1. Use of hydrogen treatment in a bypass adsorption tower to reduce excess nitrate to nitrite.” 2. Addition of concentrated HNO3 or gaseous NO2 to convert hydroxides and carbonates to nitrates and nitrites. 3. Cold trap use to reduce carbonate levels. Sodium hydroxide is formed by the reaction of the NaO0 thermal decomposition product of HTS with water. X ‘ Paul Larson12 indicated hydrogen treatment was no longer used because addition of fresh salt proved cheaper than providing for doing the reduction, Various attempts to find definitive data on thermal stability at elevated temperatures were fruitless in spite of vendor and open literature 1,7,8,13-15 These efforts are recommendations for use to 1000 and 1100°F. well characterized by one lead which reported successful use of HTS at temperatures as high as 1200 - 1300°F. A couple of phone calls located my original informant's source but produced the usual disclaimer. The use at such temperatures was occasional and involved only several pounds of HTS, which gave no trouble "if changed frequently.'" When asked what happened to the HTS, the reply was, "It gets pretty lumpy .....fast.” Up to v 850°F widespread experience indicates HTS is stable in carbon steel systems even when operated with air access to the system. Slow oxidation of nitrite to nitrate and carbonate formation may increase the freezing point slowly, but these changes apparently do not result in any difficulties. Thus, Crowley and Helms trip report16 gives the composition 7 NaNO; 9,28 NaNO, 30.3 KNO 3 48,7 K2C03™* 3.38 KC1 0.27 £91.93 * Alexander and Hindinll indicate Na,CO3 would be more likely. for bulk salt after an indeterminate period of operation (probably > 2 years). The system has remained in operation with this salt charge since April 1967 (® 24000 h @ 800°F maximum) with no requirement for composition adjustment; new salt is added only to make up for losses. Helms also reports17 8 years use of the same batch of "Hitec" at a maximum bulk salt temperature of 843°F without observable deterioration; air access was prevented by nitrogen blanketing in this application. Vendor literature recommends such nitrogen, or steam, blanket:;.ngl?”14 above 850°F, but contacts with vendor personnel uncovered no experience at such temperatures. Data for the rate of disappearance of nitrite are given by W. H. v.d. Broek18 for heat transfer salt used in finishing (drying to ~ 100 %) NaOH. In this application the composition 487 KNOj 497 NaNO: 3% Na,CrOy was used. The addition of chromate is reported (ca. 1940)19 to inhibit decomposition of the nitrite to nitrate and oxide but Rottenberg (1957)3 states that the claims had not been established, Paul Larsonl2 also indicated that duPont had not found such additions either harmful or beneficial; but some users insisted on making them. W. H. v.d. Broekl8 showed nitrite decrease rates graphically for 1 1/2 and 2 1/2 year periods of operation at maximum temperatures of 980 and 800°F respectively as shown in Fig. 1. Unfortunately only the plots for nitrite loss were given, so it is not possible (no analyses) to determine how the two possible reactions 1) 2NaNO, + O;(air) -~ 2NaNOj 2) 5NaNO» ‘A* 3NaNO3 + Na,0 + N, contributed to the observed changes. Also, during the high temperature operation, concern about the pfimp caused the system to be opened to the atmosphere frequently; this was minimized during the second run at 800°F. Still another factor may have contributed to the rapid loss of nitrite at 980°F; it is reported20 that the decomposition of alkali nitrate - nitrite mixtures is catalyzed by iron above 520°C, and the system described by v.d. Broek was made of carbon steel. Stainless steel does not catalyze Q 2 decomposition of alkali nitrate - nitrite mixtures even at 800°F. 1 Some laboratory studies in 316 stainless steel containers have been 12 carried out at 1100°F. Figure 2 shows the effect of various blanket . n gases on the melting point of "Hitec;" "melting point" in HTS parlance refers to the temperature at which any solid phase appears. Reactions considered significant in the decomposition of nitrate - nitrite mixtures (the sodium salts are indicated because they are less stablezl) include: ORNL-DWG 72-12309 S0 ——— T T CHARGE II H O \ MAXIMUM TEMPERATURE| ——— W o \ 425°C (800°F) N NITRITE CONTENT (%) N O o \QARGE [ < MAXIMUM TEMPERATURE 525°C (980°F) | | Fig. 1. 4 8 {2 {6 20 249 28 TIME (months) Nitrite Decomposition in Industrial Process Loop Using HTS* * Per W. H., v.d. Broek, see Ref. 18. 10 ORNL-DWG 72-12310 320 7 300 280 AIR / / 260 f/////,///// N / e ' p.(°C) E 220 / 200 2 ////// | / / 180 / o Y NO N2 140 0 5 10 15 20 25 30 TIME AT {100°F (days) Fig. 2. Effect of Various Blanket Gases on HTS m.p. (See Ref. 12) 11 NaNO3 < NaNO; + 302 (1) 2NaNO, - Najy0 + N203 (2) NaNO, + N203 - NaNO3 + 2NO (3) NaNO, + 2NO - NaNO3 + N30 (4) NaNO, + N,0 NaNOs; + N (5) The formation of Na,C03 and/or NaOH from reactions of carbon dioxide or water with the Na,0 formed in (2) account for the melting point increases with air, steam, and carbon dioxide blankets. Conversion of the nitrite to nitrate by reactions (4) and (5) probably accounts for the rapid increase in melting point with NO as the blanket gas. Reactions (2) thru (5) produce the rise under nitrogen, again through conversion of nitrite to nitrate. This is indicated by analysis of a salt mixture after six weeks under nitrogen at 1100°F which showed the composition to be that given in Table 2, with a melting point of 329°F.l2 Table 2. Change in "Hitec'" Composition After Six Weeks at 1100°F % Composition Compound Original Final NaNO3 7 18 NaN02 40 28 KNO3 53 52 Na,0 - 2 Analysis of gas evolved gave 98.57 N, and 1.57 N20. Gurovich and Shtokman22 and Voskresenskaya and Berul'23 also conducted laboratory investigations of HIS stability at temperatures > L000°F and found appreciable nitrogen evolution. The studies from which decomposition data are available are summarized in Table 3. After an initial period, the generally accepted overall Table 3, Long Term Decomposition Studies of Heat Transfer Salt Equivalent¥* ° Wt, HTS Exposure Original Composition Author & Ref, No. Temp. "F rate of N Type System Used Time NaN02 NaN03 KN03 NaZCrO* Observations evolution g (cc/mole NaNO,/d) Larson, 12 1100 316 ss autoclave 1000 42 d 40 7 53 - Final composition: 39 28% NaNO,, 2% Na,O 2 980 18 mo Final composition: 11 102 NaN02 v.d. Broek, 18 ASTM Al(Q6é industrial process Very large 49 48 3 800 (circulating loop) 30 mo Final composgition: 1.3 38Z NaNO 2 Gurovich 29 934 Steel (.09C, 0.9Cr, 0.24V, 145 30 4 40 7 53 - 10 ce N2/d after 5 d 12 & Shtokman -~ 0.4Mo) autoclave 1040 Ag autoclave 50 30.8 d 40 7 53 - 0.09 g N2 evolved 8.3 Voskresenskaya 23 1112 Ag autoclave 50 10 d 40 7 53 - 0.07 g N2 evolved 20 f —_— & Berul 1112 18Cr 25Ni 251 steel autoclave 25 10.2d 40 7 53 - 0.08gN, evolved 45 * Assuming over-all reaction is SNaNO2 -+ 3NaN03 + Na20 + N2 ¢t 13 decomposition reaction is 5NaNQ, =+ 3NaNOj + Na,O0 + N» . This stoichiometry was used to calculate the nitrogen evolution rate from the results reported by the experimenters; the rate was necessarily assumed to be constant, although Voskresenskaya and Berul'23 and others indicate it decreases with time. The calculated N; evolution rates are those given in the final column in Table 3 and are plotted in Figure 3. Considering the disparate conditions of the several investigations the results are pleasantly consistent although more data would be helpful in more firmly removing the possibility of coincidence. The two low points due to Voskresenskaya and Berul' are for the experiments performed in silver; they and other investigators reported lower decomposition rates in noble metal containers than in carbon and alloy steels. The data due to v.d. Broek are of particular interest; recall that they were obtained from HTS in use in an industrial process loop (dewatering NaOH) and the only temperatures given were the maximums in the loop. Thus the average temperatures were appreciably lower and the data should fall below those from the isothermal studies of Larson and Gurovich and Shtokman — as they do. The consistency in all these studies suggests that the measured decomposition rates may be reasonably accurate for an engineering operation. If one accepts then that these data may be considered reasonably representative of the tertiary loop as envisioned by Robertson5 (maximum temperature 1100°F), a decomposition rate of about 40 cc Np/mole NaNO,/d would be anticipated. For the 9 x 10° 1b of HTS (40% NaNOs,) in the tertiary loop, this amounts to 4200 moles (3400 f£3) No/d. The nitrite loss would amount to 21,200 (2200 1b) moles per day; i.e., 1.0%/d. Extrapolation to temperatures above 1100°F is questionable since 23,24 report nitrogen oxides are also evolved above 600°C. various workers These were short term studies, however, and it may be that the evolution of oxides of nitrogen would not continue after a period, as happened in the longer term studies at lower temperatures. Voskresenskaya and Berul'23 also reported substantial attack on the containers in some of their experiments, unlike cther experimenters. When this occurred with high Cr alloys, appreciable amounts of chromate were found in the melts. NITROGEN EVOLUTION (cc/mole NaNO, /o) 14 ONRL-DWG 72-12311 100 \ 50 \\‘ 4 \\ 20 A \\\\\ 2 10 *\ 5 N ) N\, \ EO LARSON-NaNO, LOSS AND NayO FORMATION - - ® V.D. BROEK-NaNO, LOSS 0.5 - ® GUROVICH-N, EVOLUTION ~ A VOSKRESENSKAYA- No EVOLUTION I | .6 °F 1100 | 800 0.2 —, 1 _ c 727 593 500 Tl 0.9 1.0 .1 1.2 1.3 .4 1.5 Y7 (oK) Fig. 3. Rate of Nitrogen Evolution from HTS vs Temperature 15 Voskresenskaya and Berul'23 also present data showing the rate of nitrogen evolution as a function of nitrite concentration. These data, interpolated from a small graph and given in Table 4, are only indicative, but they are consistent with other observations. This suggests that a mixture consisting of 8% NaNO,, 427 NaNO3; and 507 KNOj3 might result in lower N; evolution rates. The sodium nitrate-sodium nitrite ratio was chosen on the basis of Freeman'525 observation that the ratio (under omne atmosphere of oxygen) is 5.9 at 600°C. Curiously, no observer mentions the evolution of oxygen from the usual mixtures at such temperatures. Alexander and Hindin”sll data indicate a melting point of N 420°F for such a mixture, still adequately low for simplification of the steam cycle. Apparently the equilibrium NaNOj <> NaNO» + 02 is substantially shifted to the left in HTS; Freeman's data indicate PO2 would be about 0.3 atm at 600°C for the usual mixture. Presumably, however, the equilibrium would shift to the right as the percent nitrate is increased, with resultant evolution of oxygen. Perhaps a mechanism change also occurs and the oxygen appears as the oxides of nitrogen 24,25 observed by some observers at temperatures > L100°F. Table 4. Rate of Nitrogen Evolution vs NaNO; Concentration at 550°C* % Composition cc Np/100g NaNO,/hr NaNO2 KNO3 100 0 360 80 20 135 60 40 15 40 60 8 20 80 5 % , These were short term (to 20 h) studies and are therefcre only indicative. 16 CORROSION As with thermal stability, corrosion data are sparse, and then only from relatively short term tests. Considerable experience has demonstrated acceptable, economic performance with no apparent necessity for quantita- tive, long term corrosion data. The vendors recommend against use of carbon steel above 850°F; at lower temperatures it gives excellent, long 3,13,16 term service. Rottenberg3'giveS'the following description of how troubles can arise in a poorly designed maloperated, direct fired heater, "The main danger is, as usual, overheating in the stagnant film at heating surfaces. The danger is accentuated in this case by the ability of HTS to detach particles of rust and scale from the inside of the plant and deposit it as sludge at the bottom. Even more important is the fact that with gross overheating, at local temperatures above 700°C, a new reaction sets in with the rapid breakdown of nitrate to oxygen, nitrogen, and oxide. 1In the presence of a mild steel surface, the oxygen combines with it in a thermite type reaction which is self-sustaining and rapidly burns through the metal. No reaction takes place with stainless steel. In plants operating at high bulk temperatures heating sections should be constructed of stainless steel to obviate this danger.” In the operation described previously, however, v.d. Broek18 observed no corrosion of the system during 1 1/2 years of operation with a maximum temperature of 980°F. The heater section and the rest of the heat transfer loop were made of carbon steel (ASTM Al06) with nickel and "Incoloy" at the interfaces in the NaOH concentrator. Conventional use of nickel and nickel alloys for materials of construction in caustic concentrators using heat transfer salt26 and v.d. Broek's experience indicate good compatibility for these materials to ~ 1000°F. Other qualitative assessments of corrosion by heat transfer salt during extended operating periods are summarized in Table 5. Miscellaneous corrosion data are shown in Table 6, taken from the 13 While no experimental details are provided, it is "Hitec" pamphlet. probable that the results are from quite short term (a few hundred hours) tests. The two period tests at 1058°F indicate the corrosion rates of the highly alloyed materials decrease with time. Note, however, that Inconel was an exception in spite of the excellent results at 1000°F and for nickel at 1100°F. Similar results are shown by data furnished by J. Friedman27 which are presented in Table 7. The relatively short Table 5. Qualitative Assessments of Corrosion by Heat Transfer Salt Time Temp. Blanket Alloy vr. op Gas System Comment Ref, 800 Air Circulating No noticeable corrosion 27 ASTM-A-285 Grade C 3 785-800 Limited Air Circulating No noticeable corrosion 27 Carbon Steel . . . . . 2.5 875 Steam Circulating No noticeable corrosion 27 347 SS 1.3 1050 Air Heat Treat Bath Negligible corrosion 27 347 SS 2 1000 Steam Circulating No noticeable corrosion 27 Seamless Stainless 0.8 930-1020 ? Circulating No visible attack 13 Tubing L1 TABLE 6 Metals Steel—open hearth (ASME # S-17) Alloy steels 15-16% chromium iron “Alcrosil” #3 “"Alcrosil” #5 Stainless steels Type 304 Type 3041 Type 309 (onnealed} Type 309 Cb Type 310 Type 316 Type 321 Type 347 Type 446 Inconel Carpenter 20 “Hastelioy” B Monel Bronze Phosphorized Admiralty Copper Nickel CORROSION OF METALS BY "HITEC” Corrosion Rate, inches Penetration Per Month 612°F, 785°F. 850°F. 1000°F. 1058°F. 1100°F, 1st Period 2nd Period _ —_ 0.0003 0.001 to 0.002 - — _— —_ —_— 0.0000 - - —_ - 0.0002 0.001 - — — — 0.0002 0.00035 - — —_ — —_— 0.0007 - - — -— — 0.0006 -— - 0.000062 0.00001 0.0000 - 0.00110 0.00064 —_ —_ —_ —_ 0.00156 0.00094 —_ —_ - - 0.00117 0.00077 — -— - 0.0000 —— —_ — —_ —_ - 0.00111 0.00056 — — -_ 0.0004 0.00109 0.00068 —_ —_ —_— — 0.00146 0.00072 —_ — —_ 0.0000 0.00153 0.00151 —_— - —_ - 0.00097 0.00059 0.00011 0.000003 - - -— — — - - 0.0001 — —_ 0.00006 0.00008 0.0001 - - —_— 0.00006 0.00005 0.0001 - - —_— - — - 0.03 —_ — 0.01 10 0.05 0.002 to 0.006 0.001 to 0.002 0.00025 8T Table 7. Resistance of Steeis to Molten High Temperature Eutectic Salt at 900 and 1100°F Assumes adherence to parabolic rate law. Quality Testing 900°F - 482°C 1100°F -~ 593°C and Time in Loss in Wt.2 Extrapolated? Loss in Wt.2 ExtrapolatedP Cast No, Hours MGMS/CM2 Penetration MGMS/CM2 Penetration at 1 year at 1 year 24 2.0 21.4 MILD 100 3.6 7.1 39.6 31.7 18.4 STEEL ' 500 21,0 - 24 2.9 9.8 ‘ESSHETE CML 100 3.3 2.4 14.0 9.2 6.4 ESSHETE 24 2.0 9.2 CRM. 2 100 3.1 2.74 13.9 10.4 5.9 500 11.0 29.3 18.5 ESSHETE 24 1.5 4.1 CRM. 3 100 2.4 2.3 8.0 6.3 3.6 500 9.3 17.2 14.7 RED FOX 24 0.3 2.7 (Type 321 100 0.3 0.5 4.0 2.3 $S) 500 1.9 10.1 7.4 SILVER FOX 24 NIL 2.2 (Type 316 100 NIL 0.03 3.5 0.60 SS) 500 0.1 4.8 3.8 ®The loss in weight was determined after descaling in sodium hydride. 6T 20 term data presented are consistent with parabolic rate law behavior — mils penetration = k vt; the extrapolated penetrations at one year are calculated from the data using this expression. The alloy compositions tested are given in Table 8. The "Hitec" pamphlet13 (page 10) indicates copper is useable at moderate temperatures (to 600°F), but corrodes rapidly at higher temperatures — 30 mils/mo at 1000°F, see Table 6. Rottenberg3 indicates, "Copper, aluminum, magnesium, and their alloys are all attacked more or less rapidly and should not be allowed to come into contact with HTS," (See additional comment on aluminum and magnesium under Safety Pre- cautions.) The presence of sulfate is said to be deleterious to the low alloy steels like CML and CRM2 (see Table 8), and Rottenberg3 states, ''the corrosive action of HIS is greatly affected by the presence of oxide and hydroxide." Insolubles, like carbonate, may lead to erosion or localized overheating problemsul6 The oxide and hydroxide effects would require investigation since temperatures of interest in MSBR's presumably would produce appreciable oxide which would be converted to hydroxide by any steam leakage. Table 8. Analysis and Heat Treatment of Materials Analysis Per Cent (Type 316 SS) Quality and — Original Cast No. C Mn Si Ni Cr Mo Ti Heat Treatment Mild Steel 0.116 0.44 0.24 700°C 1 Hr. AC ESSHETE CML 0.074 0.53 0.54 0.92 0.52 920°C 1 Hr. FC ESSHETE CRM.2 0.114 0.48 0.24 2.26 1.01 920°C 1 Hr. FC ESSHETE CRM.5 0.069 0.48 0,30 5.03 0.56 920°C 1 Hr. FC RED FOX ° (Type 321 SS) 0.047 1.73 0.60 8.56 18.44 - 0.42 1050°C 1 Hr. AC SILVER FOX 0.05 1.66 0.37 11.26 17.28 2.79 1100°C 1 Hr. AC 1¢ 23 PHYSICAL PROPERTIES Table 9, compiled by Robertson,5 compares several salts useable in MSBR's., Hitec compares favorably with the other two coolant salts. The remainder of this section, and the subsequent one on Safety Precautions is taken from the "Hitec" pamphlet.13 Figure and Table numbers are those used in the pamphlet. "Freshly prepared "Hitec" is a white, granular solid; in the molten state, it has a pale yellow color. It is a eutectic mixture of inorganic salts having the following composition: Potassium Nitrate 537 Sodium Nitrite 407 Sodium Nitrate 77 Freezing Point "The freezing point of fresh "Hitec" is 288°F, and it can therefore be melted readily by plant steam at a pressure as low as 50 psig. The presence of a small amount of moisture sharply reduces the freezing point, and in order to determine the freezing point accurately, it is recommended that the salt be melted and dried by heating at about 480°F. Driving off water vapor in this manner may cause effervescence or frothing, depending on the amount of moisture present, and it is recommended that suitable precautions be taken to protect personnel from spattering by the hot, molten salt. Viscositz "The viscosity of molten "Hitec" was measured over the temperature range from 300 to 820°F, and these data were extrapolated to 1000°F with an accuracy believed sufficient for design purposes. The experimental points and extrapolated curve are shown in the graph of viscosity versus temperature, Figure 6. Electrical Resistivity "The electrical resistivity of molten "Hitec" is 1.7 ohms per cu.cm. This value is approximately midway between that of water, 25 x 10° ohms per cu.cm., and that of mercury, 95.9 x 10™°% ohms per cu,cm, Table 9. Selected Properties of the MSBR Molten Salts5 7LiF—Ber—ThF4—UF4 NaF-NaBF, 7LiF—BeF2 KNO 3~NaNO 2 -NaNO 3 Composition, mole 7% 71.7-16-12-0.3 92-8 66-34 44 . 2-48.9~6.92 Molecular weight, approximate 64 104 33 84 Density, 1b/ft® at 1000°F 212 117 124 105 Viscosity, 1b/ft-hr at 1000°F 41 3 29 3 Specific heat, Btu/1b-°F 0.32 0.36 0.57 0.37 Thermal conductivity, Btu/ft-hr-°F 0.67 to 0.68 0.23 0.58 0.33 Estimated cost, $/1b 57.00 0.50 12.00 0.15 Circulation required per loopb for 556-MW(t) heat load: 1b/hr 23.4 x 10° 18.3 x 10° 13.3 x 10° 14.7 x 10° gpm 14,260 19, 500¢ 13, 380 17,370 Liquidus temperature, °F 930 725 850 288 a o ° 2 Eutectic composition. b , . s Based on properties at average temperatures in MSBR system. “Based on 250°F At in modified MSBR. we VISCOSITY —CENTIPOISES 20 25 o 18 16 14 12 10 300 400 300 Fig. 6. 600 700 800 TEMPERATURE—°F. Viscosity of Liquid HITEC 900 1000 26 Densitz "The density-temperature relationship for molten "Hitec" is presented in the alignment chart of Figure 7. By drawing a straight line from the desired point on the temperature scale through the reference point, density can be read from the chart in grams per cu.cm., pounds per cu.ft., or pounds per U.S. gallon. Thermal Conductivity "Thermal conductivity measurements on "Hitec" heat transfer salt have been reported by W. D. Powers, ANP Quarterly Progress Report, September 30, 1957, ORNL-2387 (Classified). A wvalue of 0.35 B,T.U./(hr.) (sq.ft.) (°F./ft.) independent of temperature in the range from 468 to 689°F was obtained using a parallel-plate, variable sample gap apparatus such as that described in ASME paper 56-SA-31 by C. F. Lucks and H. W. Deem. Battelle Memorial Institute, in a similar apparatus, extended the tem- perature range and obtained a value of 0.33 B.T.U./(hr.) (°F./ft.) independent of temperature in the range from 400 to 900°F. (private communication from H. W. Hoffman, Oak Ridge National Laboratory and Battelle Memorial Institute). "An earlier Russian article by Vargaftik, Neimark, and Oleshchuck which appeared in the Bulletin of the V.T.I., September 1952, No. 9 had reported values in the temperature range from 302 to 932°F which ranged from 0.26 to 0.17 B.T.U./(hr.) (sq.ft.) (°F./ft.). Heat Transfer Coefficient "Heat transfer coefficients (h) for "Hitec" flowing in turbulent motion inside horizontal circular, iron pipes have been determined over the temperature range of 580 to 960°F at linear velocities up to 6 feet per second, Details of this experimental work are published in Transactions of American Institute of Chemical Engineers 36, 371 (1940). Data on the thermal conductivity of "Hitec'" were lacking, and the experi- mental results, therefore, were correlated in terms of the function hD uo. 4’ the heat transfer coefficient were reported by H. W. Hoffman and S. I. Since the publication of this work, additional determinations of Cohen in a paper entitled Fused Salt Heat Transfer — Part III: Forced- TEMPERATURE U" Gc‘ 1100 400 DENSITY #IETY #JUS. GAL, -i 100.0 i 1000 4 - 500 1 900 — 1400 o > 4 1 P 800 4 ’ /s 1100 —4 y - s 400 ! s +— 15.00 _herenence o 7 'OIN'F 700 i ~ 1 / { s /s - 7~ - s 4 i s 600 { _ / 300 F o / 120.0 —f 71600 - % 500 S + ” | . s - 7’ > / / / 400 200 300 FREELING POINT 27 NOTE: TO FIND DENSITY, DRAW A STRAIGHT LINE FROM DESIRED POINT ON TEMPERATURE SCALE THROUGH THE REFERENCE POINT, Fig. 7. Density of Liquid HITEC DENSITY Gus./C.C. f— 2.00 ""]""'IITIIIT"I]vt!'llrvtlrvuvl' N = ° o o o - o o 28 Convection Heat Transfer in Circular Tubes Containing the Salt Mixture NaNO2-NaNO3-KNO3, ORNL-2433 (1957). The latter work apparently utilized a more accurate apparatus than that used in the earlier determinations. The data was correlated using the Colburn j function, E%E' QEB Hoffman and Cohen report that the variables ranged as follows: Reynolds modulus (N ), %9 4850-24,710 Prandtl modulus (N_ ), Cpu 4,2-9.1 Pr k Average fluid temperature 552-828°F. Heat flux 62,800~-194,500 B.T.U./hr.-ft.2 Under these conditions, the heat transfer coefficient varied from 800 to 2900 B.T.U./(hr.) (sq.ft.) (°F). In heat transfer applications, "Hitec" behaves as a normal fluid and the standard correlations may be used in designing heat exchange equipment. Thermal Expansion "The coefficient of linear thermal expansion for solid "Hitec', measured by the Dilatometer method of ASTM-D696-44, is 2.85 x 10™° per °F. The coefficient of cubical thermal expansion for liquid "Hitec" is 2.016 x 10" per °F. (private communication from H. W. Hoffman, Oak Ridge National Laboratory). Specific and Latent Heat "The total heat of "Hitec' was measured calorimetrically over the temperature range of 70 to 1100°F, The slopes of the total heat- temperature lines, Figure 8, indicated that the specific heat of solid "Hitec" is 0.32 and that of the melt 0.373 calories per gram per °C (or B.T.U. per pound per °F). Latent heat of fusion appeared to be approximately 20 calories per gram (35 B.T.U. per pound). Efficiency "The efficiency of "Hitec' heat transfer salt is well illustrated by a direct comparison to hot air, or flue gas, on a performance basis. 29 Such a comparison is given in Table I, and the details of the calcu- lations used to obtain these data are given in the previously cited article appearing in Transactions of American Institute of Chemical Engineers 36, 371 (1940). 30 400 /A - :_5 00 - | uQuio - o z = 5 N - 200 / 100 ¥ - y MELTING RANGE / ----- [ y TEMPERATURE—*F. Fig. 8 TABLE { COMPARISON OF "HITEC” AND AIR AS HEAT TRANSFER FLUIDS Basis of Comparison: Fluvid Conduit: Average Fiuid Temperature: 800°F, 1" 1.D. Tubes Pressure Drop due to Friction: 3.5 Ibs. /(sq.in.)/100 ft. of tubmg Difference between inlet and outlet fluid temperatures: 50°F. "HITEC" at 3 atm.press. AlR at 10 atm.press. Press. drop./100f.. . . . . . . . . . . .. 3.5 Ibs. /sq.in. 3.5 lbs. /sq.in. 3.5 ibs. /sq.in. Lbs. /hr. /tube e e e . . . 6000 185 360 Heat corrrymg copocfly /wbe, . . . . . L 111,900 B.T.U. /hr. 2310 B.T.U. /hr. 4500 B.T.U, /hr. No. of 17 tubes required to transfer 111,900 B.T.U. /hr. at 50°F. temperature dlfhronce 1 48Y2 25 Totallbs./br. . . . . . . . . . . .. . . 6000 8970 8970 Theoretical HP required o overcome fnchon .. 0.014 241 7.2 Individual coefficient of heat transfer. . . . . . 858 28 40 31 SAFETY PRECAUTIONS Principal Hazards "The principal hazards associated with "Hitec" are those which apply to any operation involving the use of liquids at elevated temperatures. Adequate precautions should be taken to protect operating personnel from burns in case of equipment failure. There is a lesser danger that "Hitec," by supporting the combustion of other materials, may escalate an existing fire hazard. Similarly, there is a danger that prolonged contact of "Hitec" with human skin may cause dermatitis. Finally, "Hitec'" should not be ingested by operating personnel in more than trace amounts because of the sodium nitrite which it contains. "'"Hitec' itself liberates no toxic vapors, but adequate ventilation should be supplied to remove products of combustion due to accidental or process contamination of the "Hitec" system. Although the combustion products of most organic compounds are principally carbon dioxide and water, traces of highly toxic gases such as carbon monoxide, oxides of nitrogen, etc., may also be present. It is reported that, in instances where a process utilizing "Hitec" generates steam, operating personnel complain of a salty taste after being in the area for a period of time, particularly when the weather is humid. Adequate ventilation will relieve this condition. Combustion Hazards "Although "Hitec' itself is nonflammable, it will support the combustion of other materials, and salt units should not be located in the vicinity of wooden construction. Similarly, contamination of the work area with "Hitec' may increase any fire hazard created by the presence of combustible materials. These problems may be minimized by washing the work area regularly with water and by substituting non- flammable materials for combustible materials. For example, sand rather than sawdust should be used for c¢lean—up, diking, etc. A number of tests have been carried out to estimate the hazards associated with the accidental contact of flammable materials with hot, molten "Hitec," but 32 since it is impossible to anticipate all conditions under which this product may be used, it is recommended that the user carry out additional tests before undertaking any novel applications. "Combustible vapors generally react relatively slowly with "Hitec" and escape before combustion is complete. Gasoline wvapors, high in unsaturated compounds, were bubbled through the molten salt for two hours at 1100°F, and over 907 of the gasoline was recovered unchanged. The reaction between "Hitec" and ammonia was also studied under similar conditions, and less than 50% of the ammonia reacted with the "Hitec'" over a two-hour period at 1100°F. 1In both cases, changes in the salt were minor. "Flammable liquids tend to vaporize rapidly in contact with molten "Hitec" and burn at the surface in the same manner as they might be ignited on any hot surface. Crude oil, for example, burned in a similar manner on the surface of "Hitec" at 1150°F. and on the surface of molten lead at the same temperature. Motor gasoline, cracked gasoline, gas oil, and crude oil mixed with sulfur were released below the surface of an open container of "Hitec" at 1100°F. 1In all cases, the hydrocarbons vaporized and burned, and in the cases of crude oil and gasoline, there was considerable spattering of hot salt due to the rapid, sub-surface vaporization of the liquids. Again, there was little change in the salt composition, indicating that the principal reactions were between the vaporized hydrocarbons and the air above the salt bath. "Combustible solids, such as wood, coke, paper, plastics, cyanides, chlorates, ammonium salts, and active metals, such as aluminum and magnesium, offer somewhat greater hazards. Magnesium, except as an alloying agent in low concentration, must not come in contact with "Hitec." Heat treatment of aluminum parts in molten "Hitec" is a common and safe practice when suitable precautions are taken, as discussed in "Potential Hazards in Molten Salt Baths for Heat Treatment of Metals,'" Research Report No. 2, The National Board of Fire Underwriters (1946), 85 John Street, New York 7, New York. Thus, we strongly recommend that before attempting to use "Hitec'" in this type application the potential hazards as discussed in the reference article be reviewed. "In these cases it is possible for the solid combustible materials to remain in contact with the hot salt until they are completely oxidized, 33 and large volumes of combustion gases can be formed rapidly beneath the salt surface. This could, of course, result in rupture of equipment or violent spattering of hot salt. "Violent combustion resulted when ground petroleum coke was held beneath the surface of "Hitec'" at 1150°F., and in contact with iron, a thermite-like reaction took place. Similarly, mixtures of fine aluminum and iron filings reacted with "Hitec" at 1000°F. with the evolution of enough heat to burn through a nickel crucible. "Finely divided charcoal flared up and burned vigorously when placed on the surface of "Hitec" at 1200°F., but grease burned quietly under the same conditions. When an oxygen-balanced mixture of "Hitec'" and paraffin was heated rapidly tc 1100°F., the paraffin merely volatilized and burned, leaving the salt practically unchanged. However, an 80/20 mixture of "Hitec" and nylon burned vigorously at 750°F., and subsequent analysis indicated the oxidizing salts had taken part in this reaction, as they had in the case of the coke held beneath the salt surface. "All attempts to detonate '"Hitec' or mixtures of "Hitec' with combustible products, such as paraffin, petroleum, nylon, and topped Houdry condensate, at 375°F. and 25 inches of mercury vacuum, were unsuccessful using either a No. 8 blasting cap or 0.5 1b. of blasting gelatin as the initiator. The salt was similarly insensitive at 1100°F. Attempts to detonate oxygen-balanced mixtures of "Hitec" and either finely divided charcoal or topped Houdry concentrate by heating the mixtures vigorously undef confinement were also unsuccessful. It was estimated that pressures as high as 18,000 psi were attained during these tests. "Water from spray sprinklers or low-velocity fog nozzles will not penetrate the surface of molten "Hitee'" sufficiently to cause dangerous spattering, and this type of fire protection is recommended for oil-fired units and for processes involving combustible materials. Carbon dioxide and approved dry-powder type fire extinguishers can be used satisfactorily to extinguish fires in the vicinity of a salt unit, but vaporizing liquid (carbon tetrachloride), foam, and aqueous types, other than sprinklers or low-velocity fog types, should not be used. An adequate supply of clean, dry sand is useful for slagging and diking to confine the spread of escaped molten salt. 34 "If, for some reason, water is introduced into a unit containing "Hitec," the heat input should be kept low until all the water has been evaporated, and the system should be vented adequately in order to release the steam which is formed." INTRA-LABORATORY CORRESPONDENCE OAIK RIDGE NATIONAL LABORATORY — 35 APPENDIX A November 4, 1970 To: W. R. Grimes From: S. Cantor Subject: The Reaction, H(g) + (Na,K)NO3() Z (Na,K)NO»(L) + H.0(g) I concur with the conclusion of Ed Bohlmann, namely, that the equilibrium, given above can prevent tritium from reaching the feed water and steam of a molten-salt reactor. ' The basic data and considerations are as follows: Nominal Composition of HTS KNO3 - NaNO2 - NaNOj; 53 - 40 ~ 7 weight 7 49 - 44 - 7 mole 7% HTS is a reciprocal system: KNO3(2) + NaNO2(2) & KNO2(2) + NaNO03(2) Assuming that nitrate icn is larger than nitrite ion, this equilibrium is shifted somewhat to the left (the '"large-large, small-small' rule in reciprocal systems). The ratio of activities, KNO3/KNO; and NaNO,/NaNO3 would, therefore, be greater than unity. Three equilibria to consider are: (i) KNO3(2) Z KNO2 (L) + 1/2 02(g), AG%(kcal) = 27.6 - 24,4(T/1000) Ref. - R. F. Bartholomew, J. Phys. Chem. 70, 3442 (1966). (ii) NaNO3 (%) Z NaNO2(2) + 1/2 02(g), AGé(kcal) = 24,0 - 24,0(T/1000) Ref. - E. 5. Freeman, J. Phys. Chem. 60, 1487 (1956). Freeman's data reworked by L. E, Gastwirt and E. F. Johnson, MATT-98, Aug. 1961, p. 157. (iii) Ha(g) + 1/2 02(g) < H20(g), AG%(kcal) = -58.9 + 13.1(T/1000) Ref. - O. Kubaschewski, E. L. Evans, and C. B. Alcock, Metallurgical Thermochemistry, 4th Edition (1967), p. 424, Combining (i) and (iii), (iv) H, + KNO3 Z KNO. + H,0, AG® (keal) = -31.3 - 11.3(T/1000) p a at 873°K, K = 2.06 x 10'0 = —20 KNO2 Py, ZkNo, xNo, %No, = O+1s them py =5 x 10712 Py,0" Assuming that Py, should be less than 10™'% atm, then szo.s .02 atm is tolerable. Guessing that a UcCcN-430 L3 B-61) 36 W. R, Grimes 2 November 4, 1970 Similarly when we combine (ii) and (iii), (v) Hz + NaNO3 T NaNO2 + H,0, AG®°(kcal) = =-34.9 - 10.9(T/1000) P also at 873°K, K = 1.30 x 10!! = "Hz0 “NaNop Py, ®NaNOs Now assuming that aNaNoz/aNaNog = 7 (the mole fraction ratio when HTS is initially constituted), then _ ~11 Py, = 5.4 x 10 pH20 Thus at 873°K, equilibrium (v) rather than (iv) fixes the tolerable Py,0° Again, supposing that sz should be less than 10~'? atm, then szO should be maintained at less than 2 x 10~® atm, Finally, please note that the equilibrium constant for reaction (v) increases with decreasing temperature. For instance, at 800°K, the equilibrium constant = 8.25 x 1011, which for 2 aN /aNaNos = 7, means that the tolerable Pu,0 is a factor of 6 higher than at g%3°K. S. Cantor SC:mp cer E. G. Bohlmann 37 APPENDIX B Reactions of HITEC with Graphite and with 72~16-12 (1iF-BeF;-ThF4) Salt C. E. Bamberger and R. G. Ross™ A batch of 50 g of HITEC was prepared by melting the analytical grade components (NaNO,, NaNOi, KNO3) in a quartz tube under argon. After melting, argon was bubbled through the liquid for 15 minutes (in order to remove any significant amount of moisture); the melt was cooled under flowing argon and stored in a dessicator. In the first experiment a 10 g portion of HITEC was melted in an open quartz tube giving a liquid layer of 3/4" depth. A piece of 4.5 g of "MSRE" graphite hanging from a platinum wire was dipped briefly (30 to 120 sec) into the melt at successive increasing temperatures and the visible signs of reaction were noted. Following is a listing of the observations. It made no difference whether the quartz tube was open to the atmosphere or was blanketed by flowing argon. No graphite introduced 340°C: Very small bubbles formed in the bulk of the melt which on raising did not seem to disturb the surface of the liquid. The bubbles were so small that they were observed only by means of a catheto- meter. Graphite in the Melt 340°C: Large bubbles formed at the interface. 440°C: TFaster evolution of large bubbles at the interface, the melt remained clear yellow. 490°C: The melt became black 10-20 sec. after the graphite was introduced. The formation of bubbles was more vigorous than before, After removing the graphite the melt cleared up in ca. 5 minutes, %* Private communication - MSR 72-5, February 2, 1972. 38 545°C: The melt became black again, then the melt started foaming vigorously and after approximately 45 seconds the graphite piece became suddenly incandescent. At this point a sharp rise of tem- perature was recorded. The graphite was removed and after approxi- mately 10 minutes the melt cleared up again. The experiment was terminated. The experiment was repeated and the reactions were recorded on 16 mm movie film, the only difference being that the melt was heated rapidly to 440°C and then to 550°C. The observations were the same as above but the time required for the incandescence of the graphite was ca. 2 minutes. After removal and cooling the graphite used in both experiments was washed and examined visually; it revealed a high degree of pitting and enhancement of the cracks that were initially present. In a third experiment, approximately 9 g each of HITEC and 72-16-12 (LiF-BeF,~-ThF,) salt were loaded into a quartz tube (the HITEC at the bottom) and were gradually heated. At 250°C the molten HITEC bubbled at the interface with the lumps of fluoride salt and became cloudy-white, then completely opaque. In the range 350-410°C a large amount of yellow- brownish gas was evolved. The lumps of salt were no longer visible. At 430°C vigorous foaming occurred which displaced liquid and suspended material upwards. The experiment was terminated. Upon cooling, the solids were digested in hot water and repeatedly washed prior to analyzing the residue by x~-ray diffraction. The results indicated the presence of ThO, (major phase), BeO and LiF. (A few unidentified lines were also present.) This could be interpreted to be the result of reactions such as ThF4(d) + 2NO3™ + 2NO2~ z ThOz(c) + 4F + 2N,04(g) BeFa(d) + NO3~ + NO,~ 2 BeO(c) + 2F + N204(g) (or 2NOz(g)) From what is known of the oxide chemistry of U(IV) there is no doubt that U0, would be the first insoluble phase to form if fuel salt is contacted by HITEC, 10. 11. 12. 13, 14. 39 References H. P. Voznick and V. W. Uhl, "Molten Salt for Heat Transfer,” Chem. Engrg. 70, 129 (May 27, 1963). Arminio Conte and Serge Casadio, '"Potentiodynamic Study of the Behavior of Various Metals in Molten Alkali Nitrates," Ricera Scientifica, 36(6), 433 (1966). P. A. Rottenberg, '""Heat Transfer Media for Use at Elevated Temperatures,' Trans. Instn. Chem. Engrs. 35, 25 (1957). R. H. Newton and H. G. Shimp, Trans. Am. Inst. Chem. Engrs., 41, 197 (1945). R. C. Robertson, "Estimated Cost of Adding a Third Salt-Circulating System for Controlling Tritium Migration in the 1000 MW(e) MSBR," ORNL-TM-3428 (July 1971). W. E. Kirst, M. W. Nagle, and J. B. Castner, "A New Heat Transfer Medium for High Temperatures,’ Trans. Am. Inst. Chem. Engrs. 36, 371 (1940). Alfred Bruhn, ''Capital Costs Trimmed by 207 by Heat Transfer Process,' Chemical Processing 33(8), 15-30 (Aug. 1970). J. Friedman and A. A. Bruhn, ''Diluted Salt Transfers Heat," 0il and Gas J. 68, 76 (1970). Private communication available from the author or American Hydrotherm, See Ref. l4. "U.S. Heat Transfer Systems Get a European Challenger," Chem. Engrg. 77, 26 (Nev. 30, 1970). J. Alexander, Jr., and S. G. Hindin, ""Phase Relations in Heat Transfer Salt Systems,” Ind. and Engrg. Chem. 39, 1044 (Aug. 1947). Paul Larson (phone conversations), E. I. duPont de Nemours and Co., Explosive Dept., Res. & Devel. Div., Technical Service Section, Eastern Laboratory, Gibbstown, New Jersey 08027. "DuPont Hitec Heat Transfer Salt," E. I. duPont de Nemours and Co., Explosives Dept., Chem. Products Sales Div., Wilmington, Del. 19898. "Hydrotherm Molten Salt Heat Transfer System," American Hydrotherm Corp., 470 Park Ave. South, New York, N. Y. 10016. 15. 16. 17. 18. 19. 20. 21, 22, 23. 24, 25. 26. 270 40 V. W. Uhl and H. P. Voznick, "High Temperature Heating Media," Chem. Engrg. Progress 59(5), 33 (May 1963). J. L. Crowley and R. E. Helms, "Visit to Hooker Chem. Corp. of Niagara Concerning Their Use of "Hitec'" (HTS) Salt,'" MSR-70-94, Dec. 2, 1970. R. E. Helms, "E. I. duPont de Nemours and Company and Their Use of "Hitec" (HTS) Salt," MSR 71-24, March 1, 1971. W.-H. v.d. Broek, private communication to J. R. McWherter, paper presented at Royal Netherlands Institution of Engineers on operating experience with HTS by W. H. v.d. Broek. J. M. Michel, U.S. Patents: 2,166,364; 2,196,153; 2,245,155, K. Leschewski and W. Degenhard, '"The Chemical Changes of Molten Alkali Nitrates at Temperatures Between 460° and 600°C," Berichte der Deutsche Chemie Geselschaft, 72B, 1763 (1939). E. S. Freeman, ''The Kinetics of the Thermal Decomposition of Potassium Nitrate and of the Reaction Between Potassium Nitrite and Oxygen,'" J. Am. Chem. Soc. 79, 838 (1957). E. I. Gurovich and G. P. Shtokman, ''Corrosion Activity of a Nitrate-Nitrite Molten Mixture," Zh. Prikl. Khim. 30, 1547 (1957). N. K. Voskresenskaya and S. I, Berul', "Thermal Stability of a Low-Melting Mixture of Nitrites and Nitrates of Sodium and Potassium,'" Zhurnal Neorganicheskoi Khimii, Vol. I, No. 8, 1956, pp. 1867-1876. Larry L. Anderson and George Richard Hill, '"Corrosion of Zirconium and Zircaloy-2 in Fused KNO;-NaN0;~NaNOj,'" Electrochem. Technology 4(3-4), 157-162 (1966). Eli S. Freeman, ''The Kinetics of the Thermal Decomposition of Sodium Nitrate and of the Reaction Between Sodium Nitrite and Oxygen,' Thermal Decomposition of Sodium Nitrate 60, 1487 (1956). C. C. Van Soye, 'Concentrator Breaks Heat Barrier," Chem. Engrg. 66, 98 (Sept. 21, 1959). Private communication: Joel Friedman, American Hydrotherm Corp., 470 Park Ave. South, New York, N. Y. 10016. 0 = r*c:;*nzgc?=?=pig-z:b)wnn CLUHGGEmOMOO0ORI R IHBTILORI UG EOREON DA - o o B W HE O a Affel Bamberger Bauman Beall Beatty Bender E. S. E. F. 0. Bettis Bettis Blanco Blankenship Blomeke Blumberg GQ E. Bohlmann (20) Boyd Braunstein Ae B. R. E. Bredig Briggs Bronstein Brooksbank Cantor L, < < o o mmpdo:c:pr*m*d:Zm?dhdm?upd:::c..r*r*z:mr*zmuw Carter . Cavin . Claffey Clark Collins Compere Cook . Cooke Crowley Culler . DeBakker . DeVan DiStefano Ditto Doss . Dworkin Eatherly Engel Ferguson Ferris Fraas . Frye Fuller Fuller Gabbard Gilpatrick Grimes Grindell Guymon 41 Internal Distribution R . » » a & o o » < [] B T W D B C o * o @ a & e ® o . a a * e °® @ e e o ® * o o 9 1] o - P P J E H D P W p R C J S J J A T J K J K. R. E M R H R W C H H D C H L A G H A R F J @ e J m;»»garfluiniriminifiaEzhir*usz:?-g T oHOPOU G OddR SO R . Hannaford Harley Harms Haubenreich Helms Hembree Hewett Hightower Hise Hitch Hoffman . Holmes Holz Huntley Kasten Kedl Kee Kennedy Keyes . Kirslis . Klein Koger Krakoviak . Kress . Krewson it Laughon, AEC-0SR Lenhard Lin . Lindauer Long Lundin Lyon MacPherson MacPherson Martin Matthews, AEC-0SR McCoy McDuffie . McElroy McHargue McLain . McNeese . Malinauskas Mamantov HHEHZ20m > - - < o .« o o . @ H. Neill P. Nichols 42 E. L. Nicholson G. P. Smith S. M, Ohr 0. L. Smith P. Patriarca I, Spiewak A. M., Perry R. A. Strehlow T. W. Pickel J. R. Tallackson H, B. Piper 0. K. Tallent C. B. Pollock R. E. Thoma B. E. Prince L. M. Toth A. S. Quist D. B. Trauger G. L. Ragan W. E. Unger G. D. Robbins C. F. Weaver R. C. Robertson A. M., Weinberg M. W. Rosenthal (3) J. R. Weir R. G. Ross W. J. Werner J. P. Sanders M. E. Whatley H. C. Savage J. C. White W. F. Schaffer R. P. Wichner Dunlap Scott W. M. Woods H. E. Seagren H., C. Young R. L. Senn J. P. Young J. H. Shaffer E. L. Youngblood W. H. Sides Central Research Library (2) M. J. Skinner ' Document Reference Section A, N. Smith Laboratory Records (3) Laboratory Records (LRD-RC) Division of Technical Information Extension (2 ) Research and Development Division, ORO (1) External Distribution A. A. Bruhn, American Hydrotherm Corp., 470 Park Ave., S., New York, N. Y. 10016. Joel Friedman, American Hydrotherm Corp., 470 Park Ave., S., New York, N. Y. 1001s. P. A. Larson, E. I. duPont de Nemours & Co., Eastern Lab., Paulsboro, New Jersey. A. B. Palmer, E. I. duPont de Nemours & Co., Wilmington, Delaware. M. Shaw, AEC-Washington. MSRP Program Manager, AEC-Washington.