L~ DATE: SUBJECT: TO: FROM: OAK RIDGE NATIONAL LABORATORY FTOR INTERNAI USE ONLY OPERATED BY UNION CARBIDE CORPORATION NUCLEAR DIVISION 0 R N L Lol UNI CARBIDE | CENTRAL FILES NUMBER OAK RIDGE, TENNESSEE 37830 7L - 7 - 38 July 7, 1971 COPY NO. Additional Calculations of the Distribution of Tritium in the MSRE Distribution R. B. Briggs ABSTRACT Some of the calculations reported in CF 70-T7-13 were repeated, taking into account recent information on the solubility of hydrogen in molten salts and the sorption of tritium by graphite. Reasonable agreement was obtained between the measured and calculated distri- butions of tritium in the MSRE. Additional experimental data are needed to reduce the uncertainties in the calculations. Key Words: tritium, MSRE, fused salts, reactors, operation. NOTICE This document contains information of a preliminary nature and was ogepared primarily for internal use at the Oak Ridge Nationa? +-8bratory. It is subject to revision or correction and therefore does not represent a final report. The information is only for official use and no release to the public shall be made without the approval of the Legal and Information Control Depart- ment of Union Carbide Corporation, Nuclear Division. ADDITIONAL CALCULATIONS OF THE DISTRIBUTION OF TRITIUM IN THE MSRE Results of calculations of the tritium distribution in the MSRE and discussion of the methods used in the calculations were reported in CF 70-7-13, "Calculation of the Tritium Distribution in the MSRE." Since the time of publication of those results, several changes have occurred: 1. Measurements by Malinauskas and Savolainen have indicated that the solubility of hydrogen in molten salt is about 1/3 the values used previously. 2. Only a small amount of lithium was found by chemical analyses of samples of insulation from the MSRE reactor furnace, so we conclude that the tritium in the reactor cell was produced in the fuel salt and diffused through the metal walls of the reactor system into the cell. | 3. The rate of production of tritium in the fuel salt during the time that the tritium distribution was being measured is now estimated to be 54 curies per day. L, Tritium was found in graphite removed from the reactor core in a quantity equivalent to a deposition rate of about 8 curies per day. Ttems 2-4 are discussed by P. N. Haubenreich in a memorandum now in prepa- ration. The effect on the distribution of tritium of hydrogen solubility and of sorption of tritium by graphite were considered in the previous memo- randum. The calculation of the solubility effect was, however, not entirely correct (reducing the solubility causes more tritium to enter the off-gas than was reported), and sorption by graphite was considered in only two cases. Because of these differences, it seemed desirable to make some additional calculations. They were made and the results are summarized in Table 1. | Values of the reference parameters in the equations that describe the tritium distribution were listed in Appendix A of CF T70-7-13. The same values were used in these calculations except that k,, the solu- bility coefficient for T, in fuel salt, was reduced from 0.06 to 0.02, and kn, the solubility coefficient for T, in coolant salt, was reduced from 0.04 to 0.02. The complete set of calculations involves cases for the following conditions. I. With UF,/UF; = 1000 A. Reference condition without graphite, with graphite, and with graphite and hydrogen. Table 1. Summary of Results of Calculations of Tritium Distribution in MSRE Condition Tritium Distribution - Percent of Production Concentrations, molecules/cmaxIO"l Mass Transfer Metal Coolant Fuel Pump Off-Ges Graphite Case Coefficient Permeability Cooling Coolant Pump Reactor T; in T, in TF in No. UF,/UF, (x Ref.) (x Ref.) Hydrogen Air Cell Off-Gas Cell T, TF Total T, TF Total Fuel Salt Coolant Salt Fuel Salt 1 1000 1 1 8 2 0.1 17 b1 31 72 0O O 0 6 0.9 130 2 2 0.k 0 3 8 1k 22 3 70 73 1 0.2 57 3 8 2 0.1 15 37 L Iy 13 20 33 310 50 9Lo 5 0.01 T 2 0.2 11 41 L b5 14 21 35 340 83 990 7 001 2 0.7 0.8 3 8 14 22 3 70 173 420 320 1100 8 0.5 1 5 1 0.2 10 b9 34 83 0O O 0 7 1 140 9 2 0.4 0.1 L 18 20 38 3 53 56 3 0.4 85 10 6 1 0.2 11 55 5 60 10 12 22 k70 T4 1100 12 0.01 5 1 0.2 10 5T 5 62 10 12 22 480 91 1200 14 - 0.001 2 0.7 0.8 3 65 5 70 12 13 25 550 350 1200 15 100 1 1 12 3 0.2 2l 5T L 61 O O 0 8 1 15 16 8 2 0.1 17 ko 3 43 14 16 30 6 0.9 13 17 10 2 0.2 20 L8 O.4 48 17 2 19 400 65 110 19 0.01 9 3 0.3 14 53 0.5 sk 19 2 21 450 120 110 20 0.001 7 2 0.3 11 Ly 3 L8 16 17 32 6 2 13 21 * 3 0.8 1 3 66 0.5 66 23 3 26 560 Lho 130 22 0.5 7 2 0.2 15 T2 L 76 0O O 0 10 2 17 23 6 1 0.2 12 58 4 61 10 10 20 8 1 15 24 7 2 0.2 13 65 0.5 65 11 1 13 550 86 120 26 0.01 6 2 0.3 11 67 0.5 68 12 1 13 570 110 130 27 0.001 6 2 0.3 9 60 L 64 11 10 20 9 2 16 28 * 2 0.7 1 3 T 0.6 T 1k 1 15 650 420 140 Measured distribution 59 - 1 5=9 46~56 15 *Indicates hydrogen added to salt at rate of 60 times the production of tritium. Addition rate is approximately that which would be produced by com- plete decomposition of 0.5 g/day of oil in pump bowl. B. Permeability of metal reduced by factor of 100 with graphite and with graphite and hydrogen. C. Permeability of metal reduced by factor of 1000 with graph- ite and with graphite and hydrogen. IT. Repeat I with all mass transfer coefficients reduced by factor of 2. IIT. Repeat I and II with UF,/UF; = 100. Not all the results are reported in Table 1. The salt is the major barrier to the transport of tritium and reducing the permeability of the metal by a factor as large as 1000 had no significant effect in some cases, so the results are not included in the table. For cases with hydrogen, the hydrogen was added to the fuel salt at a rate of 3 x 1017 molecules/sec. This is 60 times the rate of tritium production and results in a concentration of hydrogen in the salt that is about what should be obtained from complete decomposition of 0.5 g per day of oil in the pump bowl. The measured distribution is also given in Table 1 and does not account for 10-28% of the production of tritium. Some of this tritium was dissolved in the metal, some of 1t was held in deposits in the reactor system, but I believe that most of it must have left the fuel pump bowl in the off-gas. The percentages assigned to the cooling air, to the reactor cell, and to sorption by the graphite could not be too low by such a large amount. In my consideration of the data I assign 66 to Th% of the tritium to the fuel pump off-gas. Examination of the data in Table 1 leads me to conclude that case 27 is in best agreement with the measured distribution. In this case, UFs/UF, = 100, the mass transfer coefficients were reduced by a factor of 2, and the permeagbility of the metal was reduced by a factor of 1000 from the reference values. The uncertainty in the mass transfer coeffi- cients is at least a factor of 2. Oxide on the metal surfaces or a change from Q proportional to p%/a to Q proportional to p at very low pressure might produce such a reduction in the effective permeability of the metal. In case 27 the amount of tritium sorbed by the graphite is too high, but T; or TF might not be sorbed as efficiently as was assumed in the calculations. The distributions in other cases, such as 10, 12, 16, in which the hydrogen has an effect, agree about as well with the measurements except that the flow rate to the reactor cell is too high, and that measurement seems to me to be our most reliable one. Two factors deserve additional attention: 1. The distribution is sensitive to the values assigned to the mass transfer coefficients. In all calculations to date, we have changed all mass transfer coefficients simultaneously and by the same multi- plier, on the basis that an uncertainty in one parameter in the cal- culation would apply to all the mass transfer coefficients. It appears that the distribution could be shifted considerably by adjusting individual mass transfer coefficients. A more careful analysis of the mass transfer coefficients in each of the regions might justify changing only some of the coefficilents. 2. The assumption that graphite retains both tritium and tritium fluo- ride might not be correct and could markedly affect the amount held by the graphite. Experimental data are needed to confirm this assumption or to provide the basis for a better one. The conclusions and recommendations of CF 70-7-13 are not changed much by the more recent data and the calculations reported here. The graphite did indeed prove to be a reservoir of tritium and to have a greater capacity than had been anticipated. Preliminary measurements of the solubility of hydrogen in salt and analyses of the lithium in the insulation in the reactor furnace have reduced some of the uncer- tainties in the calculations and in the measured distribution. In cal- culations that give the best agreement with the measured distribution, the permeability of the metal still seems to be much lower than one might expect, although the same effect might be obtained by adjusting the mass transfer coefficients. More of the experimental data outlined in CF T70-7-13 are needed to provide an adequate understanding of the behavior of tritium in molten-salt reactors. \O 003 Oyt W O H N RN USSR YN PN NP NI NG NI NI ENP Q0N IO NERQQHE D * P EEEE QD EE NN TR E PSRN O TnEcEHE CEENeENOpEEEC Y a n Affel Apple . Anderson Baes Bamberger Beall nder Bettis Billington Blankenship Bohlmann Borkowski Boyd Briggs ntor Carter Chapman Clark Compere Cook Cope, RDT-SSR Crowley Culler Dale Ditto kBatherly Engel Ferguson Ferris Fraas Franzreb Frye Furlong Gabbard Gibbons Grimes Grindell Guymon Harley Haubenreich Helms Huntley ouye Jordan Kaplan Kasten Kedl Kelley Kirslis Koger DISTRIBUTION 57 . 58. 59. 60. 61. 109. 110-112. 113. PoENrEDIEEED PR RS R R. B. Korsmeyer A. I. Krakoviak T. S. Kress Kermit Laughon, RDT-OSR Lundin Malinauskas MacPherson MacPherson McCoy McIntosh, AEC, Wash. McLain McNeese McWherter Meyer Moore Nicholson Perry Piper Prince Richardson 1chardson Robertson Rosenthal Roth, AEC-ORO Savage . Savolainen nlap Scott H. Shaffer Shaw, RDT, Wash. Silverman Skinner Smalley, AEC-ORO Smith Strehlow Struxness Tallackson Taylor Thoma Trauger Watson, AEC, Wash. Weinberg Weir Whatley . White . . Wichner L. V. Wilson Gale Young Central Research Library Document Reference Section Laboratory Records Laboratory Records - RC HEREECRAIEP S HEQ T GUEE DR EE P EE S E EwE HOEE 2 TOHEEREEOEQ RN SO