% % gact ar M fi. 1 4 g 1% a0 Py Carbon-14 [ Hoa o WY R - - Pty s, T - 73.‘{'; =, S S A okt fi o AR S S e SR B fams = i o s ”,“.ennfl..”m.uff..m.,f..mn iRt S bbb b A s b P R TR > AR R st 0 i Frinied n the Uimted Statas of America Availdable from | Nationaz! Tachnical Infonmation Servics : Lon epartient of Commerce ; 5285 Moy 2'?;.;3-31?.! oL hpr ingfield, Virguma 22163 ; Sricoe Pronied O Gy "sd On- M{CerlCll ?\q a0 | A Admunistration/United States Nuciear Bogulatory Comimission, nor any of e emiphoyees, noi anv of thelr cortiaciors, subcontractors, or their employges, makes any warianty, express oramphed. Or assumoes any 'egalhability Grresponainiity forthe accuracy, coripletenass or usefulness of any information, apparatus, prodoct or process disclosed, or represents that ifs wasasuid not um.ug" “'wn*o'y Owed :igu s ‘ Governmant. Neiihis: e United Sxaxm nor thy '{:.‘“gy Sesearah ana uwvelopn | ‘L______‘_mv_fl‘? o —Z ORNL/NUREG/TM-12 Dist. Category UC-11 Contract No. W-7405-eng-26 CHEMICAL TECHNOLOGY DIVISION CARBON-14 PRODUCTION IN NUCLEAR REACTORS Wallace Davis, Jr. Manuscript Completed: January 1977 Date Published: February 1977 Prepared for the U.S. Nuclear Regulatory Commission Office of Nuclear Material Safely & Safeguards Under Interagency Agreement ERDA 40-549-75 Prepared by the OAK RIDGE NATIONAL LABORATORY Oak Ridge, Tennessee 37830 operated by UNION CARBIDE CORPORATION for the ENERGY RESEARCH AND DEVELOPMENT ADMINISTRATION CARBON-14 PRODUCTION IN NUCLEAR REACTORS W, Davis, Jr. ABSTRACT Quantities of "*C that may be formed in the fuel and core structural materials of light-water~cooled reactors {(L.WRs), in high-temperature gas-cooled reactors (HTGRs), and in liquid-metal-cooled fast breeder reactors (LMEFBRs) have been calculated by use of the ORIGEN code.' Information supplied by five L. WR-fuel manufacturers pertaining to nitride nitrogen and’ gaseous nitrogen in their fuels and fuel-rod void spaces was used in these calculations. Average nitride nitrogen values range from 3 to 50 ppm (by weight) in LWR fuels, whereas gaseous nitrogen in one case is equivalent to an additional 10 to 16 ppm. Nitride nitrogen concentrations in fast-flux test facility (FFTF) fuels are 10 to 20 ppm. The principal reactions that produce “C involve N, Y0, and (in the HTGR) "C. Reference reactor burnups are 27,500 MWd per metric ton of uranium (MTU) for boiling water reactors (BWRs), 33,000 MWd for pressunized water reactors (PWRs), about 95,000 MWd per metric ton of heavy metal (MTHM) for HTGRs, and 24,800 MWd/MTHM for an LMFBR with nuclear parameters that pertain to the Clinch River Breeder Reactor. Nitride nitrogen, at 3 median concentration of 25 ppm, contributes 4, 15, and 6 Ci of "C/GW{c)yr to BWR, PWR, and LMFBR fuels. respectively. The contribution of 'O in BWR and PWR fuels is 3.3 and 3.5 Ci of "CIGW(e)-yr, respectively, but it is Jess than 0.2 Ci/ GW(e)-yr, in blended LMFBR fuel. In the HTGR fuel particles (UC: or ThO.), 10 Ci of "C) GW(e)-yr will be formed from 25 ppm of nitrogen, whereas 'O in the Th{: will contribute an additional 2 Ci/GW(e)yr. Ali 'C contained in the fuels may be released in a gas mixture (CO,, CO, CHa, ete)) during fuel dissolution at the fuel reprocessing plants. However, some small fraction may remain in aqueous raffinates and will not be released until these are converted to solids. The gases would be released from the plant unless special equipment is installed to retain the "“C-bearing gases. Cladding metals and other core hardware will contain significant quantities of ", Very little of this will be released from BWR, PWR, and LMFBR hardware at fuel reprocessing plants; instead, the contained B30 to 60 Ci/GW(e)-yr for LWRs and about 13 Cif GW(e}-yr for a CRBR, will remain within the metal, which will be retained on site or in a Federal repository. The only core structural material of HTGRs will be graphite, which will contain 37 to 190 Ci of "/ GW{e)-yr, exclusive of that in the fuel particles, if the graphite (fuel block and reflector block) initially contains 0 to 30 ppm of nitrogen. All of this is available for release at a fuel reprocessing plant if the graphite is burned to release the fuel particles for further processing. Special equipment could be installed to retain the "“C-bearing gases. 1.0 INTRODUCTION The radioactive nuclide '“C is, and will be, formed in all nuclear reactors due to absorption of neutrons by carbon, nitrogen, or oxygen. These may be present as components of the fuel, moderator, or structural hardware, or they may be present as impuirities. Most of the “C formed in the fuels or in the graphite of HTGRs will be converted to a gaseous form at the fuel reprocessing plant, primarily as carbon dioxide; this will be released to the environment unless special equipment is installed to collect it and convert it to a solid for essentially permanent storage. If the "“C is released as carbon dioxide or in any other chemical form, it will enter the biosphere, be inhaled or ingested as food by nearly all living organisms including man, and will thus contribute to the radiation burden of these organisms. Carbon-14 is formed naturally by reaction of neutrons of cosmic ray origin in the upper atmosphere with nitrogen and, to a lesser extent, with oxygen and carbon. Large amounts of '“C have also been formed in the atmosphere as a result of nuclcar weapons explosions. For the last two decades, the quantities of '*C in the environment, and the mechanisms of transfer of this nuclide between the atmosphere, land biota, and the shallow and deep seas have been the subject of many research studies.”” These studies have shown that most of the "*C is actually contained in the deep oceans, at depths greater than 100 m. The nuclear weapons tests increased the total '*C inventory of the earth by only a few percent,’ but the atmospheric content was approximately doubled. Since atmospheric weapons tests are no longer being conducted, the atmospheric concentration of '*C is now decreasing as it enters the oceans as CO; and is approaching the pretest value, Some estimates of the amounts of "“C released from or formed in LWRs, " HTGR,"""* and LMFBR" have been made previously on the basis of calculations or measurements. The purpose of this report is to present detailed estimates of the production of '“C with emphasis on those pathways that are likely to lead to the release of this nuclide, cither at the reactor site or at the fuel reprocessing plant. 2.0 MECHANISMS OF CARBON-14 FORMATION IN NUCLEAR REACTORS Carbon-i4 1s formed from five reactions of neutrons with isotopes of elements that are normal or impurity components of fuel, structural materials, and the cooling water of LWRs, The cutron-induced reactions are as follows: I (1) "C(n,y"C; (2} "N@,p)"C; 3) "N@.d)"C; (4) "O(n,'He)"'C; (5) "Om,a} C. In these reactions, standard notation has been used in which n refers to a neutron, p to a proton, d to a deuteron ("H), and v to a gamma ray. Reactions 4 and 5 will occur in any reactor containing heavy-metal oxide fuels and/or water as the coolant. Reaction | will be important only in the HTGRs, while reactions 2 and 3 will occur in all reactors containing nitrogen as an impurity in the fuel, coolant, or structural materials. To facilitate calculations, the energy-dependent cross sections of nuclear reactions are typically collapsed into a single, effective cross section that applies 1o the neutron spectrum of the reactor in guestion. Such collapsed values are known with fairly good accuracies for reactions 1, 2, and 5 for the thermal-neutron spectra of LWRs and HTGRs. Values listed in Table | for the BWR, PWR, and HTGR are taken from the ORIGEN library' and its update'® according to the latest version of the “Barn Book.”!” Because reactions 3 and 4 are highly endothermic, their cross sections are assumed to be 0.0 in thermal reactors, as shown in Table I. Unfortunately, some of these cross sections for the LMFBR are very uncertain. The following discussion concerning cross sections of reactions 1-5, as they apply to the Clinch River Breeder Reactor (CRBR), has been provided by A. G. Croff.™ Reaction | " Cn,v)"C The cross section for this reaction is not well known for nonthermal neutron energies. The assumed values were taken from ref. 19, in which the *C(n,y) cross section was calculated on the bases of a few experimental data and nuclear systematics. The cross section obtained when the data are collapsed to a single value using the CRBR neutron spectrum 15 0.5 ub (I ub = (0™ barns). The fact that the thermal "C(n,¥) cross section is only about 1 mb (Table 1) couplied with the fact that cross sections in the nonthermal energy regions are considerably smaller than thermal cross sections tends to confirm that the 0.5 ub value is realistic. Reaction 2 “N(n,p)"'C Of the five ""C-producing reactions listed, this is the only one for which the experimental data may be considered adequate. Energy dependent cross-section data for the *N(n,p)"*C reaction are available from the ENDF/B"® compilation. Collapsing these data with the CRBR spectrum gives a cross section of 12.6 mb, with an estimated error of +30%. Heaction 3 lS!\/’(n,af}MC The only cross-section data available for this reaction are some sketchy information on the angular distribution of the deuterons when the neutrons have energies of 14 to 15 MeV., This information, coupled with the fact that the reaction is endothermic (Q = -7.99 MeV), would probably lead to a value of the reaction rate in the 0.00 to 0.1 mb range. However, for calculational purposes, a value of 1.0 mb was used. Reaction 4 'O, He)"' C Of the five reactions considered, the data for this reaction are by far the least well-known. It is highly endothermic (Q = -14.6 MeV), indicating that greater neutron energies are required for the teble 1. Cross sections for formation and yields of *¢ in BWR, PWR, HTGR, and LMFER® 140 Tormation Reaction Cross section for formation of 14¢ in (curiesAper gram of parent element) No. Reaction BWR PWR HTGR IMFBR BWR PAR HTGR IMFER 1 Y20(n,v) e 1.00 mp 1.00 mb 0.416 mb 0.5 ub 1.51E-7 1.618-7 3.388-7 4. 81E-9 (3.69E+0) 2 N (n,p)*4C .48 v 1.48 b 1.02 12.6 mb 1.718-2 1.83E-2 3.84E-2 9.668-3 3 *Bn(n,a)t%c 0 0 0 1.0 mb 0 0 0 2.855.6 4 180(n,%He )2 4 0 0 0 0.05 b 0 0 0 3.82E-8 (4.53E-2)° 5 Y70(n,*Re ) % 0.183 0.183 v 0.110 b 0.12 mb 7-31E-T7 0 T.75E-7 . 1.79E-6 3. 4oE-8 (1.01E-1)" (0.878-2)" (2.25p-1)¢ (4.03E-2)C aAll of tne valuves in this table woere obtained by collapsing available neutron cross-section data to a single value, using neurvon spectra of the individual reactors, as discussed by 8el1.1 These values are mnot ecual to 2200-m/sec cross sections, such asg 0.9 mb, 1.81 b, and 0.235 b for reactions 1, 2, and 5, respectiveiy. b . . . , . Based on 10.93 MT of carbon/MTEM where HM = thorium pius uranium, CBased on 8383 g-at. of oxygen/MTHM where EM = uranium or uranium plus plutonium, present as UO2 and Pu0 & 2 - Based on 2.9094 MT of thorium/MTHM with +thorium oresent as ThO2 and uranium as UC. reaction to proceed. Information supplied by the Physics Division of Lawrence Livermore Laboratory indicates that the cross section at 15 MeV should be less than | mb, and at 20 MeV it should be less than 10 mb. By combining these “guesstimates™ with the CRBR spectrum and a theoretical expression for the availability of high-energy fission neutrons, the reaction cross section is estimated to be about 0.05 ub. The lack of information on both the high-energy cross sections and the high-energy neutron spectrum makes this value very uncertain, Reaction 5 " Ofn,a)"*C As with reaction 1, the cross-section data for this reaction are not well known. The data, which again are based on only a few experiments and nuclear systematics, were taken from ref. 19. The cross section, which is calculated and based on the CRBR spectrum, is 0.12 mb. The assumed LMFBR fuel model was the Atomics International Follow-On Design. Initial concentrations of the isotopes of importance in this case (in g-atoms/ MTHM) are: 12 ’C 33.33 "¢ 0.374 "N 1.42 "N 0.00528 "*Q 8383. "0 3.27 0 17.2 The ORIGEN code' is not capable of explicitly accounting for (n,d) or (n,'He) reactions. This difficulty may be circumvented by combining reaction 4 with reaction 5 and reaction 3 with reaction 2, since the naturally occurring isctopes are present in a fixed ratio for each element. Alternatively, since the depletion of the carbon, nitrogen, and oxygen is relatively small (<<2%j). the calculation is easily performed by hand. 3.0 CARBON-14 FORMATION IN LIGHT-WATER REACTORS Carbon-14 is formed in the fuel (UQ2), in core structural materials, and in the cooling water of L.WRs. 3.1 Formation in the Fuel Carbon-14 will be formed primarily by two reactions in the fuel: "O(n,a)"'C and “N(n,p)"'C. The quantity of "*C formed from the first of these reactions can be calculated accurately on the basis of the stoichiometry of UQ: (134.5 kg O/MTU) and an abundance of 0.039 at. 9% 'O in normal oxygen, which corresponds with 55.6 g of "O/MTU or 3.27 g-atoms of '70/‘MTU. As listed 1in Table 2, burnup of BWR and PWR fuels to 27.500 and 33,000 MW(t)d/MTU, respectively, leads to the formation of 0.098 and 0.104 Ci of *C/MTU, which corresponds with 3.3 and 3.5 Cif GW{e)-yr, respectively. Table 2. Production of '*C in core hardware and Tuel at light-water reactors (BWR and PWR) 14 C existing 150 days after Total '*C production Qua:;;ity Quantity of element in core discharge of fuel (Ci/MTU) core {g/M70) From From From Calcuiated Observed Material {kg/MTU) Carbon Nitrogen Oxygen carbon nitrogen oXygen Ci/MTU cifcvi(e)-yr® Ci/GW{e)-yr Soiling-Water Reactorb Zircaloy-2 {Grade RA=1} 316 £85.3 $25.3 1.29E-5 4.338-1 0.k33 14,5 30k stainless steel 50 <40.0 50-80 0.60E-5 (0.86-1.37)E+0 0.86-1.137 28.7-45.9 Inconel-X 3.4 <34 0.058-5 0., 000 2.0 Uranium dioxide 1135 low 10 134,500 1.718-1 9.83E-2 0.269 9.0 Med 25 L.28E-1 0.526 17.6 High 79 1.288+0 1.38 6.3 ater 216 192,000 1.408-1 0,140 L.7 &° Totals, Low 1.70 57 Med a.21 7h High 3.32 111 Pressurized-Water Reactox‘ti Zircaloy-i {Grade RA-2) 2135 <53%.5 £18.8 1.02E-5 2.74E-1 0.274 9.5 302 stainiess steel L.2 <3, L.2-6.7 0.05E-5 (0.61-0.98)8-1 0.061-0,098 2.1-3.4 304 stainless steel 7.1 £29.7 37.1-59.4 0. %8E-5 (5.42-8.671E-1 0.5k2-0, B6T 18.8-30.0 Inconel 716 12.8 1.3 C.02E-5 0. 000 0.0 Microbraze 50 2.6 0.3 c.2 1.1 0.00E~5 3,66E-3 0.858-6 0.004 0.12 Uranium dioxide 1135 tow 10 134,500 1,938-1 1,041 0.287 9.6 Med 25 4, 57E-1 0.561 i8.8 High 79 1. 37E+0 1.48 k9.5 Water 216 192,000 1.402-2 0. 1%g 5.0 ¢ Totals, Low 1.32 44 Med .77 59 High 2.87 9% Bhased on 33.5 MTU/GWie)-yr. Y)RIGEN caiculations assume 18.823 MWt)/MTU, & years in reactor, to 27,500 MHd/MTU; 2.6 wt % 2%*U. Quantities of metal in core from ref. 2%i. . 2 . : . ; . . . - . “The measured vaiuel at the Nine Mile Point reactor [625 MWie)}] was & Ci/yr; see text for comments on power density and steam/liguid water volume. dORIGEN calculations assume 30,0 MW{t)/MTU, 3 vears in reactor, to 33,000 MWA/MTU; 3.3 wt % 2%, Quantities of meta: in core from re?. 22. There is considerable variation in production of "“C from the '“N{n,p) reaction because of variations in the nitrogen content of LWR fuels. Crow’’ presented the following brief summary of a survey of five fuel fabrication plants: Maximum nitrogen allowed by specification, ppm 75-100 Maximum nitrogen reported, ppm 100 Minimum nitrogen reported, p;fim 1 Average nitrogen in reactor fuel, ppm 255 He has indicated that the 25 +35 ppm average is not a true arithmetic average but a consensus derived from discussions with representatives of fuel manufacturers. Table 3 contains the results of 2 much more extensive survey of the nitrogen content of fuels made at these same five plants. The current average nitrogen content varies from 3 to 50 ppm and the standard deviation of each average is in the range of 40 to 70% of the average. The data shown in Table 3 suggest that the median value of fuel from all plants is about 25 ppm,‘ The differences in the nitride-nitrogen concentrations in LWR fuels from the five manufacturers listed in Table 3 are due to many variables. Some of these have been described qualitatively and are discussed by Pechin et al.”* without reference to reaction times, temperatures, and concentrations. Uranium hexafluoride from gaseous diffusion plants, enriched to 2 to 4 wt % in *°U, is the starting material in the manufacture of LWR fuels. Four of the manufacturers use the ammonium diuranate (ADU) process, and one uses the direct (dry) conversion (DC) process. Powdered UQ, is obtained from both processes, cracked NH; being the preferred source of hydrogen reductant. Pellets are obtained by pressing the powder into pellet form and sintering these in hydrogen, as in the uranium-valence reduction step. Pellet pressing is performed as a dry operation (except for a little lubricant). Sintering is performed at temperatures ranging from < 1600°C to = 1750°C. After cooling, the pellets are loaded into Zircaloy fuel tubes (closed at one end), usually without any additional treatment. Before the fuel tube is Welded closed in a helium atmosphere at all plants, air is removed in a vacuum degassing step at four plants, but is left in place at one of the plants. During the degassing operation, pellets in the fuel rods are unheated in some plants and heated in others. All vaccum degassing operations are followed by filling the fuel rod with high-purity helium and closing the second end by welding in 2 helium atmosphere. Helium is added under pressure to fuel tubes at the plant at which the the vacuum degassing step is not employed. The gaseous nitrogen from 18 to 30 ¢c of air in a single fuel tube containing about 1.75 kg of UO; corresponds to an additional 10 to 16 ppm-of N, that is not included in Table 3. Because of the wide range of nitrogen concentrations, three values of '*C production from the “N(n,p) reaction are listed in Table 2. These correspond to 10, 25, and 75 ppm of nitrogen. At these three levels, "*C production for the listed burnup conditions are 0.171, 0.428, and 1.28 Ci/MTU, respectively, which corresponds to 5.7, 14.3, and 42.9 Ci/ GW({e)-yr for the BWR. Similar values for the PWR are 0.183, 0.457, and 1.37 Ci/ MTU, respectively, and 6.1, 15.3, and 459 Ci/GW(e)-yr. It may be noted that the same quantity of "C will be produced from "O(n,o) and ""N{n,p) reactions when the nitrogen content of the fuel is about 5.7 ppm for both PWRs and BWRs. The chemical form of "“C in the fuel is not known. When formed from any of the five nuclear reactions presented in Sect. 2, this nuclide might become bound to uranium as carbide, remain as impurity atoms, or be converted to carbon monoxide or carbon dioxide. A nitrogen impurity of 75 ppm corresponds to 1.28 Ci of "C/MTU in the case of the reference BWR and to 1.37 Ci of “C/MTU in the case of the reference PWR (Table 2). These maximum expected activities Table 3., Ni trogen content of U0, fuels for LWRs and of FFTF fuelg? -_— FFTF fuels®[ (U,Pu)0. ] Current production of LWR fuels (U0;) Compan pany A fuel ComEanX B fyuol Company Analyzed by Analyzed by 1 2 3 4 5 Company A HEDL Companv B HEDL No. of measurements 358 408 38 206 70 8G 10 80 10 Percent of measurements with nitrogen, ppm <10 i¢o 75 42 14 10 68 100 78 9¢ it - 20 12 53 39 1 4 17 20 - 35 9 36 16 12 5 10 >35 4 5 35 - 50 10 27 2 >50 1 486 14 Mass-weighted av nitrogen, pom 2.8 13.3 13.7 21.6 47.8 <21.6% <1p© <11.1¢ <9 2¢ Std deviation, ppmS 1.4 8.3 9.8 11.1 21.2 N.A N.A N.A N.A. aPrimarin nitride nitrogern. bFrom ref, 52. CNumerical values are based on using the many values <10 ppm as 1C0.0 ppm. It is emphasized that the distribution of nitrogen analyses is not normal. N.A. (not available) 1is used because a meaningful sta ndard deviation cannot be calculated, correspond to a ratio of about 1 "*C atom/ 200,000 uranium atoms. Ferris and Bradley™ studied the reactions of uranium carbides with nitric acid and found that 50 to 80% of the carbide carbon was converted to carbon dioxide;, the remaining carbide carbon was converted to nitric acid-soluble chemicals such as oxalic acid, mellitic acid, and other species, probably aromatics highly substituted with -COOH and -OH groups. Formation of such compounds can be reconciled with the existence of the polymeric -C-C- bonds of uranium carbides. However, at a ratio of 1 “C atom/ 200,000 uranium atoms, or even at a ratio | C atom/500 uranium atoms, which would correspond to an impurity of 100 ppm of carbon in the UQ;, there will be a very low concentration of -C-C- bonds in the UO- fuels. This suggests that a larger quantity of any carbide carbon, including that formed from nuclear reactions, will be converted to €O, in dissolving operations at the {uel reprocessing plant than the 50 to 80% reported by Ferris and Bradley®® for pure uranium carbides. An experimental program to measure C liberated during fuel dissolution is now in progress.” 3.2 Formation in Core Hardware Core structural materials include stainless steel support hardware, Zircaloy cladding, and nickel alloys used as springs and fuel tube separators. According to specifications,”” "' the primary source of “C in these materials is the nitrogen that is present in quantities listed in Table 4. The quantities of each of the types of metal (i.e., stainless steel, Zircaloy, Inconel-X) are somewhat dependent on the reactor type (BWR™ or PWR™ ) and on the year and size of the design within a reactor type. For example, Fuller ei al.”” have presented data on the fifth and sixth generation BWRs (BWR/5 and BWR/6) from which the weight ratios are calculated to be 247 and 265 kg of Zircaloy-2/MTU, respectively. Other estimates of quantities of structural hardware have been given by Griggs™ and by Levitz et al.”” However, the quantities of these metals, the contained nitrogen, and the 'C produced (as listed in Table 2} are based on information pertaining to present reactor designs provided by Marlowe’ and Kiip.™* Carbon-14 values are based on calculations with the ORIGEN code' for a BWR operated to a burnup of 27,500 MW(1)d/MTU in 4 yr and a PWR to a burnup of 33,000 MW(t)d/ MTU in 3 yr. The revised light-element library'® was used in these calculations. Most of the “C formed in these structural components will be retained within the metal when the latter is encapsulated for long-term disposal, although a very small fraction in the Zircaloy might be dissolved in fuel leaching solutions at the fuel reprocessing plant. Experiments have never been performed to evaluate this possibility. 3.3 Formation in Cooling Water Oxygen of the cooling water and nitrogen-containing chemicals in this water are sources of HC, An accurate calculation of the quantity of “C that will be formed would require integrating the flux over the volume of water in and surrounding the core. Data to perform such an integration do not appear to be readily available, but reasonable approximations can be made. Reference 34 gives values for the atomic ratio H/U of 3.74 and 4.23 for BWRs and PWRs, respectively; these correspond to 7860 and 8890 g-atoms of O (as H:0)/MTU. tuller et al.” give values of the water; fuel volume ratio of 2.52 for BWR -5 and 2.50 for BWR ;6. A water density of 0.805 g/cm’ and a UQ, density of 10 g/cm’, both at 556"F, indicate a ratio of about 13,000 g-atoms of O/ MTU for the BWR cores. Reference 36 gives a hot, {first care H,0/ UO; volume ratio (for a PWR) of 2.08, Tavlie L. Specifications for carbon and nitrogen in reactor structural and claddinrg metals Specifications {(wt %) Reactor type Carbon Nitrogen Reflerences for specifications . 27 25 Steinless steel 204 BWR <0.08 0.10-0.16 ASME SA213-73 and ASME SA-2L ‘ 27 2 304 PWR <0.08 0.10-0.16 ASME SAZ213-73 and ASME SA-2LO 29 316 IMFBR 0.040-0. 060 <0.010 RDT M73-287 5 Zircaloy=2 BWR <0, 027 <0, 008 ASTM B253-71 (ANSI N12M-1973>3 . O Zircaloy-L PWR <0,027 <0.008 ASTM B353-7% (ANST N124-1973}3 Tnconel-X RWR <0.10 Trternational Nickel Co. o- ) . ) o 31 Inconel 718 PWR =0.10 nternational Nickel Co. Nicrobraze 50 PWR 0.01 0. 0066 01 11 which corresponds to about 10,500 g-atoms of O/MTU. For the purpose of this report, it is thus assumed that the rate of reaction ''O(n,a)"'C is specified by a ratio 12,000 g-atoms of O/MTU and a natural 'O abundance of 0.039 at. % in oxygen for both BWRs and PWRs. This corresponds (Table 2) to about 4.7 and 5.0 Ci of "C/GW(e)-yr for BWRs and PWRs, respectively, from the YO(n,a)"'C reaction: it also corresponds to an initial atomic ratio H/ U of about 220 for BWRs and 175 for PWRs using fuels containing 2.6% and 3.3% °*°U, respectively. The quantity of "C formed from impurity nitrogen cannot be estimated since there do not appear to be any analyses Ifiertaining to the concentration of this element in reactor cooling water. Although its concentration may be no more than a [ew parts per million, Cohen® mentions a value as high as 50 ppm NH: in the primary cooling water of PWRs. Quantities of '*C actually released from a BWR and three PWRs, as measured by Kunz and his coworkers,'''* are listed in Table 2. From the BWR at Nine Mile Point (625 MW(e)] they observed'” a release rate of 8 Ci of "*C/yr. These authors also reported 6 Ci of *C/GW(e)-yr on the basis of their analyses of gaseous effiuents from the Ginna, Indian Point 1, and Indian Point 2 PWRs. At the PWR stations,'' over 80% of the "“C activity was chemically bound as CH, and C,H,; only small quantities were bound as CO». At the Nine Mile Point BWR station'” the chemical form of "“C was greatly different, with 95% as CO,, 2.5% as CO, and 2.5% as hydrocarbons. On the bases of the fuel isotopic compositions and burnups shown in the footnotes of Table 2 and for the assumed ratio of 12,000 g-atoms of O/ MTU, an impurity of 1 ppm.of nitrogen in the cooling water {corresponding to 0.216 g of N/MTU) would lead to the formation of 0.124 and 0.132 Ci of '4C/GW(e)~yr in BWRs and PWRs, respectively. The difference between a calculated 5 Ci of M/ GW(e)yr from the "O(n,e) reaction and the observed 6 Ci/yr at the PWR stations'' (Table 2) is probab!y well within limits of analytical uncertainty. The extrapolation to 16 Ci of ""C/GW(e)-yr from the measured 8 Ci/yr at the Nine Mile Point BWR is based on maintenance of a constant power density and a constant volume ratio H,0/ UO,. Values of this ratio tabulated for the Nine Mile Point reactor’’ and for newer, larger reactors, such as those at Brown’s Ferry,42 do not differ significantly (2.38 vs 2.43); the average power densities for the two reactors are 41 and 50.732 kW/liter, respectively. When these ratios are combined with data on the average void fractions within a fuel assembly (a measure of steam/liquid water, and having values of 0.3 for the Nine Mile Point core and 0.4 for the Brown's Ferry core), it is apparent that “C formation in a new 1100 MW(e) BWR (such as BWR/5"™) would be larger than 8 Cij GW(e)-yr, but significantly less than 16 Ci/ GW(e)-yr. 4.0 CARBON-14 FORMATION IN HIGH-TEMPERATURE GAS-COOLED REACTORS The only structural materials in HTGRs in which "C will be formed to any significant extent are the fuel containing and reflector blocks of graphite. There will be some nitrogen and oxygen in the helium coolant.*’ However, the rate of "*C formation from coolant impurities will be very small in comparison with similar rates in the fuel blocks; in addition, the helium cleanup system is expected to remove CO:, a probable form of part of the "*C in the coolant. 4.1 F"ormation in the Fuel The compositions of fertile and fissile fuel for HTGRs have not been positively established since commercial reactors are not yet being roade. However, it is highly probable* that the initial and 12 makeup (the IM stream) fuel will be in the form of about 93 wt 9% of *U as UC,, that *"’U bred from the fertile thorium will be recycled as UC; (the 23R stream), and that uranium recovered from the IM stream after reprocessing, if it is recycled as the 25R stream, will also be in the form of UC,. Similarly, the fertile thorium is expected to be in the form of ThO;. Uranium in the IM stream will have a chemical history different than that of uranium in the 23R and 25R streams. In particular, uranium for the IM stream will be received at a fresh-fuel fabrication pla,m45 as UF,, which will be hydrolyzed with steam to UOQOsF:; this, in turn, will be reduced at about 650°C with H, ( from cracked ammonia) to UO.. Subsequently, the UO: will be mixed with carbon flour, ethyl cellulose and methylene chloride. 1t will then be dried, ground, separated into appropriate sizes, and heated in a vacuum to cause the formation of UC,. Finally, it will be cooled in an inert atmosphere, which may cither be nitrogen or argon. {n these successive processes, the uranium-bearing matenal never exists as a nitrogen-containing compound, although it is exposed to N; from cracked ammonia at a high temperaiure and may be exposed to nitrogen after formation of UC;. On the other hand,' recycle uranium, both 23R and 25R streams, will pass through the uranyl nitrate [UO2(NO;):] state i a fuel reprocessing plant. These materials will be denitrated and converted to UQO, before subsequent carbonizing steps that are similar to those described {or the [IM material. The significance of the differences in histories is that recycle uranium may contain more nitrogen (from undecomposed nitrate) than does the initial or makeup 93% **U. There are limited data concerning the quantities of nitrogen in potential HTGR fuel since this fuel 1s not made on a routine basis. It is therefore assumed that all forms of UC; and ThO: contain the same quantity of nitrogen (i.e., 25 ppm) used in this report as an industry concensus for LWR fuels. On this basis, about 0.96 Ci of “C/MTHM, or about 9.7 Ci; GW(e)-yr will be formed from the '4N(n,p) reaction. Carbon-14 will also be formed to the extent of 0.225 CiyMTHM, or 2.3 Ci/GW(e)-yr, from the reaction ' 'O(n,a)"C of oxygen present as ThO: (Table 35). 4.2 Formation in Graphite Blocks Independently of the "N(n,p)'C reaction, significant quantities of "“C will be formed in graphite of fuel and reflector blocks due to the reaction "‘C(n,y)*C. Based on a lifetime average ratio of 10.93 MTC in fuel blocks; MTHM, about 3.7 Ci of "*C/MTHM, or 37 Ci/GW(e)-yr. will be formed from this (n,v) reaction (Table 5). Additional *C will be formed in reflector blocks, which are present to the extent of 16.29% of fuel blocks on a lifetime average basis. The neutron flux in reflector blocks will be about 70 to 80% of the corc-average flux, although the "“C production listed in Table 5 is based on a flux in these reflector blocks equal to the core average. The total *C formed from the ''C(n,y) reaction in fuel blocks and reflector blocks is less than 4.3 Ci/ MTHM, or less than 43 Ci/ GW(e)-yr. The amount of nitrogen present in fuel-block or reflector-block graphite is uncertain. Four samples of graphite were irradiated in the Oak Ridge Rescarch Reactor (ORR) and were subsequently analyzed for “C.* The quantity of this nuclide in excess of that calculated to be formed from the '"C(n,y)"*C reaction was ascribed to the reaction "*N(n,p)*C. On the basis of this assumption. the equivalent nitrogen impurity was calculated to be 3.2 to 84 ppm on a graphite-weight basis. The only other estimate of nitrogen content in an in-use graphite is 26 ppm."* and is used here as the basis for the value of 30 ppm of nitrogen in fuel blocks and reflector blocks listed in Table 5. Carbon-14 formed in graphite containing 30 ppm of nitrogen corresponds to 126 Ciy MTHM or 127 Ci/GW(e)-yr. _Table 5. Production of 1% in graphite and fuel al High-Temperature Gas-Cooled Reactors 14: existing 160 days after discharge of fuel Impurity content Material Quantity of element in core {C1/MIHM ) . {g/MriM) ; F From Total '*c Nitrogen Oxygen in core From rom ro < Material {ppm} {wt. %3 _{MT /MTHM ) Carbon Ritrogen Oxygen carbon nitrogen OXygen Ci/MIHM Ci/GWie)-yr Graphite in fuel blocks 107 10.5%° 1.0G3E+7 3.28E+2 3,69 12.58 i6.27 164 Graphite in reflector b . 4 tlocks 30 .77 1.77E6 3. 54E+E . <0.60 <2,0h . <2.6% <=6.6 IM uraniom (1K, } 25 0. oi;5h81 2.50E+1 G.95G G, Olli Q. L Recycle uranium (s ) 25® o.chsize’ 2. 50E+1 0.559 0. ot G, s = Thorium dioxide 25° i2.1g 0.9091'&1f 2,50B+1 1.25E+5 C.959 0,255 1.08 10.5 Total ®Rased on 10.11 MEM/CWlel=yT {eguivalent to 38.9% efficiency in converting hest to electricity). bThi: ig an estimate based on the sssumplion that no great efforts will be made tc minimize the nitrogen content, “See ref. 13. dBased cn & neuirsn flux in reflecior blocks eguai te the coresaverage flux. HNowever, the fiux in the reflector blocks will be about 70 to 8% of the core-averusge value. €Assumed to be the same as in IWR fuels. Y¥rem rer. 13 the following values are obtained: 405,08 kg {034 3°®U) TH material, 294,07 kg 23R material, 107.83 kg 25R material, and 8394.7¢ kg thorium in the lifetime average annual reload. values listed sre MY thoriwm or uranium/MTHM. €211 of this is potentislly svailable for release at the [usl reprocessing plant except asbout 0.012 Ci/MTHM {0.12 Ci/GWie}-yr) in the initislly fissile particles cof the 25R stream £ E% P P P r which are designated 25W efter digcharge. 14 5.0 CARBON-14 FORMATION IN LIQUID-METAL FAST BREEDER REACTORS T'he primary structural material of the core of an LMFBR will be 316 or A-286 stainless steel. Carbon-14 will be formed from impurities in this metal as well as in the fuel. Since no LMFBR has yet been buili, discussion presented here is based on the proposed reference design' of the Clinch River Breeder Reactor (CRBR) and on recent updating of fuel composition.*™ A core element for this reactor is shown in Fig. 1. 5.1 Formation in the fuel In common with LWR fuels, ""C will be formed by the "O(n.a) and "“N(n.p) reactions in LMFBR fuels; in both types of reactor very small quantities of "*C will be formed by the ''C(n.y) reaction. Two other reactions produce "C in the LMFBR (Sect. 2): "N(n,d) and "O(n,'He). Croff's'* estimates of cross sections and formation rates are listed in Table 1. Production of *C from reactions involving oxygen are listed in Table 6; these values are based on 8383 g-atoms of O/MTHM (in this case, MTHM is uranium plus plutonium) and 0.039 at. % of "0 in natural oxygen (corresponding to 3.27 g-atoms of O/ MTHM). The specification limit on the nitride nitrogen impurity in plutonium dioxide™ and driver fuel™ for the Fast Flux Test Facility (FFTF) 1s 200 ppm. Air in fuel rods is evacuated and replaced by high-purity helium'' before the rods are closed by welding in a helium atmosphere. The maximum fuel-pellet gas content of 0.09 cc (STP) per gram of fuel,™ exclusive of water, would correspond to 120 g of N/MTU 1if all the gas were nitrogen. Measured nitride citrogen concentrations in FFTF fuels have been significantly less than specifications, gencrally in the 10 to 20 ppm range,” as shown in Table 3. Therefore, it is assumed in this report that the concentration of nitrogen in CRBR fuel will be about 25 ppm, with a range of 10 to 75 ppin. These values were used to estimate an average and range (Table 7) of "*C formation due to neutron absorption by "N and ""N. The average value is 0.166 Ci of "“"C/MTHM. or 6.1 Ci of "C/GW(e)-yr; the values range from 0.0665 Ci; MTHM [2.45 Ci; GW(e)-yr] to 0.499 Ci/MTHM [18.4 Ci; GW(e)-yr]. Formation of "C from oxygen in the fuel, 0.00364 Ci) MTHM, and from nitrogen would be equal if the nitrogen concentration in the fuel were about 0.55 ppm. 5.2 Formation in Core Hardware As noted above, 316 stainless steel (with specifications listed in ref. 29) or A-218, is essentially the only metal in the CRBR core and may be the only metal in future commercial LMFBRs, Specification RDT M3-28T, Table 4, requires that the oxygen and nitrogen concentrations be lower than corresponding values for 304 stainless steel used in LWRs. In particular, the specification of <0.010 wt % of nitrogen in 316 stainless steel is more than a factor of 10 below the specification of 0.10 to 0.16 wt 9% of nitrogen in 304 stainless steel for LWR applications. Calculated quantities of "*C to be formed in CRBR cladding are listed in Table 7. These are based on 100 ppm (0.01 wt %) of nitrogen and on the “mass ratios” shown in Table 6. These ratios refer only to cladding plus shroud plus wire between bottom and top fuel elevations, The neutron flux decreases very rapidly with elevation away from fuel levels. For this recason, "C formation in regions above the fuel level in the upper axial blanket and below the fuel level in the lower axial blanket 1s neglected. OR&L DWG 78 - 14882 5 833-mmn (0 2307.) DIAM WIRE WRAP .84 n {C. 2D S f422-mm (0.056n) DiaM / 217 REQD N\ H.78-cm (Tia) PITCH / 1.62-cm (4.57%4n.) HEX DUCT TUBE 3.048-mm (0.120-x) WALL o Fig, 1. Reference CRBR core fuel assembly. ST Table 6. Data pertaining to **C production in the CRBR ' ORIGEN - o , 14 Specif%c Mass sfgiilgzs Mas calculated Specific production of ~ C from pPoWer of HM a.b ratio burnup _ Carbon Nitrogen Oxygen MA (L) charged®’ steel®’ (MESS) [ngt)-d (Ci \ Ci ci ) CRBR region MTHM (M) (MT) MIHM { MTHM g c) g N 100 kg 0 Inner core 113.22 1.4361 10.63 0.66 93,066 9.98E-9 1.88e-2 §.398-3 Outer core 104.63 1.2006 9.11 0.66 86,005 6.92E-9 1.328-2 5.48E-3 Upper axisl blanket 3.482 1.0361 8.L0 0.66 2,862 1.47E-9 2.85E-3 1.03E-3 Lower axial blanket 7.276 1.0361 7.77 0.66 5,981 2.66E-9 5.13E-3 1.92E-3 Radial blanket 4.302 3.0373 20,0k 0.185 3,536 1.75E=9 3.39E-3 1.24E-3 Total in reactor 32.3505 56.25 0.393 d Mass-average 30.154 24,811 83ee Rer. L8. bThe heavy metal (HM) charge is the annual charge; annually, one-third-of the core and axial blankets and one-sixth of the radial blankets are replaced, The stainless-steel mass is the total in the specified region, not Jjust the {fresh steei, The mass ratio of stainless steel to heavy metal [{MTSS/MTHM), column 5)] is the sum {cladding mass + shroud mass + wire mass) Caiculations are based on the following betwern the bottom and top fuel elevations, Fig. data for core and axisl blanket tubes (fuel pins, see Fig, 1): 1, per unit mass of heavy metal. 0D = 0.230 in.; ID = 0.200 in,; wire-rod spacer (running 4,575 in.; hex metal thickness = (.120 in.; nearly coaxially with fuel pin) = 0.055 in. diam; hex face-to-face distance fuel diameter = 0.20C in.; density of stainless steel = 8,02 g/em’; density of fuel {U0;) = 9.316 (85% of theoretical 10.96 g/cm’®). The radial blanket fuel rod dimensions are: OD = 0.520 in.; ID = 0.490 in.: are as given above. -~ From the stoichiometry of {U,Pu)0,, therc are about 134 kg O/MTHM. drnis corresponds to 36.80 MTHM/GW(ec)-yr, as used in Table 7. fuel diam = 0.485 in.; all other parameters 91 Teble 7. Production of *C in the CRBR™ Production of 3¢ in fuel from Production of 144 Nitrogen : . from nitrogen in Oxygen Low { 10 ppm ) Average {25 ppa} High { 75 pom) stainless steel CRBR region - - C1/MDHM Ci/GW(elwyT €1 /MTHM Ci/GwWle )-yr Ci/MTHM Ci/GwW{e}-yr C1i/MTHM 2i/Gw{el-yT Ci/MIHM Ci/GWie)-yr inner core 1.138.2 1.11E-1 1.88E-1 1.84E+0 4, 7TOE-1 Y 61E+C 1.h2E+0 1.33E+1 1.2LE+O 1.22E+1 Duter core 7.35E-3 7.80E-2 1.39E-1 1. LOE+0D 3.30B-1 3.50E+0 3.00E-1 1.056+1 §.738-1 Q. 27E+0 Upper axial blanket 1,39E-3 4, k31 2,85E-2 9. 09E+0 7.12E~2 2.27E+1 2. 1hE-1 6.82F+1 1.868E-1 6.01E+1 Iower axial blanket 2.58E-3 3.9kE-1 5,.13E-2 7.838+0 1.28E~1 1,96E+1 3.85E-1 5. 87E+1 3.398-1 5. 18841 Radial blanket 1.6TE=3 b, 31F-3 3.398-2 8. T6E+0 8. 4B~z 2.19E+1 2. 5hf-1 £.57E+1 £.27E-2 1.65k+% Mass-average 3, 6hE-3 1.34E-1 6,65E-2 2. 45E+0 1.66E-1 6,128+ 0 4,99E-1 1.8LE+1 3.L49E-1 1. 28E+1 ®caleulations do not include formetion of '*C in stainless steel above the top or below the bottom of the fuel. L1 18 6.0 COMPARISONS AND DISCUSSIONS Calculated quantities of “C that arc or will be produced in the four types of reactors (BWR, PWR, HTGR, and LMFBR) considered in this report are summarized in Table 8 in units of Ci/ GW(e)-yr. Ranges are given for all calculated values of "*C from all reactors except the HTGR. The ranges are due to variations in the nitrogen content of the fuel. Values spanning the full range of 10 to 75 ppm (by weight) are shown in Table 3, which is a suromary of manufacturing data. The Barnwell plant of Allied General Nuclear Sesvices is designed to process about 5 MTHM/day, or 1500 MTHM/yr, of LWR fuel. Heavy metal (HM) is uranium or uranium plus plutonium charged tc BWR, PWR, and LMFBR; HM is also uranium plus thorium charged to the HTGRs, The Barnwell design corresponds to about 45 GW(e)-yr. Similarly, reference HTGR- and LMFBR-fuel reprocessing plants are designed to process annually fuel that produced about 45 GW(e)-yr of energy. Using this factor as a multiplier for values listed in Table 8, it is appropriate to examine the total quantities of '*C that would be released from the various fuel reprocessing planis if cquipmeni is not instalied to collect and retain the gases containing this nuchide; it 15 also appropriate to examine how much will be contained withun the hardware that becomes part of the high-level waste that may be shipped to a Federal repository. Light-water reactor fuel processed in | year in a Barnwell-sized plant will contain 400 to 2200 Ci of "*C: the hardware will contain 1400 to 2700 Ci of ""C. The calculated values for "'C in the hardware are conservatively high since they are based on the assumption that all core hardware ~ not just the cladding — is in as intense a flux field as 1s the cladding. Lesser quantities of *C will be produced in LMFBR fuel. The fuel entering a reprocessing plant of 45 GW(e)-yr capacity will contain 100 to 800 Ci of “C per year while the cladding will contain about 600 Ci of "“C per year. Quantities of this nuclide in other hardware are not included in Table 8. The "'C content of HTGR fuel entering a 450 MTHM/ yr [45 GW(e)-yr] fuel reprocessing plant in | yr will be about 530 Ci if the nitrogen content of the fuel is 25 ppm. Only this “median” nitrogen content is considered because the graphiie probably will be the dominant source of "“C. In particular, if there is no nitrogen in the graphite, the "*C content [due solely to the ''C(n.y)"'C reaction)] of graphite entering the fuel reprocessing plant 1 | yr will be about 1660 Ci; the “'N(n,p)"'C reaction will add about 5660 Ci of “*C if the nitrogen content of the graphite is 30 ppm. The value of <200 Ci of ""C/GW(e)-yr shown in Table 8 for the HTGR corresponds to <9000 Ci entering the fuel reprocessing plant cach year. These maxima include C in reflector blocks as well as in fuel blocks. There is no metallic hardware in an HT'GR corresponding to cladding and other structural components of the LWRs and LMFBRs. 6.1 Comparisons of Reactor Produced and Naturally Produced "C The natural rate of ""C formation in the atmosphere from cosmic-ray induced reactions and the contribution of "*C to the total radiation dose to man are valid bases for evaluating the impact of reactor-generated quantities of this nuclide. Lingenfelter’' reported a global average production rate of 2.5040.50 "C atoms cm ~ sec ' over the ten solar cycles prior to 1963. Reference has been made to this value by Lal and Suess’ and in the UNSCEAR 1972 report.™ Using 5. 1E18 cm’ as the carth’s surface area.” Lingenfelter’s value corresponds to (4.220.8)E4 Ci of "C yr. More recently, Light et al.™ have calculated the average production rate from 1964 to 1971 to be 2.217+0.10 "C atoms Table 8, Comparison of Y40 production in different types of reactors in units of i fow(e J=yr® Cladding and core ) ' In structural 1n coolsnt Total Reactor fuel materials Calculated Cbserved calculated BWR 43.3-60.4 4.7 8 ° Low value 3.0 57 Median value 17.6 . , - 7h High value 46,3 111 PWR 30.5=41.6 5.0 6 Iow value 9.6 iy Median value 18.8 _ 29 High value k9.5 % - - . . - , o o HTGR <190 nil N.A, Median value 12.0 <200 C IMFBR 12.8 nil N.A, Low value 2.6 15 Median value 6.3 19 High value 18.5 31 aReactor paramefers pertaining to these calculations based on the ORIGEN program are as follows: BWR, 18.823 MW(t)/MTU, L years in reactor, to 27,500 MWA/MTU:; 2.6 wt % 2°°U; 33% thermal efficiency. PWR, 30.0 MW(t)/MTU, 3 years in reactor, to 33,000 MWA/MTU; 3.3 wt % 23°U; 33% thermsl efficiency. HIGR, L MW(t)/MIEM, L years in reactor, to 95,000 MW4/MTU; 38.5% thermal efficiency; see lable 5 Tor fuei compositions. IMFBR, 30.18 MWw(t)/MTHM {mass average), 75% on-stream time for 3 years, to 24,800 MWA/MTU (mass average); 35% thermal efficiency; see Table 6 for fuel-region specifications. bA velue of 9.1 Ci/GW(e)-yr is presented in the following report, issued as the present report was in the final stage of preparation: R. L. Blanchard, W. L. Brinck, H. E. Xolde, H. L, Krieger, D. M. Montgomery, S. Gold, A. Martin, and B. Kahn, Radiological Surveillance Studies at the Oyster Creek BWR Nuclear Generating Station, USEPA, EPA-520/5-T76-003 (June 1976). “§.4. = not applicable. 61 20 e sec . Based on projections of sunspot numbers for the remainder of the solar cycle, they also estimate that the 11-yr mean rate could be as large as 2.2840.10 *C atoms cm ° sec '. (The error limits on the rates apply only to the statistics of the calculation.) This value corresponds to (3.8+0.2)E4 Ci of "C/yr. Thus, to one significant figure, the 1l-yr average natural rate of production is 4.E4 Ci of "'C/yr. On this basis, the quantity of "*C in fuel annually entering an LWR fuel reprocessing plant with a capacity of 1500 MTHM/yr [equivalent to 45 GW(e)-yr and about fifty 1000 MW (e) reactors] is | to 5.5% of the natural production rate; corresponding values for '*C entering an LMFBR fuel reprocessing plant are 0.3 to 2.09 of the natural production rate. The 1660 (from graphite only) to 9000 (from graphite, oxygen, 25 ppm of nitrogen in fuel, and 30 ppm of nitrogen in ali graphite) Ci of "*C annually entering the HTGR fuel reprocessing plant, of the same 45 GW(e)-yr equivalent capacity, corresponds to 4 to 22% of the natural rate of production of this nuchde. 6.2 Worldwide and Local Radiation Doses from Reactor-Produced "C World population radiation doses from all forms of radiation and from naturally produced '*C provide a second form of comparison of the effects of discharge of this nuclide from fuel reprocessing plants. World-wide dose rates to gonads, bone-lining cells, and bone marrow due to internal and external irradiation from all natural sources in “normal” areas are about 90 mrad/yr (Table 20 of ref. 54, UNSCEAR 1972). Oakley’ reports a gonadal dose equivalent to the population of the United States from all natural sources of 88 mrem/yr. The contribution of “C to this total is about 0.7 to 0.8 mrad/yr.™ Other values of the contribution of '*C to the total have been as high as 1.6 mrem/yr.'""”™ Thus, based on the percentages histed above and a nominal | mrem/yr due to natural "“C, after this nuclide becomes uniformly distributed over the earth, additional radiation doses due to "*C will be in the range 0.004 to 0.06 mrem/yr for discharges from an LWR fuel reprocessing plant of capacity equivalent to 45 GW(e}-yr. corresponding incremental doses due to "C discharges from equivalent LMFBR and HTGR fuel reprocessing plants will be in the range 0.0004 to 0.023 mrem/yr and 0.035 to 0.19 mrem/ yr, respectively. Potential radiological impacts of annual releases of 5000 Ci of "*C on the population out to S0 miles from a fucl reprocessing plant have been analyzed by Killough et al.”” Three techniques for reducing these local population doses were: (1) use of a discharge stack up to 1000 ft tall; (2) heating of the discharged gas to obtain a large effect of buoyancy to increase the cffective stack height: and (3) use of nocturnal, rather than continuous, emissioni in order to minimize the availability of the discharged "‘C for uptake by vegetation. Using metecorological data for the Oak Ridge, Tennessee, area and a 300-ft stack, the total-body dose of a population of 10" people within the 50-mile radius was 110 person-rem/ yr; the average individual dose was 0.107 mrem/yr, and the maximum dose to “fence-post man” (who spends all his time at 1.5 miles from the stack and eats food grown only at this location) was 240 mrem/ yr. 6.3 Other Predictions of ""C Formation Rates Tahle 9 summarizes predictions of "'C formation rates in BWR and PWR fuels presented in this *7 Calenlated formation rates in BWR fuels range from 13.6 to 22 Ci/ GW(c)-yr. In the BWR coolant. from the ' O(n.a) reaction only. the range 15 4.7 to 9.9 C1, GW(e)-yr. and other reports. ) Iy Table 9. Comparisons of some estimates of C production rates? in LWRs (values are in Ci of '*C/GW(e)-yr) Source of information Region Reactor of %C Parent Bonka Kelly Fowler This type formation nuclide et al. et al.© NUREGd et al.® report BWR | Fuel RN 12.9 16,9 we® 18. . 11.5 70 8.4 2.7 NC 4. 3.3 Yo+ 0 21.3 13.6 NC 22. 14.8 Coolant AN 1.3 NC NC 0.26 NC 17 9.9 NC 9.5 8.9 B PWR Fuel 14y 12.2 10.9 NC 18. 12.2 ‘o 7.1 2.7 NC b 3.5 1'%y + 170 19.3 13.6 NC 22. 15.7 Coolant 1Ay 1.28 NC NC G.09 NC 179 9.8 NC 8 3.2 5.0 8Baged on 20 ppm nitrogen (by weight) in the UO; except for Bonka et al.,60 whose basis is not given. bref. 60. CRef. 61. Parameters in ref. 62 for the BWR and in réf. 63 for the PWR correspond to about 0.9 GW(e)-yr. ~Thus, values in this column, which are taken from these references, should be increased about 107%. €Ref. b4. Calculations pertaining to teg produced in the BWR cooling water are based omn the assumpltion that there gis noe void volume in the core due to steam. NC means not calculated. 1¢ 22 Corresponding values in PWR fuels also range from 13.6 to 22 Ci/GW({e)-yr, and in PWR coolant they range from 3.2 to 9.8 Ci/GW(e)-yr. Carbon-14 formsation rates in cooling water from the “N(n,p) reaction are small and uncertain, since data on concentrations of nitrogen are nearly nonexistent. When the uncertainties in cross-section data are coinbined with the varying choices of other nuclear parameters used by these different authors, it 1s perhaps not unexpected that the largest values are about twice the smallest. Bonka et al.*’ give "*C production rates from nitrogen in the fuel and coolant of LWRs. These authors list the 2200-m/sec cross sections for the "C(n,y)"*C, “N(n.p)"*C, and '"O(n,a)""C reactions without stating whether they used these or cross sections collapsed according to reactor fluxes. They also do not indicate the nitrogen content of the fuel or cooling water. Thus, 1t i1s not possible to comment on the agreements and differences between the values of Bonka et al.* and those of other authors listed in Table 9. Kelly et al.”' give '*C production rates 5 to 239 lower than values in this report (Table 9). These authors also present only the 2200-m/ sec cross sections for reactions 1, 2, and 5, they do not discuss collapsing cross-section data in terms of the fluxes of specific reactors. Again, no comparison can be made between their model reactors and those of this report. The U.S. Nuclear Regulatory Commission (NRC) has presented an estimate of 9.2 Ci of "C/yr formed in the cooling water of a BWR® and of 8 Ci/yr in the cooling water of a PWR.** Both values are based only on the "O(n,a)"*C reaction; formation of "*C from the "*N(n,p) reaction is considered to contribute only a siall fraction of 1 Ci/yr because of the low concentration of "N in the reactor coolant (less than | ppm by weight). The calculational procedure of the NRC reports includes use of an average flux of 3.0E+13 neutrons cm ’ sec ' and a thermal neutron cross section for 'O of 0.24 b for both BWR and PWR; the masses of water in the reactor cores are 39 and 33 MT, respectively. The product of flux and cross section corresponds to 7.2E-12 atoms of "“C per second per atom of '"O. Fowler et al.”® wrote a technical note partly to elicit comments concerning EPA calculations of “C source terms and the radiological impact of this nuclide. The EPA has already published®’ proposed standards pertaining to releases of “Kr, '“’1, and certain long-lived transuranic nuclides from nuclear power operations; no standard pertaining to '*C was proposed, because the knowledge base available (in 1975) was considered inadequate for such a proposal. Calculations in the technical note are based on assumptions of a flux of 5.0E+13 neutrons cm * sec”', an effective cross section of 1.1 b for the l4N(n,p)"’C reaction, and an effective cross section of 0.14 b for the ”()(n,a)C” reaction, for both the BWR and the PWR. This choice of flux and cross sections corresponds to 5.5E-11 atoms of "*C per second per atom of nitrogen, and 7.0E-12 atoms of "*C per second per atom of 'O, respectively, for both the BWR and the PWR. These authors™ also calculated a source term for "'C formation from | ppm of nitrogen dissolved in the cooling water. This use of | ppm is arbitrary since essentially no data are available on this concentration at operating reactors, as discussed in Sect. 3.3. The calculations with | ppm of nitrogen were made because similar sample calculations had been made in draft regulatory guides.”"’ in refs. 62 and 63 which were developed from these drafts. Calculations in this report are based on parameters listed in footnote a of Table 8 and in Sect. 3.1. From the effective fission cross sections (p. 72, Table A-l, of ref. 1), the ORIGEN code calculates average fluxes of 2.07FE+13 and 2.92E+13 neutrons cm ~ sec ' for BWR and PWR, respectively. However, the initial and final fluxes for the BWR are 2.00E+13 and 2.26E+13. and initial and final fluxes for the PWR are 2.58E+13 and 3.456+13 neutrons cm - sec . The average formation rates for a BWR are, therefore, 3.06E-i1 atoms of ''C formed per second per atom of *N However, such calcuiations are not made 23 present and 3.79E-12 atoms of “C formed per second per atom of 'O present; corresponding values for a PWR are 4.32E-11 and 5.34E-i2. Thus, the "*C formation rates calculated in this report for the “N(n,p) reaction are only 55% (for the BWRY and 799% (for the PWR) as large as values presented by Fowler et al* Carbon-14 formation for the "O(n,a) reaction rates in this report are only 53% (for the BWR) and 74% ({for the PWR) as large as values in refs. 62 and 63, they are only 54% (for the BWR) and 76% (for the PWR) as large as values in ref. 64. Cross sections listed in Table 1 are the current best estimates for application to the steady state of reactor operations {after the first few reloads). The most recent (1974) revisions (soon to be incorporated in the ORIGEN library) of '“N cross sections for use in the ENDF/B-1V library™ WL(C prc-»ented by Young, Foster, and Hale,” largely from an earlier revxew by Young and Foster.” Croff” has used this revision and the XSDRNPM c.omputcr program’' to obtain & one-group value of 1.45 b for the effective thermal cross section for the “N(n,p)"*C reaction for LWRs. This is very close to the value 1.48 b used in this report. 6.4 Comparison with Releases from Russian Reactors Rublevskii et al.”* have presented data, listed in Table 10, on measured releases of '*C from five Russian reactors. These authors combined their data with Spinrad’s’” projections concerning world-wide installed nuclear power to estimate the magnitude of *C discharges to the year 2010. Neglecting the small Obninsk and ARBUS reactors, the data in Table 10 show releases ar the reactor stations of 200 to 800 Ci of “C/GW(e)-yr. These values are far in excess of the 6 Ci/ GW(e)yr reported by Kunz et al.’' for the Ginna, Indian Point 1, and Indian Point 2 PWRs, and of the 8 Ci/GW(e)-yr for the BWR at Nine Mile Point.”” The reported releases of "C from Russian reactors are thus seen to be about of 10 to 100 times greater than corresponding releases from the four-mentioned American reactors. Such a discrepancy implies that Rublevskii et al.”* have grossly overestimated the potential releases of “C from non-Russian nuclear reactors, and that a need exists for an analysis of the origin of "“C formation in the Russian reactors. This overestimation appears in their conclusions that the daily production rates of “C in water