. " . . . T . I OFL ORNL P 1265 - . . : .: " . • . : . * . . i wote .. € . . ... ! 45 A 50 ! : i MS 6 BN 26 03: 11:16 |L25 L6 LG . MICROCOPY RESOLUTION TEST CHART NATIONAL BUREAU OF STANDARDS – 1963 LEGAL NOTICE This report was prepared as an account of Government sponsored work. Neither the United States, nor the Commission, nor any person acting on behalf of the Commission: A. Makes any warranty or representa- tion, expressed or implied, with respect to the accuracy, completeness, or usefulness of the information contained in this report, or that the use of any information, appa- ratus, method, or process disclosed in this report may not infringe privately owned rights; or B. Assumes any liabilities with respect to the use of, or for damages resulting from the use of any information, apparatus, method, or process disclosed in this report. As used in the above, "person acting on behalf of the Commission” includes any em- ployee or contractor of the Commission, or employee of such contractor, to the extent that such employee or contractor of the Commission, or employee of such contractor prepares, disseminates, or provides access to, any information pursuant to his employ- ment or contract with the Commission, or his employment with such contractor. ORNU 8-1265 CONF-65 407-5 JUN 24 1965 MASTER BEHAVIOR OF IODINE IN THE NUCLEAR. SAFETY PILOT PLANT MODEL CONTAINMENT VESSEL ****.- ***.. ... . . ......." 1. F. Parsly, L. F. Franzen, * P. P. Holz, T. H. Row, and J. L. Wantland Oak Ridge National Laboratory Abstract prinatoly owood rights; or with the Commission, or bir employment with such contractor. disseminalou, or provides accois Lo, Loy Information pursuant to his epoplojtaent or contract such employee or contractor of the Commisi'ion, or employee of such contractor preparos, ployco or contractor of the Commission, or omployee of such contractor, to the extent that As used in the above, "person acting on beball or the Commission" Inciudes uy on- use of any Information, appartus, wotbod, or procesu dlaclorod lo this report. B. Asmoa Lay liabiliues with roapoct to the use of, or for damages resulting from the of any inforbution, apparutus, method, o procesu disclosed to this report may not lofringe racy, completeness, or usefulness of tho Informatiou contatood in this report, or that the we A. Makes any warranty or representation, expressed or implied, with respect to the accu- Statos, por the Commission, nor any porno acting on baball of the Commissioa: This report me prepared as an account of Government sponsored work.. Neither the United -- LEGAL NOTICE Results of a series of five runs in which stable iodine plus 1311 tracer was released into an air-filled 1350-fts model containment vessel are reported. The iodine concen- tration and form were observed over periods of 24 to 48 hr. In two mains, the iodine was released from NaI packed into bollow UO2 cylinders by melting with a plasma torch in an air atmosphere. In two others, it was released from crys- talline Iz contained in a pyrex ampoule by melting the am- polle with the plasma torch. In the fifth run, iodine re- lease was effected by heating the furnace to 98°C and breaking an ampoule containing crystalline I2. Flowing helium trans- ported the iodine intn the model containment vessel. Iodine material balances were attempted in all runs, and the amount of iodine accounted for ranged from 50 to 80% of that initially present. Purging losses were suspected as one important contributing factor in the low iodine re- covery, and provisions to account for these were added. Significant conversion of the airborne iodine to formas other than Iz was observed. As much as 40% of the airborne iodine penetrated silver-plated screens, a 0.5-4 pore diame- ter membrane filter, and three layers of charcoal-im fiberglass filter paper and was collected on activated-char- coal beds. Presumably, this material was CH3 I. A second compound form was observed that was collected slowly by silver, but more rapidly by rubber, and which was quantitatively removed by charcoal-impregnated filter paper. There is no accepted identification of this form. The airborne iodine concentration was reduced by factors ranging from 4 to 40. The least reduction occurred in the runs where Iz was released by melting the pyrex ampoule with the plasma torch, and the greatest reduction occurred in the run in which I2 was released by breaking the ampoule. This sizggests . #THE PUBLIC IS APPROVED. PROCEDURES *Present address: Institut für Reaktorsicherheit der Technischen Überwachungs - Vereine, e.V. Koln, Germany,ENT CLEARANCE OBTAINED. RE OBTAINED. RELEASE TO Aleldrek Aponcerned dy ydttore ARE ON FILE IN THE RECEIVING SECTION. Comenutiñeinde érntract with the Union Carbide Carparation. 2 that the aerosol produced by the torch may affect the iodine behavior. However, there is no correlation between the iodine release mechanism and particle behavior. At most, 1% of the iodine appears to be associated with particles. Preliminary results of theoretical studies on diffusion in a spherical vessel are reported. These indicate qual- itative agreement with the observed time-dependent behavior. Introduction The Nuclear Safety Pilot Plant is a facility for investigating the behavior of fission products or simulated fission products in a contain- ment system that is intermediate in size between laboratory equipment and a full-scale containment vessel. An effort is being made to deter- mine the laws that govern the transport and removal processes in order to permit scaling of results from one size experiment or real contain- ment system to another. The effect of certain types of engineered safeguards can then be evaluated. It is hoped that the results will demonstrate that present criteria regarding the fraction of released fission products available for leakage are unnecessarily conservative. A flowsheet of the facility is shown in Fig. 1. Airborne fission products are generated by heating the fuel with a plasma torch in the furnace. This melting technique permits a wide range of atmospheres, pressures, and temperatures to be used. With the present furnace, fuel elements as Large as 1 in. in diameter and 6 in. long can be beated to temperatures in excess of 5000°F. The gases from the furnace, with their burden of solids and vapors, are conducted through a 3-ft-long transfer line into tbe model contain- This vessel is a 1350-ft3 stainless steel pressure vessel containing a number of sampling devices. There are also provisions for introducing sprays and for circulating the vessel atmosphere through a filter train to study means of removing fission products from the vessel atmosphere. A view of the major equipment, showing principally the furnace and model containment vessel, is shown in Fig. 2. The facility also contains provisions for purging the vessel through an existing off-gas system that has roughing filters, absolute filters, charcoal canisters, and a 100-ft stack. There is provision for decon- taminating the vessel and furnace by circulating suitable solutions. Sampling Process The sampling devices originally incorporated in the NSPP were in- stalled outside the model containment vessel. However, it was later necessary to install samplers inside the vessel because the existing flanges could not accommodate the numerous complex sampling units re- quired, and it was considered essential to minimize sampling line lengths. Since such internal samplers must be removed by remotely controlled manipulations before the vessel is decontaminated, and in some instances the samplers may have to be disassembled in a hot cell, ease of removal and disassembly is an important design criterion. For instance, the number of piping, electrical, and instrument con- nections is held to a minimum. The first samplers used were the standard commercial paper-tape air samplers mounted in pressure housings. These units have since been modified extensively to meet NSPP requirements. The present loading of these samplers comprises a nylon-reinforced membrane fil- ter tape followed by a charcoal-impregnated fiberglass felt tape. The need for handling double tapes and the complications introduced because the charcoal paper is quite thick forced most of cations. The air sampler housings are pressurized with instrument air controlled by biasing relays to maintain slightly higher pressure in the sampler housing than in the model containment vessel. Thus leakage should always be from the sampler to the vessel, and con- tamination of the sampler housing should be minimized. May Pack samplers are used for more detailed characterization of the iodine. The sampler has 12 May Pack units; each unit has a sers of different filter media, and each unit can be successively connector to a vacuum source by means of a sequential valve. The 12-unit device is shown in Fig. 3. Normally, samples of I to 10 liters are drawn at intervals during a run. The sequential valve and the vacuum pump that draws the sample are controlled from the panel board, and there is great flexibility with respect to sample size and timing. The makeup of the May Packs is also quite flexible. The loading in the most recent run is listed below in the order in which the sample contacted it: Kel-F interference ring Kel-F spacer Five Silver-plated 80-mesh copper screens Membrane filter, 0.5-4 pore diameter Stainless steel screen, 100-mesh Kel-F spacer Three ACG/B charcoal-impregnated fiberglass felt filters Stainless steel screen, 100-mesh Kel-F interference ring Kel-F tube holding above elements Four 1-inch thick charcoal beds compromising Kel-F interference ring Stainless steel screen, 100-mesh Coconut charcoal, 8-14 mesh, as required Stainless steel screen, 100-mesh Kel-F interference ring Kel-F tube container During the runs reported here, several variations of the above loading were used, and other changes are planned for future runs. The May Pack cluster was installed inside the vessel so that inlet losses were mini- mized. It was also desired to obtain data on fallout as a function of time. Originally a tape device was used, but this was judged unsatisfactory, and the carousel sampler shown in Fig. 4 was devised as a replacement. In this sampler a plate is rotated to place eleven sampling disks suc- cessively under a sampling opening. An Enercon vibratory motor, which gives very precise indexing, is used to rotate the plate. Samples can be collected on any convenient medium; membrane filter material has been used because of the ease of preparing specimens for electron micro- scope examination. It was suggested that a sampler comprising a large May Pack with a number of side streams be used, with each side stream being passed through a diffusion channel or a May Pack or both. Such a device is utilized in other programs at ORNL. It serves to characterize the iodine forms being collected by the several sections of the May Pack. However, the device described appeared to be inconsistent with the NSPP remote-handling criteria, and in its stead a sampler was built with five parallel channels, each of which consisted of a complete May Pack and a romposite diffusion tube. The diffusion tube inner surfaces are silver, rubber, and activated charcoal. In channel 1, the diffus-on tube is placed up-stream of the entire May Pack; in channel 2, the tube is between the silverplated screens and the re- mainder of the May Pack; and so on. It was expected that flow would divide uniformly among the five channels, but to date a uniform split has not been achieved. Further development work is needed in this regard. This device also permits evaluation of the precision of the samplers, since each channel deals with a portion of the same sample. A sampler comprising a diffusion tube and a simplified May Pack is installed to draw a sample of the gas from the line connecting the furnace with the model containment vessel at the point of entry to the vessel. This is used for sampling during the transfer period to char- acterize the aerosol that is discharged to the containment vessel. There are also provisions for installing up to 250 test coupons in the model containment vessel. The coupons are exposed throughout a run and the purge period, but they are normally removed before the vessel is decontaminated. With all these sampling devices in operation, the NSPP produces over 2000 samples for analysis in a run. In addition to the samplers, there is a condensation nuclei counter that is used to follow the particle concentration in the model containment vessel as a function of time. It is the General Electric Company instrument described by Skala. Sampling is inter- mittent, since continuous sampling would remove over 20% of the model containment vessel volume per day. Experimental Conditions for the Five "Iodine" Runs The experimental conditions are listed in Table 1 for the five runs (runs 3 through 7) in which some form of iodine was released to the containment vessel (runs 1 and 2 used unirradiated VO2 only). In uns 3 and 7, the simulated fuel was stainless steel-clad VO2 with iodine present as sodium iodide. Sodium iodide was selected because (1) it is the form in which 1911 is normally stored by the ORNL iso- topes division, (2) the fuel capsule end cap can be welded with no iodine evolution problem, and (3) it was judged that the sodium iodide would be completely dissociated at the temperature of release and that the iodine would be evolved as Iz if the furnace atmosphere were oxi- dizing. The experimental results confirmed this judgment. Tracer quantities of 13 I have been used with 100 to 1000 mg of stable iodine. If complete transfer is assumed, the concentration in the model con- tainment vessel should be 2.5 to 25 mg of iodine per cubic meter. These concentrations are in the range that might be expected as the result, of an actual core meltdown accident.? In the remaining three runs, the simulated fuel was crystalline iodine sealed in pyrex ampoules. The ampoules contained 100 mg of stable iodine with 1311 tracer. Thé iodine was released by melting the similated fuel element in all runs except No. 6, in which the furnace was preheated to 98°C, the ampoule was broken, and the iodine was swept into the model containment vessel by flowing helium. The rate of iodine transfer was followed with a gamma sensor mounted on the line connecting the furnace to the model containment vessel in all runs except No. 3. In runs 4, 5, and ó the transfer was practically instantaneous. In run 7, lodine transfer occurred over a period of 3 1/4 min, approximately the time required to pass the entire fuel piece under the plasma jet. Run 3 probably was sim- ilar to run 7 in this regard. Experimental Results e experimental results discussed bere give the chemical be. havior of the iodine and its airborne concentration as a function or time. The deposition coupon data obtained during these runs are being reported in another paper. The data indicate that at least three forms of iodine are present and that conversion from one form to another occurs during a rin. The three forms are: 1. A material with a diffusion coefficient in air of approximately 0.1 cm2/sec that is very readily collected op silver plate; it is pre- sumably 12. Table 1. Summary of Experimental Conditions of Runs 3 through 7 in Nuclear Safety Pilot Plant Run No. Initial Form of Container Material Mass of 1271 (mg) Release Temperature (°C) He Plasma Gas Flow (cfm) Shielding Gas and Flow (cfm) Sweep Gas and Flow (cim Fan Open Time (min) 20 100 ~2800 1.4 Air-1.2 0 100 Stainless Steel Pyrex Pyrex None None 10 au fo w UO2-NaI 12 12 12 002-NaI 1.4 1.4 ~1100 ~1100 98 ~2800 H20-0.5 100 100 1000 Pyrex Stainless Steel Air-1.2 N2-2.3 He-1.4 N2-2.3 None None 1.4 Air-0.7 10 2. A material with a diffusion coefficient in air of approxi- mately 0.02 cm2/sec that is slowly collected on silver plate, more rapidly collected on rubber, and very readily collected b There is no accepted chemical identification of this form, 3. A material with a diffusion coefficient in air of approxi- mately 0.1 cm/sec that is not significantly collected plate or by rubber but is collected by activated charcoal. This is presumably methyl iodide. In addition, the data discussed below indicate evidence of a higher molecular-weight iodine compound that penetrates to the acti- vated charcoal. It has a diffusivity of approximately 0.05 cm/sec in air. The May Pack data show that if the particle filter is upstream of the silver-plated screens, up to 39% of the total iodine is col- lected on the filter (although the percentage is usually much less) while if the particle filter is downstream, 1% or less of the iodine in the sample is found on the filter. From this it seems probable that little, if any, of the iodine is associated with particles, the 1% representing an upper limit. It is quite likely that most, if not all, of the iodine on the filters is adsorbed vapor compounds. When the particle liiter is downstream of the silver-plated screens, data on materials known to be present as particles (for example, uranium) indicates that the five screens together collect about 10% of the solids. The May Pack data on the various icdine forms show that elemental iodine normally predominates early in a run and that the rubber-seeking and charcoal-seeking forms become more prevalent later. Table 2 gives the maximum percentages of iodine found on the different collection mediums in runs 4 through 7. These values are indicative of the maximum percentages of the different forms present. The variation of these porcentages is somewhat erratic; for instance, in run 4, successive samples showed 19, 39, and 29% of the sample collected on the charcoal beds in the May Pack. Interpretation of the May Pack data is complicated by the fact that each of the collection mediums in the May Pack is effective to some extent for more than one form of iodine. For instance, the silver-plated screens apparently collect I2 with an efficiency of greater than 90% per screen, the rubber-seeking iodine form is col- lected with an efficiency of about 20%, and particulates are col- lected with an efficier.cy of approximately 2%. In addition, struc- tural components of the May Pack, such as spacers and support screens, coll.ect a significant fraction of the iodine. Usually, these bave been lumped with adjacent collection mediums in reporting results. In spite of these limitations, the May Packs are probably the most useful tool available for characterizing iodine. Much can be learned about the forms of iodine present without having to analyze an in- ordinate pumber of samples. However, anyone using May Packs or data generated by them should be aware of the complications. It has not been possible to identify the factors that influence conversion of iodine to forms other than 12. Table 2. Maximum Percentages of Different Iodine Forms Collected on Various Mediums at Different Times in NSPP Runs 4 Through 7 Run 4 Run 5 Run 7 Run 6 Collection Medium Most Probable Iodine Form Collected Amount. Time of Amount Time of Amount Time of Amount. Collected Sampling Collected Sampling Collected" Sampling Collected (min) (min) (min) (%) Time of Sampling (min) (%) 790 800 245, 365 12 96 Silver screens Particle filters 39l, e 0.3 12 Unknown 9.5 19.5 1438 - 478 178 9.5 2878 143 137.5 72.5 72.5 687.5 5 Charcoal papers 149 380 1565 Charcoal beds CH3 I 39 collected only at B level. Amounts include analyses of associated spacers and rings. At center of containment vessel. "At 5 ft below center of containment vessel. Particle filters upstream of silver screens. The combination May Pack and diffusion-channel device was used in runs 6 and 7. This unit provided data both on diffusion behavior to three different surfaces - silver, rubber, and activated charcoal - and on collection on the several components in a May Pack. Its principal disadvantage is that as many as 700 individual samples must be counted to obtain complete data on one air sample. Table 3 presents a summary of the data obtained with this device in run 7. The samples were taken at 1 liter/min over a 2-hr period starting 8 hr after meltdown. The tabulated data are arranged to inäi- cate whether the May Pack components are upstream or downstream of the diffusion tube. The total count per channel indicates that the flow did not distribute evenly among the five channels; rather, there is about a 5:1 range from the maximum to the minimum. The data show that 92.4 to 96.8% of the iodine was colleted on silver plate, either the screens or the diffusion channel. In channel 1, the deposit on the silver-plated tube showed two diffusion coefficients. I'ne more readily removed fraction had a dif- fusion coefficient of 0.09 cm/sec, which suggests that it was la. This represented 94.9% of the total on the silver-plated diffusion tube and 87.3% of the total iodine in the channel. The remainder had a diffusivity of 0.03 cm/sec. The second form has not been identified, but the similarity in the diffusion coefficient suggests that it may be the rubber-seeking form. In channels 2 and 3, only the low-diffusivity constituent was col- lected on the silver-plated tube, indicating complete removal of I2 by che silver-plated screens. This constituent; was not found on the dif- fusion tube in channels 4 or 5, indicating chat it was removed by char- coal paper. In channels ☺ and 2, a constituent with a diffusivity of 0.05 cm?/sec was collected on the charcoal-lined section of the diffusion tube. The diffusion coefficient is too low for this to be methyl iodide. Since this constituent was not present in sufficient quantity to produce a clearly defined. diffusion coefficient in channels 3, ht, and 5 and since the change made in channel 3 was to place the particle filter upstream of the diffusion tube, it seems likely that the par- ticle filter removed this constituent. Figure 5 shows data for the silver-plated screen sections in three of the channels. Data for the specers upstream of the screens are also shown to emphasize the relatively large fraction of the iodine collected on these components. The data show that the screens collect two con- stituents with different efficiencies. An attempt is being made to analyze the count versus position data to determine the amount of each constituent collected and the efficiency of the screens for each. The data on concentration in the vessel atmosphere as a function of position and time are likewise fairly complex. It has not been pos- sible in any of these runs to produce a uniform initial iodine concen- tration, although a fan is provided in the model containment vessel for Table 3. Summary of Iodine Collection Data Obtained with May Pack-Diffusion Channel Sampler in Run 7 2 33 5 93.3 1.2 92.7 0.8 3.3 96.6 0.3 1.5 1.3 0.2 0.1 1.3 Channel number Upstream May Pack Silver screens, go 95.3 Membrane filter, you Charcoal paper, ya Charcoal beds, le Diffusion tube Silver, ga 92.0 1.5 Diffusion coefficient, cm2/sec 0.088/0.03 0.026 Rubber, % 2.9 Diffi-ion coefficient, cm2/sec 0.0105 0.101/032 Charcoal, 1.3 0.8 Diffusion coefficient, cm²/sec 0.047 0.051 Downstream May Pack Silver screens, geo 0.4 Membrane filter 0.4 0.1 Charcoal paper, 0.5. 0.1 Charcoal beds, go 2.6 0.9 Total disintegrations per mimate 713,000 2,940,000 1.5 0.020 1.6 0.021 1.5 0.3 0.2 0.6 0.1 0.1 0.7 600,000 2.0 1,030,000 1,920,000 Percentage of total iodine in channel. نودي ت نهن أسب: پیتر :: .متهم ميامنننانس مغنم همه مهر و مه بر سینه ناحیه Vy .*- vVANANJiMTi: this purpose and it was operated in all runs except No. 3. Comparison of the data from run 3 with those from later runs indicates that the fan is partially effective. Figure 6 shows partial data on total iodine concentration versus time in run No. 6. Data from some of the samplers have been omitted in the interest of clarity. The data show a wide variation in the airborne iodino concentration with time at the loca- tions sampled. All the air samplers that collect samples at the vessel wall show a significantly higher iodine concentration than the May Pack samplers, which are located at a radius of 20 in. from the vertical cen- ter line of the vessel. Also the data indicate two or more maximums. The first maximum occurs approximately 100 min after meltdown. Later maximums occur at approximately 500 min and at about 2400 min, although these latter are not too strongly defined and may not be real. Figure 7 presents results obtained with the air sampler on the northeast side of the vessel at the top for all runs. Erratic behavior is indicated for runs 3, 4, and 7 and more orderly behavior for runs 5 and 6. Figure 8 presents data collected by the May Pack at the center of the vessel for runs 4 through 7 only. Since these results are more or- derly than those from the air samplers, deposition velocities can be calculated. - '.-4.-•-... LI. ,...-.-'. . .-.-Sav.. Theoretical Analysis of Iodine Behavior in Containment Vessel ...- 1 -..- . MENA.A Y .V SLES In order to gain some understanding of the processes that might be i operating in the model containment vessel, a theoretical study of the concentration-time history that might be anticipated as a result of dif- fusion was undertaken. The study was limited to spherical geometry to simplify computation. The following cases have been investigated to date: 1. A mass, M, of a diffusing constituent is introduced at the center of the vessel as t = 0, and the vessel wall is considered to be a perfect sink. 2. A mass, M, of a diffusing constituent is introduced at the center of the vessel at t = 0, and the vessel wall is considered to be a perfect reflector. 3. A mass, M, of a diffusing constituent is uniformly distributed through the vessel volume at t = 0, and the wall is considered to be a perfect sink. General solutions have been obtained in terms of three dimension- less parameters: 1. Concentration, C, which is the ratio of the concentration to the average initial concentration, M/V, i.e., C = CV/M. 2. Radius, R = r/A, where r is the actual radius at the point being considered and A is the vessel radius. 3. Time, T = Dt/A, where D is the diffusion coefficient of the substance being studied ard t is time. ALL rzienin ndekilen U.. . . i K - - ! ' .' 9.45 , , . 12 Figure 9 presents the results of this study for cases 1 and 2 for relative radii of 0.1, 0.5, and 0.9. For case 1 (C on Fig. 9), the dimensionless concentration rises to a maximum and then decreases ex- ponentially. At R = 0.1, the maximum occurs at T < 0.01 and does not show. The exponential decay curve has the same slope at all radii, which should have_the value 72D/A2. For case 2 ( Ton Fig. 9), the curves are identical to those for cesc I for short times where the influence of the boundary is negligible, but they subsequently break off and asymtotically approach C = 1. So far, it has not been possible to solve the real case in which the boundary behaves in a manner intermediate between cases 1 and 2; that is, some form of phase equilibrium exists at the boundary. How- ever, much information about the behavior can be inferred from the solutions discussed above. Ato short times, the case I curves should be applicable. At very long times, if the phase equilibrium laws were known, it would be possible to calculate the asymptotic value of C. The equilibrium law is not known at present. This would leave only a fairly short transition period in which concentration-time behavior could not be predicted analytically. The behavior predicted by the above analysis shows considerable similarity to the results obtained by Megaw and May+ and by Croft and Iles, as well as to the NSPP data. Analytical solutions of the dif- fusion equation for cases similar to 1 and 2 but with the mass M intro- duced on a shell R = Ro, not 0, are available, and this extension of the analysis is now being computed. The theoretical analysis indicates that the deposition velocity concept is valid only in the time period where case 1 behavior governs.... Where an investigator reports a high initial deposition velocity fol- lowed by a lower value at later times, he is reporting approximations to case ? behavior. Corresponding times for different containment systems will vary as 7a, where 1 is the appropriate characteristic dimension for the vessel in question. Turning baok to the experimental results, it appears that the observed behavior is qualitatively consistent with a diffusion-con- trolled process from a nonuniform initial distributio analytical investigation should enable the appropriate model to be more clearly defined. cks of the parameter Dt/A2 indicate that for large systems, if molecular diffusion controls, very long times will be required to approach equilibrium. In the NSPP, if the radius of a sphere of equal volume is taken as an appropriate characteristic length, periods of 28 to 86 hr might be needed to attain equilibrium if the final value of C is in the range 0.01