ML20012C932
| ML20012C932 | |
| Person / Time | |
|---|---|
| Issue date: | 08/31/1989 |
| From: | Hurt R NRC OFFICE OF NUCLEAR MATERIAL SAFETY & SAFEGUARDS (NMSS) |
| To: | |
| Shared Package | |
| ML20012C912 | List: |
| References | |
| REF-PROJ-M-32 NUDOCS 9003260134 | |
| Download: ML20012C932 (20) | |
Text
{{#Wiki_filter:_ _ _.... _... Review of West Valley Demonstration Project Vitrification Off-Gases R. Davis Hurt West Valley Project Manager Advanced Fuel and Special Facilities Section Fuel Cycle Safety Branch August 1989 s $4 $, -)4L'; 9003260134 900318 PDR PROJ PDC t4-37 l l l
s t*
- .s*j
^ ..r ~ ' 1. H-3 Estimated inventory to melter: 0.65 C1 (1992) Confidence of estimate: moderate Most of the approximately 100 Ci of H-3 contained in the HLW in 1987 was in the form of tritiated water (HTO). As the supernatant is processed, a majority of the water originally in the HLW will reach Buttermilk Creek through t!.e Liquid Waste Treatment System (LWTS) and the lagoon network. Most of the rest U will be'used to hydrate cement for solidifying the-decontaminated salt slurry. The only.H-3 that will reach the melter will be from any residual supernatant in the water left as cover for the sludge, which should be a very small amount, or from hydroxides in the sludge. The H-3 inventory in the sludge is somewhat uncertain. -The moles of-hydrogen in the sludge amount to less than one percent of the moles of hydrogen in the supernatant, so an estimate of less than one curie as the H-3 sludge inventory seems reasonable. On the other hand, the reference sludge composition compiled in 1986 by Rykken includes 7.3 C1 (1987) of H-3, which would decay to 5.5 Ci by 1992. Rykken's report does not discuss how this value was obtained, nor is it clear how the WVDP's estimate of 0.65 Ci- ~ (1992) can be reconciled with the Rykken figure. ' Analysis for H-3.in the new sludge samples to be taken in 1990 may clarify the situation. For now we can say that the-inventory of_H-3 to reach the melter will probably be less than one curie and almost certainly less than ten curies. 10'CFR 20 airborne release limit: 4 x 10~5 pCi/mL Relative significance as an off-cas component: low Virtually none of the H-3 that reaches the melter is expected to dissolve in the glass. It will all enter the off gas system, probably mostly as HT0 vapor, perhaps to some extent as H2 or other gases, although tests at Savannah River formation in their melter.9 Most of the Laboratory (SRL) showed very little H2 HTO vapor will be condensed along with the other vapor, recirculated, and 1
m .- j x eventually discharged as-liquid-. effluent through the LWTS. A very_small-percentage'of-the water vapor produced at the melter will be exhausted from p the' plant' stack. This quantity, plus'any-H-3 forming H2 or another gas in the 4 melter, will be the only airborne H-3 to consider.- Until more information on-the H-3 content of the; sludge and on the chemistry of hydrogen in the melter off gases'is'available, a maximum of 0.1 Ci of H-3 discharged from the_ main stack during vitrification would be a conservative estimate. Dilution by 1016 mL_ of air throughput (which comes from multiplying the designed air flow of-l 50,000 standard cubic feet per minute by an estimated processing duration of .l 11,000 hours) will result'in an average stack concentration of around 10-10 1 - pCi/mL. This is 4.x 10s times'below the Part 20 limit'for H-3 concentration at-the site boundary, i 4 i E
- (
? 2
e 2.. C-14 Estimated inventory to melter: 1.3 Ci Confidence of estimate: moderately low Heasurement of supernatant samples in 1985 yielded a C-14 inventory estimate of 137 Ci and showed that the main form of carbon in the supernatant was dissolved carbonate. C-14 analysis was not performed on the 1985 sludge samples, but the carbonate inventory was measured and indicated that the sludge contains about 5 percent as much carbonate as the supernatant. The reference sludge inventory published by Rykken in 1986 lists 0.1 Ci of C-14, with no discussion of the basis for this value. Carbonate ratios would suggest a C-14 value for the sludge of around 7 curies (5 percent of 137 Ci). A Thorex C-14 inventory of 0.13 Ci was based on analysis of three Thorex samples in 1985. It is not helpful to check measured C-14 values against theoretical ones that might be calculated with ORIGEN or other burnup codes. C-14 is formed from neutron activation of N-14 impurities in the Inconel, stainless steel, and (to a lesser extent) Zircaloy structure of the fuel, and from N-14 and 0-17 in the fuel itself.10 It.is not practical to try to estimate what small fraction of the C-14 formed in the spent fuel might have made its way into the HLW. It is not clear at this time how the WVDP's estimate of 1.3.01 to the melter is to be reconciled with the measurements reported by Rykken. Only C-14 in the sludge or Thorex waste will reach the melter. Summing Rykken's values for these two gives 0.23 Ci, but the inconsistencies described above recommend against having too much confidence in this or any other estimate for the time being. When new sludge samples are analyzed in 1990 a firmer estimate can be made. 3
10 CFR 20 airborne release limit: 10-6 pCi/mL as CO2 Relative significance as an off-cas component: low Any C-14 reaching the melter can be expected to form C0 or 00, neither of which t will be removed from the off gas. If the WDP estimate of 1.3 Ci is correct, the concentration of C-14 at the stack will be about 10-9 pCi/mL (dilution by 1015 mL of air throughput), This is three orders of magnitude below the Part 20 site boundary limit, so even if the C-14 inventory of the sludge turns out to be considerably higher than reported by Rykken, C-14 should not be a problem. h 4
y . l ~ 3. Sr-90/Y-90 Estimated inventory to melter: 13.1 x 108 Ci (1992) Confidence of estimate: fairly high i Sr-90 is by far the major source of activity in the Tank 80-2 sludge. The inventory estimate is based on two core samples taken in 1985. The WVDP deter-mined that the relative uncertainty in the Sr-90 inventory estimate attributable to random counting error was only 1.0 percent at the 95 percent confidence level. At the same time, they acknowledge that the representativeness of the samples is open to some question. Both samples were taken from the same general part of Tank 8D-2, and the soft sludge from the upper parts of the cores was probably lost during retrieval. This potential sampling error affects estimates of all radionuclides found mainly in the sludge, and can only be addressed by taking more samples. In the case of Sr-90, burnup code calculations agree closely with 3 the 1985 sample analysis results, so confidence in this inventory estimate is reasonably strong. Further sludge sampling will be done in 1989, with most analysis in 1990. 10 CFR 20 airborne release limit: 2 x 10-10 pCi/mL (insoluble) Relative significance as an off-gas component: high St-90 is not volatile at molten glass temperatures, so losses to the off gas system will be exclusively by entrainment. Even with low air flow rates through the overhead space above the molten glass, large amounts of radioactivity can be entrained into the off gas system as small droplets of feed material, as bits of ash from the cold cap (a crusty pile of slag-like material that tends to accumulate on top of the molten glass), or as tiny particles of molten glass sputtered into the air. The amount of entrainment experienced seems to depend on the exact details of the melter design, the rate and steadiness of the feed, and amount of N0x evolved.4'8 5
r: i Experiments with nonradioactive strontium have been run at the WVDP and at Pacific Northwest' Laboratories (PNL).4 The WVDP tests have used a full-sized melter exactly the same as the one that will be used in hot operations.2 PNL-used a smaller melter of the same general type (i.e., a slurry-fed ceramic melter) but with design differences that may have affected some of the results. Both organizations used the same downstream off gas treatment components: a submerged b'cd scrubber (SBS) and a high-efficiency mist eliminator (HEME). The results are summarized below in terms of decontamination factors (DFs), defined as the ratio of input strontium content to output strontium content. In other words, an off gas DF of 100 means that one part in a hundred of the material entering the piece of equipment exits in the off gas. PNL WVDP i i Melter off gas DF 430 11 SBS off gas DF 1,300 900 HEME off gas DF M 50 i Combined off gas DF 2.5 x 107 5.0 x 105 i The most interesting feature of this is.the low melter DF at West Valley. The melters at West Valley and PNL are sufficiently.similar in design that j such a large difference in entrainment (39 times more at West Valley) is j surprising at first, but it may be possible that small design details can have 1 a big effect'on entrainment losses. The lower position of the off glass l ' plenum relative to the molten glass surface has been suggested as a partial I explanation for the poorer' melter performance at West Valley. It should be pointed.out that some of the DF figures shown above are averages. For example, melter DFs at PNL were measured over nine separate campaigns, with_ individual strontium DFs varying from 75 to 1,200. The WVDP melter DF for strontium-is an average of two individual campaign values, 10 and 12. i The PNL staff was disappointed in the initial performance of their HEME.4 They had problems maintaining the dew point of the filtered gas low enough, with the l 6
- result that their filter did not stay dry as expected. They were subsequently able to improve the HEME's performance by lowering the gas velocity, but this experience suggests that it is a tricky piect of equipment to operate and that periods of poor performance may be encountered during hot operations. The Savannah River Laboratory expects a DF of arour.d 40 on their HEME.8 In hot operation, the off gas will probably pass through two HEPA filters after the HEME, one inside the cell and one outside. (The design of this system is not final.) It will not be possible to test the in-cell HEPA once hot operation commences and the in-cell filter may have to be replaced f airly of ten. The outside HEPA will be accessible for ordinary types of in place testing. Since the SBS and HEME are most efficient in removing large particles, most of the particles reaching the HEPAs will be small, probably less than 0.3 microns in diameter according to PNL experience. A DF of 200 is normally considered conservative for well-installed and well-maintained HEPA filters for 0.3 micron particles. Because of the smaller particle sizes and the inability to test the first filter in location, a somewhat lower DF may be a reasonable assumption for the WDP. Assigning a value of 100 to each of the two HEPAs, for an overall HEPA DF of 104 woo d probably be conservative. A total system DF for Sr-90 for the melter, SBS, HEME, and HEPAs would be about 5 x 108 based on West Valley's experience so far. This DF, combined with dilution of the off gas by 1015 mL of air throughput, would give a Sr-90 concentration at the stack of 3 x 10-12 pCi/mL. This is comfortably below the Part 20 site boundary limit (2 x 10-20 pCi/mL) and is based on fairly conser-vutive assumptions, so it appears that the WDP off gas system will be adequate to control Sr-90 emissions. 7 1
l i 4 Tc-99 [ Estimated inventory to melter: 14 Ci + Confidence of estimate: moderate 7 j Sample analysis in 1985 indicated approximately 1,600 Ci of Tc-99 in the Tank 8D-2 supernatant. Nearly all of this material will reside in the cement drums when supernatant processing is finished. Tc-99 analysis was not performed on the 1985 sludge samples, but burnup codes predict that 1,600 Ci is almost exactly the amount of Tc-99 that should have originally been present in the spent fuel ( received at West Valley, so there is no reason to suspect significant quntities in the sludge..The only Tc-99 to reach the melter will be from the Thorex waste -and from whatever small quantity of supernatant remains in Tank 8D-2 when supernatant processing is complete. 10 CFR 20 airborne release limit-7 x 10-8 pCi/mL (soluble) Relative significance as air off-gas component: moderate .i Tc-99 that reaches the melter can be expected to evaporate to a substantial 't -extent, depending on its oxidation state. Evaporated Tc-99 will condense in the air. space above the glass and much of it will probably enter the off gas system as fine particles. Experiments at Savannah River Laboratory (SRL) confirm-this behavior.0 Experience with cesium, another semi-volatile element, has shown that material that evaporates from the glass and condenses in the overhead space is likely to form finer particles than entrained material. So j the_.off-gas system DF will probably be somewhat smaller for Tc-99 than for f 8 Sr-90 (5 x 10 ), but there will be so little Tc-99 going into the melter that ' it will still not be a problem. With dilution by 1015 mL of air throughput, f 'the average Tc-99 at the stack would only be around 10-6 Ci/mL, even with no Tc-99 DF at all. So only a very modest DF will be needed to reduce the stack concentration of Tc-99 below the Part 20 site boundary limit. + 8
r 'i* 5. Ru-106/Rh-106 Estimated inventory to melter: 8.4 Ci (1992) Confidence of estimate: moderate Ru-106 concentration has been measured in the Thorex waste and in the Tank t 80-2 sludge, showing 1987 inventories of 1.25 Ci and 260 Ci, respectively. A measurement of Ru-106 concentration in the Tank 8D-2 supernatant was attempted, but no Ru-106 was detected. From the sensitivity of the technique, WDP deter-L mined that the supernatant could not contain more than 5.6 Ci of Ru-106 (as'of 1986), If the total HLW Ru-106 inventory in 1987 was a little over 260 Ci, the 1992 inventory would be approximately 8 curies. An inventory of 260 Ci in 1987 is fairly close to theoretical predictions; burnup codes would suggest that the HLW in Tank 80-2 should have contained about 100 Ci of Ru-106 at that date. 10 CFR 20 airborne release limit: 2 x 10-20 pCi/mL(insoluble) Relative significance as an off-aas component: moderate Data on the behavior of ruthenium in slurry-fed melters have been collected at SRL and'PNL. The extent to which ruthenium will volatilize depends on its oxida-tion state; Ru0 is the most volatile form. Any ruthenium in this state would 4 be expected to evaporate from the molten glass and condense into fine particles of Ru02 or ruthenium metal in the overhead space.8 Such fine particles would -not be removed as effectively by the downstream equipment as would entrained material such as Sr-90. But most of it should still be removed. PNL tests achieved average DFs of 29 in the melter, 16 in the SBS, and 23 in the HEME, for a combined DF of 104 prior to the HEPA filters.4 The two HEPA filters togetiier would be likely to achieve as much again, despite the sma11' particle size. Combined with dilution by 1025 mL of air throughput, a total system DF of 40 would reduce the Ru-106 concentration at the stack exit to the Part 20 site boundary limit. A DF several orders of magnitude larger than that will almost certainly be obtained. If vitrification startup is delayed a few years, there will be even less need to worry a'sout Ru-106, with its 1 year half-life. 9
t L 6. 1-129 Estimated inventory to melter: 0.19 Ci Confidence of estimate: low I-129 concentration was measured in the 1985 Tank 8D-2 supernatant samples and extrapolated to an inventory of 0.21 Ci in the supernatant. This material will end up in the cement drums in the course of supernatant processing. The Tank 8D-2 sludge was assumed to contain no I-129 and no measurement was attempted on the 1985 sludge samples.. An 1-129 measurement was attempted on samples from the Thorex waste, but no I-129 was detected. Based on the sensitivity of the technique, the I-29 inventory in the Thorex waste was concluded to be less than 0.18 C1. The estimated inventory to the melter of 0.19 Ci is apparently the sum of the maximum possible Thorex I-129 and residual amounts from the Tank 80-2 supernatant. Burnup codes indicate ~that the spent fuel processed at West Valley originally contained approximately 4 Ci of I-129. A substantial percentage of this may have escaped into the off gas system during dissolution, and it is possible that some=I-129 may have evolved out of the HLW during twenty years of storage. Nevertheless, the burnup codes suggest that the Tank 8D-2 inventory of I-129 could be higher than 0.21 Ci, implying that there may be a curie or two of I-129 in the sludge. The 1990 analysis of new sludge samples will include I-129 measurements and should clarify this issue. 10 CFR 20 airborne release limit: 2 x 10-11 pCi/mL (soluble) Relative significance as an off-cas component: moderate The WVDP assumes that all I-129 reaching the melter will convert to gaseous -form and: escape the off gas system entirely. Were this to be the case, and assuming that the WVDP estimate of 0.19 Ci is correct, the average 1-129 con-centration at the stack would be about 2 x 10-10 pCi/mL, ten times the Part 20 site boundary limit. As discussed above, the WVDP inventory estimate may be low. On the other hand, it is very conservative to assume that all I-129 reaching the melter will escape as airborne contamination, which is equivalent 10 L
N4. W w1 u c,a j.3,, / $ 5 = ?. " 4- "i 7%j.y ,g.. jj ~ ' 7 .':x ^ m .(
- l.
- 1 k
E .to assuming that all 1-129 reaching the melter will: convert to a gaseous form. l e [,
- Tests st PNL indicate that as much as half of the iodine may be retained in the glass; If the 1990 sludge analysis shows significant amounts of I-129 in j
f the West Valley sludge' the issue'of its chemistry in the melter will have to j a e be revisited.. I h ar1 i I p c-. .t I a s wy- .l 1 e -i4 t 5 I 9 J if (! 1 + 5 r 4 s t _s+, e -{ \\ .y I - e t F f f .11 --I -a :;w 1 N'{5 3;-- l s l 1,. ? ~.,
y ,a t i 7. Cs-137/Ba-137M i w Estimated inventory to melter: 13.5 x 100 Ci (1992) Confidence of estimate: high Virtually all of the Cs-137 in Tank 8D-2 is dissolved in the supernatant, where its concentration has been accurately measured. The Cs-137 inventory of both i Tanks 8D-2 and 8D-4 is known with higher confidence, in terms of percentage I uncertainty, than any other radionuclide. The relative uncertainty attributable i to random counting error was only 0.2 percent at the 95 percent confidence level .in the 1985 sample analysis. And sampling errors for the fairly homogeneous supernatant, while not quantifiable, were certainly small compared to sludge sampling errors. 1 10 CFR 20 airborne release limit: 2 x 10'8 pCi/mL (soluble) Relative significance as an off-aas component: high There are two mechanisms by which significant amounts of Cs-137 can enter the off gas system:
- 1) entrainment of bulk feed slurry, glass droplets, or ash from the cold cap-and 2) evaporation of Cs-137 directly from the molten glass.
-Cesium has a fairly high vapor pressure at 1175 C, the temperature of the molten glass. Evaporating cesium will condense in the overhead space, mostly into particles smaller than entrained feed material. Keeping the off glass plenum temperature as low as possible encourages rapid condensation and minimizes-the escape of Cs-137 from the melter. SRL concluded that Cs-137 losses from their hot test melter were "primarily" by evaporation, as opposed to bulk entrainment.8 The ratio was not quantified, but-their entrainment losses were generally low, less than one percent. PNL found cesium losses from the melter ranging from 1.5 times to 8.8 times higher than strontium losses in nine test campaigns, an average of 4.4 times higher.4 Since entrainment losses of cesium and strontium should be nearly equal, this gives an estimate of the extent and variability of cesium evaporative loss. 12
I I i,- .,.o In two cold tests, the WDP found cesium losses to exceed strontium losses by 1.5 times and 2.1 times.2 Bear in mind that the WDP melter showed much higher absolute rates of entrainment than the PNL melter. Factors that affect cesium volatilization have been studied in considerable detail by PNL and SRL. The PNL staff believes that halogen concentration in the molten glass-is, of the things they have studied, the most important factor controlling cesium volatilization; more important than plenum temperature, feed rate, or waste loading.6 Fortunately, the WDP melter feed should contain only low levels of halogens. Sodium cf.loride is a significant constituent of the Tank 8D-2 supernate, but no chlorides or fluorides have been measured in the sludge or the Thorex waste; and only trace amounts of supt:rnate should reach the melter. The SRL staff is more concerned about borates. They have found evidence that cesium evaporated from their test melter primarily as cesium metaborate (CsB0 ).7 The Thorex waste contains 2 about 500kg of boric acid (H B0 ), which is more than enough boron, in 3 3 stoichiometric terms, to react with all of the cesium in the West Valley high-level waste. Cesium that evaporates and condenses in the overhead space generally forms smaller particles than entrained cesium; thus it is to be expected that the SBS and HEME will perform less efficiently on Cs-137 than on Sr-90. In their nine campaigns, PNL found that the SBS worked an average of 200 times better on strontium than on cesium, with individual campaign results varying between 7 times and 400 times. In the one test campaign for which they collected such i data, the WDP found that their SBS worked 7.5 times better for strontium. (SRL's scrubber is of a completely different design, so their results are not especially relevant.) For the HEME, PNL found strontium removal 2.3 times better than cesium removal. The WDP found that their HEME worked 3.6 times better on strontium. The experi-mental DFs for Cs-137 at PNL and the WDP are summarized as follows. I 13
,3,,, l-PNL WVDP g Melter DF 97 6.5 SBS DF 6 120 r HEME OF 19 14 7 Combined DF 1.1 x 10'* 1.1 x 10-4 i. \\g PNL's combined cesium DF is over 2,000 times poorer than their strontium 0F, i The WVDP's combined cesium DF was poorer by only 45 times because their L strontium DF was so much. lower to begin with. HEPA filter efficiencies for Cs-137 will probably not be much lower than for Sr-90, by the argument that most of the particles reaching the HEPA trains will be equally small for both radionuclides, most of the larger particulates ~ having been effectively removed in the SBS and HEME. An overall value for the two HEPAs of 104 would probably be reasonable and conservative'for Cs-137 as well as Sr-90. So the overall Cs-137 DF prior to exhaust would be a little over 108 Combined with dilution by 1016 mL of air throughput, the average'Cs-137 .:c concentration at the stack would be about 1.3 x 10~10 pC1/ml, one order of magnitude' below the Part 20 site boundary limit for soluble Cs-137. l. 1 14
- 4 8. Pu-238 - Estimated inventory to melter: 6,700 Ci (1992) Confidence of estimate: moderate Analyses,in 1985 indicated that 98 percent of the plutonium in tank 8D-2.is in the sludge. As discussed previously..there is considerable uncertainty as to the representativeness of the sludge samples, so estimates of sludge-inventories-1 should be viewed cautiously. Burnup code calculations of the amount of Pu-238 that should be present in Tank 8D-2 are four times lower than the inven-tory estimate based on the sludge samples. The new sludge sampling program scheduled for'1990 will-shed further light on this question. In the meantime,. l the estimate of 6,700 Ci based on sludge analysis can be considered as possibly J on the high side. 10 CFR 20 airborne release limit: 10-22 pCi/mL (insoluble) Relative significance as an of f-aas component: moderate l Plutonium is non-volatile at molten glass temperatures; its entry into the off gas system will be by entrainment only. For obvious reasons, cold tests cannot include plutonium or other transuranics. Cerium was used at the WDP to simulate transuranics in cold campaigns.2.The WDP's off gas system was found to have a cerium DF more than ten times better than its strontium DF. The mean-ing, if any, of such a difference between two elements believed to be completely f non-volatile is not clear. Nor is it clear whether this has any implications for plutonium DFs. PNL and SRL do not have any direct information on plutonium DFs. 1 - If we assume an overall Pu-238 0F of 5 x 109, the same as for Sr-90, the-concentration of Pu-238 at the stack would be 1.3 x 10-25 pCi/mL, compared to a site boundary limit of 10-22 pCi/mL for insoluble Pu-238. N 15
pm 3 v f- ;.,.* \\; ' 9.< Pu-239 i:. '. - f 4 Estimated inventory to melter: .1,700 Ci p L i Confidence of estimatei fairly high f r - As? for Pu-238, 98 percent of the Pu-239 in Tank 8D-2 is; believed to be in the
- sludge, Therepresentativenessofthe1985sludgesamplesissufficihntly 4
k: uncertain to recommend some conservatism in judging inventory estimates based [ on_these samples. 'In the case of Pu-239, however,.the burnup codes support y the estimate based on sludge sample analysis fairly closely. i 10 CFR 20 airborne release limit: 10-12 pCi/mL(insoluble) Relative significance as an off-cas component: moderate As an off gas constituent Pu-239 should behave exactly the same as Pu-238. . Since there is somewhat less Pu-239 (in activity terms), and since the two
- 7. 7; have_ the same Part 20 site boundary limits as airborne ~ contaminants, the Pu-239 can be' viewed as slightly less important than the Pu-238.
-1 r u l', p. l g-I 16 p
m. 7,,, ,,, o _ g,e j L
- 10. : Pu-240 t
Estimated inventory to melter: 1,300 Ci Confidence of estimate: fairly high' The same observations as for Pu-239 apply. In the case of Pu-240, the burnup codes predict a-slightly lower inventory (about 1,000 Ci). p 10 CFR 20 airborne release limit: 10-12 pCi/mL (insoluble) ' p Relative significance as an off-cas component: moderate The inventory of Pu-240 is nearly the same as for Pu-239, in activity terms. The Part 20 limits and chemical properties are essentially identical, so the relative importance of Pu-240 as an off-gas component is about the same as Pu-239. L I k 6 i k 1 17 a:L
N W ',. ' ',, ;s ' 51 ' 11. P -241 Estimated inventory to melter: 68,000 C1 (1992) Confidence of estimate: fairly high L The same observations as for the other plutonium isotopes apply.. The Pu-241 inventory predicted by burnup codes is slightly lower (66,000 C1). 10 CFR 20 airborne release limit: 10-9 pCi/mL (insoluble) l i Relative significance as an off-aas component: moderate The fact that the inventory of Pu-241 is ten times larger than Pu-238 is o ti compensated by the fact that its Part 20 site boundary-limit is a thousand times higher. Pu-241 actually verges on being of low significance by the definition purveyed in the cover note. That is, dispersion would reduce its concentration almost to the Part 20 level at the site boundary even with no. ~~ off gas treatment. L '1 18 l
y w lc (' l l: 12. Am-241 4 1 Estimated inventory to melter: 71,000 C1 (1992) [ Confidence of estimate: fairly high r i [ Virtually all of the Am-241 in Tank 8D-2 is believed to be in the sludge. As L discussed above, sampling uncertainties encourage some caution about sludge . inventory estimates, although the measured Am-241 inventory agrees closely with p ' predictions based on burnup codes. Uncertainties in Pu-241 inventory contribute to uncertainty in future Am-241 inventories since Pu-241 decays to Am-241. In the next few years Am-241 decay will approximately balance Am-241 ingrowth from f i L decaying Pu-241. L i 10 CFR 20 airborne releasn limit: 4 x 10-12 pCi/mL(insoluble) Relative significance as an off-aas coteponent: high As with plutonium, it is not possible to include americium in cold tests. The WDP used cerium to simulate americium in cold tests. As discussed for Pu-238, f the WDP' tests showed-cerium DFs about ten times better than strontium DFs. The significance of this is not clear. PNL and SRL do not have any data on -Am-241 DFs. l A reasonable assumption for now would be that the off gas system will achieve an Am-241'DF of 5 x 109, the same as for Pu-238, Sr-90, and other non-volatile isotopes. Dilution by-1015 mL of air throughput combined with this DF would give an Am-241 concentration at the stack of 1.4 x 10'14 pCi/mL, compared to a Part 20 site boundary limit.of 4 x 10-12 pCi/mL for insoluble Am-241. i l e 19 I -., -.}}