ML18331A325

EDF-1239, In-Line Rupture Disks for TMI-2 Canisters During Drying

ML18331A325
Person / Time
Site:	07200020
Issue date:	09/30/1999
From:	Lockheed Martin Idaho Technologies Co
To:	Office of Nuclear Material Safety and Safeguards
Shared Package
ML18331A337	List: ML18331A317 ML18331A318 ML18331A319 ML18331A320 ML18331A321 ML18331A323 ML18331A324 ML18331A325 ML18331A326 ML18331A331 ML18331A336
References
CLN190154
Download: ML18331A325 (32)

Text

{{#Wiki_filter:Document ID: EDF-1239 Revision: 0 IDAHO NATIONAL ENGINEERING & ENVIRONMENTAL LABORATORY In-Line Rupture Disks for TMl-2 Canisters During Drying September 1999 412.14# 03-97 Rev. #01

INELFORM L-0412.15# (08 Rev. #00) DOCUMENT MANAGEMENT CONTROL SYSTEM (DMCS) DOCUMENT APPROVAL SHEET Document Type: EDF (Analysis) Document Identifier: EDF-1239

Title:

In-Line Rueture Disks for TMI-2 Canisters During Drying Author: A. J. Palmer 1¥v~ q/2.*/'I~ Phone: 6-8700 Document Owner: J. H. McDaniel Phone: 6-3935 REVIEW CONCURRENCE AND APPROVAL SIGNATURES Denote R for review concurrence, A for approval, as appropriate Type or printed name RIA Date Organization Mailing Address Discipline Signature J. H. McDaniel J.lJ,,/l, A 6310 3114 ~A,,~ ~,.

• ~

""* I q-t9--99 Project Management (/ I (Requester) R. G. Rahl A 41AO 3760 vA J~~) vt(-wl°l°\\ Applied Mechanics C.A.S~ I /) A 4130 3765 f-2'1-99 Mechanical Engineerjng ~ *'/lvl- <;f- ~ I I

SUMMARY

Too many variables remain unknown to confidently predict whether or not the canister seals will leak during drying as result of blocking the canisters' drain ports with a rupture disk. Theoretical considerations would imply leakage to be likely, but in the two tests conducted (which were admittedly non-prototypical) no evidence of leakage was observed. In the Author's opinion, the majority of the evidence leans toward the conclusion that no leakage would occur with a blocked drain port, but again, only a true prototypical test, or series of tests, will provide conclusive answers. Flow calculations indicate the potential leakage to be no greater than 20% of total flow. In a related issue identified during this study, there appears to be a real chance of relaxing the bolt torques on the fuel canisters due to o-ring thermal expansion. It is believed that this will not alter the canisters' containment boundary function, however the project should evaluate whether there are any other issues associated with potentially loose head bolts on the fuel type canisters.

In-Line Rupture Disks for TMI-2 Canisters During Drying EDF-1239, Rev. 0 p. 2/15 CONTENTS

1.0 BACKGROUND

....................................................................................................................... 3 2.0 0 -RING INTEGRITY............................................................................................................... 5 2.1 Theoretical............................................................................................................................. 5 2.1. l Current Condition of 0-Rings........................................................................................ 5 2.1.2 0-Ring Performance During Canister Heatup............................................................... 7 2.2 Test Data............................................................................................................................. 11 2.2.1 Vectra Testing.............................................................................................................. 11 2.2.2 INEEL Testing.............................................................................................................. 11 2.2.3 0-Ring Geometry......................................................................................................... 11 2.2.4 Test Configuration vs. Actual...................................................................................... 12 3.0 MAXIMUM LEAKAGE......................................................................................................... 12 3.1 Pressure Generated.............................................................................................................. 12 3.2 Gap Size.............................................................................................................................. 13 3.3 Flow Through Gap.............................................................................................................. 13 4.0 MISCELLANEOUS ISSUES..........................................................,...................................... 14 4.1 Effects on Knockout and Filter Canisters............................................................................ 14 4.2 Potential for Relaxation of Bolt Torque on Fue] Canisters................................................. 14 5.0

SUMMARY

............................................................................................................................ 15

6.0 REFERENCES

........................................................................................................................ 15 APPENDIX A - ANALYSIS PLAN........................................................................................ Al APPENDIX B - CALCULATIONS......................................................................................... B 1

In-Line Rupture Disks for TMI-2 Canisters During Drying EDF-1239, Rev. 0 p. 3/15 IN-LINE RUPTURE DISKS FOR TMl-2 CANISTERS DURING DRYING

1.0 BACKGROUND

The TMI-2 canisters are dried in a vacuum furnace prior to being placed in dry storage. While in the furnace, the canisters are fitted with sintered metal filters to minimize the quantity of fugitive radioactive material available to contaminate the furnace and downstream vacuum equipment. In the basic configuration "by-pass" rupture disks are installed just below the filters to provide pressure relief in case the filters plug. This configuration is shown in Figure 1 below. Figure 1. "By-Pass" Rupture Disks on Both Canister Ports

In-Line Rupture Disks for TMI-2 Canisters During Drying EDF-1239, Rev. 0 p. 4/15 In the basic configuration, the rupture disks provide a by-pass path around the filters should one or both of the filters plug. An overpressure protection evaluation was performed for this configuration [1] and it was determined that the canisters were protected in accordance with ASME Section VIII rules; and in fact adequate protection is provided by a rupture disk on only one side. The project elected to install rupture disks on both sides to provide an additional level of assurance. After a single use the two filters have the potential of being contaminated to such a degree that they cannot be reused without an expensive cleaning process. Because of the large number of canisters to be dried (335), and the high replacement costs of the filters, means of reducing the filter costs have been sought. One proposal is to place an in-line rupture disk on the drain side of the canister and a by-pass filter on the inlet side. This configuration is shown in Figure 2 below. IN-LINE DISK ALL STEAM MUST EXHAUST ' ....--------THROUGH THIS 1 /8 ORIFICE Figure 2. By-Pass Disk on Inlet, In-LineDisk on Drain Side

In-Line Rupture Disks for TMI-2 Canisters During Drying EDF-1239, Rev. 0 p. 5/15 With an in-line disk blocking the path to the filter on the drain side, the drain side filter remains clean and can be reused. Figures I and 2 above represent a "fuel" canister type, one of the three types used in the cleanup operations, and the type which will be most affected by this alternate configuration. For the fuel canisters, all the steam produced during drying must be exhausted through an 1/8" hole in the canister head. This will raise the pressure in the canister during the drying process. The purpose of this EDF is to determine the consequences of this pressure increase with respect to contamination of the furnace and vacuum equipment. This evaluation is not intended to be a "worst case", or bounding analysis; but rather a best estimate of the likely consequences. 2.0 0-RING INTEGRITY The fuel canisters are fitted with a removable head which is sealed to the body by means of a large o-ring around the periphery and a small o-ring which seals the drain port penetration. Both o-rings are made of EPDM (Ethylene Propylene) elastomer. If these o-rings leak during the drying operation, while the canisters are pressurized, unfiltered steam (as well as radioactive gases and possibly particulates) will escape through the joint between the canister head and body. The following is an attempt to evaluate the likely performance of the o-rings during the drying operation. 2.1 Theoretical The Parker o-ring Handbook (2) provides technical information relative to the performance of o-rings under various environmental conditions and is one source used to estimate the current conditions of the canister o-rings as well as their likely performance during canister heat up. 2.1.1 Current Condition of 0-Rings The first TMI-2 canisters were loaded and shipped to the INEEL beginning in 1986 and have been stored underwater in the T AN-607 storage pool over the 13 year intervening period. 0-rings are subject to deterioration due to age and environmental effects. The EPDM compound used in the TMI-2 canisters is about average among elastomers with a recommended shelf life of 5-10 years (2, p. A3-14). However, Parker (2, p. A3-13] explains that storage conditions are actually more important than time. The most important beneficial storage conditions are listed below along with a discussion of how they apply to the TMI-2 canisters.

1. Ambient temperature below 120F.

The TMI-2 canisters have been stored underwater at temperatures below 80F virtually their entire lives. Even during shipment to the INEEL the temperature likely did not exceed 80F because the massive shipping cask tempered any high day time temperatures.

2. Exclusion of oxygen.

Since installation the o-rings have not been exposed to air. Their exposure to dissolved oxygen in the water has also been minimal because of the relative isolation of the o-ring grooves from

In-Line Rupture Disks for TMI-2 Canisters During Drying EDF-1239, Rev, 0 p. 6/15 the rest of the water in the canisters.

3. Exclusion of Contamination Contamination in this case most reasonably translates to fluid compatibility. Most o-rings swell considerably after long periods of water immersion. For static applications this can be quite acceptable, even beneficial. Of the compounds tested, EPDM is the most resistant to swelling [2,

p. A2-16].

4. Exclusion of Light Since installation the TMI-2 o-rings have experienced complete light exclusion.

5. Exclusion of Ozone Generated by Electrical Devices Since installation the TMI-2 o-rings have been isolated from ozone generated by electrical devices.

6. Exclusion of Radiation The TMI-2 o-rings are in a radiation field. However, the effects of this radiation may not be as drastic as first assumed. In 1994 radiation readings were taken at contact with the canister heads on some of the highest loaded TMl-2 canisters. The highest reading was 400 mR/hr. Since the o-rings are below the head which offers some shielding, we may infer that they see something on the order of 1 R/hr. Since there are roughly 10,000 hrs in a year, the o-rings have seen about 1.3 X 10 5 Rads over the 13 year period. According to [2, p. A2-13] all the standard o-ring elastomers can withstand 1 a6 Rads with negligible damage and EPDM can remain useable up to 10 7 Rads. The evidence suggests that radiation alone has likely not rendered the seals useless.

In summary, with the exception of radiation, the o-rings have been stored in a very benign environment. Radiation alone has probably had little effect on the seals. However, it's possible that the long time period coupled with radiation may have caused some compression set to take place. Any compression set could very well be counterbalanced by swelling due to water immersion. There is a single piece of test data somewhat relative to the current condition of the o-rings. According to C. A. Seaquist, fuel canister D-388 has been dewatered and refilled twice while underwater in the TAN pool vestibule. During the dewatering process canisters are pressurized with air at about 30 psig. This provides a modest leak check on the canister and no bubbles were observed from D-388 during dewatering (bubbles have been seen on other canisters due to a loose Hansen fitting). It should be noted that canister D-388 is lightly loaded and so its radiation field has been quite low. In the absence of additional test data, the Author's best estimate is that the majority of the canister o-rings are likely providing an effective seal.

In-Line Rupture Disks for TMI-2 Canisters During Drying EDF-1239, Rev. 0 p. 7/15 2.1.2 0-Ring Performance During Canister Heatup The temperature rating for EPDM for continuous service is only 300F [2, p. A3-4], which when compared to the maximum furnace operating temperature of 800F - 900F, would lead one to expect no possibility of maintaining a seal during canister heat up. However, as shown in Figure 3, elastomers have the capability of withstanding much higher temperatures for short periods of time. 1000 l\\ \\"-. 900 5 X 001\\. "'~ r\\. ~ ~ z aoo a: X C ... 7 ~ a: s 600 0 I ~ 500 i 400 ~ .. 3 00 00 0 '~~ ~~ ~ ..._ ---r-:::::::--- Sil/CON£ ~ FLIJOROELASTOM(R HYlENE PROpyl I EN( 4 Nro;;;r- ~(HIGH TEMPER J ENE NITRllE (LOW T ATURE TYP[ EMP£RATURE TYPE) 0.1 0.5 1.0 5.0 10 EXPOSURE TIME-HOURS fl GUM AU SEAL LIFE AT TEMPERATURE GENERAL TEMPERATURE LIMITS OF BASIC ELASTOMER COMPOUNDS IOO 700 500 200 100 50 100 500 1000 This chart is intended only as a rough guide. It cannot be used for precise predictions o( seal life. Results will vary with compound and fluid medium. Figure 3. General Temperature Limits of Elastomer Compounds (from Fig. A3-6 of [2]). A layout of the vacuum furnace is shown below in Figure 4. A time-temperature history for temperature probes 2 and 4 was evaluated from the first (System Operations) SO test conducted on the furnace with four canisters present. The control system set-point temperature is based partly on the vacuum level achieved in the furnace, which is in tum based on the rate at which steam is evolving from the canisters. As lower pressures are attained (as the canisters become dryer), the furnace control temperature is increased. During SO Test#! probes 2 and 4 recorded maximum temperatures of about 750F and 450F respectively for the first 12 or 13 hours. Over the next few hours the maximum temperatures seen were on the order of 850F and 550F for these two probes. In the last stages of drying, as furnace pressure dropped to very low levels, temperatures as high as 900F and 600F were recorded for probes 2 and 4. The design of the control system and the SO test data would indicate that during the times of significant steam evolution, and therefore canister pressurization, the maximum temperatures at the locations of these two probes are 850F and 550F.

NOTE O**RINGS ARE OF COURSE NO T EXPOSED AS SHO\\./N, BUT THE IR RELATIVE LOCATIONS ARE AS ILLUSTRA T£. D HERE. TEMPF.RA TlJRF PROBE ii 2 NEAR BOTTOM OF FURNACE 0 0 LARGE 0-R!NG 0 0 In-Line Rupture Disks for TMI-2 Canisters During Drying EDF-1239, Rev. 0 p. 8/15 L r uRNACE SHELL 0 TEMPERATURE PROBE >>4 .NEAR TOP or ruRNACE 1 Ml CANISTER (J rn* 4 SHOWN) 0 0 ~CAL-ROD HEATERS 0 0 Figure 4. Quarter Section of Vacuum Furnace with Canisters A finite element thermal analysis model was created [3] to determine the temperature of the canister head during the time period in which significant steam is evolving. Radiant temperature sources were applied as shown in Figure 5. The head was modeled as a solid slab of stainless steel of uniform thickness.

0 0 0 700F, ' 0 0 0 + 0 In-Line Rupture Disks for TMI-2 Canisters During Drying EDF-1239, Rev. 0 p. 9/15 0 0

---

, - __ )

Figure 5. Radiant Heat Sources Applied to Canister After 15 hours of heating using these radiant sources, the canister head temperature profile is shown in Figure 6.

In-Line Rupture Disks for TMI-2 Canisters During Drying EDF-1239, Rev. 0 p. 10/15 .~~~~~ ~ GeoStar 1.75A: C: \\ COSMOSPROBS\\CANHEAD. MOl~~ ~~~~~~~~~~~~~~~~ Figure 6. Canister Head Temperature Profile After 15 Hours. Te mp - 73 3. 78 = 724. Bi; =7 15. 93 .,l.tt .::;\\.* 7(9 7, lll© = -= G9B.&B -= .. 689. 15 =- 68&. 22 = 67 1. 3lll 662, 37

In-Line Rupture Disks for TMI-2 Canisters During Drying EDF-1239, Rev. 0 p. 11/15 Figure 6 shows a maximum temperature on the southwest surface of the canister of 734F. According to the guide provided by Parker in Figure 3, a rough estimate for the life of an EPDM seal at this temperature is about 10 minutes. On this information alone it appears there is a strong possibility the o-rings will not maintain a seal at this point in the drying cycle. However, as explained below, additional information is offered by testing experience. 2.2 Test Data 2.2.1 VECTRA Testing During the process for developing the vacuum furnace concept, VECTRA (now Transnuclear, West) dried a single fuel canister by strapping a series of band heaters to it and covering the whole assembly with insulation. During this testing the canister shell reached a temperature of 700F or higher for 12 hours and throughout this time the o-ring seals held a vacuum and near the end of the drying cycle achieved levels less than 2 Torr [4]. It should be noted that the exact temperature the canister head (and hence o-rings) achieved was not specifically measured, but was probably roughly equal to the shell temperature. The actual head temperature is obviously a critical issue, and if it were much less than the shell temperature, the results of this test would not be particularly applicable. After the canister was allowed to cool, the vacuum system was again employed to verify dryness, but the canister head seal was found to be leaking and had to be replaced before a vacuum could again be achieved. Theo-ring seal used in this test was a Buna-N (Nitrile) compound ratlier than the higher temperature rated EPDM used in the TMI-2 canisters. 2.2.2 INEEL Testing As part of the series of vacuum drying SO tests conducted at the !NEEL during the fall of 1998, efforts were made to determine whether cesium compounds could migrate from the canisters to the vacuum furnace. The main focus of these tests was to determine whether the addition of sintered metal filters was effective in preventing cesium from passing through them, but efforts were also made to determine whether cesium was escaping through the canister head to body joint. No cesium was found on the surface of this joint and from this evidence it may be concluded that the o-rings maintained a seal during the period the canisters were pressurized. Of course, during this testing, the drain side filter was open to flow and so the peak internal pressure generated was very low, probably on the order of 3 psi, versus the 30 psi which could be seen with the drain line filter blocked (see Section 3.1). After these tests, when the canister heads were removed, the o-rings were either very cracked and brittle, or literally falling apart; indicating again that the o-rings may perform well through one temperature excursion, but once allowed to cool lose all integrity. 2.2.3 0-Ring Geometry A big difference between the tests described above and the production drying configuration is

In-Line Rupture Disks for TMI-2 Canisters During Drying EDF-1239, Rev. 0 p. 12/15 that the o-ring geometries are different. During the tests a true o-ring was used, i.e., the elastomer had a circular cross section, while the o-rings in the TMI-2 canisters are of a special cross section. Theo-rings used on the TMI-2 canisters were specially designed to fit tight in the groove to allow for remote installation. The consequence of this design is that the o-ring nearly fills the groove when cold and upon heating, because of the high coefficient of expansion of EPDM compared to SST, exceeds the volume of the groove (see calculations in Appendix B). It appears there is a definite possibility the expanded o-ring could force a gap between the canister head and body and be partly extruded into the gap'. Whether this phenomenon will result in canister leakage which was not seen in the testing is not known. 2.2.4 Test Configuration vs. Actual These two tests indicate that the o-rings can potentially maintain a low pressure seal for a single thermal excursion to temperatures far beyond the guidelines. However, caution is still in order. There are several differences between the conditions experienced during these tests and the production drying campaign.

a. The seal geometry used in the loaded TMI-2 canisters is substantially different than that used in the testing. The canister seals are so tightly confined they may force open a gap between the heads and bodies resulting in leakage.

b. The seal material in the canisters is EPDM rather than the Buna-N used in both the tests discussed. EPDM has a higher temperature rating and should theoretically perform even better than Buna-N, however this is not guaranteed and any differences can yield non-prototypical results.

c. The seals used in the tests were fresh new material, while the canister seals are 13 years old and have been in a moderate radiation field.

d. The production canisters will have a lot more water in them than the canisters in either of the two tests and so significant steam generation may continue further into the drying cycle and thus challenge the seals at a higher temperature.

e. The pressure difference in the Vectra test was one atmosphere and was outside-in rather than inside-out as will be seen in the drying campaign. The pressure generated in the INEEL test, while inside-out, was of a very low magnitude. The production canisters may see pressure differentials of 30 psi or higher if the drain side filter is blocked.

3.0 MAXIMUM LEAKAGE In evaluating the feasibility of this rupture disk arrangement, a bound on the maximum leakage is a useful piece of information. This calculation requires an estimate of the pressure generated in the canister during drying and as well as the average gap between the head and body should the seal completely decompose. 3.1 Pressure Generated Reference [1] performed pressure vs. flow calculations for the TMI-2 fuel canisters. Using the equations developed in [ 1, Appendix A], the following table was generated.

Canister Internal Specific Volume Pressure (PSIA) Saturated Steam (ft"3/lb) 0.5 633 1.0 333.6 2.0 174 3.0 118.7 4.0 90.6 5.0 73.5 10.0 38.4 15.0 26.3 20.0 20.1 25.0 16.3 30.0 13.7 In-Line Rupture Disks for TMI-2 Canisters During Drying EDF-1239, Rev. 0 p. 13/15 Inlet Port Flow (lb/hr Drain Port Flow (lb/hr steam) steam) 0.26 1.94 0.51 3.78 0.99 7.40 1.47 10.97 1.95 14.50 2.42 4.73 7.00 9.25 11.48 13.72 Reference [5] calculated the maximum steam generation rate per canister to be 14 lb/hr when the furnace is filled with 4 fuel canisters. This agrees well with the maximum water removal rate seen in [3]. According to the table above, a steam flow rate of 14 lb/hr results in a canister internal pressure of about 30 psia if the drain port is blocked. (For comparison, if both ports are open, the internal pressure is a little more than 3 psia.) 3.2 Gap Size In the absence of an o-ring, the flow path is a result of the fit between -the canister head and the mating flange on the canister body. Both are machined parts. The mating surface of the head has a.002" flatness call-out (INEEL dwg 346328) while the mating surface on the flange does not have a flatness call-out (INEEL dwg 346342), but is probably of about the same accuracy. Given this information, a uniform gap of.001" is a reasonable and probably conservative estimate of the fit between the two surfaces. The depth of the gap, or the distance the steam must travel between the two surfaces to exit the canister is not uniform, but rather varies from a minimum of about.75" to 2.5" because of the "square in circle" geometry of the flange. The average gap depth is about 1.7". The width of the gap is the perimeter of the leak path which is roughly the perimeter of the bolt circle for the bolts holding the head to the body, or 10.25"*3.1416 = 32.2". The overall gap geometry then is a "duct" about.001" high, 32.2" wide, and 1.7" in length or depth. 3.3 Flow Through Gap Based on the gap geometry discussed above and a canister internal pressure of 30 psia, the calculated steam flow through the gap is 3 lb/hr, or approximately 20% of the total estimated flow at this pressure (see Appendix B). Again, the gap flow is of course completely dependent on the assumed gap geometry which is believed to be fairly conservative.

In-Line Rupture Disks for TMI-2 Canisters During Drying EDF-1239, Rev. 0 p. 14/15 As a point of interest, gap flow calculations were also made for a canister internal pressure of 3 psia, the expected canister pressure if both ports are left open. In this case the gap flow would be about 0.044 lb/hr or.3% of the total flow. This again assumes that the o-ring is completely destroyed. There are certainly scenarios in which there is no leakage at the lower pressure (insufficient pressure to break through), but significant leakage at the higher pressure. 4.0 MISCELLANEOUS ISSUES 4.1 Effects on Knockout and Filter Canisters The knockout and filter canisters do not have removable heads, but do use EPDM elastomers on the "Thaxton" plugs in the process flow ports. Degradation of these seals does provide a potential leak path for steam and radioactive materials. However, because the knockout and fuel canisters do not contain the hard to dry Licon, the high steam generation period is expected to occur earlier in the cycle, probably before the canister heads reach full temperature. Additionally, the inlet ports on the knockout and filter canisters are much less restrictive than the fuel canisters and have flow capacities about the same as the drain port on the fuel canisters [6]. Therefore, whether or not the drain port is blocked, these canisters will see internal pressures of a few psi only. The same conclusion would apply to the possibility of one of the Thaxton plugs being ejected from a knockout or filter canister during drying, i.e., because the pressure is not raised significantly by blocking one port the likelihood of this occurring is not substantially increased. 4.2 Potential for Relaxation of Bolt Torque on Fuel Canisters. Note that the following discussion is unrelated to whether an in-line rupture disk is placed on the canister drain port or not. These issues surfaced as part of the evaluation and so are documented here. The fuel canister heads are attached to the bodies using %" Inconel 625 bolts torqued at TMI to 50 - 60 ft-lbs. The Inconel bolts have a lower coefficient of expansion than the 304L SST heads and mating flange. According to the calculations of Appendix B, the pre-load combined with the load due to differential expansion of the bolts vs. the members is not sufficient to yield the bolts at an operating temperature of 700F. However, an additional deflection of only.001" induced by the constrained o-ring will yield the bolts and completely relax the pre-load upon returning the bolts to room temperature. As discussed in Appendix B, predicting whether, and by how much, the bolts will be stretched due to the constrained o-ring is a difficult proposition because of the number of ~nknowns involved, one being the fluidity of the o-ring while at temperature. The range of consequences vary anywhere from retention of essentially all of the bolt pre-load upon cooling to stretching the bolts sufficiently to allow a gap of three or four thousandths between the head and canister body. It is the Author's opinion that relaxation of the bolt pre-load is very likely. A gap on the order of a few thousandths does not materially alter the canisters' function as containment vessels. With the quick connect fittings removed, an indirect path with a diameter of.125" is opened into the canister internals. The size of particles that could pass through this path are at least a factor of 20 or 30 greater than could pass through any path which could be

In-Line Rupture Disks for TMl-2 Canisters During Drying EDF-1239, Rev. 0 p. 15/15 opened up between the mating surfaces of the canister head and body. In serving as containment vessels the purpose of the canisters is to prevent gross redistribution of the fuel rubble. This small gap does not alter that function, given that the.125" diameter hole in the fuel canister head has been accepted previously as well as an implicit acceptance of the potential leak path through the canister to body joint once the o-ring is destroyed. However, in addition to these technical issues, there may be other program related concerns with leaving these bolts loose during handling, transportation and storage. Therefore it is left to the project to determine whether any effort should be made to investigate this issue further. Some question might be raised as to the acceptability of exceeding the ASME Section VIIl [7] stress allowables for these bolts while the canisters are at temperature during drying operations. Appendix S of Section VIII specifically addresses this situation and acknowledges it as an acceptable condition. The Appendix does caution against repeated stretching of flange bolts, but the TMI canisters are expected to be heated only once. 5.0

SUMMARY

Too many variables remain unknown to confidently predict whether or not the canister seals will leak during drying as result of blocking the canisters' drain ports with a rupture disk. Theoretical considerations would imply leakage to be likely, but in the two tests conducted (which were admittedly non-prototypical) no evidence of leakage was observed. In the Author's opinion, the majority of the evidence leans toward the conclusion that no leakage would occur with a blocked drain port, but again, only a true prototypical test, or series of tests, will provide the conclusive answer. Flow calculations indicate the potential leakage to be no greater than 20% of total flow. In a related issue identified during this study, there appears to be a real chance of relaxing the bolt torques on the fuel canisters due to o-ring thermal expansion. It is believed that this will not alter the canisters' containment boundary function, however the project should evaluate whether there are any other issues associated with potentially loose head bolts.

6.0 REFERENCES

1. A. J. Palmer, Overpressure Protection for TMI-2 Canisters During Drying, Rev. I, EDF-766, August 1999.

2. Parker 0-Ring Handbook, ORD 5700, Parker Seal Group 0-Ring Division, Lexington, KY, 1992.

3. Unpublished test data in Author's possession. Test conducted by Chris Johns of VECTRA Corp., November 1996.

4. R. E. Spears to Distribution, COSMOS/M ANALYSIS COMPUTER CODE -

VERIFICATION/VALIDATION DOCUMENTATION -RES-04-99, September 7, 1999.

5. R. G. Ambrosek to A. J. Palmer, TEMPERATURE AND VAPOR GENERATION RESPONSE FOR TMI-2 FUEL STORAGE CANS IN FURNACE - AMB-06-98, February 23, 1998.

6. A. J. Palmer, TMI-2 Storage Canisters Maximum Drying Temperature, EDF# DCSP DWS, File Number INEEL/INT-97-00862, September 1997.

7. American Society of Mechanical Engineers, Boiler and Pressure Vessel Code, Section VIII, Division 1, 1995 Edition.

8. Metals Handbook, IO'h Ed., Vol. 1, ASM International, 1990.

9. Flow of Fluids Through Valves, Fittings, and Pipe, Crane Technical Paper No. 410, 1988.

A. B. C. D. E. F. G. H. I. J. K. APPENDIX A Analysis Plan For In-Line Rupture Disks for TMI-2 Canisters During Drying ReQuester: J. H. McDaniel Performer: A. J. Palmer Deliverables: Evaluation of the consequences of installing an in-line rupture disk on the drain side of the TMI-2 canisters during drying operations. Document in the form of an EDF released through document control. Purpose of Analysis: The current plan is to dry the TMI-2 canisters with a "by-pass" rupture disk installed on both the inlet and drain ports. Costs of replacement filters may be reduced by replacing the by-pass disk on the drain side with an in-line disk to protect that filter from contamination. This analysis should evaluate the consequences of this alternate configuration with respect to potential for contaminating the furnace and down stream vacuum system components. It is expected that a definitive prediction of the exact consequences (i.e., amount of contamination) will not be possible. This EDF is therefore to be a review and synthesis of the limited data available with a best estimate as to the possible results. An evaluation of the overpressure protection is provided by another EDF. Description of Item to Be Analyzed: TMI-2 canisters (specifically the canister seals) within vacuum furnace as configured during drying operations. Applicable Documents: TMI-2 canister drawings, vacuum furnace SO test data, VECTRA canister heating data (all currently in performers possession). Design Requirements, Operating Conditions, Applicable Codes: No national standards are applicable to this evaluation. Operating conditions to be supplied by SO test data. Safety Category & Quality Level: This analysis is not safety related. Because the analysis is not intended to be definitive, but rather a synthesis of the limited data available and a best estimate of consequences; Quality Level is level 3. Analysis Verification: Review and approval by a competent individual other than the performer. Approval by requester. Cost: 60 hrs (including reviews) Schedule: Begin August 23, 1999 - Complete September 24, 1999. Change Control: Signatures equal to those of the original. Software Verification: Any software that is used will be verified on a case by case basis to support the specific application using hand calculations or by comparison to known solutions per MCP-2374. This verification will be documented within the EDF as an appendix. Analysis Plan Approval: Requester <P,n,aJ6t£-i~ 9 /JtJj,~J Performer ~ 9/1c/9~

In-Line Rupture Disks for TMI-2 Canisters During Drying EDF-1239, Rev. 0, Appendix B Flow Through Gap. APPENDIX B CALCULATIONS Loads Applied to Bolts While at Temperature Effect of Thermal Expansion on the Constrained 0-Ring p.Bl p.B6 p.B10

.,..,...~---3,. 2 ;,,, _____________ / 4==_/: ,oo, ~ 77 . (.) 'S -z.z.;... i 1t ' z.o "Z. i "'- 't F=,-: r.s.} c~c..J.,,,_;_,._ ~"'...,_0 ld?. ~ ;t;. <J 1Re_ = / "Z,.;,. 9 c( ve I .cp =.J.;.fr.t,t... u b ii-!wu.-

...... J.t.,.11'-i i,; ~~r.,..a..(

pn: ~.s '""~ t..,J {:,t.,. ~-1-u-J, ~ d. V '.;; ~.u.-(', . ~ I l'i) I <,;f * <.,*f'. ~ VQ iM.-

' 'i"'f/

,r'( b vJ: i::-c)w rr..£ i- !~ Su.. C,e2-.._~ ~ , '3, 'Z- '7f:. vdoc.:A,1 (ff-1~.J V. I I r~ _I.* AA_ ;:-. l~C.0$, f7,_ ~"'1*(><')*';.L '-'.r Bl

t' i O,,.,, !.,r -II w :: \\MA-' ~ f~..s r.-{, LD.J_fu,. 11,~,H-'J x ~ -:::., 0381 -f-fJ I-Ir ( ~ 3<t,oo ~ec.. ~ \\J- <:-~ *{* ' 11-IJ,4__.., C:: <>,( ~ l <. 4 *..J ~ I SJ ~ :e.::>..,.... f'-;. ~'t

. ~'f

-.=,'2.7 KA-. Z3', . '2-1 (/,7)

• 001...

2. 3 0 Y=-.718 6p,

.:::::. ')7

r.

• sz

53 Y=.7/a ( 7:;,;;-) ( /3.7) 1)00 I,'~ l~. ..,, -~. w-I', If I I."-'"""'~ -\\-1..o.....i,.,. Le/~ c::.. IO ' I ll = tJ~.j*h*t.~._, Qo... '....;i::;u -...~O ~,'".,fv,,r ~ - I

_.._, - "D-~w,-.lt,c""-3'\\. t(.~,o~*,

,-,,;1 :; o.

o. --

3 7, Z-u

• A ::- a..,-ct,.

a is. r __, "t. 1,.. "f'f" \\. / OJu,~t) ft7

---

J-.-------- - *'-----------'------ - ---- - - f--""'*_s_

l't ft5;;.c..,~ fl().,J n,_-bi is -h.>o l;..;0 1,... 1 lo~r p....,_..J n.h w1*rJ. .&e.er<c1-l-(?...- ~ ~ ~ lnc.,....~~<....fn*: ft*.,..,._ J:1...:h..,. r-<-L<./......, -tf,....(.lovJ, .F:. ~ :: <-()'.? /,y I

Loa.J.s P~-Loa.J R-CC-Drt2i"'-Cf .fu ~ ~f~S~"'j A..../.,ve.. :fa_~ rfr... ~j_>~"'- -p()...c.-k.otJI!- .fo.,. <.J a.. ~c ~-isf..r (1)-Z.'lJ'o) Cr..1-,rrcv.J..!1 i.-. -!-I~. o..v...+f...or'l po:Sc.1-;h*,~. I -P"'<- -hr..{-h wc.r-"-. -h:,7v..1t& -i-o 5"'6- '7o Pf... (:.,, 1 t..1/./-i._ s-s f'f.([,J k_;() ~ :}.~t (/tL~ 0;,t. & -fo,l t,,..L w,-t!. Kt:-l. h - 1 I -- 1::.~ ~' "( ~ ,1-4,-,_ L ~, T ~ t:...-::: -k7;_ c..:,e/fic:e_/ ~. ~ f;,, 2tiiD.O. ~= J:.~.,_.,. cfL.,._:-+/-;.r f:: ~~ l.<l.-4.fo.. ~ ~s-f-t. (b 't 1 '-too lb Ct<>)(.75;"")(1Ff) r:;r - l ~ I 0 k,..,tl ; s Y\\,t,'v,,,,;V\\.,J~ f' :=,.$0 Mp3.; @ 1 0 r .(;-o.,..,,._ f)S/YtlE-Sc., 7.Zl Au~.. '(. I

7; 'too lb ~

• ni.;"'"i.

W~er<:.. ~ -- 1 rt'f i.2-0 tl..... ( '?.. i...... ) b7 ~ J.J~.J.J~f (,a(jr= *7oFJ= ~so~ Cl O -(, 1,~ r NO '""!- 0"k.,.... Qol.( l. SS7) is -~~-- t)'J ~ t.o (f s. CJ.~c,,/:- -ftJr -//..is S,'f~.1~*->~ ; B 7ot:>,:, ir-e ;2,,1.;... ~) C 2--?),7 1ror.. r-:") ~ 'f,,/; -z ~,o~ 1,-:"' ,( 1' 111. B7

I k.lM; ~C-L<>rJ.;J -h ~L-t) lt.J '/ "L c~, ~.J. e-,,, w.'?L,,.. Gr;.{" ,to J.;~~-lr rA..l*o cJ 3::...-~ - 2 *" 7 i ...,r;.... -= 1.3 ~ +or 8"<'.lo ~ii!\\ S'i7Q,aoF fu,_ [ ~'fV\\c Afp~-L~. 13 ( z Y, a ),,Io<- pi,)(. 7~-..... J k..""- =- Z. '-t2 Xlt:,7 H, /. / 1¥1 -r~ bfrij,o;t af -+L. V\\AL...,_rQA.rs ,*s ~(.../-t...- S i,,._-; ~ ~;"'")..-*~~,J.+L.. h\\) (-4-. (~r<<-;~._, +L-. --ho lt ~pu,111~ ~;,~,-/-e....._

B07,

<if -+l.... ~Pa.... -t-;.>/i. ~r ~,--f'l,~:-S, a<.S <>~ ¢:><>L.f.. e.1pv-;.e..""us e.,....+i*I'{_ Jt~l.ec.+RVI. : . o0z..73;... (z q,_ 7 . ) -- u, >c.to ~p,)1

z. :....

l1u..2f1-<,;[ica.h'".... {i; B-'1'1,) -tor -ft,,s ?*1t~f2r,°A,f d,,tS" hal-,:.rov;J"'- f 1 /! I ( ' I. I et. l/ti_.;.e,y (E.,.-?'-'~;.,*i*:.,./~,W'ljyijMv.wi y,d/.S-rrc."dw I nOI.A)e._ue,,..- +i\\4.-::; u-b.z-e.;.-f.iw...i"/a / fr'o.,..,.__ ~ - PfSMG oJj_" v..J~0. sf~ ss fa..,!..ti,:..._ /;r,_;pJ.. £.~ ]I:.,

1i"4 'Y'\\1vt~l'l'\\1A-- v"'m).- u~~ra..-f...>-lf-<.. y,e..(J). s.k~r ~ ~~ Co"'2-S- {') Coo l'-~i. T~ A-s!Y\\.li.sks, a.)lo...;~~~ Fo.- -U..;.r...__,0.o}

• -J:;.c:.........

30~~ ~ --4o :Z~ '=:..si, 1"'-f~iJ k~~s -b'C) ~r Sa.......&- rr.,r*ho.... -"[(.... 4L~J ~-1-..r.J o...._ ~ p~u<*-n,....,. P"-<y- :s a,...f s(.11-~ o~ ~VI.ti! Lw/- h K-/A.-<,W"-- -h, µ So-u...JLJ tJ,t,... -:,e.::...Al#c. /II..¢ aC-G4>.__J (..A.)~J ~,-._+o-) +he.se ko w.4..-~is~ t:P:<< ;,i,o/.. '¥f9 ladr si..H.c.,e.J- ~ yie..<J. +k_ ~o c+s, a.;{~!,. ~y wH( b~ O"'- -rt..L 1.urv~ a.} ~ie.CJ.1:Ac

• U.

t, ( / l _ A L ti !

• r[*I TLow~l.),er1 ;so_.... 1w..-.,-r,o"'--

V\\,\\..1(/IA-r .I>'=:'... ra...,.s12&.. (A..s ,o 't°t.v a.~tJ+tx.'o*,, of -U:(L:l..di,tj ~ MMe ~-1-,o...... \\[ilL_..S4rt.u c...Ck<Ao...~~r to.r ~~e ~o (+.s '-'-'4- ~ .a.. C,o~,*.~~o..... s. Ap~1-J.iic S oj ~c..-fio~JiilI-s~,_;f:;~l~ o...J.&.re.~-e.J +~;:, sr./-tA_c.J.ia-a..J. a,::;.~~..J~.d~~J it as a.~ a.c.~;r/*'~ ~.'h'ovt. 1k_ ~~ii< ~es ui-to...._ °f)C,.,*A-if, r~fC?a.~J 5-l-rt--h:.L*J of fl.,.~~ f, o f h I b ~-i -fl-- -r'f/l/lZ c.av,.,i'!, tu-s C\\.-<. ~ <:..../c,i -h ~ ;..,.,_+, j

t(,,.Q_ <:.ros :; ~.,.,._ oJ. +L ~-r~~O\\ s v...~ 1*.A. +l.v Y.M 7

• 2....

C:L....i..s-h.cr G Q.J'(_ no.f c..;.r~.o..v-Co~ -Hu. lv-'r'" o-11"\\~ i ) 0J.1L~1c-la..~J "'-~) 1 bJr a.#-._ !f'O-fur-.+-~ I J.A..u.-~ v..-( ~ ~ ~ ~ ~ v.l""- ~~W W ~<<.. (..,_ u.)O..t d2o.1,*t 4 ~~~f:.*c..* J~ 4t, s~'1 i- {p{o_<- o..~ +l... c..a.-.;~r ~r l-4>~.~<:.. TYP

---

c----.-- *** ---*-.... --****"**...........

I

t!!.t"

.--... 84i1 ---i* OIA (L J.c.oc.. .15(,.' ~~~~-' +/-.oat.. .2.IG. SECTlON 0- (;~ I ~ t.o<><..

• .oesr I

P1< 10212-l

• NO 5C.ALE.

h lf.>>..,J s. ~ u-oc..>-::,,uJ.ia/1.4..( o.....-e....... WO...f d2,r1..~,,._ o., AvJnc.AO a..v.../J.. -+-4e. (!) ~ ~:f *

• C A J o, i MO,w. _

2 1 ?* 3-u, t,;,..{" 8/0

r r ( T~e_ 0)<,f,a.... :.,o.,c. c.oeff ~., ftv_. S5 r ~i~ c..,..-p -nu.. c'r,.>c,i_µ_. Is fro'-~~) ex'.=,.<o"l(\\{(),., T~ ~.-.J vo (I.A.~ aJ ~ (9 T~..._ f"J ~+v..,.-<- Q

• r; 1q (r~tw -fu +Lt.

~ r't\\(.)Ve.) ; ~ : '-1 I ' 5 ( ) l._t,oso} .95"8i,.? I. I I in 3 ,( J.~ * ( j ~,/..... ' --1 c.,~ ------1-'J. I I 0

l.

I I R* :: /,2!'t5 i l 7.71 (., l l \\__ ____ J~ W~... c;...r ~ +'- J (A. Si~

z '.:: f.n~ ro..JJ.; ~

/,./lif,lf":) , c;, tJ,*vi '1 /, /3~ L/j(),.... LJ +~ C)... f'i-'"", rJ !Jou.....__._

• ~6t 1 VO~\\,\\;\\..L of

~ .e.,.,,.u.:_,rj '-;; --tfu. j rootK... ~! t-S,c; (-

,J~-

I ( ti ;k A

• ('tr/IA

-4,).1 ~ QU.c,-i,.,....oJ~ .f--l,._ 0- v-i""_j W~<-- Ll it- +:...._ o-r-,~ ~.f..~J S**h*o..~vJ. <Af>. R,s,-p\\:ts ""e:"'j S.11($"';... -,,~11;.... ::, o,, ;II\\. ~ft a.i fe.-rcJ..:l~ a..JJ.. i.,,.J;{( ~ o.........a~* o,,,...J ~w,.J~ ~ ('c.sJ....f.;,._J ~ ~+, 0 ~r* h,..~ j 1 'M.CA-u..~...f-1-... £,., { ~ 5' C<. x.. ~ ~O.OJ re' cl lo_._+ -ftd t i' c ( j be.~,(...+k O* cri'j 1~~- i's a...ff (;-!? }) o..., 1.1.kl-u',,;......,/ AW,'~"'- w,; _ _,( re.s!A..,(.(-,_ st.-(:_{cl.*:) of.4-fµ.. ~a U-s. I~ rHu_ to (h s.f.-,.e:+c.-L e"e.- 1 ,I I A II f /'I . oa ;,11 a..ll ~ pre_-loo.1. w/l( bt.. ru:e..v-e&. a...u:v +"'-A., *;. o.-tJ wil/ f3IZ.

,___ Ntrtt., ~3 ~J _µoAQ.Sc ~us °'-"'* '1p.S.,~..Q

• K'>>~l7 J..o,,: f- ~ ;~4 os... :M"!>t <<,..,;~-:;;: r.s *

.fL*i*o...:..-C ,a-oJlt,\\.4-f¥i'r:.- j.,~{74 / o.... ~Q 3<&z.ca fk-1 ~ prrw;~ ~ ~ .. r c>-,,-i'() ~ 1'b ~fa.~J wt'~ t;{..f:u +!,.... AeJ. T~ II'(_ '<-~ s J/Y\\.L Cl '8 ""'ov--!f c....,.... + l!:>,r,j o*0041<. *.j:)~ c..o~r t)()~ !' rp-rov.of.L. a_ /e.-te+ pa.-ft... /;;,,,-- 4 o-r"i'U W; ( U> "'"f Lr' c. J-. ~YI.} s u..c.L as ~ o "'<2 cY...-s ~ s S.<.J o...fw ()-fl. p ~st!.,J flu_ ~( ~{t..<4.""Ui.> of ~/,j a.. ~,s-fv ~ t.AJ:.tt. f0.. <:..c>~.f;,a;y.,c.J; o *ri,,-">. C<Je.. ritlf de/:;;iz:&. ~ ;:>,,..eJ, c..1-.-..bG. u_.J.,, (.+he.. J 4.c.. /*/vi'f7 , *5 a.e-f.va.(t; ~-J.,,,._c:f;.j, I11t..-h-.'.j.,*-.K.(, ;.j. a.fF'4"S ~r~ Co*tc...;"'- f-l..~J-* -f.ev.. D-r;... 0 <'..¢~ lcl.,J~..f.L.... lvi.a..JJ o~ "- )t:.<.J.J!~u.so-l fl.r ~i -J.l_ wo~ ~~ 1.J..s.t...(:f.,~-R -H... s~Ci G*~f t'll.r~J. *}}