ML20138F532
| ML20138F532 | |
| Person / Time | |
|---|---|
| Site: | Rancho Seco |
| Issue date: | 04/28/1997 |
| From: | Redeker S SACRAMENTO MUNICIPAL UTILITY DISTRICT |
| To: | NRC OFFICE OF INFORMATION RESOURCES MANAGEMENT (IRM) |
| References | |
| MPC&D-97-073, MPC&D-97-73, NUDOCS 9705050430 | |
| Download: ML20138F532 (10) | |
Text
.
(m)suun
~~
SACRAMENTO MUNICIPAL UTluTY DISTRICT O 6201 S Street. P.O. Box 15830, Sacramento CA 958521830.(916) 452-3211 AN ELECTRlC SYSTEM SERVING THE HEART OF CAllFORNIA MPC&D 97-073 April 28,1997 r
U. S. Nuclear Regulatory Commission Attention: Document Control Desk Washington, DC 20555 Docket No. 50-312 Rancho Seco Nuclear Station License No. DPR-54 RESPONSE TO REQUEST FOR ADDITIONAL INFORMATION, UPDATED CASK DROP DESIGN HASIS ANALYSIS
References:
- l. R. Dudley (NRC) to S. Redeker (SMUI') letter dated May 10, 1996, Request for Additional Information - Proposed License Amendment No.192, Updated Cask Drop Design Basis Analysis
- 2. S. Redeker (SMUD) to S. Weiss (NRC) letter MPC&D 96-034, dated March 18,1996, Proposed License Amendment No.192 Attention: Seymour Weiss Enclosed is our response to your request for additional information specified in the Reference 1. Please inform us of any additional concerns or questions in writing so we may expeditiously provide you with the desired information.
Members of your stali requiring additional information or clarification may contact Mr. Jerry Delezenski at (916) 452-3211, extension 4914.
Sincerely,
- I w r ,1 1 r.
proy l Steve J. Redeker JL=2 8 Manager Plant Closure & Decommissioning cc w/ Encl: E. W. Merschoff, NRC, Arlington, Texas R. Dudley, NRC, Rockville hp,mgi}Q.P,. $-
9705050430 970428 PDR ADOCK 05000312 U PDR RANCHO SECO NUCLEAR GENERATING STATloN O 14440 Twin Cities Road, Herald. CA 95638-9799;(209) 333 2935
i 1
1 I
l ENCLOSURE l l
l I
l l
l 1
l l
1 f
9
Page 1 of 8 District Response to NRC Request for Information NRC Reanent:
Please analyze a possible accident scenario involving a drop which results in tipping of the spent fuel cask and ejection of fuel from the cask into the spent fuel pool. Determine 1 whether a critical spent fuel configuration might result and address whether any actions are necessary to mitigate the consequences of such a scenario.
District Resnonse:
Even though the District does not consider tipping a fuel transfer cask and ejecting spent i fuel assemblies from the cask into the SFP to be a credible event (refer to the discussion beginning at the bottom of page 6 below), the District provides the following analysis of such an event as requested.
i j
Analysis of Cask Drons that Result in Cask Tinning and Fuel Ejection:
The District performed several cask drop analyses (District Calculation Nos. Z-DRY-N0062 through Z-DRY-N0065) that assume a fuel transfer cask fully loaded with 24 spent fuel assemblies tips on an edge in the FSB and falls into the SFP, either discharging all 24 assemblies into the SFP or retaining the assemblies in the cask. Since the final geometry for a cask tipping event that results in ejected fuel assemblies is indeterminate, the District studied a number of hypothetical cask tipping and fuel ejection scenarios that are not realistically achievable for an actual loaded cask tipping accident.
The following is a description of the criticality analyses performed for various, hypothetical FSB cask tip scenarios that the District studied. The K,, value result listed for the following analyses include a total statistical K,, uncertainty factor of 0.012.
- 1. All the fuel pellets from 24 spent fuel assemblies spilled from a cask collect in a pile on top of the SFP storage racks. The fuel assembly cladding and hardware, control components, and cask and canister components are segregated from the fuel pellet pile. The fuel pellet pile is arranged in a four sided rectangular pyramid,166 cm by 164 cm on the base sides, with a height of 112 cm. The fuel pellets are from once burned (most reactive) fuel assemblies. The fuel pellet pile is on top of fuel storage racks that contain no Boraflex and fuel assemblies with no control components. The SFP water is at 70 F and contains 1200 ppm Boron.
The resulting K,, value for this analysis is 0.876.
- 2. All the fuel pins from 24 spent fuel assemblies spilled from a cask collect in a pile on top of the SFP storage racks. The fuel assembly hardware and control, cask, and canister components are segregated from the fuel pin pile. The fuel pins are
l .. .
. . i'
- Page 2 of 8 j- District Response to NRC Request for Information i arranged in alternating perpendicular layers,20 layers high with 250 fuel pins in !
i each layer. The fuel pins are in contact in the vertical plane and are separated by j O.36 cm in the horizontal plane. The fuel pins are from once burned (most . l j reactive) fuel assemblies. The fuel pin pile is on top of fuel storage racks that j j contain no Boraflex and spent fuel assemblies with no control components. The l SFP water is at 70 F and contains 1200 ppm Boron.
l' I
l The resulting K,y value for this analysis is 0.869 i f
3.- All 24 fuel assemblies are spilled from a cask and collect, undamaged, in a pile on j top of the SFP storage racks. All the fuel assembly control, cask, and canister i i components are segregated from the fuel assembly pile. The fuel assemblies are i
! arranged in a three layer pile, with 10,8, and 6 assemblies per layer, and are in )
contact on the vertical plane. The fuel assemblies are assumed to have a one inch l l separation between each assembly on the horizontal plane. No more than six fuel i assemblies have K,y greater than 0.99. No more than 12 fuel assemblies have K,y l greater than 0.92. The average K,y is no more than 0.94 (the average K,y for all ;
i fuel assemblies in the SFP is 0.8771). The most reactive fuel assembHes are 1 i clustered in the center of the fuel assembly pile in an arrangement that gives the
! highest overall reactivity. The fuel assembly pile is on top of fuel storage racks
- that contain no Boraflex and spent fuel assemblies with no control components.
The SFP water is at 70 F and contains 1200 ppm Boron.
l l
t l The resulting K,y value for this analysis is 0.944 i
! 4. The same as analysis 3 above, except that the fuel assemblies have a two inch i separation on the horizontal plane instead of a one inch separation.
- i
- The resulting K,y value for this analysis is 0.873
! 5. All 24 fuel assemblies spill from a cask that was loaded in an administratively ,
controlled configuration, with the least reactive fuel assemblies placed in the
- center of the cask and the most reactive fuel assemblies are placed on the ;
periphery of the cask. The fuel assemblies collect, undamaged, in a pile on top of :
the SFP storage racks. All the fuel assembly control, cask, and canister
- components are segregated from the fuel assembly pile. The fuel assemblies are arranged in a three layer pile, with 10, 8, and 6 assemblies per layer, with no separation between the fuel assemblies. The most reactive fuel assemblies are assumed to re-arrange themselves from their periphery position in the cask and j end up clustered in the center of the fuel assembly pile in such a way as to
- maximize the reactivity of the pile. An exclusion zone established in the fuel
- storage racks ensures the resulting fuel assembly pile will not interact with other l spent fuel assemblies stored in the SFP. Also, the fuel assembly reactivity i distribution meets the following cask loading criteria
b i
- _ - . . ,-- s_ ,
.. l Page 3 of 8 1
- District Response to NRC Request for Information
- 1. No more than three fuel assemblies have a K,y greater than 0.95; e
f 2. No more than six fuel assemblies have a K,y greater than 0.90; and ;
- 3. No more than 12 fuel assemblies have a K,, greater than 0.85.
i The resulting K,, value for this analysis is 0.948.
! 6. The same as analysis 5 above, except the fuel assemblies are assumed to deform j from the impact between themselves and on the SFP storage racks, with a resulting fuel pin pitch of 96.5% of the normal fuel pin spacing. This fuel pin
- spacing maximizes the fuel assembly pile reactivity. Also, the fuel assemblies are
, assumed to arrange themselves in the pile such that the most reactive fuel l assemblies are uniformly distributed around the periphery of the pile, with the f least reactive fuel assemblies in the center of the pile.
i.
} The resulting K,y value for this analysis is 0.876.
4 l 7. The same as analysis 6 above, except the cask tips and falls into the SFP without
- ejecting the fuel assemblies. The 24 fuel assemblies remain in the cask in their
- loaded configuration, and the fuel assemblies are deformed as specified in analysis l 6 above.
j The resulting K,y value for this analysis is 0.886.
The same as analysis 7 above, except the highest reactivity fuel assemblies in a 8.
l 4 cask are arranged around the periphery of the cask and canister in a manner that maximizes K,y. ;
1 The resulting K,y value for this analysis is 0.894.
1 Comparing this result to analysis 7 above (K, = 0.886) indicates that the order or 1
sequence in which the highest reactivity fuel assemblies are placed around the cask l and canister periphery has very little impact on these deformed fuel criticality
- anlysis result.
i The assumption that has the largest impact on reactivity is the way spilled fuel assemblies i are assumed to be arranged in a pile. As seen in analyses 3 and 4 above, increasing the fuel assembly separation distance one inch significantly decreases the resulting reactivity
, (K, = 0.944 versus K, = 0.873). Analysis 5 assumes no separation between fuel i assemblies. It is not credible to assume spilled fuel assemblies will come to rest in a neat pile with little or no separation between assemblies. Even with these ultra conservative,
. non credible " neat pile" assumptions, spilled fuel assembly analyses 3 through 6 above l result in a K,y that is less than 0.95. For any credible spilled fuel accident, the resulting
. - - . - - - - - - -. - - _- -.- - - - - - = _ -
i .. .'
?
1 .
Page 4 of 8 i
District Response to NRC Request for Information fuel assembly, pin, and/or pellet pile would be less orderly and not result in the optimum moderator to fuel ratios studied above. Actual K, values would be well below the values calculated above.
The hypothetical cask tip accident that results in the highest reactivity is summarized in analysis 5 above. The following is an evaluation of the many conservative assumptions and cask and canister loading restrictions (i.e., fuel assembly reactivity cask loading criteria, fuel assembly ALARA canister loading configuration, and SFP storage rack exclusion zone) contained in this reactivity calculation.
1.- The calculation assumes the SFP contains 1200 ppm Boron, while the actual SFP Boron concentration is approximately 1280 ppm.
- 2. The District will load each canister in an administratively controlled configuration, with the most reactive fuel assemblies placed around the periphery ,
of a canister and the least reactive fuel assemblies placed in the middle.
- 3. The spilled fuel assembly scenario assumes all 24 fuel assemblies, without fuel assembly, caik, or canister components, come to rest in an orderly array of three layers of ten, eight and six undamaged assemblies on top of the SFP storage racks.
- 4. Though the fuel assemblies well be loaded into each canister as noted in analysis assumption 2 above, this calculation assumes the most reactive fuel assemblies re-arrange themselves from their initial periphery position in the cask and end up .,
clustered in the center of the resulting fuel assembly pile in such a way as to maximize the reactivity of the pile.
- 5. No separation is assumed to exist between the fuel assemblies in the pile.
Assuming no spacing between spilled fuel assemblies achieves the maximum k,.
- 6. The calculation does not take credit for the Boraflex reactivity control material that is contained in the SFP storage racks.
- 7. The District will maintain a spent fuel assembly exclusion zone in the SFP storage racks that will ensure fuel assemblies spilled from a tipped cask will not come to rest over spent fuel assemblies that are being stored in the SFP.
- 8. The calculation assumes the reactivity distribution for the 24 fuel assemblies <
ejected from a hypothetically tipped cask meets the following cask loading criteria:
- No more than three fuel assemblies may have a reactivity (K,) greater than 0.950;
.i l
4 !
Page 5 of 8 District Response to NRC Request for Information No more than six fuel assemblies may have reactivity greater than 0.900; and No more than twelve fuel assemblies may have a reactivity greater than 4 0.850.'
i l NOTE: Of the 493 fuel assemblies stored in the SFP,56 fuel assemblies have a
- K, greater than 0.950, 64 fuel assemblies have a K, between 0.900 and 0.950,115 l fuel assemblies have a K, between 0.850 and 0.900, and 258 fuel assemblies have i a K,less than 0.850. Therefore, the District can meet the above fuel assembly j reactivity cask loading criteria for loading 21 fuel storage canisters in a cask.
l-l 9. The conservatively calculated reactivity for analysis 5 above (K, = 0.948) includes a total statistical K, uncertainty factor of 0.012.
{
NUREG-0612 Criticality Compliance:
4
- The District committed to NUREG-0612 for the Rancho Seco fuel off-load project. In ]
- accordance with NUREG-0612, paragraph 5.1, item II, the District took steps to ensure .
i a cask drop "does not result in a configuration of the fuel such that K # si larger than 0.95." The (bliowing summarizes the SFP Boron status and the administrative controls that ensure compliance with NUREG-0612, paragraph 5.1, item II.
l The District has not had to add Boron to the SFP since Rancho Seco permanently shut down reactor operations and off-loaded the reactor core to the SFP in 1989. The SFP loses an insignificant amount of borated water through nominal SFP liner leakage (less ;
than two gallons per hour) to the leak chase system. Evaporative water volume losses, ;
with subsequent make-up to restore evaporative losses, have no net afTect on SFP Boron concentration.
Currently, the SFP water Boron concentration is approximately 1280 ppm. The District analyzes the SFP Boron concentration monthly in accordance with plant chemistry 1 procedures. Over the last two years, the SFP Boron concentration has averaged approximately 1,300 ppm. Plant administrative requirements require the SFP Boron i concentration be maintained above 1200 ppm during cask handling activities. Also, plant operating procedures address addition of Boron to the SFP should the need arise. l' The District intends'to maintain its current SFP Boron management program, complete with administrative monitoring, controls, and minimum concentration requirements, until the District completes removing all the spent fuel assemblies from the SFP.
Additional administrative controls related to loading cask and canisters for transfer to and storage at the Rancho Seco ISFSI are as follows: ,
m---p.-
! .. .- I 4
1 O l ,
Page 6 of 8 1 l District Response to NRC Request for Information
- 1. Cask loading procedures will require maintenance of a spent fuel assembly l l exclusion zone in the SFP storage racks that ensures fuel assemblies spilled from a l 4
tipped cask will not come to rest over spent fuel assemblies that are being stored
- in the SFP. 1 i- l, l 2. The cask loading procedures will require and verify that the fuel assembly l 1 reactivity distribution for the fuel assemblies loaded into a cask and canister meet i i the following cask loading criteria:
j -
No more than three fuel assemblies may have a reactivity (K,) greater ;
- than 0.950; I i
! =
No more than six fuel assemblies may have a reactivity greater than 0.900;
- and i No more than twelve fuel assemblies may have a reactivity greater than i 0.850.
1
- 3. Cask loading procedures will require and verify each canister to be loaded such j that the most reactive fuel assemblies will be placed around the periphery of the 1 cask and the least reactive fuel assemblies will be placed in the middle of the cask. I The conservatively calculated reactivity for the non-credible, hypothetical cask tip scenarios analyzed above is below regulatory limits (i.e., K,less than 0.95). This analysis result supports the conclusion that a fully loaded cask can be handled in the .
vicinity of the SFP without a criticality concern. l Cuk Tin and Eiection of Fuel Incident Not Considered Credible As part of a NUREG-0612, " Control of Heavy Loads at Nuclear Power Plants,"
compliance evaluation, the District determined a drop that resulted in the tipping of a spent fuel transfer cask (an edge drop) and ejecting spent fuel assemblies from the cask into the Spent Fuel Pool (SFP) is not a credible event. NUREG-0612 identif'ies the .
various causes of crane failures and load drops. The greatest cause of crane load drop i accidents is crane operator error in either the control of the crane and its load or the selection of appropriate rigging. The District evaluated the NUREG-0612 crane failure i and load drop causes and concluded that dropping a transfer cask on an edge in the Fuel '
Storage Building (FSB) is an extremely unlikely event and is not a significant hazard.
. Page 7 of 8 District Response to NRC Request for Information First, the cask handling and Gantry Crane rigging and operation training program meets NUREG-0612, ANSI B30.2, and 10 CFR 72 requirements. The District will train and certify all personnel involved in cask handling operations and Gantry Crane rigging and operation in accordance with the approved training program. In addition, the District will procedurally control all cask movements and will follow pre-determined safe load paths; thereby eliminating unplanned or unsafe moves.
The District will pre-determine safe load paths using markings, targets, and pointers.
This approach removes human judgment from determining whether the crane operator has properly positioned a cask prior to commencing a move. Cask movement
, procedures require use of markings, targets, and pointers. These procedures ensure (1) proper cask positioning before each move and (2) operator adherence to established safe i
load paths.
4 The District specifically designed and dedicated Gantry Crane cask handling rigging for
- each cask handling task. The Gantry Crane rigging design is in accordance Jth the single failure proof criteria addressed in NUREG-0612 and ANSI N14.6-
- Also, procedures require verification of proper rigging installation prior to moung a cask.
These design and procedural controls eliminate rigging failures.
As a further safeguard against operator error, the District has designed Gantry Crane controls such that interlocks allow only one Gantry Crane motion when the crane operates in the Fuel Storage Building. Therefore, when the crane operator drives the crane horizontally to move the cask over a SFP edge, the hoist machinery is locked out i and both hoist brakes are set.
j Each Gantry Crane hoist brake is rated at 150% of the 185 ton hoist capacity (total braking capacity of 555 tons). Since a fully loaded cask will weigh approximately 126 tons, this brake design provides a total cask load handling safety margin of 440% (220%
for each hoist brake). Furthermore, the District tested a single strand of the new crane hoisting rope to a breaking strength load equivalent of 743% of the 185 ton Gantry Crane hoist capacity. The ANSI B30.2 and Crane Manufacturers Association of America standard for this breaking strength test is 500% of hoist capacity. The 743% of I
hoist capacity wire rope breaking strength is equivalent to a tensile strength that exceeds 10 times the cask lond the crane will actually handle. Using a wire rope proof tested to
~
more than 10 times ine actual cask load meets the wire rope single failure proof design criteria specified it NUREG-0612 and ANSI N14.6. Therefore, the robust Gantry Crane design provides reasonable assurance that the crane hoisting machinery and control failures discussed in NUREG-0612 will not occur.
In addition, the District (1) proof tested the hoist machinery, trolley, and bridge center span with a 231.19 ton test weight, and (2) proof tested the bridge cantilever section with a 162.5 ton test weight. Given the margin of safety between the test weights and the cask weight, a structural failure leading to a load drop with the hoist machinery inactive is not credible.
l ~
Page 8 of 8 District Response to NRC Request for Information Many load drop events discussed in NUREG-0612 resulted from improper or inadequate crane maintenance. In preparation for the Rancho Seco fuel olT-load project, the '
District completely refurbished the Gantry Crane with a new, improved plow steel hoisting rope and a new, state-of-the-art control system. Also, the District performed ,
NDE examinations of the primary load carrying elements, such as sheaves and load blocks, and recertified these elements through inspection and test. >
A desig 1 basis earthquake during cask handling activities would impose maximum credible loads on the Gantry Crane structure and hoist machinery. The Gantry Crane design considered earthquake stresses while handling the design loads on the Gantry Crane center span (185 tons) and cantilever (130 tons) sections. District Calculation No. ;
Z-CHS-M2555 verified that a design basis earthquake would nol exceed Gantry Crane
, design allowable stresses and overturning moments.
i l During cask handling activities, the crane will carry a newly loaded cask over a SFP or i
- Washdown Structure edge for less than a minute as the crane moves the cask from the
- SFP to the Washdown Structure. Except for safety reasons, procedures will not permit
! deliberately stopping Gantry Crane motion while the cask is over a SFP or Washdown Structure edge.
4 Finally, additional design features, such as (1) electrical limit switches and mechanical i stops, which limit east-west trolley travel, (2) limit switches and bridge stops, which limit
, north-south main bridge travel, and (3) a fuel assembly exclusion zone in the fuel storage ,
rack locations adjacent to the area where casks will be handled, prevent a transfer cask i
. from:
i j' l. Moving over the fuel storage racks in the SFP; or I l 2. Possibly dropping or tipping and spilling fuel assemblies onto spent fuel
- assemblies stored in the SFP.
L
! In keeping with the defense in depth philosophy of NUREG-0612, paragraph 5.1, the i District has taken several other significant design and administrative steps to ensure a cask drop does not occur during the Rancho Seco fuel off-load project. Both the Safety Analysis Report and No Significant Hazards Consideration submitted to thd NRC in District letter MPC&D 96-034, dated March 18,1996 (Proposed License Amendment No.192), discusses these steps and associated analyses addressing cask drop issues.
Based on the above analysis, the District considers a cask drop during the time a cask is over a SFP or Washdown Structure edge to be (1) highly improbable, and (2) not credible.