ML20214T027
| ML20214T027 | |
| Person / Time | |
|---|---|
| Site: | Diablo Canyon  |
| Issue date: | 06/01/1987 |
| From: | Ferguson R GRUENEICH, D.M. (FORMERLY GRUENEICH & LOWRY), Sierra Club |
| To: | |
| References | |
| CON-#287-3682 OLA, NUDOCS 8706100093 | |
| Download: ML20214T027 (48) | |
Text
r :-
. 3&#
00ME iiT uwK 1 Edwin F. Lowry Dian M. Grueneich
'2 GRUENEICH E LOWRY '87 JUN -8 P3 :23 380 Hayes Street, Suite 4 3 San Francisco, California 94102 gn . .
(415) 861-6930 00C5D C " ?d 4 Attorneys for The Sierra Club W H'"
5 6 UNITED STATES OF AMERICA 4 7 NUCLEAR REGULATORY COMMISSION 8 BEFORE THE ATOMIC SAFETY AND LICENSING BOARD 9
)
10 In the Matter of: ) Docket Nos.
) 50-275 and 50-323 OLA 11 PACIFIC GAS & ELECTRIC COMPANY )
)
12 (Diablo Canyon Nuclear Power )
Plant, Units 1 and 2) )
13 )
1
- 14 TESTIMONY OF RICHARD B. FERGUSON 1
! 15 This testimony is presented on behalf of the Sierra Club, i 16 Santa Lucia Chapter, and addresses the contentions raised by the i 17 Sierra Club and set for hearing in this proceeding.
18 The Licensee, Pacific Gas and Electric Company (PG&E), has j 19 made no serious attempt to inform or solicit views from the l
20 population of San Luis Obispo County regarding either the i 21 proposed reracking or alternatives to the proposal. No public i
l 22 meetings have been held, no proposals circulated for public j 23 review and comment, no environmental impact statements have been
} 24 prepared, and no review by the County of San Luis Obispo i
- 25 Environmental Coordinator's Office has been requested.
26 The Sierra Club, Santa Lucia Chapter, regularly reviews I- 1 i
~'
0706100093 870601 PDR ADOCK 05000275 -
T pyg I
I 1 proposals for projects which may affect the environment of San 2 Luis Obispo County. The Club first learned of the proposed 3 raracking at Diablo Canyon several months after the Reracking 4 Report had been submitted when a proposed no significant hazards 5 determination was published in the Federal Register (Reference 6 1). To the best of my knowledge, no member of the public of San 7 Luis Obispo County had been contacted by PGGE or by the USNRC 8 regarding the proposed reracking prior to this announcement.
9 Since that time, there have been occasional news articles
~
10 in the local ~~ media regarding ~ the current ' licensing proceeding 11 before the NRC, and a prehearing conference was held in San Luis 12 Obispo County on May 13, 1986. In addition, some materials 13 regarding the reracking have been placed in the public document 14 room at the library, California Polytechnic State University, 15 San Luis Obispo. There has been, however, no attempt to inform 16 the general public about the raracking in any meaningful way.
17 The Roracking Report (Reference 2), submitted by PG&E in 18 support of their license amendment application, considers only 19 three alternatives to the currently proposed roracking, namely:
20 shipment of fuel to a reprocessing or independent spent fuel 21 storage / disposal facility; shipment of fuel to another reactor 22 site; and not operating the plant after the current spent fuel 23 storage capacity is exhausted.
24 The licensee has failed to consider other alternatives 25 including:
26 a) other means of expanding existing storage facilities, 2
= m __ _____ . ____ _ _ . . m_.. __
1 including high-density spent fuel storage racks which would not i
2 be expected to impact with other racks or with the walls of the 3 spent fuel pools during seismic events; 4 b) construction of new or additional spent nuclear fuel 5 storage facilities; 6 c) acquisition of modular or mobile spent nuclear fuel 7 storage equipment, including spent nuclear fuel storage casks; 4
8 d) such other technologies for spent nuclear fuel storage 9 as may be approved by the Commission for use at Diablo Canyon.
4
~'
(Reference 3)
~
10 11 In my opinion, it is the view of those living in proximity 12 to the reactor at Diablo Canyon that on-site storage of spent 13 nuclear fuel shall employ the safest of reasonable storage 14 technologies. Since no attempt has been nade to demonstrate that 15 the proposed reracking represents the safest of reasonable 16 alternatives, the reracking is not consistent with the views of
! 17 the public.
18 RERACKING REPORT AND SAFETY EVALUATION i
19 The Reracking Report issued by PG&E in September, 1985, 20 contains several inaccuracies which raise serious doubts as to 21 the veracity of claims that the proposed high-density fuel racks 22 would be able ,to withstand the safe shutdown earthquake ("SSE").
23 The Roracking Report lists the maximum impact force on a rack as ,
24 71,400 lb. (Reference 4) The report fails to mention that this 25 value is predicted for one corner only, and that the total 26 impact force on the rack could be four times this amount, or 3
i
I I nearly 300,000 lbs.
2 Furthermore, other calculations indicate that the maximum 3
force on one corner could be as much as 105,000 lb, for a total 4 of over 400,000 lb. (Reference 5) There is even some indication 5
that the maximum computed impact force on one corner could be as 6 high as 125,000 lb. (Reference 6), with a maximum force of 7
perhaps 500,000 lb. To my knowledge, the Reracking Report has 8
never been amended to reflect these higher values, even though 9
they may represent an increase of 75% over those reported.
~
^ 10 The Reracking Report also fails to mention the likelihood 11 of collisions between the fuel racks and the walls of the spent 12 fuel pools, although the potential for such collisions is now 13 recognized. (Reference 7) It is now known that in response to 14 the SSE some of the racks are indeed expected to impact the 15 walls of the pools (Reference 5). Again, the Reracking Report 16 has never been amended to reflect this significant fact.
17 This testimony will address collisions between the racks 18 and the pool walls in more detail later. It should be pointed 19 out, however, that as late as November 21, 1985, two months 20 after the submission of the Reracking Report, engineers for PGGE 21 did not know that the fuel racks were expected to impact the 22 walls. In a memo to E.C. Connell, Sean Dhattacharya writing for 23 J.
K. McCall said (Reference 8):
24 "The girdios [ sic] bars, of racks H, N and E will 25 reduce the distance to the Spont Fuol Pool wall to loss that the postulated maximum deflection of the racks. Joseph Oats [ sic] should address and verify 26 that the actual displacement of thoue and other 4
4
1 peripheral modules is such that there is no impact to the pool walls."
2 3 The Reracking Report is, therefore, an incomplete document.
4 As mentioned above, it fails to discuss the full range of 5 alternatives to the proposed reracking. It substantially 6 underreports the expected impacts between racks and it also 7 fails to mention impacts between the racks and the pool walls.
8 The S_afety Evaluation (reference 9), was issued by the 9 USNRC on May 30, 1986, simultaneously with the issuanco of the 10 license amendment for the high density reracking of the spent 11 fuel pools at Diablo Canyon. It contains as an appondix a 12 Technical Evaluation Report, the principal author of which was 13 R. C. Herrick, now deceased, of the Franklin Research Conter.
14 The date of his report is April 30, 1986. Mr. Herrick also filed 15 an affidavit in the 9th Circuit Court of Appeals (Reference 10) 16 regarding an analysis of the Hosgri seismic data made by this 17 author. In his affidavit, Mr. Horrick implies that my analysis 18 is incorrect. Soveral others, including ongincors at PGGE, had 19 checked my work and found it to be accurato (Reference 11). As a 20 result, one must question either Mr. Iforrick's professional 21 qualifications, or his objectivity. In either caso, the occuracy 22 of his Technical Evaluation Report is suspect.
23 Mr. llorrick's review was performed approximately six months 24 after the Roracking Report was submitted. In the intervening 25 timo, consultants for PGGE had completed moro calculations 20 showing clearly that the rack impact forces woro higher than 5
t ,
I those reported in the Reracking Report. (Reference 5) Mr.
2 Herrick chose to ignore these higher values, citing only the 3 material in the Roracking Report. (Reference 12) 4 Mr. Herrick does mention the' predicted rack-wall collisions 5 (Reference 13), and raises no questions concerning these-6 impacts, although, as we have shown, only a few months before, 7 PGEE's engineers were concerned that they not occur. This 8 testimony will address the acceptability of such collisions in 9 detail later.-
10 Mr. Herrick reports on the theoretical analyses used to 11 predict the behavior of the racks without significant criticism.
12 Although the analysis makes use of theoretical results beyond 13 their demonstrated region of applicability, Mr. Herrick approves 14 the extrapolation without question. He says (Reference 14):
15 "While this is opposite to the conditions that prevail for spent fuel rack modules, the technique is based 16 upon well established principles in fluid mechanics and serves to provide a lower bounding estimate of the 17 fluid coupling for rack module analysis."
18 He provides reference neither to the "well-established princi-19 ples" nor to any evidence that the technique does provide a 20 " lower bounding estimate".
21 Most striking, perhaps, is Mr. Herrick's failure to discuss 22 critically the methodology used by PG&E and its consultants to 23 model the complicated behavior of the sixteen fuel racks in the 24 spent fuel pools. This methodology is addressed later, but it 25 should be pointed out that another consultant for the NRC, Mr.
26 J. DeGrassi of Brookhaven National Laboratories, did raise 6
I substantive questions about this methodology when he reviewed a 2 later reracking application for the Byron nuclear power station.
3 (Reference 15) 4 The seismio analysis for the Byron plant was similar to 5 that to the Diablo Canyon analysis, performed by the same 6 consultants. Mr. DeGrassi neems to have recognized, where Mr.
7 lierrick had not, that the simplifying assumptions made in the 8 analysis had not been demonstrated to be conservativo. That is, 9 there was no evidence that tho impact and other stresses on the 10 racks to be expected in actual practico were less than those 11 computed by the soismic analysis. In recent months, as a result 12 of questions raised by Mr. DeGrassi and other NRC consultants, 13 additional computations for the Diablo Canyon raracking have 14 been performed in an attempt to provido this evidence,. (Refor-15 onces 16 - 18) 16 That Mr. Herrick's review was inadequato was demonstrated 17 in February, 1987, nearly a year aftor the Safoty Evaluation was 18 writton. Even though this document had boon accepted and a 10 liconao iocued many months beforo, hearings before the Atomic 20 Safety and Licensing Doard woro dolayod an additional throo 21 months at the requent of NitC staff in ordor to gathor additional 22 information.
23 While the NRC ntaff in to be commanded for ito incronood 24 attontion to tho Diablo torncking problem, the biotory of the 25 proconding illustratos twd saliant factn. Tho work of Mr.
20 Horrick and tho Safety Evaluation banod upon it in simply not 7
I credible; even the NRC's own staff have found it an inadequate i
2 review. In addition, the technical issues raised in the proposal 3 to rarack the Diablo Canyon spent fuel pools are evaluated 4 differently by different reviewers.
5 These technical difficulties arise from the seismicity of 6 the Diablo Canyon site. The original fuel racks were anchored to ,
7 the concrete floors of the pools, but the proposed new high-8 density racks are free-standing and expected to undergo signifi-g cant motions in response to the safe-shutdown earthquake ;
10 (Reference 4). The issues raised lately by the NRC staff and 11 consultants (References 16 - 18) and raised earlier as conten-12 tions by the Sierra Club (Reference 19) relate primarily to the 13 potential damage to the fuel racks resulting from this motion.
14 The technical issues will be dealt with in greater detail 15 later. Given the complexity of these issues, however, it is 16 clearly not unreasonable for the public to expect that alterna-17 tive storage technologies in which the racks or other equipment 18 are secured against seismic motion be considered. ,
19 OT POSITION i 20 The Sierra Club has pointed out in its " Motion for Summary 21 Disposition" (Reference 20) that currently approved kinematic '
22 acceptance criteria for spent fuel storage systomo prohibit l 1
23 racks from colliding with each other in response to coinmic l
24 excitation. It also pointed out that the analyson used by the 25 liconno applicant to study the bohavior of the proposed upont 26 fuoi ayatom at Diablo Canyon in rooponoo to the pontulated 8 i
1 Hosgri event (PHE) do predict collisions between the fuel racks 2 and between the racks and the walls of the spent fuel pools 3 (References 2, 5, 9). The proposed raracking at Diablo Canyon 4 would seem, therefore, to represent a prima facie violation of 5 the current acceptance criteria as recorded in the Standard 6 Review Plan, NUREG-0800 (SRP) (Reference 21).
7 The SRP includes spent fuel racks in the list of Category I 8 seismic equipment, equipment which is required to remain g functional during earthquakes. According to the SRP, all 10~ Category I equipment is prohibited from sliding and overturning.
11 Safety margins against these motions are required to be main-12 tained in the design of the spent fuel system. The margins vary 13 from 1.1 to 1.5 according to the severity and likelihood of the 14 earthquake. (Reference 21 ). The SRP was revised in 1981 to 15 include updated guidelines specified in the "OT Position for 16 Review and Acceptance of Spent Fuel Storage and Handling 17 Applications" (Reference 22 ). The OT Position refers to these 18 as " factors of safety against sliding and tilting." (Reference 19 23 )
20 Since oliding and tilting motions woro prohibited by the 21 critoria of the SRP, thoro was no nood for the SRP to discuos 22 the analysis of impacto botwoon racko and/or betwoon racko and 23 the wallo of tho opont fuel pools during colomic evento and no 24 auch discuoolon occuro. All 011 ding and tilting wan prohibited 25 and auch bohavior was thoroforo procumod not to occur. Critoria 20 for the analyoin of impacto from externally and internally 9
I generated missiles, dropped loads, etc., are discussed in some 2 detail in various sections of the SRP, but no treatment was 3 included for the analysis of impacts cased by interrack col-4 lisions or by rack-wall collisions.
5 In order to satisfy the prohibition against sliding and 6 tilting, como spent fuel storage systems include anchoring 7 dovices imbedded in the foundations of the pools to prevent 8 motion of the racks. The storage system originally constructed 9 at Diablo Canyon was of this type. The low-density racks were 10 boltod to " imbeds" anchored in the floor of the pools. During 11 the summor of 1986, the original racks woro removed and rein-12 stalled in the Diablo Canyon Unit I pool. The racks in this 13 pool are currently fastened to the imbods by means of welds. If 14 the wolds and imbods are sufficiently strong, the racks comply 15 with the SRP restriction against sliding and tilting.
16 During the 1970's, it became apparont that the utility 17 companios owning nucioar reactors would be requirod to store 18 spent fuoi at the reactor sitos for an indefinito length of 19 timo. On-sito storago was necessitated by the failuro of tho 20 fodoral govornmont to provido spent fuel storago and reprocess-21 ing facilities. In responso to the nood for expanded on-sito 22 storago, technology becamo availablo which in cortain circum-23 utancos enabled the undorwator storago of spent fuoi olomonts in 24 moro donsoly packed configurations than had previously boon 25 feasible.
26 t,oadod fuoi racks in the newor configurations tended to bo 10
l 1 .,
1 more massive than before, and technical considerations made it 2 difficult, if not impossible, to anchor the new racks to the 3 existing pool foundations. Although free-standing racks have 4 certain technical advantages over anchored racks, a major 5 disadvantage is the potential for sliding and/or tilting motions 6 to occur during seismic events.
7 Unlike some other Category I equipment, spent fuel racks 8 are not connected to electrical, gas, or other systems. Since 9 the racks are surrounded by water to remove heat produced by the IO fuel, small rdovements by the racks are not likely to lead to 11 catastrophic accidents. The major concern arising from rack 12 motion is the potential for racks to collide with each other or 13 with the walls of the pools and to become damaged by the 14 collision impacts. So long as rack movement does not lead to 15 collisions, the complete prohibition against sliding and tilting 16 might be safely relaxed. Free-standing racks could then be used 17 in spent fuel storage systems if it could be shown that inter-18 rack collisions and collisions betwoon the racks and the pool 19 walls could not tako placo.
20 FUEL IMPACTS 21 If spent fuel racks are to 60 froo-standing and able to l
22 undergo sliding and tilting motions, however, it is necessary to 23 havo guidelinos as to how the amplitudo of thoso motions shall 24 be predicted so as to verify that the amplitudos of the motions 25 will bo " minimal". Mathematical models for predicting seismic 20 motions of the pools themselves had becomo moro or less stand-11 i
l t _
(. .
1 ard, but those used to predict resulting rack motions were not.
2 In high density spent fuel racks, the fuel elements, which 3 consist of bundles of fuel rods, are held vertically in metal 4 sleeves. There is a space between the sleeve and the fuel 5 element called a " gap" or a " clearance" which is needed for 6 several technical reasons. In the racks proposed for Diablo 7 Canyon, for example, this gap is approximately 1/6 inches wide.
8 Although the bottom of the fuel is fixed in place, the rest of g
the element can " rattle" back and forth in the sleeve if the
-:c2 . -
g g:g,_.haken s during an e'arthquake. The resulting impacts 11 between the fuel and the sleeves not only have the potential to 12 damage the fuel elements, but also to influence the motion of 13 the rack differently than if the fuel were held rigidly in 14 place. Regulators at the NRC were aware of this problem, of 15 course, and sought to ensure that these impacts were fully 16 accounted for by rack designers when predicting rack motions.
17 Thus, according to the OT Position (Reference 24 ):
18 "Due to gaps between fuel assemblies and the walls of the guide tubes, additional loads will be generated by 19 the impact of fuel assemblies during a postulated seismic excitation. ... It should be demonstrated 20 that'the consequent loads on the fuel assembly do not lead to a damage of the fuel."
21 also:
22
" Details of the mathematical model including a 23 description of how the important parameters are obtained should be provided including the following:
24 the methods used to incorporate any gaps between the support systems and gaps between the fuel bundles and 25 the guide tubes;..."
26 Mathematical models of fuel and rack behavior have become 12
~
w f., '
I more sophisticated in recent years by the incorporation of a 2 phenomenon known as " fluid coupling" (References 25 -
29 ).
3 When~ : an object under water moves, some water must also move.
4 Fundamental la'ws of physics require that if the object acceler-s .; 5 ates the water,' the water must exert a force on the object. The 6 object therefore moves differently underwater than it would out 7 of water, since the motion of the object is " coupled" to that of 8 .the fluid. Some progress has been made in predicting how these 9
forces influence the behavior of fuel and fuel racks during 10 seismic events. The theories ~are, however, relatively new and 11 have not been experimentally verified. Fluid coupling effects
~11 iM' . 12 have been included in the analysis of rack motion at Diablo i
a 13 Canyon and perhaps at other plants. (Reference 30 ). The OT i,
14 Position cited above requires that mathematical details like 15 fluid coupling b6~ adequately described.
"\ 16 ACCEPTANCE CRITERIA 17 So long es adequate attention was focussed on the problem 18 of fu'el impacts and the necessity of avoiding interrack and 19 rack-wall collisions, the NRC could relax the prohibition 20 against - all - rack movement. The Structural Acceptance Criteria 21 promulgated in the OT Position do so by making two exceptions to 22 the SRP positions as follows (Reference 31 ):
"When considering the effects of seismic loads,
'23 factors of safety against gross sliding and overturn-24 ing of racks and rack modules under all probable L. 25 service conditions shall be in accordance with the Section 3.8.5.II-5 of the Standard Review Plan. This 26
, position on factors of safety against sliding and '
,pp tilting need not be met provided any one of the 13
4 1 following conditions is met:
2 "a) it can be shown by detailed nonlinear dynamic analyses that the amplitudes of sliding motion are 3 minimal, and impact between adjacent rack modules or between a rack module and the pool walls is prevented 4
provided that the factors of safety against tilting are within the values permitted by Section 3.8.5.II-5 5 of the Standard Review Plan.
6 "b) it can be shown that any sliding and tilting motion will be contained within suitable geometric 7
constraints such as thermal clearances, and that any impact due to the clearances is incorporated."
8 9 Exception (a) deals only with sliding displacements and 10 requires any sliding displacements to be " minimal" but does not 11 define this word. Evidence from the literature indicates that 12 " minimal" has been taken to mean less than half the equilibrium 13 spacing between racks (Reference 32 ). If we use this defini-14 tion, the first part of the exception is nearly identical to the 15 second, that " impact between adjacent rack modules or between a 16 rack module and the pool walls is prevented..." For if sliding 17 motions are less than half the distance between racks, then 18 impacts will be prevented so long as the racks do not undergo 19 tilting motions. Thus exception (a) is clearly written to 20 prohibit rack collisions for situations in which rack motion is 21 expected to be due entirely to sliding. (It should be noted 22 that the OT Position refers here to " tilting" motions, rather 23 than " overturning". There is some confusion between these terms 24 in the SRP and they seem to be used interchangeably. In the 25 present case there can be no misunderstanding as to the authors' 26 intent, since the language refers specifically to " tilting".)
14
.- . .= . - . .
1 Exception (b) was included as guidance for situations in 2
which both sliding-and tilting motions can be expected. This 3
clause also prohibits rack collisions by requiring that all rack 4
motions be contained within suitable geometric, as opposed to
- 5 physical, constraints.
As an example of such - a geometric 6
constraint, they cito " thermal clearances", by which they mean a 7
gap designed to accommodate thermal expansion of a rack. Their 8
choice of the word " clearance" is perhaps unfortunate, since the 9 second condition in exception (b) also uses the term:
10
~ ^ ~ incorporated."
" . . 7 and ' that any impact due to the clearances is 11 12 It is my opinion that this phrase refers to the previous '
13 discussion regarding the impacts of the fuel elements as they 14 rattle in the rack sleeves due to the " gaps" or " clearances" l 15 between the fuel and the sleeves. These two terms are also used 16 interchangeably as we see, for example, in the OT Position:
is11 caps (clearances or expansion allowance) and 17 sliding contacts should be indicated." (Reference 33 )
18 The language of exception (b) should therefore be interpreted to l- 19 read:
i 20 "..., and that any impact due to gaps between fuel '
21 assemblies incorporated."
and the walls of the guide tubes is 22 Whether or not one agrees with this interpretation, it 23 should be noted that the " impact due to the clearances" is to be 24 incorporated in addition to the requirement that rack motion be 25 contained within suitable geometric constraints. It is, there-26 l fore, inconceivable that exception (b) was intended to permit 15 m.._..,.
I rack collisions to occur.
2 There is considerable evidence that both the utilities and 3 rack designers have interpreted the OT Position as a prohibition 4 against rack collisions. K.P. Singh was the vice-president for 5 engineering at the Joseph Oat Corporation, manufacturers of the 6 high density racks for Diablo Canyon. Alan Soler is a consul-7 tant to Joseph Oat Corporation and author of the report " Seismic 8 Analysis of High Density Fuel Racks for Pacific Gas & Electric 9 for Diablo Canyon Nuclear Power Station" (Reference 5).
10 Together they published a paper entitled " Seismic Response of a 11 Free Standing Fuel Rack Construction to 3-D Floor Motion" in the 12 professional journal Nuclear Engineering and Design 80 in 1984 13 (Reference 27). This paper discusses certain techniques for 14 evaluating the effects of fuel elements rattling in a rack 15 during seismic activity upon the behavior of the rack. Regard-16 ing their criteria for acceptable rack motion they state 17 (Reference 35 ):
18 "In addition to stress limitations, adjacent racks must not impact during a seismic event. In the 19 simulation herein, virtual mass effects from gaps between racks have been included based on adjacent 20 rack separation equal to 3" (76.2 mm). Therefore, assuming the worst motion of adjacent racks, inter-21 rack impact is precluded if the maximum corner deflection of the rack in either direction is less 22 than 50% of the rack spacing." (Emphasis added.)
23 There is some evidence that the acceptance criteria and 24 the prohibition against collisions were discussed by representa-25 tives of PG&E and the NRC. A meeting on December 5, 1985, uas 26 held to present the reracking proposal to the NRC. According to 16
1 brief notes which' serve as.a summary of this meeting (Reference 2 36 ), PG&E referred to the acceptance criteria:as:
3 " Acceptance Criteria
-4
- Kenematic [ sic]
tilting)"
(No inter-rack impact, 1.5 against 5
6 No further notes on the discussion of this topic at that meeting have been provided. It is evident, however, that the 7
8 reference is to exception a) of the OT Position, which prohibits inter-rack. impact, and the SRP safety margin of 1.5 against 9
gross tilting motion.
- jo-There is also evidence that owners of other. nuclear-plants 11 12 have recognized the prohibition against collisions. K.R. Singh and T.
i 13 Lee-Ng, also of Joseph Oat Corp., authored a review of 14 several spent fuel facilities entitled " Planning Study for 15 Experimental Measurements and Analytical Correlations of Fluid 16 Drag of Fuel Assemblies in-Fuel Rack Storage Locations" (Refer-ence 37 ). Regarding collisions, they cite reports from the 17 18 Crystal River Unit 3 and the Ocobee nuclear plants:
Crystal River:.
19 "It has been concluded, therefore, that the gaps
- 20 provided between storage racks (1 inch minimum) - and 21 between storage racks and pool structure (7/8 inch
. minimum) are sufficiently ample to preclude any rack-to-rack or rack-to-pool structure - collisions 22 under the SSE event..."
23 also:
24 "The free standing concept would preclude the need for 25 any pool structure / liner modification if it could be shown other that and 1) the with thepool racks will not and structure, collide
..." with each i
26 i
17
~
l I
, . _- ~ . -_. . . . - _ _ . . __
l 1 Ocobee:
2 "The racks rest freely on the pool floor and are 3
evaluated to ensure that under various loading conditions they do not impact each other, nor do they impact the pool walls."
4 5
In the Diablo Canyon reracking review, there is no indica-6 tion that standard structural acceptance criteria were not used.
Indeed, 7
the Safety Analysis issued by the NRC specifically 8
states that the SRP criteria were used by the NRC (Reference 38):
9 1
10 "4.2 Applicable Codes, Standards and Specifications
- - "The staff has evaluated. the acceptance criteria for 11 reracking used by the licensee with respect to those in ' Staff Position for Review and Acceptance of Spent I
12 Fuel Storage and Handling Applications' (Ref. 34).
These criteria were previously reviewed and accepted 13 by the staff during other spent fuel pool expansion 14 applications and, .therefore, are also acceptable for this reracking design."
15 In other words, no attempt has been made by the NRC staff 16 to. formulate new acceptance criteria for the Diablo Canyon 17 reracking. Staff has used the same standards as they have for 18 other rerackings. The technical review performed for - the NRC 19 Safety Analysis by their consultants also lists the SRP and the 20 OT Position as " Applicable Criteria" (Reference 39 ). Neither 21 the NRC Safety Analysis nor the PG&E "Reracking Report" (Refer-
.22 ence 2) mention any. alteration or modification of these criteria 23 for the Diablo Canyon reracking.
24 Curiously enough, the Safety Analysis does not claim that i
25 the proposed reracking satisfies the acceptance criteria in 26
!. " Staff Position for Review and Acceptance for Spent Fuel Storage 18 t-
1 and Handling Applications". The above quotation states only 2
that the licensee used the same criteria. Nowhere in the Safety 3
Analysis is any reference made as to whether or not' applicable 4 criteria are, in fact, met.
5 The only indication that the applicant has used different 6
structural acceptance criteria occurs in the Reracking Report 7
(Reference 40 ):
8 "6.5 STRUCTURAL ACCEPTANCE CRITERIA 9
"There are two sets of criteria to be satisfied by the rack l modules:
10 "a. Kinematic Criterion-11-12 "This criterion seeks to ensure that the rack is a physically stable structure. Diablo Canyon racks are 13 designed to - sustain certain . inter-rack impact at designated locations in the rack modules. Therefore,.
14 physical stability of the racks is considered along with' .the localized inter-rack impacts. Localized permanent deformation of the module is permissible, so 15 -
long as the minimum flux-trap gap in the. active fuel region is not violated.
16 i
"b. Stress Limits 17 i
"The stress limits....."
l- 18 19 No citation is given for their " criterion", nor is the 20 language of the " criterion" specified, except for a vague 21 reference to the necessity of maintaining a physically stable 22 structure.
No mention is made of the criteria detailed in the 23 OT Position nor is there any guidance as to how collisions i
24 l-should be modeled in order to calculate " inter-rack impact."
25 This appears to be an ad hoc criterion (unpublished, unreviewed, l
26 and not subj ected to scientific dissemination, inquiry, and i
19
1 criticism) which negates the restrictions of the Standard Review 2 Plan and which does not provide useful guidance concerning rack 3 collisions.
4 If the.above acceptance " criterion" refers to the criterion 5 published in the OT Position and the'SRP, the authors of the 6 Reracking Report have seriously misrepresented the OT Position, 7 because the OT Position does much more than " seek [ing] to ensure 8 that the rack is a physically stable structure" --
it forbids 9 impacts, as we have shown.
10 The ' applicant freely admits that the proposed spent fuel 11 racks are expected to collide during the PHE. (References 41 -
12 44) Indeed, the estimation of the magnitude of the forces 13 generated by these collisions is a major objective of the 14 seismic analysis. (References 25, 26, 34 ) The proposed 15 reracking at Diablo Canyon represents, therefore, a clear 16 violation of the only existing NRC guidelines and regulations.
17 If existing acceptance criteria prohibiting rack collisions 18 are ignored, a new set of criteria must be devised for deciding 19 when rack collisions shall be deemed acceptable. This is an 20 enormously complicated task; to my knowledge no one has at-21 tempted it in any systematic and scientifically valid way.
22 Certainly no new documents purporting to deal with acceptance 23 criteria for spent fuel rack collisions have been cited in the l
24 current proceeding.
l 1
25 The technical analyses presented by the NRC and PG&E and 26 their consultants do, however, make use of certain ad hoc 20 i
l
(
1 " criteria" for their claims of safety (see above). However, 2 these ad hoc criteria are inadequate, in that they ignore 3 certain phenomena which are important to rack integrity and in 4 that certain assumptions have been made which are demonstrably 5 non-conservative. In addition, there has been no attempt to 6 obtain experimental or other empirical evidence as to the 7 adequacy and validity of these criteria.
8 STRUCTURAL ANALYSIS -
9 It may be argued that the requirements of the OT Position
' ~ ~
'10 regarding'se'ismic loads can be ignored if an adequate engineer-11 ing analysis indicates maximum loads remain below allowable 12 values. The difficulty with this interpretation is that there 13 exist no guidelines other than the OT Position to indicate what 14 might constitute an adequate analysis. As pointed out earlier, 15 the author of the Safety Analysis evidently believed that the 16 analysis submitted with the Reracking Report was adequate.
17 However, later reviewers clearly did not. Such disagreement is 18 not at all surprising, since existing NRC regulations prohibit 19 the behavior the s.31smic analyses attempt to justify.
20 It is my opinion that the seismic analyses, including those 21 done in recent months, fail to adequately demonstrate that the 22 proposed spent fuel storage system would safely survive an 23 earthquake of the magnitude postulated for the Diablo Canyon 24 site. This opinion is based on the following:
25 a) The seismic analyses make use of a new theory regarding 26 the interactions between the motions of the fuel racks and the 21 l
..-m. - a . . , _ ..m J
1 motion of - the . water in the spent fuel pools, the so-called 2 " fluid coupling theory". The accuracy of the ' seismic analyses 3 and therefore the safety of the proposed storage system depends 4 critically upon the validity of this fluid coupling theory. In 5 spite of the importance of this theory, its applicability to 6 situations like those encountered in the spent fuel pools has 7 never been experimentally verified. The lack of experimental 8 support for the fluid coupling theory used in the seismic 9 'analyues raises serious questions about the analysis' reliabil-10 ity.
11 b) The seismic analyses necessarily involve simplifying
'12 assumptions to reduce the complexity of the fuel rack movements
.13 to mathematically manageable levels. Such assumptions are common 14 in analytical work. It is essential, however, that the simpli-15 fled models used for the analyses be demonstrably conservative.
16 That is, it must be demonstrated that the values of the stresses 17 obtained using the simplified model are less than those which 18 would be obtained from more complete realistic models. In spite 19 of the recent additional analytical work which has been per-20 formed, there remain several areas in which such conservatism 21 has not been adequately demonstrated.
22 FLUID COUPLING THEORY 23 The fluid coupling theory used in the seismic analyses are 24 based on the work of R. J. Fritz which was published in 1972.
25 (Reference 29) Fritz recognized that if two objects separated by
~26 fluid move relative to each other, the fluid must also move.
22 i
1 That is, the movement of the intervening fluid is " coupled" to 2 that of the objects. .The result of this coupling is that the
'3 fluid ~ exerts forces on the objects which influence the movement 4
of the objects. These fluid coupling forces can play an impor-5 tant role in determining .the- response of the objects to other 6 disturbances like seismic events.
7 Fritz's work was explicitly limited to relative motions 8 which are small compared to the spacings between the objects. He 9 was able to derive formulas for object's of various shapes and to configurations which enable one to evaluate the effect of fluid 11 coupling forces in a' straightforward way.for situations in which 12 his. original assumption of small displacements is valid.
13 These formulas were subjected to experimental tests, also 14 performed by Fritz and reported in the same publication. The 15 reported agreement between experiment and theory was good enough 16 to lend credibility to the theoretical results.
17
'The authors of the seismic analyses for the Diablo Canyon 18 spent fuel storage system, A. Soler and K. Singh, attempted in 19 1982 to apply Fritz's theory to the problem of fuel rattling in 20 a rack. (Reference ~ 28) They realized, however, 'that Fritz's 21 small amplitude assumption did not apply in their case. A fuel 22 element in a high density rack is customarily surrounded by only 23 a small water-filled gap a fraction of an inch thick. The fuel 24 element can undergo motions large enough to close this gap, 25 causing the fuel-to strike the walls of the-container. Indeed, 26 the calculation of the magnitude and consequence of the result-23
1 ing impacts is one of the goals of the analysis. Since the 2 amplitude of the motion of the fuel cannot be considered small 3 compared to the gap, Fritz's formulas are not directly appli-4 cable.
5 Soler and Singh derive more general expressions which 6 purport to describe the motion of the fuel elements including 7 the effects of large amplitude " gap closure" motions. However, 8 no experimental verification of Singh and Soler's work has been 9 performed. The authors recognize that acceptance of their 10 theories by the scientific community is dependent on experimen-11 tal verification of their results. They close their article with 12 the following statement (Reference 45):
13 " Experimental work is currently planned to verify the analysis presented here. Once the analysis has been 14 matched with experiment, for both channelled and unchannelled fuel, the accurate inclusion of fluid 15 damping effects should become an accepted feature of the 3-D dynamic analysis of high density fuel storage 16 racks."
17 Experimental verification of their " gap closure" theories 18 has evidently not yet been obtained. Accordingly, their theories 19 regarding fluid coupling forces with large amplitude motions are 20 still to be' considered tentative and unreliable.
l 21 In a later work, however, Soler and Singh propose to use 22 the' fluid coupling theories to predict the rattling behavior of 23 fuel elements, in spite of the lack of experimental verifica-l 24 tion. They do not attempt to make use of their earlier work, but l
[ 25 continue to ignore the small amplitude restriction with the 26 following logic (Reference 46):
24
1 '"Herein, we'do not include the effects of-gap closure on the fluid inertia terms since there is some 2 preliminary evidence ... that neglect of these effects is conservative."
3 4 While Soler and Singh are correct in not using their large 5 amplitude results which lack experimental verification, they err 6 in continuing to use Fritz's small amplitude results in situa-7 tions where Fritz's assumptions are clearly not satisfied. As 8 the authors themselves point out, there was only preliminary 9 theoretical evidence that the large amplitude effects are 10 conservative. Experimental tests may well show that in realistic 11 situations the neglect of these effects is not conservative.
12 That is, after all, why we require experimental verification of 13 theoretical'models.
14 In this later work, Soler and Singh also apply small 15 amplitude fluid coupling theory to the interactions between 16 spent fuel racks. Unlike the fuel elements, Soler and Singh 17 recognize in this article that the fuel racks should not undergo 18 collisions as a result of seismic excitations.
19 "In addition to stress limitations, adj acent racks must not impact during a seismic event. In the 20 simulation herein, virtual mass effects from gaps between racks have been included based on adjacent 21 rack separation equal to 3" (76.2 mm). Therefore, assuming the worst motion of adj acent racks, inter-22 rack impact is precluded if.the maximum corner deflection of the rack in either direction is less 23 than 50% of the rack spacing."
24 Even though at that time Soler and Singh considered impacts 25 between the racks to be forbidden, they have extrapolated the 26 small amplitude theory one step farther to include the interac-25
( .
1 tions between the moving fuel racks themselves. However, in the 2 Diablo Canyon analysis the fuel racks are expected to undergo 3 large, not small, amplitude displacements. Even R. C. Herrick 4 commented on the extrapolation in the Safety Analysis (Reference 5 47):
6 "Fritz's work, as applied to the rack modules, is limited in that his experimental work, which compared 7 well with his theory, was accomplished for infinitesi-mal vibratory displacements and .for relatively large 8 fluid spaces between the vibrating body and the fluid boundary wall. While this is the opposite to the 9 conditions that prevail for spent fuel rack modules, .
'10
~ ~
11 Herrick might well have pointed out that Fritz's theoreti-12 cal work also assumed small amplitude vibrations. The Safety 13 Evaluation ends its discussion, however, by concluding:
14 "The technique is based upon well established princi-ples in fluid mechanics and serves to provide a lower 15 bounding estimate of the fluid coupling for rack module analysis."
16 17 The claim that . the fluid coupling theory used is " based 18 upon well established principles" seems to excuse the fact that 19 the theory is being applied to a situation it which it is 20 completely untested. Soler and Singh themselves understand that 21 acceptance of their work by the scientific community must await 22 experimental verification. Herrick's additional claim that the 23 extrapolation provides a conservative lower bound is without 24 foundation.
25 In recent months, Singh, et. al have further embellished 26 the fluid coupling theories in an attempt to include the effects 26
l
. i 1
of neighboring rows of racks and even the effects of wedge-2 shaped gaps between tilting racks. (References 16 - 18) As this 3 testimony ' discusses later, even these embellishments fail to 4 accurately account for all the interactions between the racks 5 and the' intervening fluid. It is enough to state here that all 6 fluid coupling theoretical work done since Fritz, however 7 sophisticated, is lacking in experimental support.
8 In conclusion, I reiterate my belief that the use of 9 experimentally unverified fluid coupling theory in critical 10 safety calculations is an unacceptuble practice.
11 MULTI-RACK EFFECTS 12 Each spent fuel pool at Diablo Canyon contains sixteen 13 racks, each of which is capable of many different motions in 14 response to seismic excitations. In order to make the seismic 15 analysis manageable, Soler et al, as consultants to PG&E, made 16 several assumptions among which are (Reference 48):
17 a) in each of several different simulations, only the motion 18 of a single rack is considered; 19 b) the single rack is assumed to undergo motions with eight 20 degrees of freedom, namely, three-dimensional translations, 21 three-dimensional rotations, and two-dimensional translations of
! 22 a fraction of the fuel mass; i
23 c) in order to allow for the interaction of the rack under i 24 consideration with its neighbors, the rack neighbors are assumed
! 25 to undergo translational and rotational motions which are the 26 opposite of the motions of the rack under study.
27
1 The last assumption is referred to as an assumption of "out 2 of phase" motion of the neighboring racks, and it is stated that 3 this assumption conservatively bounds the possible interactions 4 of a rack with its neighbors, although no evidence for that 5 argument was forthcoming until lately. The out of phase motion 6 assumption is tantamount to assuming that there exists an 7
imaginary vertical plane separating the rack from its neighbors, 8 located halfway between and moving with the pool floor. Since 9
the interior racks are located close to each other with a 10 spacing of the order of 1/4 inch, the out of phase motion 11 assumption is the equivalent of assuming an imaginary box 12 surrounding the rack, located 1/8 inch away, out of which the 13 rack is assumed never to move.
14 It is assumed, therefore, that whenever the rack has moved 15 a distance relative to the pool floor of 1/8 inch, a neighboring 16 rack is ready to cause a collision. The rack therefore is 17 assumed to rebound in its box where it moves until it encounters 18 another wall. The model is thus one of , fuel rack vibrating 19 within closely spaced imaginary walls in response to seismic 20 excitation.
21 It is true, in one sense, that the model is conservative, 22 in that if a rack undergoes a collision at high speed the model 23 assumes that the neighboring rack also has high speed, a 24 condition which is not necessarily true. It is clear, however, 25 that each real rack is not confined within an imaginary box and 26 that there is nothing to prevent an entire row of four racks 28
1 from moving together, so that a rack might not undergo a 2 collision until it had moved a distance of considerably more 3 than 1/8 inch. The motions of the racks are, after all, a result 4 of the seismic motions of the spent fuel pools, and the motion 5 of the pool floor is assumed to be the same under each rack. One 6 can reasonably expect the motions of neighboring racks to have a 7 high degree of correlation.
8 There is no a priori reason to think that the impact forces 9 computed by use of the imaginary box model are higher than those 10 which may arise from a more realistic model. Therefore, we 11 cannot regard this model as conservative. Until recently, 12 however, this lack of demonstrated conservatism of the imaginary 13 box model had been completely ignored. The Reracking Report-14 makes the claim of conservatism and the Safety Evaluation seems 15 to ignore the question altogether (Reference 49).
i i
16 The issue of multi-rack effects was raised as a contention 17 by the Sierra Club in the current proceeding. Recently the 18 over-simplicity of the imaginary box model was also pointed out 19 by J. DeGrassi of Brookhaven National Laboratories. (Reference 20 15) Mr. DeGrassi was acting as consultant to the USNRC to review 21 a reracking proposal for the Byron Plant, as mentioned earlier.
22 The Byron proposal is similar to the Diablo Canyon reracking and 23 the analysis of each has been performed by the same people, 24 Singh and Soler, et al. (Reference 50) As a result of Mr.
25 DeGrassi's review, the lack of demonstrable conservatism of the 26 original Diablo seismic analysis was recognized by the NRC 29
1 staff.
-2 . Consequently, hearings before the Atomic Safety and 3 Licensing Board scheduled for March, 1987, were postponed at the-4 request of NRC staff, and additional analytical work was 5 requested of Singh, et. al. (References 16 -
- 18) As this
, 6 testimony discusses later, the lack of demonstrable conservatism
- 7 was ignored by the original reviewers of the seismic analysis.
8 The Safety Analysis was accepted by the NRC, a "No Significant 9 Hazards Determination" and a license amendment was issued by the
^ '
-10 agency despite a critically serious' flaw in the analysis.
11 RECENT WORK ON MULTI-RACK EFFECTS 12 In response to questions raised-about the conservatism of e
13 the single-rack, imaginary box model, Singh, et. al have in the ,
- 14 last - few months performed additional computations which are f
15 . intended to show that multi-rack effects will not produce 16 stresses on the racks larger than those obtained by the earlier 17 model. (Refs. 18, 51) There are several problems with these 18 calculations, however, with the - result that the single rack 19 model is still not demonstrably conservative. '
20 The new work attempts to study the simultaneous behavior of 1.
21 four fuel racks in a row. Because of the added complexity, 22 however, the study restricted the motion of each of the four 23 racks to lie in a single two-dimensional plane. Each rack 24 therefore moves with only four degrees of freedom, instead of 25 eight as with three-dimensional motion. A row of four racks 26 therefore has sixteen degrees of freedom, a system of consider-4-
1
' 30
(
t 9y c+e----n-v s
,,.,.,-r-.y - _ - - - - - - . . . - , ,..,,--------,e,.-_.,---,- - - . . , , . , - -
1 able complexity.
2 The two dimensionality of the model has serious limita-3 tions, however. As Singh and Soler themselves point out (Ref.
4 52):
5 " An accurate picture of the results can only be obtained using 3-D nonlinear time history analysis 6 regardless of the rack modelled. A large contribution to the maximum rack horizontal displacements can be 7 made during an instant when the rack is only supported on one foot and the seismic loads cause a. pivot of the 8 rack about the only remaining contact point."
9 In other words, a two-dimensional model does not capture enough 1'J realism to realistically portray a spent fuel rack. This fact 11 was recognized by NRC staff, PGEE staff and consultants at a 12 meeting in San Francisco on February 18, 1987. No one has yet 13 attempted a' full three-dimensional multi-rack analysis adequate 14 to verify conservatism of the single . rack model used in the
- . 15 Reracking Report.
16 The two-dimensional study attempts to compensate for this 17 lack of realism by first studying a single rack in 2-D. Subse-
}
18 quently, more complex two-dimensional multi-rack results are l 19 computed and are compared with the single rack results. The 20 assumption is that if a more realistic two-dimensional model l.
I 21 fails to produce higher stresses that those for a two-22 dimensional single rack model, then a more realistic three-l l
23 dimensional model would not produce stresses higher than those 24 computed in the licensing basis work, either. (Ref. 53)
L 25 The comparison of two- and three-dimensional results was 26 made more complicated, however, by the decision of Singh, et. al 31 l
l
1 to change the rack parameters used in the 2-D work from those 2 used in the 3-D licensing basis. The purpose of the 2-D studies 3 was to afford as direct a comparison between the multi-rack and 4 single rack models as possible. Using rack parameters for the 5 2-D study which are different from the licensing basis parame-6 ters makes the comparison less straightforward. An improved 7 methodology would have been to perform two sets of 2-D calcula-8 tions, one with the old parameters and one with the new, so that 9 side by side comparisons could be made.
10 The authors explain their approach by claiming that the new 11 parameters are "more realistic" than those used in the licensing 12 basis studies. They claim that the licensing basis used 13 different, less realistic parameters in an attempt to make the 14 licensing basis work more conservative. They attempt to
, 15 demonstrate this conservatism by performing a second 2-D single 16 rack calculation (Case 2) with the modified rack parameters.
I 17 The results of the second calculation show the behavior of 18 a modified rack, case 2, to be quite different from that used in 19 the licensing basis, case 1. (Ref. 54) Although they found the 20 maximum impact force to be some 13% smaller using the "more 21 realistic" parameters, they also found that the "more realistic" 22 rack is expected to undergo much larger displacements than the 23 licensing basis rack (Case 1). The former is expected to impact 24 with the walls of the pool, whereas the latter is not. The fact 25 that the 2-D single rack studies predict significantly different 26 behavior for the "more realistic" rack from that used in the 32 i
1 licensing basis raises serious questions about the licensing 2 basis work itself. The "less realistic" parameters used in the 3 licensing basis work in an attempt to introduce conservatism may 4 well have altered the behavior of the racks under study so that 5 phenomena important to the safety of the racks are missed.
6 The authors point out that the impacts predicted with the 7 pool walls result in smaller forces in Case 2 than the rack-rack 8 forces in Case 1, for which no rack-wall impacts were found.
9 They therefore conclude that the "less realistic" parameters 10 'were indeed co'nservative. In my opinion, this evidence is less 11 than convincing; the fact that the expected behavior of the 12 racks using "more realistic" parameters is quite different from 13 that found in the licensing basis is disturbing.
14 The 2-D study then continues to examine several different 15 cases using the multi-rack model and the "more realistic" 16 parameters. In two of these cases they find that the multi-rack 17 impact forces are greater than the Case 2 single rack study, 18 with which the multi-rack results are to be compared. There is, l 19 therefore, the clear implication that the multi-rack effects are
- 20 expected to increase the maximum impacts between the racks.
l 21 (Ref. 55) 22 The authors dismiss this implication in one of the cases 23 (Case 7) by pointing out that the maximum impact is still i 24 smaller than that of Case 1, which used the "less realistic" 25 parameters. Thus, while admitting that the multi-rack effects 26 may be important, the authors believe tnat the conservatism of 33 i
I the "less realistic" parameters is enough to provide an adequate 2 margin of safety. This is a dubious approach at best, since the 3 full implications of the "less realistic" parameters are in 4 doubt.
5 More seriously, in Case 5 the multi-rack impact forces are 6 found to be 374 higher than the comparable Case 2 single rack 7 value! This result is even 19% higher than the supposedly 8 conservative Case 1 results. This is strong evidence that the 9 multi-rack effects can be expected, at least in some cases, to 10 result in forces larger than those found in the licensing basis 11 study. Although the forces calculated in Case 5 are less than 12 the claimed allowables, one must recall that the forces calcu-13 lated with the 2-D models are unrealistic and are to be used for 14 comparison purposes only.
15 The 2-D studies were undertaken specifically in an attempt 16 to ascertain whether or not the multi-rack effects ignored by 17 the licensing basis studies could lead to forces larger than 18 those reported in that work. The answer, according tc Singh, et.
19 al, is clearly yes.
20 This result is most important, since it demonstrates 21 clearly the limitations of the model used in the licensing basis 22 study. The single rack model is inadequate as a basis for the 23 seismic analysis of an interacting group of spent fuel racks.
24 Unfortunately, and fatal to the analysis, Singh et al. dismiss 25 the implications merely with a vague reference to what they 26 perceive to be the conservative nature of the model. (Ref. 56) 34
1 MISCELLANEOUS OUTSTANDING PROBLEMS IN THE SEISMIC ANALYSIS 2 The above discussion of fluid coupling effects was related 3
primarily to the important role they play in the seismic-4 analysis and the need for experimental verification of the 5 theories. In the seismic analysis by Singh, et. al, it is 6
assumed that the fluid effects always act to impede the motion 7 of the racks, resulting in an increased " hydrodynamic mass" 8 which, in general, tends to reduce the seismic response of the 9 rack system. When considering the possible motions of surround-
- - 10 ing racks, however, it is clear that is assumption is not always 11 valid.
12 The seismic analysis has neglected the effect of water 13 expelled from between two racks on their neighbors. Since the 14 racks are closely packed and the motion of the water is expected 15 to be largely two-dimensional, (Ref. 57) two colliding racks 16 will force water into the space separating a neighboring pair of 17 racks. Depending on their relative motion, the effect of this 18 water movement may increase or decrease the resulting movement 19 of the racks involved. Thus the forces on colliding racks may be 20 increased by such many-body fluid coupling effects in some 21 collisions. The seismic analysis performed to date ignores these 22 effects. This neglect is expected to be non-conservative at 23 least some of the time.
24 Some of the debate over which rack parameters are appro-25 priate, i.e. what is more or less realistic, could have been 26 avoided simply by making the necessary measurements on the racks 35
I themselves. As in the fluid coupling theory discussion, PGGE and 2 its consultants have failed to make these measurements, although 3 they admit the importance of the parameters in their work.
4 Consequently, it is not known how well the values used in the 5 studies to model the racks accurately represent the real fuel 6 rack modules. The absence of experimentally verified data 7 reinforces doubts as to the reliability of the structural 8 analysis, 9 In the early stages of the analytical work, Singh, et. al
^ ' 10 decided to ignore the flexibility of the fuel racks and to treat 11 them as rigid objects. (Ref. 58) The 3-D single rack models 12 were thereby reduced from 32 to 8 degrees of freedom, simplify-13 ing the problem considerably. Several justifications for this 14 simplification have been offered. (Ref. 59) 15 However, real racks are not perfectly rigid as assumed in 16 the analysis. Real racks would be expected to flex and bend 17 slightly in response to the fluid coupling forces. These 18 effects have been consistently ignored, even in the earliest 19 work, but they significantly modify the expected fluid coupling 20 effects. The fluid coupling forces are highly sensitive to the 21 size of the gaps between the racks. The flexing of a rack wall 22 by even 1/8 inch could significantly alter the behavior of the 23 rack.
24 Inasmuch as rack flexure acts to decrease the fluid 25 coupling forces on a rack, this effect is expected to be 26 non-conservative. It has been assumed throughout the analysis 36
1 that increasing the fluid coupling forces decreases the expected 2 impact forces on the racks. The neglect of rack flexure could, 3 therefore, increase these impacts to unacceptable levels. The 4 evaluation of these effects is one of the primary examples of 5 the necessity of experimental verification of the safety of the 6 spent fuel system.
7 CONSEQUENCES OF RACK DAMAGE 8 As a result of my review of the seismic analysis, I believe 9 that an earthquake of the magnitude of the postulated Hosgri 10 event could result in significant damage to the spent fuel 11 racks. Such damage has the potential to increase the criticality 12 constant above accepted levels, at least locally, with the 13 concomitant generation of excessive amounts of heat and radia-14 tion. Such damage could therefore result in the release of 15 unacceptable amounts of radioactive material into the environ-16 ment and pose a threat to the public health and safety.
17 The proposed fuel storage system fails to satisfy regula-18 tory, technical and scientific standards for reliable and safe 19 operation during seismic events postulated for the Diablo Canyon 20 site. Storage of spent nuclear fuel at Diablo Canyon in these 21 propor,ed unsecured high density fuel racks poses a significant 22 danger to the public health and safety and to the environment.
23 COMPLIANCE WITH NEPA 24 Prior to making a licensing decision concerning a proposed 25 action which might significantly affect the environment, The 26 National Environmental Policy Act of 1969 ("NEPA") requires the 37
j preparation of an environmental impact statement ("EIS")
2 regarding the possible environmental effects of the proposed 3
action. 42 U.S.C. S 4332(2)(c). It is undisputed that no such 4 EIS was prepared for the Diablo Canyon reracking; nor is any planned. Instead, the NRC has merely prepared an Environmental 5
6 Assessment ("EA") which summarily concludes that the Diablo 7 Canyon reracking will have no significant impact on the environ-8 ment. (Environmental Assessment and Finding of No 'Significant g Impact - Spent Fuel Pool Expansion, Diablo Canyon Nuclear Power 10 Plant, Units 1 and 2, dated May 21, 1986.) (Ref.- 60) -In this-93 case, the EA relies on a hopelessly outdated, seven-year old 12 Generic Environmental Impact Statement ("GEIS"). (Ref. 61) Thus, 13 the summary conclusions of the EA are equally out of date and g defective.
15 The GEIS upon which the EA relied was completed in 1979.
16 Although the passage of time itself would not necessarily g invalidate the conclusions of the GEIS, they do so in this case 18 because (a) there have been changes in the law, and in possible g alternatives which were not considered in either the GEIS or the 20 EA; and (b) the GEIS is based on assumptions which are neither g applicable to Diablo Canyon nor distinguished in the EA.
( g Three years after the GEIS was completed, for example, 1
23 Congress passed the Nuclear Waste Policy Act of 1982, 42 U.S.C.
5 10101 et seg. This law established, for the first time, that 24 l 25 the federal government has the responsibility to provide for interim storage of spent nuclear fuel for nuclear powerplants l 38 a
I that "cannot reasonably provide adequate storage capacity at the 2 sites of such reactors when needed to assure the continued, 3 orderly operation of such reactors." 42 U.S.C. S 10151( a )(3 ) .
4 Thus, federal responsibility for storage of spent nuclear fuel 5 at Diablo Canyon is a critical alternative which should be 6 addressed under NEPA. Yet, the GEIS' discussion of alternative 7 federal responsibility for Diablo Canyon spent fuel is almost 8 nonexistent, and this deficiency was not addressed in the EA.
9 Similarly, the EA assessment contains only cursory consi-10 deration of alternatives which should be thoroughly analyzed in 11 a detailed EIS. The EA asserts that trans-shipment of the fuel 12 from Diablo Canyon to another reactor "would entail potential 13 environmental impacts greater than those associated with the 14 proposed increased storage at the Diablo Canyon Site" (Environ-15 mental Assessment, pp. 3-4), yet it does not identify which, if 16 any, trans-shipment locations were considered and provides no 17 basis whatsoever for its conclusion that this alternative is not 18 feasible.
19 The GEIS does not discuss the seismic hazards of Diablo 20 Canyon --
hazards unique in the field of atomic power geneha- ,
21 tion. The EA does not remedy this deficiency, as it similarly l 22 contains no site-specific seismic analysis and simply adopts the 23 GEIS conclusion that the " environmental impact costs are 24 essentially negligible, regardless of where such spent fuel is 25 stored." Environmental Assessment, p. 2.
26 There are further deficiencies in the GEIS which render it 39
1 inadequate as the required "hard look" under NEPA. It assumes 2 that the cost of reracking will be between $1.9 and $2.2 million 3 and will involve between 868 and 833 assembly spaces. PG&E has 4 estimated the cost of the Diablo Canyon reracking at $16 5 million; it involves 1324 spent fuel assembly spaces at each 6 reactor. The GEIS looks only to spent fuel storage until the 7 year 2000; the spent fuel will be stored.at Diablo Canyon until 8 at least the year 2008, with no guarantee that alternative 9 storage will be available then. Similarly, the GEIS assumes 10 that both the original and reconfigured racks will be bolted to 11 the floor of the spent fuel storage pool. The reconfigured 12 racks at Diablo Canyon are free standing. The EA ignores each 13 of these factors which render the GEIS, and, hence, the conclu-14 sions of the EA, obsolete.
15 CONCLUSION 16 The currently proposed reracking of the spent fuel pools at 17 the Diablo Canyon Nuclear Power Plant has not been demonstrated
. 18 to be safe according to accepted scientific standards. The 19 theories underlying the seismic analysis are untested and <
20 unverified and the analysis makes assumptions which are either 21 demonstrably non-conservative or which have not yet been l 22 demonstrated to be conservative. In addition, the proposal 23 appears to violate current NRC acceptance criteria for seismic 24 motions, according to the OT Position.
{
~25 Because of these technical and regulatory considerations, t 26 ordinary prudence as well as common sense dictate that, if spent 40 '
l f
I.
I
c s,
1 nuc ear fuel must'be stored at the Diablo Canyon site, alternate 2 means of; storage should be considered. Unfortunately, no such
- 3 study of alternate on-site storage methods has been performed.
4 It is my opinion that the population of San Luis Obispo 5 County desires on-site storage to be as safe as is reasonably 6 possible.. Because the current. proposal fails to meet regulatory 7 and technical standards for seismic safety, it cannot be 8 considered to be~ consistent with the protection of the public i
9 health and safety and the environment, nor with the views of the 10 population surrounding the reactor.,
The proposed reracking of 11 Diablo Canyon should be rejected.
12 13
[f 14 15 16 17 18 19 20 21 22 23 24 25 26 41
o 1
REFERENCES
- 1) 51 Federal Register.1451 (1986)
- 2) Pacific Gas and Electric Company, "Reracking of Spent Fuel Pools, Diablo Canyon Units 1 .and 2", September, 1985,
'("Reracking Report"), p. 9-2 3). 10 CFR S 53.13
- 4) Reracking Report, p. 6 - 27
- 5) " Seismic Analysis of High Density Fuel Racks", for Pacific Gas & Electric for Diablo Canyon Nuclear Power Station, TM Report #779, A Soler, 1/29/86 p. II - 62
- 6) Summary, Diablo Canyon Project Telephone Calls, 3/8/85 -
-6/20/86, p. 3 (" Calls")
7). Ibid. pp. 2ff
- 8) Ibid. p. 7
- 9) Safety Evaluation by the-Office of Nuclear Reactor Regulation Relating to the Reracking of the Spent Fuel Pools at Diablo Canyon Nuclear Power Plant, Units 1 and 2 . . . , USNRC, May 30, 1986. (" Safety Evaluation")
- 10) Affidavit of R. Clyde Herrick Regarding the Intervenor's Application for.a Stay, June 30, 1986
- 11) Calls p. 14
- 12) Safety Evaluation, Appendix A, p. 33
- 13) Ibid. pp. 49, 72
- 14) Ibid. p. 23
- 15) Meeting Summary - Structural Analysis of Spent Fuel Pool Expansion, Byron Station Units 1 and 2, USNRC, 1/12/87
- 16) Response to NRC Staff Questions Raised at the May 6, 1987-Reracking Meeting DCL-87-115, May 18, 1987
- 17) "Three Dimensional Studies of High Density Spent Fuel Racks (Acorn 10.and Acorn 12)" PG&E DCL-87-082, 4/23/87
- 18) " Additional Information on Rack-Rack Interactions" PG&E 4/7/87 [ PROPRIETARY] (" Additional Information")
42
- 19) Contentions of the Sierra Club,-April 24, 1986
- 20) Sierra Club, " Motion For Summary Disposition", before the USNRC Atomic Safety and Licensing Board, 12/15/86
- 21) USNRC, " Standard Review - Plan for the Review of Safety Analysis Reports for Nuclear ' Power Plants, NUREG-0800",
Revised July, 1981 ("SRP")
- 22) USNRC, "OT Position for Review and Acceptance of Spent Fuel Storage ant Handling Applications", January 18, 1979 ("OT Position")
.23) Ibid. p. IV-5
- 24) Ibid. pp. IV-3 and 4
- 25) Reracking Report,.pp.6-7ff
- 26) Safety Evaluation, Appendix A, pp. 23ff
- 27) A.I. Soler and . K.P. Singh, " Seismic Response of a Free Standing Fuel Rack Construction to 3-D Floor Motion", Nuclear Engineering and Design 80 (1984) 315-329 (" Seismic Response")
- 28) A.I. Soler and K.P. Singh, " Dynamic Coupling in a Closely Spaced Two-Body System Vibrating in a' Liquid Medium: the Case of Fuel Racks", Proc. of the Third Conf. on Vibration in Nuclear Plant, British Nuclear Energy Soc., (1983), pp. 815 -
834 (" Dynamic Coupling")
- 29) R.J. Fritz, "The Effect of Liquids on the Dynamic Motions of Immersed Solids", Jour, of Eng. for Ind., February, 1972, pp.
167 -'173 ("Fritz")
- 30) K.P. Singh and T. Lee-Ng, " Planning Study Experimental Measurements and Analytical Correlations of Fluid Drag of Fuel Assemblies in Fuel Rack Storage Locations", EPRI Project No. RP2124, Joseph Oat Corp., November, 1982. (" Planning Study")
- 31) OT Position, p. IV-5
- 32) Seismic Response, p. 327
- 33) OT Position, p. IV-1
- 34) A.I. Soler, " Seismic Analysis of High Density Fuel Racks for Pacific Gas & Electric for Diablo Canyon Nuclear Power Station", TM Report #779, Revised September, 1986
- 35) Seismic Response, p.323 43
- 36) Summary-Notes, Meeting of December 5, 1985, PG&E and NRC et al., Enclosure 6
- 37) Planning Study, pp. A.2-1 ff, p. A.4-14
- 38) Safety Analysis, p. 10
- 39) Ibid., Appendix A, p. 3
- 40) Reracking Report, p. 6-15
- 41) Ibid., p. 6-3
- 42) Ibid., p. 6-15
- 43) Ibid., p. 6-23ff
- 44) Ibid., p. 6-27
- 45) ' Dynamic Coupling, p. 826
- 46) Seismic Response, p. 318
- 47) Safety Evaluation, Appendix A, p. 23
- 48) Reracking Report, Ch.6
- 49) Ibid., p. 6-13
-50) Telephone Conference with J. DeGrassi, 5/12/87
- 51) K.P. Singh and A.I. Soler, " Design Basis, Planar Motion and 3-D Analysis with Cross-Coupling Coefficients for Diablo
-Canyon High Density Rack Structural Evaluation, Holtec Report HI-87117, [ PROPRIETARY]
- 52) Seismic Response, p. 328
- 53) Additional Information, pp. 5, 24 t
- 54) Ibid., p. 15
- 55) Ibid., p. 29
- 56) Ibid., p. 17
- 57) Ibid., p. 12
- 58) Reracking Report, p. 6-14
- 59) Safety Analysis, Appendix A, p. 22 44
60)' Environmental Assessment and Finding of No Significant Impact
-- Spent Fuel Pool Expansion, Diablo Canyon Nuclear Power Plant, Units 1 and 2, dated May 21, 1986.
I
- 61) USNRC, " Generic Environmental Impact Statement On Handling and Storage of Spent Light Water Power Reactor Fuel", August 1979 i
l 45 l
1
_ _ _ - . - -, . . . . , , - , . _ . . . . _ _ , . . _ _ . , . - _ , _ . , , - - _ _ . _ _ , . _ ~ . ~ . . . - - . - . . - _ _ .
_ ~ _ _ . . , . . . _ , . . _ _
1 STATEMENT OF QUALIFICATIONS: DR. RICHARD B. FERGUSCN 2 I am a professional physicist. I received a Bachelor of 3 Arts degree, cum laude, from Amherst College, Massachusetts in 4 1960 and a Ph.D. in physics from Washington University, St.
5 Louis, Missouri in 1967. I taught on the physics faculty of the 6 University of California at Los Angeles from 1967 until 1970 and 7 on the physics faculty of the California Polytechnic State 8 University at San Luis Obispo from 1974 to 1984. I recently 9 resumed my faculty position at the California Polytechnic State
~ -
10 University.
11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 i
2 46
DECLARATION OF RICHARD B. FERGUSON I, Richard B. Ferguson, declare:
I was principally responsible for preparation of the attached testimony and know its contents.
The facts stated therein are true, except for those which are stated on information end belief, which I believe to be true.
I am competent to testify, and if called to testify would do so consistent with the attached testimony.
I declare under penelty of perjury that the foregoing is true and correct.
Executed at Creston, California, on June [ , 1987. ,
6/ n-
' Mchard Bdergusp .
~
47
., t-PROOF OF SERVICE I, Deborah M. Hunt, declare that on June 4, 1987, I deposited copies of the attached Testimony of - Richard B. Ferguson in the United States mail with postage thereon fully prepaid and addressed to the parties listed below:
Dr. Jerry Harbour Mr. Leland M. Gustafson, Administrative Judge Federal Relations Manager Atomic Safety & Licensing Board Pacific Gas & Electric Co.
U.S. Nuclear Regulatory Comm. 1726 "M" Street, NW, Suite 1100 Washington, D.C. 20555 Washington, D.C. 20036-4502 Glenn O. Bright Philip A. Crane, Jr. sq.Eb Administrative Judge Pacific Gas & ElectrictCo Atomic Safety.& Licensing Board P.O. Box 7442 E4' U.S. Nuclear Regulatory Comm. San Francisco, CA .94120 i h! di Washington, D.C. eg 20555 5- 03 1:'
Mr. Gordon A. Silver? o Henry J. McGurren, Esq. Ms. Sandra A. Silverhj LJ Lawrence J. Chandler, Esq. 1760 Alisal Street ' 65 Office of the Exec. Legal Dir. San Luis Obispo, CA 9340l" U.S. Nuclear Regulatory Comm.
Washington, D.C. 20555 Ms. Laurie McDermott, Coordinator Atomic Safety & Licensing C.O.D.E.S.
Board Panel 731 Pacific Street, Suite 42-U.S. Nuclear Regulatory Comm. San Luis Obispo, CA 93401 Washington, D.C. 20555 Mrs. Jacquelyn Wheeler B. Paul Cotter, Jr., Chairman 2455 Leona Street Administrative Judge San Luis Obispo, CA 93400 Atomic Safety-& Licensing Board U.S. Nuclear Regulatory Comm. Dr. R.B. Ferguson Washington, D.C. 20555 Sierra Club / Santa Lucia Chapter Rocky Canyon Star Route Atomic Safety & Licensing Creston, CA 93432 Appeal Panel U.S. Nuclear Regulatory Comm. Ms. Nancy Culver Washington, D.C. 20555 192 Luneta Street San Luis Obispo, CA 93401 Bruce Norton, Esq.
Pacific Gas & Electric Company P.O. Box 7442 San Francisco, CA 94120 1 am, and was at the time of the service of the attached paper, over the age of 18 and not a party to the proceeding.
I declare under penalty of perjury that the foregoing is true and correct.
/
TbYI Deborah M. Hunt