ML20102C235
| ML20102C235 | |
| Person / Time | |
|---|---|
| Site: | Perry  |
| Issue date: | 02/27/1985 |
| From: | Hiatt S OHIO CITIZENS FOR RESPONSIBLE ENERGY |
| To: | Atomic Safety and Licensing Board Panel |
| Shared Package | |
| ML20102C230 | List: |
| References | |
| OL, NUDOCS 8503050402 | |
| Download: ML20102C235 (4) | |
Text
.
February 27, 1985 DCdfrp:
UNITED STATES OF ANERICA US,y c NUCLEAR REGULATORY COMMISSION ,
Before the Atomic Safety and Licensing Board bd? -f P2:57 In the Motter of ) CEFI
} CCC$"ri f g E(fiApv THE CLEVELAND ELECTRIC ) Docket Nos. 5$A%6ibE ILLUMINATING CO. ET AL. ) 50-441 OL
)
(Perry Nuclear Pouer Plant, ) '
'~
Units 1 ond 2) )
OCRE RESPON5E TO APPLICANT 5' MOTION FOR SUMMAPY DISPOSITION OF ISSUE 16 2
I. INTRODUCTION On February 5, 1985, Appliconds moved for summary disposition'in its favor on Issue M16 which states:
Applicont hos not demonstroted that it con reliably generote emergency on-site power by relying on four Transomerico Delavol diesel generatcrs ['TDI DGE'], two for each of sts Perry units.
The basis for this contentzon is the extensive adverse operatins experience of TDI diesel engines in nuclear, stationory, and morine service and the poor or totolly locking TDI quality assurance program, os reveoled by the findings of
_5 toff and Applicants. See, e.g., Boced Notifications BN-83-160, BN-83-160A, BN-84-018, BN-84-020, BN-84-021. BN-84-024, and BN-
- rs $ 84-051. Virtuolly every moJor TDI engine component
.NOL I '(cronkshorts, pistons, cylinder heads, connectins rods, push gg rods, b l'o c k , base, bearings, fuel lines, turbocharger, jocket water pump) has o history of failure, Despite this adverse experience, Applicants have continually
! hh
! #EO clozmed that the TDI DGs are reliable and fit for nuclear i i
a f
service, and now seek to prevail on the issue, osserting that no i 2
issues of material fact exist for hearing. Intervenor Ohio Citizens for Responsible Energy ("0CRE') OOPoses Applicants' motion. The discussion below demonstrates that there are genuine issues of material foch to be heard and that the TDI DGs are still inherently unrelioble, despite all the poperwork generated by Applicants and their consultants in o desperate attempt to prove otherwise. Applicants' motion must be denied.
II.' STANDARD 5 FOR
SUMMARY
DISPOSITION The burden of proof lies upon the movant for summary disposition, who must demonstrate that no genuine issues of material fact exist. .In fact, the record and pleodings must be viewed in the light most favoroble to the opponents of summary disposition. Public Service __Co. of New Hampshire (Seobrook Station, Units 1 ond 2), LBP-74-36, 7 AEC 877 (1074).
In on operating license proceeding, where significont health and safety or environmental issues are involved, o Licensing Soord should grant a motion for Summary disposition only if it is convinced thot the public heoith and safety or the environment will be satisfoetorily protected. Cincinnott Gas and Electric (Wm, H. Zimmer Nuclear Storion), LSP-81-2, 13 NRC i
36, 40-41 (1981).
It is improper to grant summary disposition of a safety issue before the issuonce of the Starr*s SER on that issue.
Duke power Co. (Wm. B. McGuire Nuclear Storion, Units 1 ond 2),
LSP-77-20,'5 NRC 680 (1977).
i
Finolly, 10 CFR 2.749(d) states thot (t)he presiding officer shall render the decision sought if the filings in the Proceeding, depositions, answers to interro9atories, and admissions on file, together with the stotements of the porties and offidoVits, if a n y ,, show thot there is no genuine issue os to any material fact and that the moving party is entitled to o decision.os o mother of low, For the reasons stated below, Applicones' motion rails.
They hcVe not met their burden of aemonstrating the obsence of o genuine issue of material fact, and they have not demonstrated tho the public health and sofety will be satisfactorily protected.
III. DISCU55 ION
- 4. Applicable Regulatory stondores It is oxicnotic that nuclear licensees meet all cf the Commission's regulations. See, e.g., Mosne Yankee Atomic Pcwer fgt._ (Moine yonkee Atomic Power storion), ALAB-161, 6 AEC 1003 (1973), one Vermont yonkee Nuclear Power Corp. (Vermont yonkee Nuclear Power storion), ALAB-138, 6 AEC 520 (1073) (it cannot be argued thot a facility not meeting the Commission's regulations is safe anyway, as such orgument is on impermissible challen9e to the regulations). It is thus appropriote to oddress the regulations opplicable to this issue.
The opplicable regulations, which Applicants have violated, ore Appendix B to 10 CFR Port 50 and General Design Criterio 1 and 17, Appendix A to 10 CFR Port 50. GDC 17 requires the Perry racility to have on onsite electric power system twhich the DGs ode to supply] with the copocity to ensu"e that fuel design
limits and reoCtor Coolont pressure boundary design conditions are not exceeded and that the core is cooled and containment integrity and other virol functions are maintoined in the event or on occident. Appendix B and GDC 1 recuire that the DGs be designed and monuroctured to oppropriate quality stoneords one in ocecedonce with on appropr:ote cuolity assuronce progror, Appliconts' noncom 0lionce with these regulations is detailed below.
- 1. GDC 17 The inherent unreliability or the TDI DGs is such thot the Commission hos determined that facilities using these DGs are not in ccmpliance with GDC 17. Long__ Island Lighting Co.
(shoreham Nuclear Power Storion), CLI-84-5, 19 NRC 1154 (198a)
(GDC 17 must me complied with even for low-power operation, and snorenom, using TDI DGs, did not comply). 5ee also BN-84-02'
($ECY-34-34), which stat.es that "the operating history of TDI engines and the QA program or the manufacturer , , , [ coll] into
- uestson the reliodility or oil TDI diesels,'
5ee also Exhibit 1, in which these principles are applied to the Grand Gulf facility, ("On the basis of the problems ossocioted with TDI dsesel engines , , , the onsite electrical supply systems at Grand Guir do not meet GDC 17'). Grand Gulf utilizes TDI's D5RV-16 engines, one cr which served as the prototype for the Perry DGs. It thus nust be concluded thot the Perry onsite electrical power system is likewise in
noncompliance with GDC 17.
Of further relevonce is the commitment Applicants have made with regard to their onsite power system and specifically the DGs. FSAR Section 8,3.1.1.3.2 states that the DGs have a continuous rating of 7000 kw, with o short time roting of 7700 kW. The procurement specification, SP-562-4549-00, supplies additional, more specific requirements. For example:
the minimum continuous rating of the engine shall be 7000 kw net outputs the engine with its generator and exciter shall be engineered and designed os o complete Unit and shall be free of all deleterious critical speeds or torsional vibration for any operoting speed within the range of 90% to 110% of rated speed at any load from O. tp 100% of rated outputs the engsne shall be able to operate at 110% or its continuous rating (7700kW) for o period of 2 hours2.314815e-5 days <br />5.555556e-4 hours <br />3.306878e-6 weeks <br />7.61e-7 months <br /> out of every 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> without offecting the normal life of the unit. SP-562, Section 2.06.1 011 equipment and services offered by CTDI] shall be of such quality as to make the equipment safe with high ovoilobility.
To this end, all items offered. including all occessories, shall be of proven reliobility. SP-562, section 1.04.
equipment supplied under this Specification shall be in accordance with opPlicable codes and standords. EP-562, Settion 0.05. The codes and standards cited include NRC Regulatory Guides, ANSI standards, ASME code, and the standards of DEMA, IEEE, AW5, NFPA, and others.
It is thus clear that these commitments are the OPpropriote criterio by which to judge whether compliance with the regulations hos been ochieved, since they comprise Applicants' own standards for determin'ing DG reliability.
- 2. GDC 1 and Appendix B Applicants imposed quality assurance program requirements on TDI through Ottochment specification SP-706-4549-00, ottached to SP-562-4549-00, the procurement specirication for the DGs.
However, Applicants own documents, such as DARs, nonconformonce reports, and surveillance reports demonstrate that this standard was never me t, Applicants admit that 28 DARs have been written to document significont deficiencies discovered with respect to the TDI DGs, Christionsen offidovit at 19. Mony more nonconformance reports hove been written to document conditions of noncompliance, OCRE hos ottoched os Exhibit O o summary of some of the more significant nonconformance reports ('NRs') concernin9 the TDI DG5 Most of the deficiencies have involved poor welding proctices, poor olignment of equipment, lock of documentatien or identification, and domoge in shipping ond hondling. Exhibit 3 is Appliconts' list of NRs generated as a result of the engine inspection 05 part of the DR/QR effort,
!!ony of these latest findings are similor to the deficiencies discovered earlier. It also oppears' that some of them should hove been discovered and corrected earlier. That these engines, which would have been used'os is were it not for the DR/QR in5pection, contained so many deficiencies at this late dote is evidence of the ineffectiveness of TDI's and Applicants' QA progroms. It must be noted that Applicants also foiled to ensure that TDI imposed QA requirements on the manufacturer of exhouse silencers for the PNPP DGs, contrary to specification requirements. Applicants were not even owore of this until the Staff discovered it in its vendor inspections, reported in EN-
see Applicants' response (March 8, 1984) to OCRE Interrogatory 11-2.
These breakdowns notwithstanding, Applicants claim that the large number of discovered defici.encies proves that their QA program'is working. Applicants' motion at 14. However, there is o Point beyond which this argument is involid. Even if all deficiencies are corrected, there remains the question of whether there has been o pervasive QA failure of sufficient dimensions so os to raise legitimate doubt as to the overall integrity of the DGs. See Union Electric Co. (Colloway Plan t) ,
ALAB-740, Slip op. at 2.
Applicants' own documents indicate they they themselves reolized that that point Was reached some time 090 (Their concern was unfortunately too late to do any good.) Exhibit 4 is DAR 139, which expresses some concern about having 10 DARs on TDI in the lost 3 yects. Ekhibit 5 is on audit of TDI dated April 12, 1982, which outlines TDI's history of noncompitance and concludes:
the audit team feels that the quality assurance progrom in effect at the time work was performed for CEI and the one presently in place at Delovol does not meet the requirements contained in 5P-562 and SP-706. Even if the program described in the O.A. Manual was effectively implemented it would not meet the requirements of SP-562. The ottitide towards quality assurance is one of tolerance, not support, It is evident from review of the contract history presented above thoe this has been the cose since the contract's inception.
This conclusion is remarkably similar to that of the NRC
_g_
5 toff in its Inspection Report Nos. 99900334/83-02 and 83-03 (BN-34-021)-
It is apparent from the results of these and Prior inspectsons thor serious deficiencies have extsted in the implementation of your committed quality assurance progrom for manufacture or emergency diesel generators. Whoe concerns us greatly is thor certain of these findings are of a noture which brings into question both the odequacy of existing manufoeturing process controls and the level of complianceby manufacturing and Quality control personnel. When reviewed in the context of the numerous deficiencies which have been identified to the NRC in 10 CFR Port 21 and 10 CFR Port 50.55(e) reports, we believe that significant concern is worronted with respect to CTDI DC
/ reliability].
It is thus clear that both Appliconts and the NRC Stoff
- ensider TDI to be in noncompliance with GDC 1 ond Appendix B.
This situation hos si9nificonce beyond that of re9ulotory noncompliance. As discussed below, the poor TDI QA renders invoisd the ossumptiOns of the DR/QR erfort which is supposed to compensote for this problem, B.'The TDI owners Group P(An As explained by Applicants in their motion. they and other utilities owning TDI DGs have rormed on owners group with the purported 9o01 of oddressing the re9Ulotory concerns about IDI DG reliobility. This Owners Group hos formuloted a program plan by which it hopes to resolve these concerns. The program consists of 4 elements; Phose I, the resolution of 16 'known problems *; Phase II, Design Review and Quality Revolidotton
("DR/QR') of all components deemed important to DG relsobilitys engsne inspection ond testing; and maintenance and surveillonce Programs.
_9 In this section OCRE oddresses the odequacy or this plan, the Storr's evoluotion or this plan, and the true motives or the Owners Group, Specific rindings With respect to Phose I ond II errorts are discussed in subsequent Sections.
- 1. The Stoff's 5ER On August 13, 1984 the NRC Storr issued its SER on the owners Group Program Pion. See BN-84-150. (The Starr oiso zntends to issue on SER on each of the Phase I components, but hos Ot yet done so.) It is the Storr's opinion that the Owners Group Program Plan incorporotes the essential elements needed to ensure compliance with GDC 1 and GDC 17. These essential elements include Phase I resolution, Phase II DR/QR, oppropriate engine inspections and testing, and maintenance and Survei11once programs. SER at 6.
The Starr otso outlined on interim basis for licensing those plants which have not completed all the elements or the Program Pion. However, this is not opplicable to Perry, os this is a contested proceeding in which the Starr connot argue that something 9
less than rull compliance with the regulations is also In any event. Applicants have not sore. Vermont Yonkee, supro, requested any exemptions from the regulations for interim licensing, and have committed to implementing the entire plan, including rull pre-operationoi test 2ng, before plant licensing and operation. Motion at 12. See also Exhibit 6, from Appliconts' January 17, 1935 submittoi or their TDI Progrom
x
-/0 -
Plan.
Exhibit 7 is Applicants' schedule for implementing the Program Plon, osain ' rom the January 17 submittol. Note that results from engine inspections ofter the pre-operational testing will not be avoilable until June. Thus, it as not G
possible to determine whether the Perry DGs meet the standards the Stoff hos set forth for regulatory compliance until that time.
Engine testing and inspection is the key to verifying engine reliability. See Battelle Pacific Northwest Laboratory evoluotion of the Pro 9Pom Plon, part of BN-84-152, at 11.
However, PNL considers the tests outlined by the Owners Group to be insufficient. PNL recommends that o *1eod engine' ee operated at qualified loco for 10 million (1E7) cycles. This is Engine equivalent to 750 hours0.00868 days <br />0.208 hours <br />0.00124 weeks <br />2.85375e-4 months <br /> for on engine speed of 450 rpm. * ,
a disonsembly and inspection is to follow. If any key compnent should foil the test,,the root cause should be identified, corrective action taken, and the component should be retested for another 1E7 cycles.
The testing should oiso include 10 modified starts to at least 40% of qualified lood, 2 rost starts to qualified lood, and one 24 hour2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> run at qualified 100d. These tests are in addition to those required by Re9. Guide 1.106.
Applicants' pre-operational testing is described in the Leidich offidovie. Basically, Applicants have committed to the
.p-testing requirements of Reg. Guide 1.109, with a few ceditions, survey, such as o torsiograph test and engine vibration reliance on the ' lead engine concept' and Appliconds ploce great the lead engine, at Comanche Peak, for 100 on the OPerotion of (not leod engines) for hours, and of the CotoWbo V-16 engines 1600 hours0.0185 days <br />0.444 hours <br />0.00265 weeks <br />6.088e-4 months <br />, they con only First, if Applic'onts are to be consistent, which coes take credit for the one leod engine, Comanche Peak, Secondly, to be not have the required 750 hours of operation, in accordance with PNL's standards, the cited operational hours os If not, they are meaningless, must have been foilure-free, is needed, App 12 cones do noe retesting to another 1E7 cycles In fact, the PNFP cicim hhot these tests have been successful.
component tracking system cites failures of components at fuel and Cotowbo, such as cylinder heads, Comanche Peok injection pumps, turbocharger, and subcovers, of ' lead engines, following engines
- is The whole concept flowed.
It assumes that there is sufficient consistency 2 n.
manufacturing, and ossembly among engines to extend design, the operation of one engine to all favoroble findings from I,e,, it ossumes on effective QA program, engines of that type, since the poor QA at TOI is one or the causes of OG unreliability, it is most inoppropriate to base the of consistent Quality, on on assumption requalification progrom are not identico1, The 2 Furthermore, all TDI V-16 engsnes
Perry Unit i engine 5 have crankshafts supplied by 2 different companies. The only logical opproach is to treat each eng:ne os o lead engines each engine must run for 750 hours0.00868 days <br />0.208 hours <br />0.00124 weeks <br />2.85375e-4 months <br /> without failure. Only then con a finding of regulatory compliance be
- mode,
- 2. The TDI DG Owners Group In evoluoting the owners Group Plon, the true noture of the Owners Group must be discerned. Applicants have portrayed the owners Group ond its consultants os on independent, disinterested entity devoted to on importial evoluotion of TDI DG reliability. The truth is that the Owners Group is more of a (
political body, driven by economic considerations, hoving as its goal NRC occepeonce of the TDI DGs and the ovoidance of licensing delays, The true noeure of the Owners Group is illustroeed by Exhibits 8 through 13. Exhibit 8 is a portion or the minutes or' the Owners Group meeting held November 29, 1983. Note that the Group is to m0ke no decisions that could offect DG manufacturer competition in the future. This would necessarily preclude o finding that TDI DGs are unrelioble.
Exhibit 9 is o memorondum to owners Group members from the ,. ,
ihcreham opplicant. This memorondum illustrates ':ne Owners Group's motn concern, obtaining ropid NRC occeptance of the DGs, Exhibit 10 is o memoromdum to the Owners Group from its technical program director. This document shows the Owners
~{ 3 -
the NRC Stoff to reject the Group's efforts to get c o n s u l t o n't , PNL, and to develop recommendations of its technical
- reolistic' DG loading curves to replace the 'ultro-conservative
- FSAR commitments.
Executive. Committee meeting E'thibit 11 is the Owners Group minutes for January 9, 1985. Described therein is the success Exhibit the Owners Group hos had in influencing the NRC Stoff.
The Stoff is willing, no doube 12 similarly cites this success.
to due to the political pressure exerted by the Owners Group, relax the 185 BMEP interim licensing restriction and to relax from the DR/QR the maintenance and surveillance restrictions reports.
Exhibit 13 is o proposal for closure of the TDI Owners Group.
Note the concerns about raising additional generzc concerns and the ' visibility of the Owners Group twhich] sets the TDI diesel generators oport from other make diesel generotors and other plant equipment as needing special consideration.'
Appliconts' ottitude closely parallels that of the owners Group.
Cpmpare Applicones' response to OCRE's Interrogatory 11-11(d), in wnich Applicones state that there is no number or type of foilure or quality deficzency which they consider unocceptoble for the DGs, and that it is not o purpose of the conclusions with respect to the fieness of owners Group to draw ony porticular TDI DG or TDI DGs in general.
-I4-Because or the political nature or the Owners Group and the success it has had in influencing the NRC Starr, the rindings or neither the Owners Group and its consultants nor the NRC Stoff of Nor should CEI's empicyment should be uncritically accepted.
- on independent engineering consultant, Southwest Research a disinterested Institute' (motion at 10) be considered as verification of the Owners Group findings. Ex ibit 14 indicates that Southwest Research Institute was hired for advocacy purposes.
The nature of the Owners Group's " lobbying' activities demands that on independent evoluotion be made of the Perry DGs.
Both Staff and Applicants must be held to their commitments and For example, the DR/OR report is to the strictest standards..
soic to form the 60515 for concluding that the TDI DGs are in the capable of performing their safety function as ' described DR/QR Report or 1-1.
Appendix II or the DR/QR ,
Perry FSAR.
Report 'contoins o comprehensive set of maintenance and for each component.' That program surveillance recommendations is supposed to maintain the qualification of the DGs for the Applicants have DR/QR. Report at 2-6, life of the Plant.
committed to surveillance incorporating all the DR/QR maintenance and recommendations. January 17, 1985 submittol ce 22. Because of influence with the NRC, the danger the Owners Group's political is that these Commitments are hollow promises to be quickly
-/5r-rescinded ofter licensing.
It must be ensured that App:.icants will be held to their commitments, or it must be assumed that F54R requirements will not be met.
- 3. Relionce Upon TDI It is important that any evoluotion of TOI DGs be . ..
independent of TDI's own commercial interests and influence.
evoluotions were preformed Appliconbs claim that all technical independent of TDI, and that the Owners Group program was R
While it is
' independent from TDI's QA program. Motion et 8.
not clear what role TDI Played in the directzon of Owners Group policy (TDI officiols have attended Owners Group executive relied extensively meetings), it is clear that Appliconts have on TDI, and continue to do so.
Relionce upon TOI is particularly dangerous, os TDI hos See Exhibit 15, in supplied false information to Appliconts. /
which TDI refers to the ' successful' operation of TDI engines in stationary ond marine service.
Unfortunately, Applicants have relied upon TDI's Specific recommendations in implementing u the DR/QR program.
examples are oddressed in subsequent sections devoted to that program. This uncritical reliance on TDI makes it all the more imperative that on independent evoluotion of the Perry DG5 be performed.
C. Phase I
' Phose I*
is that portion of the Owners Group Program Plan
-( & e s-
)
V which deals with the 16 known, generic Problem components.
These components i nclude piston skirts, connee. ting rod bearing shells, rocker orm cop screws, air stort volve capscrews ,
' IJ cylinder head studs, push rods, high pressure ruel lines, crankshort, turbocharger, connecting rods, engine base and bearing cop 3, cylinder heads, cylinder liner, cylinder block, engine-mounted electrical cable, and Jocket water pumps. Most Phase I onolyses were conducted by Foilure Analysis Associates
('FOAA") for the Owners Grocp.
.The proper stondord by which to evoluote the Phase I reports is that outlined by PNL in Exhibit 16. -The , essence or this stardord is that the analysis address the problem in a manner that is logicol, complete, thorough, ond technically correct.
The Owners Group has failed to accomplish this with the most critical components, s
b N'
- OCRE has addressed below the most significant components.
Inspection and DR/QR results pertaining to these components are also oddressed heroin, and not in-the subsequent section on the DR/QR,
, 1. crankshort
' Applicants claim that the FoAA onolysis of the V-16 I cronkshort demonstrates the adequacy of the PNPP-eronkshort, ucod Art.idovit at 80, In octuality, FoAA did noe reach such o conclusion. FoAA conducted on ave uotion or crankshorts at Shoreham Gnd Grdnd Gulf (Rtt: FoAA-84-3-16). The
-/7."'
Statement or Applicobility ror that report specifically cautions:
This-report addresses the structural integrity of the crankshorts in Transomerico Delaval Inc. 05R-48 engines at ene Shoreham Nuclear Power Station and D5RV-16-4 engines at the Grand Guir. Nuclear Power Storion. In view or possible dirrerences in generators, flywheels, and engine operating conditions, the results may not necessarily apply to other engines or the some model. These plant-speciric dirrerences, whcre they exist, will be evoluoted in separate reports.
- - - It-- is no t clear whe ther o ' separate report *~was eVer issued for Perrys OCRE is not aware or any. In roct, ese christiansen Arridovit identifies the May 1984 DSRV-16 crankshort Phase I report as being opplicable to Perry. (The FoAA report referenced above is dated May 1984.)
PNL ogrees with FoAA that cronkshort onalyses 'opply only to engines of the some type that are rated for the some load, and-that are equipped with generators and riywheels with the some torsional vibration chorocteristics." BN-84-152. PNL Report at a
7.
It is not clear what dirrerences exist between Perry and Grand Guir. engines, but Perry hos a larger size Flywheel than its designoted lead engine at comonche Peak (90 inch diameter at Perry, 68 inch or comanche Peak). Other dirrerences undoubtedly exist Which ofrect the cronkshort analysis.
The claimed success or cotowbo torsiograph tests (Wood Arridovit at 80) is therefore totally irrelevant to the adequacy of the Perry crankshorts.
The-report on Grand Gulf does not give much assurance that
c
(
- f f .-
V-16 crankshorts are suitable ror nuclear service. ,This report ,
identifies 3 railures or V-16 crankshorts in non-nuclear
- stationary' service. The roilures were attributed to torsional ratigue crocks initiating in the oil holes in main journal
' The Perry component Trocking System oiso numbers 6 or 8.
identifies-other V-16 cronkshort railures. In the V-16 r
Stationary engine at Glencoe, MN, cracks Were found in the No. 5 crankpin. Ita is postulated that.the crocks may have resulted from manurocturing riows. The V-16 stationary engine at St.
Cloud, FL surrered a broken crankshort, supposedly due to other failures.
The railures identified in the FoAA report resulted in design changes by TDI. The railed engines had a 4th order 4
critical speed at 446 rpm, very close to the operating Speed or 450 rpm. counterweights were odded to the cronkshort which moved the 4th order critical speed coWn to about 430 rpm. Gron'd Guir is said to have c 4th order critical speed or obout 430 rpm. Applicants admit that the Perry 4th order critical speed is 438 rpm, even closer to the OPeroting speed. Wood Arridovit at 79.
There is oiso a 3-1/2 order critical speed in V-16 engines which creates larger stresses than the 4th order, said to be in '
the 500 rpm range. See Exhibit 17, from BN-84-182. Note that this speed i s below the oversPeed trip setting or 518 rpm, and that the DR/QR maintenance recommendations for the overspeed -
p- - -
e--%-p e- +--w.a- .%ww c,- wy, ,%wq5eyr--,-,i,--ryw---g ww-y-wwi,- g- y -
,-tw yr wwwww y v wy,r g ,i-yry- -4 e -r-Alae' -- e--
f-I
[j -
w trip call for running the engine at no lood up to this trip s
point _at every refueling outoge.
The DR/QR Report for the governor mentions o critical speed.
' of 496 rpm, DEMA recommends that no harmful torsional Vibratory stresses occur Uithin 5% above and below the roted speed. For a 450 (Pm engine, the range in question is 427.5 to,472.5 rpm.. The Perry DG procurement specification hos even more stringent standards, e-
'that-the-DGshall'be' free of'oll deleterious critical sFeeds'or torsional vibrations within 10% obove and below the rated speed at any load from 0 to 110%. SP-562 Section 2.06.1. This speed range is 405 to 495 rpm.-
Applicants claim thot the 4th order stresses are not.
'hormful' because the components from the right and left banks f
almost concel, and. assuming a one degree delay in right bank ,
timing, the stresses are below the DEMA allowables. Wood 2 i
! Affidavit-at 78-79. The Owners Group's analytical ossumptions, hoWeVer, should not be accepted Without scrutiny.
i' FoAA developed-o torsional model of the V-16 crankshft for the Grand Gulf report. The hormonic loading on the crankshaft 1-
?
' is admitted to be sen'sitive to firing pressure, reciProcoting inertio, and, frictional loods. However, the firing pressures A used in the analysis were those measured for the Shoreham The peak riring pressure there was about 1600 osig.
{ engine.
Applicants admit that peak firing pressures for the Perry 1
I a
- ~ , ,- -,.e w y,- .n,,me- ew -,,ywea,m- .-
n.n-w_ -w , m g m _ .m pr., y, ,..,mn-y.ym- ,,,-.,.-mw--,-w--- m
-N- '
l engines may reach 1700 psig. Wood offidavit at 24. FoAA then
-ossumed firin9 Pressures to be the some, except for the timing difference, between the 1 e ,f t and right banks.
Factory tesb dato for the Perry engines indicates that this is not c conservative assumption. Exhibit 18 is the record of factory test runs for the Unit 1 engines. Exhibit 19 is OCRE's onalysis of the firing pressure variations exhibited in these tests. Note the considerable variotion in firing pressure among the cylinders. This variotion is Within TDI's alloWoble range, Which permits o difference in moximun and minimum pressures of 150 psi. See Exhibit 20, from TDI's instruction manual. Note also that the overage firing pressure for.the two banks con vary considerably.
FoAA 'found' 'the reciprocoting moss of the V-16 connecting rod and piston to be 820 lbs for each connecting rod. This e
number is remarkably identical to the reciprocating mass used for the in-line crankshaft analysis, despite the differences between connecting rod designs. See Exhibit 21, from the TDI Instruction manual, which gives the oppedximate Weights and configurations of the V-16 articulated ccnnecting rod design.
The moster and link rods do not weigh the some.
FoAA ogrees that these differences between the 2 banks ofrect the imbolonce driving the 4th order critical. See Exhibit 22, from the June 22, 1984 meeting between the NRc and the Owners Group. It is not clear that these effects have been
)
-2 l -
considered, nor that the suggested ' cure-all", the torsiogroph test, Will odequately evoluote them.'s The factory test dato showed erratic behavior with regard to firing Pressure differences. A torsiograph test may or may not
'cotch" the Worst-case situotion. Indeed, it was found for the Son onorre V-20 engine that the initial position of the crankshaft had a significont effect on stresses, in that for P
some positions stresses are in phase. See BN-84-182 ot 65-75.
Presumably the timing of DG loading could h0Ve the some effect.
Appliconts state that the horsiograph test Will be performed
~~"
on'only one~PNPP' engine ot'0%, 25%, 50%, 75%,~and 1007 of nomeplate roting. CEI January 17, 1985 submittoi ot.18, 110%
lood will apparently not be considered, even though it is o FSAR and SP-562 requirement. Nor is it clear that transient 100 ding conditions will.be considered.
e It is these conditions that may be the most toxing for the DGs.
FSAR 8.3.1.1.3.2 states that ' sequencing of large loads at 5
~
second intervals ensures that large motors will hoYe attained roted speed and that voltage and frequency Will hove stabilized before succeeding loads are applied. The decreases in frequency and voltage have been verified by qualification testing to be not greater than 5 and 20 percent of nominal, respectively '
Since -the frequency of an Ac generator is directly proportional to the speed of its prime mover, a 5% decrease in frequency means that DG speed will drop to 427.5 rpm. Every time a large lood is added to the DGs, they Will pass through the 4th order
_12-critical speed of 438 rpm. ,
Actual looding conditions may be even more severe than those p'o s t u l a t e d by the FSAR. The NRC's Integrated Desi9n Inspection found that there is no specification requirement that large motors reach rated speed within 5 seconds. In roet, the ESW5 pump considered by the inspectors requires 9 seconds to reach robed speed. BN-85-02 ot P. A-277. The result is that the DG moy have to supply motor storting currents for more than one motor at once, which heavily loads the DG and lengthens the time spent at the critical speed. The octual conditions will not be known until actual DG operation with actual loads.
More severe conditions than those Onalyzed by FioAA would also be Produced if the engine were to misfire. See Exhibit 23 from the June 22, 1984 NRC-Owners Group Meeting. Engine misfiring is a serious problem for the V-16, is likely to occur, e
and is considered in marine opplications. Exhibit 24, from TOI's Instruction Monual, also illustrates the large number of factors ofrecting cylinder bolonce. Misfiring thus should be evoluoted.
The only appropriote standards by which to evoluote crankshofe design are those of the ship classification societies. These stondords, of which Lloyd's Register of Shipping is the most conservative, consider a lorge number of inputs, including engine misfiring. See Exhibit 25, from the Joint Testimony filed by suffolk County on July 31, 1984 sn the
- 2 3-Shoreham proceeding. An independent evoluotion must be performed according to these conservative standards.
Only on ind'ependent evoluotion and realistic testing of the most severe DG operating conditions (for both DGs) will determine whether the Perry U-16 crankshorts are suitable for nuclear service.
- 2. Ristons The Owners Group has identified 6 runctional attributes or pistons. See Exhibit 26. Most or these have not been veriried by the Owners Group, which has only actively addressed the ratigue cracking or the piston skirt stud ottochment boss oreo.
Other ptston problems which could odversely ofrect engine performance have not been evoluoted.
For example, the strength of the piston crown hos never been analyced, even though the Component Tracking system has identified piston crown crocking on on engine in nuclear service (Kuosheng, Taiwon) and on the N/V Gott. A hole, said to be caused by ' secondary shrinkoge', was discovered on on engine piston crown. These failures, their causes, and actions needed to prevent rurther roilures, were not addressed by the Owners Group.
Similarly, there hove been numerous instances or rretting between the skirt and crown. This hos not been evoluoted by the Owners Group.
i Piston rings and pins are known to be susceptible to floking
r e
of their chrome plating. This has been observed at Perry (pin, see Exhibit 27) and Comanche Peak (pin, see Exhibit 28). No onolysis of the root cause of this problem hos ever been performed, nor is there.ony solution for preventing the problem.
Chrome flakes con cause scoring of the cylinder liner, resulting in piston blowby, which con eventually cause piston sei:Ute.
The PNPP DR/QR report for pistons is written for the wrong i skirt type, AH instead of the AE skirts actually used.
However, Applicants'do take credit for the FoAA report on AE piston skirts. These onolyses are seriously deficient, as demonstrated by Exhibit 29, from the Suffolk' County testimony in the Shoreham proceeding, summari:ed, the testimony shows that FoAA neglected a number of signiricant foetors, including use of underestimated peak firing pressure (ocknowledged to be 1700 p5ig by Applicants; 1670 woe used in the evoluotion); use of Skirt-to-crown gap (and other dimensions) unverified by measurement, except on o sample basis (see Exhibit 30): use of ide01 assumptions, such as isotropic material and uniform skirt temperaturess reliance on Kodiok engine (1000 psig firing pressure) and R-5 experimental engine i
operating experience, with only o sample of AE skirts; neglect i of piston side thrusts and the effects of tin skirt plating, Whtch con collect detritus which will cause liner scoring, leading to Piston blowby ofid possibly piston sei:ure. Note that I
- ~w-* ,
P- ~'- *' N g - e-tpw+W+77--wegy&m-------eiF-*+ey-w weiy,ppr- t-em-= rummPW&M-g M tt- P7 -y--*+e~- e'"'F W" TT ' '*
TDI hos not demonstrated on ability to produce defect-free piston skirts. Linear indications have been found by'2iquid penetront and magnetic porticle testing on new Perry AE skirts
.(see Exhibits 30, 64, and 66) and on Comanche Peak AE skirts (Exhibit 65). Although these indications were removed, there is no assurance that subsurface flows are obsent, since the inspection techniques are only copoble of detecting surfoce or near-surface indications. ,
contrary to Applicants' belief, reasonable assurance that the AE , piston skirts ' ore adequate for unlimited life under full load conditions' simply does not exist. A thorough evoluotion oddressing all the piston ~ functional ottributef"identfTied by' .
the Owners Group has never been conducted, and must be conducted i
by on independent, disinterest'ed entity before the DGs are considered acceptoble for nuclear service, s
- 3. Cylinder Heads FoAA hos divided cylinder heads into 3 groups, depending'On i
when they were cost. Heads cost before October 1978 (Group I heads) were not stress relieved and are subject to fatigue crock growth in thin sections ond/or from fabrication-induced defects.
l-Heads cost before September 1980 (Group-I and II) were subject to core shift, inadequate control of solidification, and i
inadequate control of the Stellite volve seat weld deposition i
i
,- . - - - . . . . , ..-,._,,m__,--~,--.--.---.------, e,..,,,-,...~,---,-.~,.--~.--,-.w,.~, -,------,-,,,,-c-.-----ym,m
1
)
process. Heads cost ofter september 1980 (Group III) are ;
-supposedly free of these defects. FOAA cy'linder Head Report at ii,-1-2. l The Perry DGs are using Group I heads, which Were returned Wood Affidavit at 44, stress relier,
.to TDI ror stress relief.
however, does not solve problems such as core shift and inadequate control of solidification. In fact, stress relieved heads have crocked in marine service (M/V Go t t) , according to the PNPP component Tracking System.
Nor is it clear that TDI's manufacturing ceilities have improved. Exhibit 32 is a portion of a report or on inspeceion conducted by Applicants at TDI facilities for the purpose of witnessing inspections on reworked cylinder heads. Five out of 19 heads were rejected for lock of fusion, hot tears, and inclusions, discovered by ,mognetic particle testing.
It should be noted that other nuclear facilities (Shoreham,'
Comanche Peak) have replaced Group I or II cylinder heads witn Group III heads. Nothing less should be expected from Applicants.
The DR/QR report for cylinder heads states that a design review is not necessary due to the FoAA report. It also states khot, for increased head reliobility, the engine should be
' blown-over* ofter each operation of the engine, and the fuel injection Port visually inspected for witer leaks during the monthly engine run. This oPProach is flowed on several grounds.
l t
First, FoAA never conducted a design review or TDI cylinder hoods. See p. 3-1 of FoAA's report, which states "no ottemPt was made to perform o detailed design review of the cylinder head." The reason for this-was that the head geometry is too complex ror such on onolysisi Instead, the rire deck only was modeled os o riot plate. The numerous riows in this onolysis are detailed in Exhibit 33, from the Surrolk county testimony in the Shoreham proceeding.
The Shoreham testimony also demonstrates that_there is no assurance that.the Group III heads will'be reliable eithert indeed, the head design is inherently defective, and TDI's costing process has not improved. Nearly all heads case in 1980-83 had defects TDI's inspection techniques will not detect subsurroce crocks. .
Wide variotions in rire deck thickness make cracking more l i k e l y ' t o " o c c u r.~ TDI ignores'the' maximum-rire deck tNTekness, ,
Which should be 0.515 inch except between the intoke volve .
ports, where 0.765 (nch is required. Exhibit 34, a record or UT inspection or PNPP cylinder head fire deck, illustrates the wide variation in fire deck thickness and the routine violation of '
the maximum thickness stondord.
There is inadequate evidence to support the claim that Group III heads have never crocked. TOI never evoluoted its files in the post 2 years to determine whether head railures have occurred, f
4
~
-LS-Cylinder head cracks con occur during cold shutdown of the engines and may not be detected before engine stort is attempted. Cylinder head cracks con lead to cotostroPhic DG failure, turbochorger domoge, and 'oir-lock' of the head water passages. Uoter leakage into,the cylinder is very dangerous; even small omounts of leakage con impair cylinder lubricotton, Compare Exhibit 35, from TDI's Instruction Monuol, which worns of the serious consequences of water in cylinders, The barring-over or blowing-over (see Exhibit 36 for a description) Procedure is inodequate for detecting the cresence of water in cylinders, Performing this procedure ofter each engine run will not detect leakage occurring ofter the test but before the next. engine start. Obviously such procedures connot be performed prior to on emergency OG store, Given the extensive history of TOI cylinear head cracking in nuclear, stationary, and marine service (see Exhibit 37, from the PNPP Component Tracking system), the derective Owners Group onalyses, the inherently flowed head design, the lock of assurance that manufacturing Problems have been solved, the severe consequences of head cracks, and the inability to detect crocks before they cause domoge, the use of TDI cylinder heads for nuclear service connot be justified,
- 4. Connecting Rods A large number of connecting rod failures hos been observed in TOI engsnes, see Exhibit 38, from the PNPP Component i
~29-Tracking System. Due to this adverse experience, connecting rods were evoluoted under the Owners Group Phase I Program.
Exhibit 39 is the Owners Group's Tosk Description for connecting rods.
The Owners Group has failed to fulfil its commitments in that task description. For example, buckling strength, while ,
examined for the in-line rod, was never evoluoted for the articulated design used at Perry.
Similarly, the Owners Group has not odequately addressed wrist pin bushing failures, FoAA's report on in-line connecting rods included on analysis of wrist pin bushings (since extensive cracking was discovered at Shoreham), but this analysis may not be conservative (no indication of the peak firing pressures used was given) or opplicable to V-16 engines (due to inertial lood differences in the exhaust stroke).
An independent ondlysis is ,
e necessary to ensure the suitability of TDI Wrist Pin bu,shings.
The extensive cracking observed at Shoreham, even in new bushings, is yet further' evidence of the inherently POUR quality and unreliability of TDI engines. The recommended NDE inspections for wrist pin bushings (LP testing) in the Perry DR/QR report will not detect subsurface flows that could Propogote to foilure.
The problem most thoroughly oddressed by Applicants is that of rod box or bolting failures, However, their own analysis shows that even the supposedly superior 1-1/2 inch bolt e----- L::L_-__ ._r_-_-_____ _,
_3 o .-
configuration used at Perry is marginal (foetor of safety = 1.08.
But we are told not to worry because of favorable field experience with 1-1/2 inch bolts. Wood Affidovit at 84.
The truth is that field experience hos not been favorable.
FoAA conducted a survey of Vee engine connecting rod experience.
Out of a population of 148 connecting rod assemblies, two 1-1/2 rods have crockeds four 1-7/9 rods have crockeds six 1-7/9 rod bolts have crocked 'with no opporent domoge to the rods.' FoAA-84-3-14 DSRV-4 connecting Rod Report at 1-3 to 1-5. No statistical analysts Was performed to determine whether the dato 1
could support a conclusion as to the superiority of the 1-1/2 inch design. Applicants admit that there must be ' substantial operating experience ond/or experimental dato to confirm the design integrity.' Wood Affidavit at 81. Neither exists.
The results of inspections of the Perry connecting rods a
likewise do not inspire confidence in their reitobiltty.
Exhibit 40 is o nonconrormance report describing go11ing or the rod bon threads. Note that the cause of this problem has not been determined, and no efforts to prevent recurronce are in progress. Exhibit 41 is o nonconformance report describing fretting on the connecting rod rock teeth, Ho cause of the problem was determined, and the rods were used os-is based on TDI's disposition. This is on example or Applicants' continued uncritical reliance on TOI for technicol evoluotions.
FoAA believes that 1-1/2 inch rods are acceptoble for use e
provided rods with pre-existing flows are eliminated. FoAA Report at 2-15. The use of the fretted and galled rods violates this criterion.
The Perry DR/QR Report, Appendix II, the maintenance matrix, states that if connecting rod rock teeth fretting is found on engineering evoluotion should be performed. None was done here.
It must be concluded that no assurance exists that the Perry connecting rods are suitable for nuclear service. The Owners Group's own onolysis, which may not be conservative, demonstrates that the rods are marginal. Adequate experience of successful rod operation does not exist. Indicotions found during inspection of the PNPP rods have not been dispositioned i
in a technically volto manner.
5, connecting Rod Beoring shells Exhibit 42 is the PNPP Component Tracking system for connecting rod bearing shells. Note the wide variety of f o i l u r'e mechanisms involved. Foilures have been'ottributed to bad alloy makeup, connecting rod fretting, and loss of bearing crush.
Many of the failuees identified did not give the cause.
Applicants have~en_y addressed the type of failure cecurring at Snorehom.
While the DR/QR report recommends o number or NDE tests (rodtographs, eddy current, liquid penetront, visual, and dimensional verification), no verification of alloy type is required. In addition, Appliconts are using 05-15 bearings
which have failed the FoAA inspection criterio. Relying instead on information from TDI, they accepted bearings with linear indications and scoring and galling. See Exhibit 43, inspection results for connecting rod bearing shells, Note rhot even though the acceptance criterio clearly stated that surface cracks, linear indications, scoring and galling are unacceptable. Applicants occepted bearings exhibiting such flows because TDI's service representative accepted them.
It cannot be concluded that the Perry connecting rod bearing shells are acceptobte when they include obvious nonconforming indications, the cause of which is unknown, Not all of the failure mechanisms found in field experience for connecting rod bearings have been studied. Nor has it been demonstrated by on independent disinterested onolyst that the evoluotion of the Shoreham bearing failures is conservative. Reasonable assurance that the PNPP DG connecting rod bearing shells are suitable f o r' nuclear service does not exist,
- 6. Engine Base and Bearing Cops Appisconts claim that the engine base and bearing cops are There adequate for nuclear service, based on FoAA's evoluotion.
is no evidence that this analysis is conservative, Os the FoAA report is on inscrutoble document, For example, the FoAA report states that the primary function of the base ossembly 15 to 0119n, support and react the crankshoft loods at the bearing saddles, and to react the firing
- " ^' '~~
_g3 forces transmitted by the through bolting. The loads imposed by the cronkshaft include crankpin, piston.' and rod inertio and the firing loods. FoAA-84-4-1 Rev, 1 at 5, But it is never explained Just how these loods are modelled in the stress the firing analysis. It connot be dete'rmined whether e,9,,
pressures are conservative.
Applicants state that the nut pocket failure was due to impurities in the costing materiol, Wood Affidavit at 16. FoAA this was TOI's conclusion. 'TDI reported never determined this that this foilure was due to impurities in the costing material that reduced the en9ine base strength. They reported that the impurities were traced to non-ferrous components amon9 the scroP iron used for the costings,a FoAA Report at 10, FoAA's analyses apparently assumed that base materials were of specified composition and Strength, which, given this failure, e
is not a conservative assumption, A complete analysis of the engine base and bearing cop failures hos never been performed, and needs to be performed by on independent disinterested entity before the suitability of this component con be determined, The DR/QR report otso fotis to ensure the adequacy of these
~
componentt, If a failure has been attributed to costTng material impurities, then o logical approach to revolidating the engine would include o determination of the chemical composition and material properties of the base for each en9ine. The OR/QR
-3 4 -
report does not recommend this; the only inspections required ore o verification or bearing cop stud nut torque and visual and LP inspection or only the No. 5 main bearing soddle area on only one engine.
The Perry DGs railed even this limited test, See Exhibit 44, which shows that linear indications have been round on the No. 5 saddle or the Unit 1 Division 1 engine, Applicants necepted these indications os-is ond have not tried to determine the cause or these indicokions, whether they Will propogote in service, or, through expanded inspections, whether other Soddles, nonconforming conditions exist in other In light or these rindings, there con be no assurance that the Perry engine bases are suitable for nuclear service,
- 7. Turbocnorgers The Elliott 90G turbochargers used at PilPP have had unrovorable experience. See Exhibit 45, from the PNPP Componen't Trockir9 System. Note that the problems encountered include thrust bearing lubrication, excessive vibration, surging, and no: le ring vone breako9e. The Owners Group has recused only on the thrust bearing and no::le ring problems. Other. turbocharger problems have been ignored.
FoAA's analysis or no::le ring vones did not prove that these components are acceptable. FoAA concluded, based on vone railures observed at Shorehom, Kuoshen9, Comanche Peck, and Crond Guir, that the no::le may indeed experience vone roilures, L-m-
- 3 Ji -
But, because these failures have not yet resulted in engine shutdown, FoAA concludes that vone failures should not significantly ofrect turbocharger operation. FoAA recommends that Elliott consider its findings an future no::le ring designs, PNL opparently disage,ees with the conclusion that vone failure is harmless, Appendix M of SSER 6 for Grand Gulf, NUREG-0831, is PNL's evoluotion of the reliobility and operability of the Grand Guir TOI DGs, Therein (p.19) PNL states 'there is o high probability of domoge to the turbochorger if the vone breaks in service,' The Owners Group 0150 reali:es that loose and broken ports con domoge the turbochorger. See Exhibit 46.
The Division 1 engine turboChorger Wos found to be severely domoged upon inspection and was replaced, See Exhibit 47. The o
cause of the domoge was not determineds it was merely attributed to the factory testing (only a few hours). There is no guarontee that domoge will not occur during subsequent engine o p e r a t i o *i ,
PNL ottributes excessive vibration at Grand Gulf to turbochorger misolignment. SSER 6. Appendix H ot 17.
Misolignment is a problem at Perry as well, see Exhibte 48.
Note that Applicants have again relied exclusively upon TOI for technical infornation, ' Mo analysis" of the~ proposed d1Tposition (elongoting bolt holes) was conducted to determine the effect O
-3G - l upon turbocharger operation.
The Perry turbochorgers are inadequate. They are likely to experience no :le ring vone breakage, which could domoge the unit, have been domoged by on unknown mechanism, and do not align properly With their mounting brackets.
S. cylinder Blocks and Liners TDI cylinder blocks have experienced numerous failures.
These failures include ligament cracking, stud-to-stud crocking, stud-to-end cracking, circumferential cracking at the liner counterbore lip, and creep and thermal distortion. Applicants hove only evoluoted the cracking problems) no analysis of the distortion and creep was performed.
Applicants imply that cylinder block failures at Perry con be ovoided if the blocks are found to have acceptable material properties and microstructure and if liner proudness is reduced.
First, it,hos not yet been determined that the PNPP eagine blocks are free of substondord material. Secondly, FoAA never recommended reducing liner proudness:
tiodification of liner collar counterbore vertical fit (liner proudness) to the reduced level currently specified by TOI will result in reduced probability of circumferential cracks.
Quantification of the exact impact of this mcdification on stress perpendiculot to ligament, stud-to-stud, and stud-to-end cracks hos not been performed. Therefore, no recommendation 15 offered regarding the overall desirability of reduced liner proudness. FoAA-84-9-11, ' Design Review of TOI R-4 ond RV-4 Series Emergency Diesel Generator cylinder Blocks', Dec. 19G4 oc v.
In fact, it would seem that re'ducing liner proudness would lead to loss of liner crush, which hos occurred in TOI engines. Loss
-37 of crush could cause liner fretting or leakoge of exhaust gases into the coolont or leokoge of cooling water into the cylinder.
The deleterious effects of these events h0Ve been described above in the discussion on cylinder heads.
FoAA's evoluotion predicts that cracks will initiate in the cylinder blocks. even those of normal microstructure. FoAA concluded that cracking of the block top may initiate ot.1007; nomeplate load due to high frequency fatigue or ofter 100 starts to full power due to low cycle fatigue. ' Ligament crocks are more likely to initiate thon s tu d- t o-s t u d or stud-to-end crocks, but once ligoment cracks form the other cracks are more likely to occur. FoAA report at 5-2.
The cause or this propensity for cracking is the inherently defective design of the blocks. See Exhibit 49. the M/V Columbio Engine Rebuild Report. Which concludes that a circumferential cracks will be recurring due to the number of cousative factors ,
including high comoressive stresses on the counterbore lips localized stress from the sharp internal lip corners nearby drilling for Waterjocket or studs termination of stud treading at the some levels creep deformations and fatigue.
'Because the design stresses were so high, there was no forseeable Way to prevent foilures from occurring Without a signicont redesign of the liner-block landing surfaces.' Ex. 47 at II-9.
Similarly. FoAA found that increasing the rodiol cleoconce L.
- . _ ~ . . - _ . . _ _ - . . . . _ . -- - ~ - . - _ _ - - . . _ . _ . - . - . . , . ~ - . ~ _ ,.
4 I
I
-3 g , ,
i between the block and liner will reduce the block top stress and !
reduce the likelihood of cracking (the liner exponds thermo11y l more than the block). Apparently realizing the need for ,
t redesign, TDI has modified block design in its R-5 engines. !
These modificottons include increasing the block top thickness from 2.5 to 3 inches, providing deeper stud bosses and stud hole i
threads, and upgrading the block material from Class 40 to Class
! 45 cost iron. FoAA report at 1-6.
i TDI hos otso increased the radial cieoconce between the block ond liner, from 0.008/0.0045 to 0.0105/0.0070 (upper 9o9: .
l .
i lower gap was otso increased). FoAA Report Figures 1-6, 1-7, i
{ ond 1-0. The PNPP engines, however, do not have the benefit of l i these improvements.
l FoAA does not predict unlimited block lire. Rother, o i '
cumulative domoge index hos been developed to determine the i
4 length of time on engine con operate. This calculation utilizes [
a engine OPerotional history (time at lood), and predicted ;
f operational chorocteristics, included those for o LOOP /LOCA r
event. The Perry DR/QR report references Exhibit 50, apparently for the purposes or this calculation. Note that Exhibit 50.was !
- prepared in response to Exhibit 10. Which bemoons 'ultro- i conservative' F5AR load dato and urges arealistic' load f j prorites, perhaps towing'credte for operator oceton in removing
)
1 loods from the DGs.
i Thus, the loods assumed in Exhtbit 50 may not be 2
I 'l i
conservative. Nor is it clear that this lood profile was assumed for LOOP /LOCA only and not oiso for normal DG testing.
If so, this contradicts draft Technical Specification commitments to test the DGs at least once every 31 days by running at 7000 kw for 60 minutes, and oe least once every is months by running at 7000 kw for 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br />. See Exhibit 51.
Also, it is not clear that the calculation has considered the possibly more severe starting conditions discussed above, concerning the ability of large motors to reach roted speed before additional loods are added. There thus is no assurance that the cumulative domoge algorithm wsil conservatively predice safe operatin9 life of the cylinder block.
Exhibit 52 is FoAA's flowchort for applying the cumulative domoge procedure. Note that for o crack-free block, continued ,
operation is dependent upon the block material being typical for gray cost iron, the block and liner dimensions being
- 10tisfactory", and the engine having significant operational history. It has not yet been determined Whether the PNPP engine block material is acceptable. The engines otso lock significant operational history (only a few hours of factory testing). No
.accetance criterio are given for 'sotisfactory* block and liner dimensions. These are given, however, in Exhibit 53.
Incredibly, the occeptable dimensions are all thole in use by l
TDI, including the early dimensio.ns for liner / block radiol gop which increase t
-4o -
the likelihood of cracking, A conservative ossumption Would be that there are no satisfactory standards for these dime,nsions.
The flowchort for Perry thus leads to the conclusion "no analytical basis for continued operation,'
Additionally, FoAA's enetre methodology with respect eo block cracking, including the cumulatsve comoge calculation, ss roulty. This is thoroughly discussed in Exhibit 54, from the Suffolk county testimony in the Shoreham proceeding, FoAA also believes that any crocks that may initiote are hormless. This too is.o rollocious assumption Which is rebutted by the Suffolk County testimony. This testimony shows that all types of. observed cracking are very dangerous and con lead to Cotostrophic engine failure, Appliconts identify 'the ability to withstand reactive side forces Without eVeessive Wear or scuffing' os a functional ,
attribute or the cylinder liner, Wood Afridovit at 55. This hos never been analy:ed for the TDI liners, despite adverse experience With scurring and distortion due to piston side thrust, Applicants may hoVG created a new problem offecting cylinder liner reliability. See Exhibit 55. Appliconts failed to ensure that liner and block matchmarks were oligned When installing the Itners. Alignment is necessory as there is o cutout at the liner bottcm for connecting rod travel, The misoignment is os great as 7/16 inchi however, Appliconts accepted this situotion
-4(-
os-is because TDI's service representative accepted it. TDI occepted 7/16 misolignment even ofter recommending that 1/4 inch be the maximum misalignment. This is yet another example of Applicants' continued relsonce on TDI for technical input.
There is no assurance thoe engine domoge will not resule from this situation.
There is no evidence that the PNPP DG cylinder blocks and liners will perform their functions without failure. An
\
odequate onolysis of block crocking and its consequences has never been performed, and other observed failure modes have not been analy ed at all. Ik must be concluded that these components are not suitable for nuclear service.
D. Phase II OR/QR The DR/oR program is supposed to verify the reliability of d
engine components other than those evoluoted in Phase I. 171 components were selected for the Perry engines however, only is
-,s of these received a full design review. The others were deemed not to require design review because of supposed similarity to the lead engines at Comanche Peak.
The heavy reliance on the lead engines is illustrated by the following statements from Applicants
- January 17 submittal of their DG Program Plon:
Upon completion of the OR/QR Program on the lead engine, the results . . . were foetored into the follow-on engines such as PNPP. A separate design review was not required, for example, on o common component for follow-on engines. However, on
3-I inspection moy.be required to verify that the component is actually the some os the one reviewed for the lead engine.
tIR'eWise,' quality revolidotion' inspections ~compIrted'cn o'*
component in o lead engine would serve os o basis for either increasing or decreasing inspections for o follow-on engine, depending on the lead engine inspection results. (P. 5)
Expanded testing or inspections will generally be rocused on the
- 1ead" engine, with less stringent requirements for following engines of the.some type if worronted by preceding ~ results. (P.
6)
This concept is inherently flowed due to the poor or totally s
lacking quality assurance at TDI. Using the lead engine concept for design reviews might be justified if TOI's design control were oceeptoble. However, this is not the cose . - An example of this is the front gear case bolting, component #02-335B. The Perry DR/QR repcrt for this component states:
o QC inspecion or,5heoron Harris (CP&L NCR 94-1777) showed two bolts internal to the georcose that were not evident on the parts'monual drowing. . . The CP&L inspection found that no positive means of locking was provided. Considering the domoge these bolts could cause if they loosen, it is recommended that these bolts be inspected at Perry and positive locking features (bent tab or lockwire) be odded. . .
Were it not for the fortuitous inspection or another site, , a these bolts Would hove gone undetected, os TOI's pares manual a failed to identify them. TOI*s QA is such that other unidentified differences between the parts monuol and the engines (or among engines) exist. They Will not be detected unless complete inspections are performed. Thus, on inspection must be done for each component to ensure 'that it is identical to that on the leod engine.
Any reloxotion in OR inspections or testing based on w
i
-q3-roverable results at lead plants is likewise improper and without technical justification. Because or TDI's inerreceive ,
QA program, rovoroble results at one' engine does not guarantee
, rovorable.results for all engines. Inspections on a sompling ,
basis are similarly meaningless, i Hony components (listed below) received no QR. OCRE i
3 believes that these items must be inspected and/or tested to ensure their reliobility, Components receiving no QR include:
F-068. Intercoolers 02-3804. Exhoust Manifolds 02-8050, Flex Connectionss00-420. Lube Oil Pressure Regulating Volves 02-307A B,D. Lube Oil Fittings: 02-420. Lube Oil Pump (engine-driven): 02-465A. Lube Oil Liness 02-467A, Turbocharger-Lube Oil Fitting Pipings02-540 A.C. Lube Oil 5. ump Tonk. Stroiner, &
l_
Hardwcres 02-540B _ Lube Oil Sump' Tank misc, rittings, gaskets, e pipe, and boltings 02-717F. Aux, Sub Bose Oil & Water Piping; 02-717I. Aux, Sub Bose Piping supports, mounting hardwares 02-
-820B, before & orter Lube Oil Pumps.02-820G, Lube Oil Heat Exchangers 02-310C. Thrust Ring Beoring: 02-315D, Jacket Water l
Honifold Pipings ~
02-315G, Cylinder Block / Liner seois and Goskets: 02-441A, Storting Air Manifold Piping. Tubing, & Fittings: 02-4410.
Storting Air Monifold supports; 02-835B. Storting Air Tonk( 02 '
345C, Fuel Pump Bose Assemblys 02-350B. Comshort Bearings 02-I 390E, Bushings,s 02-330A, Flywheels 02-330B, Flywheel Boltings 02-4100, Overspeed Trip Vent Valves 02-6958. Engine Shutdown Equipments 02-316C, Jocket Water Inlet Manifolds 02-435A, Jocket Water Fittings: 02-435B, Jocket water Supports: 02-437. Turbo l Water Piping & Fittings: 02-7004, Jocket Water Standpipe pipe, fittings & gaskets: 02-7000, Jocket Water Standpipe supports 02-700F, Jocket water standpipe bolting: 02-717V, Aux, Sub Base Pipe couplings, rittings, etc,s 02-7170, Aux, Sub Base JW goskets & bolgings 02-717E, Aux, Sub Bose Supports: 02-810B. JW Heat Exchangers 00-621A. Fuel Oil Drip Tonks 02-3650. Fuel Injection Supports: 02-450A, Fuel Oil Header pipings 02-717J.
Aux, Sub Base Fuel Oil Piping and Fittings 02-717L, Aux, Sub Base Fuel Oil Bolting & Gaskets: 02-8254. Fuel Oil Day Tonk5 02-650A, Generators 02-650C, Generator Short-& Bearings: 02-550.
L Foundation Bolts 02-717A, Aux, Sub Bose, The above is not on exhoustive list or items receiving no
. _ _ _ _ . , _ . . _ ~ . , _ _ _ _ . , _ _ _ _ , _ . . . _ . _ , _ . _ _ _ . _ _ _ _ , , _ , _ . , _ _ _ , _ . . _ _ . _ _ ,
+ .p4-QRs excluded were items wuch as volves, filters, and other components not produced by TDI which are of stondord design ond-are used extensively in the industry, and electrical components . . . . , -
such as switches, wiring, thermocouples, terminals, etc. Items such as these do not need QR.
Speciric comments on the components in the DR'/QR Report are i given below.
- 1. 02-500A, Control Panel Assembly cabinet / System .
The DR/QR Report for this item add'resses Georgio Power Company *s Vogtle Plant and not PNPP.
' 2. 02-717C, F,\ ond I Aux. Sub Bose Piping, Supports, and
]
Mounting' Hardware These components are part of the Jocket water and lube oil systems. These systems are required to be designed and built in occordonce with ASME Code Section III, Class 3. SP-562. -
9 Sections 2.06.7 and 2.06.8.
2 The DR/QR Reports for these components all contain the fcilowing statement:
The lead engine report does address site specific modifications
< to the skid piping ond/or supports. Generic application of r
these modifications is not required for Perry since the Comanche
. Peak modifications were not required for piping operability.
The lead engine modifications were' recommended in order to meet the intent and philosophy of the ASME Code for the boundary conditions and assumptions used in the Owners Group Analysis.
These boundary conditions and assumptions may be somewhat different from those used in the original manufacturer's analysis.
Since the lube oil and jocket water systems must be designed and built to the ASME coda, ir modifications are required to y -, - -y-. w-y-y, ,~y-, ..-,y-- v~,-1,v . - . - - .-,,-y e- -- p i.g , -
ye . , -
, .,r
- 4 5-meet the ASME code then they are required Gt Perry. Whether they are required for' operability is irrelevant.
- 3.02-317 A&B, 02-435A,02-437, 02-467A, 02-467A. 02-717C, 02-717F These components are lube oil and Jocket water fittings.
T'iese items are to be visually inspected for leaks, and should they occur, the existing Dresser Style 65 couplings should be replaced With Dresser Style 90 couplings equipped with Viron gaskets. The reason for this is that the maximum suggested operating temperature of 150 de9 Pees-F may be exceeded. The Style 90 coupling con Withstand 212 degrees-F.
7 It is not conservative to wait until leaks occur before taking corrective action. The couplings should be replaced ,
before plant operation.
- 4. 02-360B, Intoke and Exhaust Volves o
The DR/QR Report admits that the primary odverse experience associated With the volves has been Chrome plate floking, scuffing, scoring, and exhaust gas blowby due to lack of concentricity of volve and seat. However, no modifications are required for the volves, since the problems are not' expected to
- noticeably ofrect engine performance
- due to the small number of hours the DGs are expected to operate between inspections.
The maintenance recommendations for the volves in Appendix II of the DR/QR Report include 3 items. One is o one-tim'e-only inspection for evidence of exhoust 90s blowby after 500-600 aw~**' "*******=me-e-. .-.e-du-- **
-q s- '
hours of operation. The other 2 inspections, o visuoi inspection for pitting, distortion, concentricity or other obnormalities, and measurement of volve head thickness, are to be performed once every 5 years. This is not frequent enough to
'~~
detect'the problems"Which hove ~ occurred; - ~ ~ ~
The design reviews for the lead engines (the volves did not receive o PNPP-unique DR) opparently did not determine the root cause of these problems, nor were any corrective actions formuloted.
The consequences of these volve problems are for from benign. Chrome flakes from volve stems could be drown into the cylinder, cousing liner scoring, piston blowby, and eventually piston sie:ure. Blowby post te volves will result in a loss of cylinder power, and the hot combustion goses may cause further thermal stresses in the already substandard cylinder heads, a
thereby acceleroting crocking.
The adverse operating experience for the volves demonstrates that,they are not suitable for nuclear service.- The DR/Or Program.hos foiled to properly evoluote and r e v o l:i d o t e these components.
- 5. Componenes Improperly Excluded From Review (o) 02-530A.B.c,&D, Platforms, Ladders, Stairs, and suppores These items received no DR or QR. Evoluotion should have been' made os to whether the failure of these items during a seismic event could adversely ofrect other components necessary l
. .--~~~- . . . . . _ _ . . ,
,y - m--,r - - - - , - , - --,#-- - y- -e,, - - -
- 4 7-for proper engine operation, (b) Foundation This item was not even included in the component Tracking System (and thus received no DR or QR) desPite adverse PNPP experience, which is evidenced by Exhibit 56. After the Unit i engine foundations were grouted, it was discovered that some of the chock plates did not meet TDI's 85% bearing requirement.
Some plches had as little as 10-15% bearing. Applicants accepted the condition as-is on the basis of a calculation performed by TDI, and provided that crankweb deflection checks are performed prior to stortup, and at 20 hours2.314815e-4 days <br />0.00556 hours <br />3.306878e-5 weeks <br />7.61e-6 months <br /> and 168 hours0.00194 days <br />0.0467 hours <br />2.777778e-4 weeks <br />6.3924e-5 months <br /> of operation.
TDI's Instruction manual states that the foundation is to be constructed to the highest occuracy, and that the engine must be oligned before grouting, and that the engine weight should be e distributed evenly on all sole plates. See Exhibit 57.
Engine misolignment con cause excessive eronkweb deflections, which TDI admits con cause cotostrophic cronkshaft failure. See Exhibit 58. Another example of failures caused by insufficient contact between the engine and chock plates is given in Exhibit 59, from the PNPP component Tracking System.
Note that the corrective action was to grind the chock Plates to .
ochieve full contact.
A ' loose' foundation will also lead to excessive vibration,
[
which con cause other casualities. See Exhibit 60, the report
~
l l
__ .__ _. - ~ , _ _ _ _ _ _ _
(
of a Coast Guard investigation into failures of TDI engines on the icebreaker Northwind. Loose foundation bolts caused the following vibration-related failures:
fractured 4 turbocharger mounting bracket bolts: froctured 3 exhoust bellows; many lube oil and water pump failures, with associated piping failures; one exhaust pipe failure; 3 governor failures. Ex. 60. Finding of Fact #46.
(Note also that this investigation found.that cronkweb - - ,. . . _ - . , .-
~
deflection measurements, to be meaningful, must be toRen when the engine is hot, i.e., ofter running 8-12 hours under load.
Finding of Fact #26. It is not clear that Applicants are aware of this requirement.) ,
The Coast Guard's investigation of the Northwind foundation has produced evidence that costs doubt on the conservatisms, if any, in TDI's calculation. Exhibit 61 indicates that TDI doesn't " pay much attention to creo of chock vs. bolt torque",
in contrast to the Classification Societies and the chock supplier. Exhibit 62, information from the chock supplier, indicates that the maximum possible chock areo should be used, and that, for the resin chocks in question, compressive stress should be limited to 500 Psi, despite its compressive strength of 19,000 psi. This results in a factor of sorety of 38. TDI's
' calculation assumes o factor of safety of 2.
Exhibit 63 is o calculation occording to Lloyd's Rules of chocking and bolting criterio. It is not clear whether these criterio are specific to chock type; i.e., allowable loods may be higher for the steel chocks used at PNPP. However, given the
-4 9-inherent coni .votism of the Classirication Societies, as discussed above, the adverse experience resulting from l
J insufficient chock bearing (Ex. 59), and the potential for cotostrophic crankshaft failure, it is prudent'to determine the adequacy of the as-built PNPP roundation accordib9 to Lloyd's Rules. Until proven otherwise, this should be considered on unocceptable condition.'
IV. CONCLUSION From the discussion above, it is clear that serious problems exist'with the PNPP'TDI DGs. They do not comply with GDC 1 and
'GDC 17. The Owners Group program has failed to resolve the significant design and quality problems. Applicants have failed to demonstrate the obsence of a genuine issue of materici fact.
Their Motion for Summary Disposition of Issue M16 must be denied.
r Respectfully submitted.
A Susan L. Hiort OCRE Representative 8275 Munson Rd.
Mentor, OH 44060 (216) 255-3158