ML20056F856
| ML20056F856 | |
| Person / Time | |
|---|---|
| Site: | McGuire  |
| Issue date: | 08/23/1993 |
| From: | Mcmeekin T DUKE POWER CO. |
| To: | NRC OFFICE OF INFORMATION RESOURCES MANAGEMENT (IRM) |
| References | |
| 93-06, 93-6, NUDOCS 9308310215 | |
| Download: ML20056F856 (20) | |
Text
{{#Wiki_filter:______________ 11 1 \\ Dde h>urr Csmpany TCKumn AfcGuwe helm Generatwn Department Vice hesident ' 1:!U3 Hopen ferry Road ( AtG01A) (704)h!S 4800 I!antenu!!c NC2%!8 M5 (704)S75-4RO9 tax b DUfWPOWER August 23, 1993 U.S. Nuclear Regulatory Commission Document Control Desk j Washington, D.C. 20555 l L
Subject:
McGuire Nuclear Station Special I'. aport Number 93-06 l Problem Investigation Process Nos.: 1-M93-0283, 1-M93=0244 Gentlemen: Attached is Special Report Number 93-06 which documents the various l debris items discovered during the McGuire Nuclear Station End of L Cycle Outage IEOC8. This report has condensed the information contained in the in depth Engineering Package (attached) and organized the information into the standardized format. The Engineering Package should be referenced in conjunction with this report for in depth discussion of specific areas of interest. l Very truly yours, vl }l Lv { / l T.C. McMeekin TLP/bcb Attachment xc: Mr. S.D. Ebneter INPO Records Center Administrator, Region II Suite 1500 L U.S. Nuclear Regulatory Commission 1100 Circle 75 Parkway 101 Marietta St., NW, Suite 2900 Atlanta, GA 30339 Atlanta, GA 30323 J) i) Mr. Victor Nerses Mr. George Maxwell I "[ U. S. Nuclear Regulatory Commission NRC Hesident Inspector Office of Nuclear Reactor Regulation McGuire Nuclear Station 'j Washington, D.C. 20555 9308310215 930023 97 pf'I I l ~%iw PDR ADOCK 05000369 b S PDR [$
bxc: B.L. Walsh R.C. Futrell (CNS) P.R. Herran j R.C. Norcutt M.E. Patrick (ONS) G.H. Savage l G.B. Swindlehurst i I H.B. Tucker R.F. Cole D.B. Cook G.A. Copp-C.A. Paton 7 M.E. Pacetti j P.M. Abraham W.M. Griffin NSRB Support Staff (EC 12-A) 1 i i i t I l i l l l i i i
1 s t McGUIRE SAFETY REVIEW GROUP SPECIAL REPORT 1-REPORT NUMBER: 93-06 2. DATE OF REVIEW: April 5, 1993 through August 12, 1993 3. SUBJECT DESCRIPTION: This special report investigates the circumstances relating to the incident described on Problem Investigation Process (PIP) report 1-M93-0283, Discovery of Debris on the McGuire Unit 1 Lower Core Plate (LCP). The specific purpose of this report is to document the actions taken to retrieve and identify the foreign material. 4. EVALUATION AND COMMENT: Abstract: 2 PIP 1-M93-0283 documents the discovery of significant amounts-of debris on the LCP. The major portion of the debris appeared to be small wire 1 like objects, ranging in length from approximately 3 to 6 inches. During a subsequent, detailed video inspection of the LCP, another object, approximately 10 inches in diameter and 12 to 16 inches in length was located. Four other unidentified objects were also noted during the video inspection of the LCP. The identity of these objects and their specific origin (s) have not been reconciled. Additionally, while attempting to identify-the source of the wire-like debris, a number of small stainless steel chips were discovered in the Emergency Core Cool..g System (ECCS).
Background:
Wire Debris: The sleeving for Steam Generators (SGs) at McGuire Nuclear Station was subcontracted to Babcock and Wricox Nuclear Technologies (BWNT). The
~~ is 4 J .i I DPC/MNS l Special Report No. 93-06 j PAGE 2 BWNT kinetic sleeve designs are described in the Topical Report BAW 43-l 2045PA-00, June 1988, entitled " Recirculating Steam Generators Kinetic Sleeve Qualification of 3/4 Inch OD Tubes". l s The BWNT kinetic tubesheet sleeve is designed to repair 3/4 inch OD SG j tubes which exhibit defects near the secondary face of the tubesheet. i The sleeve is nominally 0.545 inches ID with a 0.041 inch wall. j Material for the sleeve is thermally treated stainless steel Alloy 690. j Two a)eeve lengths have been designed. Due to the SG bowl dimensions in i r the periphery, an 11 inch sleeve length has been designed to permit access to all SG tubes. The lower end of the sleeve is positioned betwar.n 3.2S and 5.0 inches below the secondary f ace of the tubesheet. A se ad sleeve, 17.5 inches long, can be installed in approximately 93% of ';:e tubes. The lower end of the sleeve is positioned in the same location as the 11 inch long sleeve. The longer sleeve will span both j the secondary face of the tubesheet and the flow distribution baffle plate. '( 1 The freespan sleeve / tube joints are produced by a kl.netic welding l process. The kinetic tubesheet sleeving provides repair options of l kinetically welded, mechanically rolled joint configurations, or a l combination of the two. The rolled joint is produced following the l ) stress relief of the upper joint, while the lower kinetic we3d ::+ccurs simultaneously with the freespan weld. During the welding process, both the upper and lower kinetic weld devices (KWD) detonate simultaneously, producing the sleeve to tube welds. The KWD 1 tstic holding the I enarge(s) breaks up during the welding process. The plastic, lead wires, and most of the connector wires, are removed from the SG tube along with the nylon insertion tube following welding. Some of the connector wire travels up the SG tube, over the U-bend, and into the opposite SG bowl. The SG bowls are vacuumed after sleeving to remove any residual welding debris. The first step of sleeve installation is to clean the SG tube where the j kinetic welds will be performed. Next, a sleeve with a prepackaged KWD(s) is inserted ir.to the tube e,nd detonated. Following the formation of the freespan weld (s), a stress relief heater is inserted into the { sleeve / tube and a stress relief. process is performed on the freespan weld. If the tubesheet sleeve is of an all welded design, the i i
-i ' i l i
- i h
DPC/MNS t -Special Report No. 93-06 PAGE 3 installation process is completed at this time with no further action required. Otherwise, the optional tube sheet rolled joint is performed. i Finally the sleeve is inspected using Eddy Current techniques to provide process and NDE records for the tube / sleeve combination. An Eddy Current probe can also be used to rapidly run through the SG tube to i ensure no residual wire remains in the tube. l I [ Thermal Sleeve ( For Westinghouse Pressurized Water Reactors (PWR), thermal sleeves have l been installed at those reactor coolant system branch line nozzles where I mixing of flows with different temperatures exists, such as the nozzles -l of safety injection lines, pressurizer surge lines, and charging lines. -{ Design evolution of the thermal sleeves has resulted in a total of five { generations. f i In 1982, cracking and failure was experienced in thermal sleeve l attachment welds of certain operating Westinghouse reactors that used f the " Generation 3" thermal sleeve design. Subsequently, McGuire Nuclear l Station, Unit 1, was shutdown on June 23, 1982 for purposes of Eddy Current testing (EOT) of all SGs. Pursuant to evidence of the degradation ci thermal sleeve components in .4e reactor coolant system at the Trojan Nuclear Plant, Duke Power Company (EPC) initiated an. i I inspection of all thermal sleeves in the McGuire Unit I reactor coolant (NC) system utilizing radiographic techniques. The NC system contained i the followir, 4 armal sleeves: (4) - 10 inch accumulator nozzle cold leg, (2) - 3 Targing nozzle cold legs, and (1) - 14 inch pressurizer surga t.ozzle hot leg. Radiography of the 10 inch accumulator nozzle thermal sleeves began on l July 1, 1982. On July 5, 1982, the examination of Loop B revealed that l the sleeve was detachvd and missing. On July 6, 1982, the missing j sleeve was confirmed by a remote visual inspection using a small TV camera inserted through the upstream check valve on the 10 inch line. similar inspection of all other connection thermal sleeves verified that .f the snaining sleeves were in place with their welds intact. The l intwe. 'n indicated that the Loop B 10 inch line welds located at the t<. . e4 sleeve had failed at the interface to the nozzle wall with possu t3 ly a small portion of one weld remaining on the nozzle wall. .The nozz.o eall showed no indication that the sleeve had broken apart e
- ~ l l i DPc/MNS l Special Report No. 93-06 l PAGE 4 prior to being e) cased. As a result of the monitoring of the lower Reactor Vessel area for loose parts during low and normal NC flow i conditions, system experts concluded that the missing 10 inch thermal sleeve was 1+>cated in the lower reactor internals.and that the small impacts during low flow conditions were of a minor nature. No movement l was detected when full NC flow was present. This was accepted as objectiva evidence that the sleeve was parked and remained fixed in j place. t i As a result of the discovery of the detached thermal sleeve, a safety l evaluation was performed on the effects of loose and missing NC piping l 1 nozzle thermal sleeves. The results of this evaluation were documented i in the letter dated July 13, 1982, which resides in Nuclear Licensing Services File number Mc801.01. { i In 1983, all remaining sleeves were removed from McGuire Unit 1 and j McGuire Unit 2 started operations without sleeves installed. By License I Condition 2.C (4) of McGuire Unit 2 Facility Operating License NPF-17, the Nuclear Regulatory Commission (NRC) required that "By December 31, } I 1983, the licensee (DPC) shall provide, for'NRC staff review and j approval, justification for continued operation with the seven thermal j sleeves removed from selected locations in the NC system." The required a justification was submitted to the NRC by letter from DPC dated December 14, 1983. I The Westinghouse recommendations and supporting analyses for plants with " Generation 3" thermal sleeves were reviewed by the NRC staff. The NRC ] 5 review concluded that plant operation without thermal sleeves or with ) sleeves removed is acceptable, provided a program is implemented through ) requirements in the Technical Specifications to monitor the occurrence i of injection flow transienLS. and evaluate the fatigue usage factors for i i the affected nozzles. i. Following proposals by DPC to satisfy the above stipulations, the NRC concluded that McGuire Nuclear Station met the conditions resulting from the review of plants with the " Generation 3" thermal sleeve design, and could operate safely with the affected seven sleeves removed from Unit 1 and without thermal sleeves installed in Unit.2. i o__,- e--.
l j j l ? i DPC/MNS .f Special Report No. 93--06 i PAGE 5 l Stainless Steel Chips: l Nuclear Station Modification (NSM) MG l-2343 involved the installation of an addit _onal mini-flow line on each Residual Heat Removal (ND) pump { to protect against strong pump / weak pump induced damage. The installation of this piping employed the wet tapping method for the connections to the main ND piping. { t f Description of Incident i McGuire Unit 1 was taken off line at 0607 on March 12, 1993, for a l projected 84 day refueling outage. The NC system was cooled down and [ depressurized with no abnormal occurrences. On day 7 of the outage, the f Reactor Vessel head was removed and removal of the upper internals and fuel was begun. During defueling operations, video observations of the Lower Core Support Structure revealed wire like objects on the support plate. The bulk of the material, however, was positioned within the l individual fuel assemblies at the lower nozzle and within the.first grid strap. A subsequent inspection also revealed a large, pipe like foreign [ object located below the lower core support. A team of engineers was assemb)ed to review the information, determine the origin (s) of the l foreign material and to resolve concerns associated with the presence of this material. A detailed review of the maintenance, modification, and inspection j history of the NC system, as well as other ECCS components was :cade in an attempt to identify the source of the wire fra pents. The source of the wire was subsequently determined to be.related te the SG tube sleeving installation conducted in the 1EOC7 refueling outage. s t i Inspections were performed on the Reactor Vessel, Lower Internals, Upper j Internals, A, B, C & D Cold Leg NC piping and C & D Hot Leg NC piping. f The inspection equipment consisted of underwater color cameras and 2 'I inch diameter black and white cameras. The color cameras were used for general area inspections of the Lower Internals, Upper Internals and Reactor Vessel. Black and white cameras were used to inspect the LCP, Tie Plates, and downstream side of the Lower Former Plate. A vendor oupplied remote control underwater tracked vehicle (OSCAR) was outfitted with a color camera and an underwater vacuum nozzle to inspect and clean the NC piping sections. j 1
.4 i DPC/MNS Special Report No. 93-06 PAGE 6 An inspection was performed in the Reactor Vessel lower hemisphere. Metallic and non metallic debris was discovered. The non metallic material consisted of a small quantity of paper, paint chips, and rust flakes. Paint chips and rust flakes are acceptable in the typical f quantities found during refueling outages per the station Chemistry Department. Paper, such as small pieces of tape or pieces of housekeeping tags is also acceptable. Debris locations within the Lower Internals included approximately 25 debris items on the Lower Core Support Plate and approximately 7 debris items on the Upper Tie Platt. The 7 debris items on the Upper Tie Plate -i i were not removed due to limited access for underwater vacuuming. An evaluation has determined that no adverse conditions will be experienced as a result of the material remaining in this location. An evaluation was performed on the flow velocities of the Lower Former Plate flow holes. There is sufficient velocity to move debr.is into the annulus between the Lower Former Plate and the number 2.-Former Plate. Once in the annulus area, the flow velocity significantly reduces. If j debris was found on the Lower Former Plate, the debris would remain trapped there due to the low flow velocities. Four locations on the i axes of the Lower Internals were chosen for inspection above the Lower Former Plate. No debris was identified during this inspection. t Inspection of the Upper Internal Guide Tubes was considered due to the potential affects of control rods scramming the reactor. The major i concern was the possibility of debris entering the guide tubes and affecting movement of the control rods. It was concluded that an j inspection of the Upper Internals Guide Tubes was not necessary, since no debris was found above the second fuel grid strap. The Upper Internals did, however, receive a routine inspection of the top side of the Upper Core support Plate and a random inspection of the Upper Core from the top side. The results of the inspection were satisfactory and no additional debris was found. i Inspection of the NC piping was made after flushing the ND system. The flush was initiated in response to suspected drill shasing material in the Residual Heat Removal (ND) system. No debris was found. Additionally, inspections of all B SG bowls upon initial opening showed i - + - -.,
+ l i DPC/MNS Special Report No. 93-06 PAGE 7 the bowls to be clean. No other locations outside of the Reactor vessel, Internals,-and Fuel were inspected because high velocities within the NC system made finding debris outside of the Reactor Vessel f highly unlikely. Cleaning of the Reactor Vessel, Lower Internals, Upper Internals, A,B,C & D Cold leg NC piping and C & D Hot leg piping was accomplished by underwater vacuuming and/or mechanica] removal. Cleaning was not deemed necessary in the NC system outside of the Reactor Vessel. i A large cylindrical object was identified just below the LCP while searching for debris in the bottom of the Reactor Vessel. An initial inspection of the object showed it had the same general geometry as a thermal sleeve that had been missing since 1982. A detailed inspection and sizing of the object could not be performed due to the location and t orientation of the object. Subsequent field measurements and inspections proved that the object matched the geometry of the thermal i sleeve. This cylindrical object was determined to be the missing thermal sleeve. The original missing thermal sleeve was from the Loop B { cold leg ND 10 inch injection nozzle. l The thermal sleeve was located just below core location K-7, wedged -j between Butt columns J-7 and K-6. The LCP has 9.5 inch dismeter holes j below half the core locations. As seen from above, the thermal sleeve profile viaually appeared to block appr Jximately 60% of the LCP K-7 flow hole. The other end of the thermal sleeve profile completely blocked the J-6 location. The thermal sleeve in this orientation did not prevent flow from entering the flow holes due to the thermal sleeve cylindrical profile. The thermal sleeve was cocked approxLmately 10-15 degrees from the horizontal with the lower end at K-7. The ti;ermal sleeve, having a diameter of approximately 8.5 inches, was wedg?d in a space approximately 8 inches. A vendor was contracted to provide the thermal sleeve removal services. The vendor provided a tool with a long arm equipped with a flexible wrist. The wrist had a two degree range of motion in the vertical and horizontal planes. An expander was attacued to the wrist. The expander consisted of three 10 ton hydraulic jacks mounted between two plates which were aporoximately 2.5 inches wide. The outer surfaces of the i w ws, 1
4 DPC/MNS Special Report No. 93-06 PAGE 8 plates were profiled to match the curvature of the thermal sleeve and apply a uniform load while minimizing the propagation of the cracks. The arm was mounted on a vertical column. The column had a vertical trasel of 12 inches and could change the angle of the arm +/- 30 degrees. The vertical column was mounted on,a "V" shaped platform with wheels at the end of the legs. The base ofI ae column and platform were attached to a custom saddle that mounted on CAR. The vendor performed 5 qualifications tests. The tests were designed to determine the load required to "ovalize" a pipe to a shcrt diameter of less than 8 inches, to determine the strength of the expander, and to determine if the cracks would propagate. Test results concluded that under the loading required to ovalize the thermal sleeve sufficiently, the cracks would not propagate. The arm was positioned to enter a narrow path between core location F-1 and E-2. The hydraulic expander was inserted into the thermal sleeve, expanded and removed. The thermal sleeve was removed and deposited on the refueling canal floor and remained there until an appropriate removal time. The contact radiation doce rate of the thermal sleeve was approximately 7 Rem. The location of the area whers the thermal sleeve was parked was video inspected by quality Assurance. The results of the examination indicated that no damage had occured as a result of the thermal sleeve and the as found condition of the area was acceptable. The pinch point of the two Butt columns was discolored but did not appear fretted or degraded. A bolt at core location K-7 was in contact with the thermal sleeve but was not damaged. The pinch points of the Lower Internals were evaluated and found acceptable. While initially searching for the source of the Reactor Vessel debris by reviewing the ECCS, a routine inspection of the Containment Spray (NS) Heat Exchangers (Hx) identified a number of small stainless steel chips. An in depth review of this finding revealed the existence of a flow path to the NS system from the site of a modification on the ND system. The modification was implemented during IEOC7 using a " wet tap" prccess. The modification involved the installation of an additional mini-ficw i
9 4 DPC/MNS Speci.al Report No. 93-06 PAGE 9 y line on each ND pump to protect against strong pump / weak pump induced ~ damage. Due to the possible existence of chip material in the piping leading to l the Cold and Hot Legs, a conservative decision was made to restore the drained B Train of ND to service and conduct a high flow flush of both trains while the Reactor Vessel Lower Internals were removed for the 1 thermal sleeve removal effort. This decision took advantage of the removal of Lower Internals since the Reactor Vessel Cold Legs could be [ inspected with them removed. Temporary flush procedure *T/1/A/9700/99, I was developed and implemented. Babcock and Wilcox provited services to inspect all four Cold Legs and B & C Hot Legs. No chip material or j debris of any kind was found. This provided assurance that no material was transported to the NC system where it could cause fuel failures. Additionally, the likelihood of future migration of _ y chip material not removed in the subsequent clean up was greatly reduced. In 1 addition, since only a small amount of material was found in the 1B ND Hx, it was decided to forgo inspection of the 1A ND Hx. This decision l 2-was based in part on the fact that material present in that Ex inlet bowl would not migrate since no migration had occurred during the high flow flush. i Four debris items / pieces were found on the LCP after fuel unload. The j items were unique from the general debris found distributed over the 4 LCP. Item one was 1.5 inches long and 0.625 inches in diameter, made of ] two conical shapes joined at their bases with the longest cone at 1.0 l inches. Item two was 0.5 inches long and 0.1875 inches in diameter, shaped like a hollow cylinder with a wall thickness of approximately i 0.03125 inches. The third item was 0.625 inches long. A O.625 inch diameter thin wall hollow cylindrical cup with a 0.625 inch long by 4 0.125 inch diameter hollow rod was attached to the outside center base of the cup. Item four was roughly wedge shaped and was approximately d 0.5 inches long, 0.5. inches wide and 0.250 inches at its thickest point. l l The four debris items were found in different quadrants on the LCP. 8 The four debris items were removed from the LCP. The items were initially scheduled to be forwarded off site for analysis. It was then decided to store the items in the Spent Fuel pit in a retrievable basket in a spent fuel trash container. Further discussions have lead to -ya+- -~-? J 9e m
= DPC/MNG Special Report No. 93-06 PAGE 10 decision that off site analysis of the items would be of limited value i since each item is unique and share no commonality. Additionally, the wedge shaped item would be unshippable, due to radiological concerns. l f OONCLUSION: A cause of Design Deficiency, Fabrication Deficiency, (Vendor) is assigned to this incident because the selection of the short radius U bend tubes employed during procedure qualification was not, in fact, the j limiting case. The pieces of copper wire found in the NC system originated from SG tube sleeving conducted during the IEOO7 refueling outage. Debris is routinely generated as a result of the tube sleeving process. Wire conductors were used to carry an electrical signal to the j detonators. The conductor wire gauges ranged from 20 to 22 AWG which j correlates to the wire debris diameters of 25 and 32 mils. The wire i pieces created during the kinetic welding process varied in length from 0.5 to 12 inches. Most pieces were in the 3 to 6 inch range which again corresponds to the general debris lengths. i Qualification testing conducted by BWNT on the assumed worst case U j tubes (inner rows / tighter U bend) showed that the material above the l bottom charge was ejected to the opposite SG bowl where it could be j retrieved after all sleeving was completed. The material around the bottom charge was pushed downward into the plastic insertion tube used t to position the charges. This tube, with the wire fragments and associated welding debris, was then removed from the SG tube. It now is concluded that the worst case U tubes are those on outer rows where the i U bend has the greatest length and these tubes should have been employed in the qualification testing. The wire debris was found in several locations on the Lower Internals and in the Reactor Vessel. The debris was initially found on the LCP at the completion of fuel unload activities by routine video inspection. This debris was randomly distributed over the entire surface area of LCP, between the 2.75 inch flow !. oles. A small percentage of the total debris was located on the LCP and Upper Tie Plate. Relatively small quantities of wire and other debris were also noted in the lower id -w w
~ DPC/MNS Special Report No. 93-06 } PAGE 11 I hemisphere of the Reactor Vessel. Based on the assumption that the copper wire debris found on the fuel' f assemblies and LCP resulted from SG tube sleeving during the lEOC7 l i outage, the wire would have originated in the SG tube U bends. Upon -l initial filling and venting of the NC system, individual bumping of NC pumps to vent the U tubes would have moved the wire down the tubes to l i the SG Cold Leg bowls. Smaller reverse flows in opposite loops may have initially moved some wire pieces backwards into the SG Hot Leg bowls, and perhaps into the Hot Leg during single NC Pump bumping; however, it is judged that most wire pieces would initially have been pushed to the l Cold Leg bowls, into the Crossover Leg and through the NC Pump bumped j for venting. Normal NC venting consists of 20 second bumps, and 1 and 5 i minute runs. i once all four NC pumps were in simultaneous operation, the wire pieces j j would have easily been swept towards the Reactor Vessel Lower Internals and Fuel Assemblies which acted as efficient filters to catch the vast majority of this debris, as evidenced by inspection. During the 1EOC7 refueling outage, initial kinetic weld sleeving of the { first 10 rows of sleeves processed was followed by "freepathing" ( the passage of an Eddy Current probe through the tube) in order to verify { that no wires remained in the SG tubes following welding. No' debris was j found which confirmed the development test results. It has been generally concluded that a substantial amount of debris did clear the U-bends on the outer row tubes since a large amount of debris was cleaned l from the opposite SG bowl during this work. However, based on fuel and { LCP inspections, it was concluded that approximately 300 feet of wire (in pieces) was left in the SG U bends. Calculations performed by i Engineering personnel indicated that this amount correlates to the l amount of debris from roughly 30% of the sleeves. j t i 4 The majority of copper wire that went into the NC system was ultimately I deposited on the fuel assemblies. Detailed inspections and analysis ino.cated that the copper wire did not impact fuel rod integrity. A much smaller amount will return to the core for cycle 9 and is not expected to impact fuel rod integrity. l 4 i
I i DPC/MNS Special Report No. 93-06 PAGE 12 j Residual copper wire debris will not cause fuel defects. Other non-l s copper debris could cause defects if present, but the amount of non-q copper debris for cycle 9 should not exceed the typical very small level i of metallic debris within the NC system for any cycle. Also, RCCA j motion in relation to fuel assembly guide thimbles will not be impaired l l by copper debris. This cause is also assigned to the portion of this incident involving i the missing thermal sleeve due to the unanticipated cracking of the thermal sleeve weld resulting from high cycle fatigue caused by flow t induced vibrations ( see letter dated December 14, 1983 from H.B. Tucker i to H.R. Denton [ Nuclear Licensing Services File Mc801.01] for discussion). l A cause of Inappropriate Action, Inadequate Work Practices is assigned l to the portion of this incident involving the miscellaneous debris j discovered in the Reactor Vessel because personnel failed to exercise adequate control of tools and materials in and around established { cleanliness zones as required by approved station policies / directives. Four debris items / pieces were found on the LCP after fuel unload. These items are unique from the general debris found distributed over the LCP. { The source of the debris items is unknown. The four debris items were found in different quadrants on the LCP. k i The four debris items were removed frem the LCP by mechanical means. The items were initially scheduled to be forwarded off site for l analysis. It was then decided to store the items in the Spent Fuel pit f in a retrievable basket. Further discussions have lead to decision that off site analysis of the items wnald be of limited value since each item f is unique and share no commonality. Additionally, the wedge shaped item would be unshippable, due to radiological concerns. f l A cause of Installation Deficiency, Deficient Communicatirn is assigned to the portion of this incidenc involving the introduction of drill j chips into the ND Hxs because personnel who implemented NSM MG 1-2343 had been advised that the system was under pressure when in fact it was not. Had the individual been aware of the actual system condition, appropriate steps would have been taken to prevent introduction of drilling residue into the system, as evidenced by the workers actions on l subsequent drilling operations. -l t r-m
l l i DPC/MNS Special Report No. 93-06 [ PAGE 13 During initial investigation to determine the source of the debris found in the Reactor Vessel and fuel assemblies, System Engineers for the ECCS systems reviewed past modification and major maintenance work which ~ + might have created the " shaving like" debris. One modification l completed at the 1EOC7 refueling outage was identified as a source of l shavings. Discussions with personnel involved indicated that drill chips were likely left in the ND system. l l i Nuclear Station Modification MG l-2343 involved the installation of an additional mini flow line on each ND pump to protect against strong pump / weak pump induced damage. The installation of this piping included i t wet tape for the connections to the main ND piping. Later, formal j discussions with the crew that completed the work, as well as technical l I support personnel indicated that much of the chip material was left in l the ND system based on the following summary of events. The first hole was wet tapped on the discharge of the IB ND Pump. The i crew appropriately assumed that the pressure in the ND system would be j sufficient to push back the drilling material. When they breached the 8 j inch ND line, they reported hearing air versus water. After completing-f the drilling, closing the gate valve and removing the wet tap rig, they { realized that most, if not all drill shavings did not come cut _with the rig. The crew reported that they opened the valve carefully and { inspected the tap but saw few chips. Little water reportedly came out. It is assumed from this description that all drill shavings from the 1-l 7/8 inch diameter hole in the 1/4 inch pipe wall remained in the system. l The second hole was drilled on the suction of the IB ND Pump. Since j there was no pressure present on the first hole, the crew removed the wet tap rig several times to scrape shavings out of the pipe before the i ND pipe wall was breached. Once the drill penetrated the ND line I though, no more chips were retrieved. This hole was also 1-7/B inches j in diameter. The remaining 2 holes on A Train were completed like the second hole on B Train with craft personnel scraping out shavings until 1 the ND line was breached. ] Several important points to note are:
- 1) The drilling process for wet-tapping is solely manual; therefore, chips and shavings produced are small.
Long spirals, as produced by a drill press are not produced in wet tapping.
- 2) All holes drilled were 1-7/8 inches in diameter.
i ) l
-m l j t DPC/MNS Special Report No. 93-06 PAGE 14 l The best estimate of total chip volume left in ND is as follows: i L Trains (1st hole) 2.76 x 0.25 wall = 0.69 cu in j (2nd hole) 0.69 cu in .096 cu in (cone volume of material removed before breaching pipe wall assuming 135 degree bit) = 0.594 cu in A Train: Total = 2 X O.594 = 1.19 cu in of roughly 1 inch cube 4 made into small chips. It is noted that the Unit 2 equivalent of this NSH was done in two ways. One train was wet tapped with a 1 inch drill but crews felt that most chips flushed out since ND was pressurized. Later, on this same train, [ the hole was enlarged using a 1-7/8 diameter hole saw. However, this l work was done with the system drained and the " donuts" created were retrieved. The second train had holes made with a 1-7/8 inch hole saw with the system drained. Both donuts were retrieved. Based on this account, no similar concerns of drill shavings (of any significance) left in Unit 2 ND exist. Before debris was reported in the Reactor Vessel, Component Engineering l personnel reported to System Engineering personnel that chip / shaving like material was found upon initial opening of both NS Exs. This material was bagged for future inspection. The material was located i primarily on the inlet bowl of each Hx and was described as "about one handful in volume". f i t A portion of this material was sent to UWNT for analysis along with a debris sample from the Reactor Vessel. The analysis showed the low f activity chip material to be Stainless Steel 304, the type used in ND piping. High activity material from the NS Hxs was likely radioactive due to resin in the sample. Based on evidence that drill shavings / chips existed in NS and likely ND, a decision was made to inspect the ND 1B Hx, which was drained at the time. Inspection was made through the manual, upstream isolation valve. Video inspection of the inlet bowl showed some chip material (several larger pieces and 8 to 20 smaller pieces). Later af ter tha Hx was filled, small chip material was discovered to have migrated into the Hx drain valves and prevented tight isolation of those valves. Eventually, 4 E 't
DPC/MNS 9pecial Report No. 93-06 PAGE 15 I the Hx was drained again and valves were repaired / replaced. One item of planned work on ND B Train was to repair 2 valves on the auction of the IB ND Pump. After disassembly of IND-45, debris was reported lodged in the valve body near'the seat opening. This valve is a Kerotest packless globe valve and the material could not be easily removed. A flush of the valve body removed lodged material. Subsequent inspection showed at least one piece of the material to be a drill a shaving. Based on location of this drain tap relative to the wet-tap on the suction side of the pump, migration to this point was highly predictable. The transport route from the ND system to the NS Hxs was studied and it has been concluded that a probable path exists. When ND was placed in l service after modification, procedure TO/1/A/4206/09 was used to sweep air from most of the ND system (and especially the unventable U tubes in the ND Hx). This procedure utilizes IND-35, a nanual isolation valve which allows ND to recycle to the 24" Refueling water (FW) header. i J ? Chips moving into the FW header would be swept back down towards the ND pump suction; however, instead of turning back towards the ND pumps through 1FW-27A, they likely tumbled out of the flow stream down towards l the suction line leading to the NS pumps. Subsequent running of the NS pumps would have moved the chips into the NS Exs. To provide additional information about the potential for this drill chip material to migrate in the system, a test rig was designed to measure the flow velocities required to move various sizes of this debris. Several holes were drilled in stainless steel plates to simulate the chip material produced. Testing was conducted to visually determine at what velocities various debris moved horizontally and f vertically. The testing showed that for smaller drill chips, a velocity of 4.5 ft/sec was required to move the material vertically. Larger chip shavings required less velocity, with the largest requiring only 2 ft/sec for movement vertically. Thin wire strands (close ir diameter to that found in Reactor Vessel) were also tested at this time and they required 3 ft/see to migrate vertically. 4 Velocities in the ND, FW, NS and NC systems were calculated at various i
~ e DPC/MNS Special Report No. 93-06 PAGE 16 1 operational and testing flow rates. This information demonstrated that velocities were suf ficient to move chip debris throughout the system. However, due to the known system evolution immediately after the wet tapping modification was complete, it is concluded that most chip material either wenu into the ND Hxs or through IND-35 wnere it was moved to the suction of the NS pumps. This theory is based on review of procedure TO/1/A/4206/09 which is used to startup the ND system after it has been drained for outage related work. In this procedure, the ND Hx cross connects (ND-18, 33, 34) downstream of the pump (s), which are aligned open, the Cold and Hot Leg injection isolation valves are i aligned closed, and recirculation valve IND-35 is aligned open. The pump is started against a tightly throttled pump dincharge valve and flow is then slowly increased to minimize water slugging the downstream portion of the system, until full pump flow is reached. In addition, a review of the likelihood of debris migration towards the 3 f " piggy back" isolation valves 1ND-58A and INI-136B was conducted. Since the ND pump mini flow lines connect from these lines, flow is usually present during pump startup and until mini flow closes off.
- However, 6
this flow rate is less than 400 gpm with corresponding velocities of 2.5 ft/see which are marginal for moving all but the largest chips vertically. The largest chips would not pass through the ND Hx since Hx tube diameter of 0.625 inches would screen them. Later use of these lines during ECCS pump testing produced flows of 600-700 gpm through each line; however, this testing was conducted well after system startup described above and a period of normal ND system operation. Therefore, it is concluded that no debris of any consequence migrated into these 8" lines supplying the NI and NV pump suctions. 2 Chip material was removed from the ND IB Hx by vacuuming. Chip mat 3 rial was removed from the pump suction drain valves and Hx drain valves either through valve repair or replacement. All chip material and 4 debris was removed from both NS Hxa. f ( i i
~ ~ -- ) i DPC/MNS Special Report No. 93-06 } PAGE 17 l S. CORRECTIVE ACTIONS: Immediate: None Subsequent: 1. Inspections of the LCP, Lower Vessel, and fuel assemblies were performed by System Engineering and l Operations personnel to determine the extent of l
- debris, j
2. Efforts were initiated by System Engineering, Operations, and vendor personnel to recover and identify the various debris items including the i thermal sleeve and miscellaneous foreign objects in addition to the copper wire fragments. l 3. The majority of copper wire was " cleaned" from fuel assemblies by System Engineering, Operations, and j vendor personnel. I 4. Efforts were initiated by System Engineering personnel to determine the source and extent of 304 Stainless Steel chips discovered in the ECCS system. f S. Operations personnel conducted a high velocity flush t of the ND system. 3 I s i Planned: 1. Site Directives will be revised to incorporate the j INPO " Good Industry Practice" of temporary covering any system opening that is left unattended regardless ? a of cleanliness level specified. This revision will also strengthen material accountability in and around open systems. i 1 l i y -
- ~. '..,4 [ t D[*C/MNS Special Report No. 93-06 PAGE 18 7 2. Mechanical Engineering personnel will evaluate and implement the following actions as appropriate: l A. Create a Foreign Material Exclusion (FME) procedure as an attachment to each work order involving pipe / equipment and/or system openings. i B. Investigate special tooling requirements during modification design phase to prevent foreign .{ material induction as part of the design consideration. C. Make FME considerations part of the modification f 10 CFR 50.59 evaluation. i i NOTE: During the timeframe corresponding to 1EOC7, BWNT experienced occasional difficulty inserting { heaters in the sleeves following welding. The insertion problems occured on both the single and double weld designs. The problem was confirmed to be caused by a small amount of acrylic from the weld device remaining in the cleeve following welding. The tight clearance between the heater and sleeve made insertion with this acrylic difficult. The temporary fix for this problem was to run the heater at a low temperature to dissolve the acrylic, then proceed with normal insertion. A subsequent I decision was made to freepath all sleeves / tubes to ensure that this acrylic was removed. Therefore, beginning with the January 1992, l McGuire Unit 2 sleeve installation, all 1 tubesheet sleeves were freepathed. As e result, the only installation of these sleeves without freepathing was the 1EOC7 and similar debris is 4 not expected to be found in Unit 2.}}