ML20247F819
| ML20247F819 | |
| Person / Time | |
|---|---|
| Site: | Limerick  |
| Issue date: | 03/29/1989 |
| From: | Kemper J PECO ENERGY CO., (FORMERLY PHILADELPHIA ELECTRIC |
| To: | NRC OFFICE OF INFORMATION RESOURCES MANAGEMENT (IRM) |
| References | |
| NUDOCS 8904040050 | |
| Download: ML20247F819 (67) | |
Text
, -- - - - - - - - - - - _ - - - - - - - - - - --
=
PHILADELPHIA ELECTRIC COMPANY 2301 M ARKET STREET P.O. BOX 8699 PHILADELPHIA. PA.19101 (215) 841 4 So O
.... 3,t"tha'imm.,
MAR 291989 U. S. Nuclear Regulatory Conmission ATTN: Doctsnent Control Desk Washington, D.C. 20555
Subject:
Limerick Generating Station, Unit 2 Independent Construction Assessment, Inspectton Report No. 50-353/88-202
Reference:
Letter from G. C. Lainas to G. A. Hunger, Jr.
dated January 23, 1989, same subject Gentlemen:
Your. letter of January 23, 1989, forwarded the results of the NRC inspection of the Limerick 2 Independent Construction Assessment CICA) J being performed by Stone & Webster Engineering corporation'(SWEC).
Further, tiie referenced letter identified a ntsnber of discrepant conditions which PECo was requested to address.
Provided as attachments to this letter see PECo respoases to each of these discrepancies. Where the discrepancy fell within the scope of the SWEC.ICA, a reference is provided to the applicable ICA observation report nurber. For discrepancies outside the scope of the ICA, a complete response is provided to be consistent with the IDCA process for Unit 2. The responses to the NRC discrepancies have been fonnatted similarly to the SWEC IDCA observation report format.
The appilcability of the identified discrepancies to Limerick Generating Station Unit I will be addressed in a separate submittal.
Very truly yours,
$Fgf
/
WWP/sjf/3288901 Attachment Copy to: T. J. Kenny, LGS Senior Resident Inspector .
W. T. Russell, Region 1 Administrator R. J. Clark, LGS Project Manager M
I \
8904040050 890329 PDR ADOCK0500%@3 Cl
NONDESTRUCTIVE EXAMINATION (LK-C-1901)
SECTION 4.2.2.2 Concern The NRC team reviewed the radiographs for LPCI N17B-255 and LPCI N17B-315, nozzle-to-safe-end welds. The team verified root concavity indications on these radiographs that SWEC identified during their Construction Assessment. SWEC documented their findings in CAI-022. G.E. responded that the root concavity was found acceptable after a boroscope inspection.
However, t.,7FE cod Section III, NB-4000, required that root concavity j be evaluated by .22m density comparison with the base material. G. E. l should review the subject radiographs and evaluate the root concavity in accordance with the ASME Section III requirements.
Response
Concur With Concern This concern has been resolved in accordance with Stone and Webster l Construction Action Item CAI-022 and Construction Observation Report l COR-07. i 1
NONDESTRUCTIVE EXAMINATION (LK-C-1901)
J SECTION 4.2.2.3 Concern SWEC assessed the radiographs of the nozzle-to-shell welds for the "B" ;
loop RHR heat exchanger. They identified film densities that did not l comply with the density requirements set forth in ASME Section VIII, j UW-51. The density requirements stated in part that the density in the area of tnterest (i.e. the weld) shall be within -15 percent to +30 percent of the density through the penetrometer. The NRC team reviewed the radiographs of the nozzle-to-shell welds for both RHR heat exchangers, welds 35002TlN1 and N2, and 35002T2N1 and N2. The team verified SWEC's finding and identified the same condition with the "A" RHR heat exchanger radiographs.
Failure to use adequate shim thickness under the penetrometer or to use more than one shimmed penetrometer was the root cause of the problem.
A secondary contributing factor was the low density in the weld area caused by the weld reinforcement. Because of the large variation in density between the welds and the penetrometer, it was impossible to assess the quality of the welds.
Response
".oncur with concern This concern has been resolved in accordance with Stone and Webster Construction Observation Report COR-35.
NONDESTRUCTIVE EXAMINATION (LK-C-1901)
SECTION 4.2.3 Concern The NRC team concluded that SWEC's findings in the RHR system, and in particular those findings concerning the LPCI nozzle-to-safe end welds and RHR heat exchangers, may be generically applicable to similar components elsewhere in the plant.
1
Response
Concur with Concern The generic implications of these issues have been addressed in our response to Stone and Webster Construction Observation Reports COR-07 and COR-35.
h i.
i i
1 l
MECHANICAL COMPONENTS /HVAC SYSTEM (LK-C-1902)
SECTION 4.3.2.1 Concern Three of four anchor bolts for the lower support columns on RHR heat exchanger 2A did not have washers as shown on anchor bolt Drawing C-615.
Response
Concur With Concern Cause The type 'C' anchor bolt detail on Drawing C-615, Sheet 1, required the use of a washer. Detail '3' on Drawing C-196, which shows this connection, did not indicate a washer.
Extent of Condition This is addressed in our response to the generic concern later in this section regarding the fact that various pumps and unit coolers did not have jam nuts installed in accordance with Note 3 on Bechtel Design Drawing C-626, " Reactor Building Equipment Foundation".
Significance A review of the loading at the bolts per the design calculation, as well as the standard hex nut dimensions and bolt hole size concludes that adequate contact area is provided without the washers to transfer the loads for these bolts.
Corrective Action Clarification of this issue has been provided through the issuance of Field Change Request CC-1561-C which makes the use of washers for anchor bolts AB-93C1 optional.
Action to Prevent Recurrence This is addressed in our response to the generic concern later in this section regarding the fact that various pumps and unit coolers did not have jam nuts installed in accordance with Note 5 on Bechtel Design Drawing C-626, " Reactor Building Equipment Foundation".
l l
l _ _ _ _
i I
MECHAWICAL COMPONENTS /HVAC SYSTEM (LK-C-1902)
SECTION 4.3.2.1 Concern Three of four anchor bolts for the lower support columns on RHR heat exchanger 2A did not have washers as shown on anchor bolt Drawing C-615.
Response
Concur With Concern Cause The type 'C' anchor bolt detail on Drawing C-615, Sheet 1, required the use of a waoher. Detail '3' on Drawing C-196, which shows this connection, did not indicate a washer.
Extent of Condition This is addressed in our response to the generic concern later in this section regarding the fact that various pumps and unit coolers did not have jam nuts installed in accordance with Note 5 on Bechtal Design Drawing C-626, " Reactor Building Equipment Foundation".
Significance A review of the loading at the bolts per the design calculation, as well as the standard hex nut dir.iensions and bolt hole size concludes that adequate contact area is provided without the washers to transfer the loads for these bolts.
Corrective Action Clarification of this issue has been provided through the issuance of Field Change Request CC-1561-C which makes the use of washers for anchor bolts AB-93Cl optional.
Action to Prevent Recurrence This is addressed in our response to the generic concern later in this section regarding the fact that various pumps and unit coolers did not ,
have jam nuts installed in accordance with Note 5 on Bechtel Design Drawing C-626, " Reactor Building Equipment Foundation".
MECHANICAL COMPONENTS /HVAC SYSTEM (LK-C-1902)
SECTION 4.3.2.1 Concern Paragraph 4.4.7.1.1 of General Electric Company (GE) Design Specification 21A9227AT for the RER heat exchanger stated that installation requirements shall include the following requirements for bolts: " Clearance...to allow for radial thermal expansion of the heat exchanger". Note (1) of Figure 1 in the vendor (Delta Southern Company) technical manual indicated that the difference between support and mating structure bolt circles was to allow '
clearance for thermal expansion. However, GE and Bechtel design drawings for the lower support and Bechtel installation work plans did not specify any required clearance. Consequently, the existence or amount of any clearance in the installed position was unknown. Zero clearance could j result in overstress or failure of support structure components during heatup. This concern must also be addressed for Unit 1 installations.
Response
Do Not Concur With Concern The design drawings showed the location of the bolt centered in the over-sized bolt hole, but did not specify any required clearance or tolerances.
To aid in the evaluation of the cited condition, an inspection was performed of the stud-to-hole clearances for both Unit 2 RHR heat exchangers and the as-built dimensions provided to Project Engineering. Calculations were performed assuming a conservative 0-inch clearance between all the hold down bolts and heat exchanger feet. The worst case reaction loads on the structural steel have been evaluated and the conclusion is that the support steel and the bolts remain functional when assuming 0-inch clearances.
During the review of this concern, Engineering questioned clearances between the heat exchanger foot and lateral bumpers. Construction Engineering provided as-built dimensions to Project Engineering which indicated that the Unit 2 bumper gaps were larger than those used in the support steel design. Calculations were performed to evaluate the larger l gaps whereby some studs resist the lateral seismic loads in addition to thermal expansion loads.
Engineering' assessment, based upon an elastic analysis for inelastic behavior (i.e.: thermal and seismic stresses absolutely summed without credit for self limiting stresses caused by yielding) concludes that the RHR heat exchanger support steel remains functional since the steel remains within the ultimate capability of the framing system. However, in order to satisfy the original design calculation (without inelastic behavior),
shims have been added to the Unit 2 heat exchangers to limit the gap between the heat exchanger foot and the lateral bumpers.
i i
MECHANICAL COMPONENTS /HVAC SYSTEM (LK-C-1902)
SECTION 4.3.2.1 Concern Additional loads from attached pipe supports and electrical equipment were not reflected on the design drawings for the lower support structure for either RHR heat exchanger. The NRC inspection team inquired about the existence of an alternate method for assuring that designers would be aware of all loading on these structures for future calculations.
The licensee indicated that no such methods were in use.
Response
Do Not Concur With Concern Generally, in the early stages of the structural design of power plants, principal loads for major equipment and structures are reasonably well defined. Other loads,-such as pipe supports, are not known with any degree of accuracy. Therefore, the design process generally allowo for these loads either by conservatively estimating for various additional loads or by selecting member sizes slightly larger than the required for the known loads. This approach is necessary since final locations and exact design loads are unknown for the pipe support loads until their detailed analyses are performed in later stages of design. Sheets 16 and 38 of Calculation 101.73, Revision 1, show that significantly larger steel sections were used to allow a margin for additional pipe supports and other loads.
Limerick Project Specification 8031-P-401 (Design Criteria for Design and Documentation of Pipe Supports and Restraints for Pipe 2-1/2 inches and Larger) and 8031-P-366-2 (Specification for As-Built Reconciliation) address the coordination of pipe support loads with the civil discipline.
Specification P-410, Section 4.2.4.2, states that all large pipe hangers shall be coordinated with the civil group via a coordination print.
Specification P-366 addresses the procedure for coordination of small pipe hangers with the civil group. Section 5.4 of Specification P-366-2 states that small pipe supports need to be coordinated only when the attach-ments are to the members W10 or smaller, including channel sections.
However, it further states that the pipe support group shall exercise good engineering judgement to coordinate other attachments when it appears that there may be some impact to the civil structure.
I i
I l
Methods do exist for the coordination of multiple commodity supports as discussed in Project Specifications 8031-G-21 and 8031-G-28. In addition, a comparison of heat exchanger reactions and existing pipe attachment loads clearly indicate that the design engineers exercised sound engineering judgement because the incremental load in terms of percentage of the original load is about 1%. It is important to note that in cases of significant pipe support loads, civil calculations were made to in-
'estigate their effects on the structure. Another key consideration is to view the impact of small pipe support or conduit support reactions in their true relative importance. The RHR heat exchanger supports are not specifically shown to be designed for these small attachment loads, but the calculations do not ignore them; they are implicitly considered in the design process.
l l
l l
l l
l
_ - _ _ _ - _ _ _ _ _ _ _ _ _ . _________ _ i
MECHANYCAL COMPONENTS /HVAC SYSTEM (LK-C-1902)
SECTION 4.3.2.1 Concern Piping and arrangement design drawings for the RHR pump shaft seal assemblies indicated that the drain port was to be plugged with a steel plug. However, che NRC inspection team found that plastic shipping plugs were actually in- l stalled in pumps 2A, 2B, 2C and 2D. Subsequent discussions between Bechtel Engineering and GE indicated that in addition to removing the shipping plugs, drain lines should have been piped to radwaste drains. A Field Deviation Dis- ,
I position Request (FDDR) was generated to request design details for this piping and for approval to continue startup testing with the present drain port configu-ration. This concern must be addressed for Unit 1 installations also. The licensee initiated NCR 13878 and FDDR Request M1102 in response to the teams
. finding.
Response
concur with concern Cause The cited condition is the result of a Construction Engineering and Quality Control oversight with recpect to one attribute of the overall RHR pump seal installation and inspection process.
Extent of Condition The condition is limited to the four Unit 2 RHR pumps. The Unit 1 RHR pump shaft seal assembly drain ports contain steel plugs per current design require-ments. The Core Spray pump's seal is similar in design to the RHR pumn and both were manufactured by Ingersoll Rand and supplied by General Electric. .'he Unit 1 and Unit 2 Core Spray pump seal drain connections are piped to a floor drain in accordance with Bechtel design documents.
Significance The seal flush and drain lines are only required to be used in a corrosive environment to maintain clean water in the seal. The RHR pump shaft seal fluid is clean demineralized water; therefore, clean water is maintained in the seal and the plastic plugs in the seal drain connections would not adversely affect '
pump performance, function, reliability or safety.
Corrective Action NCR-13878 has been issued and dispositioned to install steel plugs in the Unit 2 RHR pump seal drain connections per the vendor manual; this work is complete and the NCR is closed. G.E. FDDR HH2-8772 does not require drain lines piped to radwaste.
Action to Prevent Recurrence Both Mechanical Construction Engineering and Mechanical Quality Control have received formal training to reemphasize the need for manual / drawing review prior to performing field installations and inspections. Additionally, per Quality Control requirements, a Corrective Action Report was issued against the respon-sible Quality Control Engineer who performed the inspection and acceptance documenting probable cause and resolution.
MECHANICAL COMPONENTS /HVAC SYSTFM (LK-C-1902)
SECTION 4.3.2.1 l
Concern Bechtel Design Drawing C-631, " Equipment Foundations", specifies the use of jam nuts or staked threads on the foundation bolts for the standby liquid control (SLC) pumps. Jam nuts were not installed, nor were threads staked on any of the foundation bolts on SLC pumps 2A, 2B, and 2C. The licensee initiated NCR-13877 based on this NRC finding.
Response
concur with Concern The cited condition is addressed in our response to the generic concern later in this section regarding the fact that various pumps and unit coolers did not have jam nuts installed in accordance with Note 5 on Bechtel Design-Drawing C-626, " Reactor Building Equipment Foundation".
MECHANICAL COMPONENTS /HVAC SYSTEM (LK-C-1902)
SECTION 4.3.2.1 Concern Washers were not installed on any of the eight attachment bolts for unit cooler 2FV211 as shown on the American Air Filter Final Assembly Design Drawing 107D-153890-G. In addition, two of the angle clips through which these bolts were mounted had been enlarged, as authorized by a Field Change Request (FCR). Standard structural practice as detailed in the American Institute of Steel Construction (AISC) Manual dictated that washers be used under nuts wherever holes are exposed. Also, one of the vibration absorber pad mounting bolts for the fan unit was not fully engaged. NCR-13906 was initiated by the licensee to document and resolve this discrepancy.
Response
Concur With Concern Cause The cited condition is attributable to an oversight on the part of Con-struction Engineering and Quality Control.
Extent of Condition All safety related unit coolers (20 total) were subsequently reinspected for similar deficiencies; 3 were found to be missing washers and 3 absorber pad mounting bolts were found not to have full thread engagement.
Significance (1) Regarding Missing Washers:
American Air Filter (AAF) drawing 107D-153890-G, Section AF-AF specified 3/4" plain washer (Typ) by AAF. In general, the use of a washer as specified in the AAF drawing is a good engineering practice. However, the necessity for the use of a washir depends on the details of a connection as related to size of hole, bolt type, and bolt pretension requirements.
The specifications for Structural Joints using ASTM A325 and A490 (AISC manual, eighth edition) stated on sheet 5-221; " tests have shown that washers play only a minor role in distributing the pressure due to bolt tension, except where oversize or short slotted holes are used." Also, on sheet 5-227 of the AISC manual, it was stated; "The use of flat circular washers is not required with A325 bolts installed in standard holes."
Based on the above statements, it is concluded that where standard holes are used, the absence of wishers is structurally acceptable without affecting the connection capability of transferring the applied loads.
~
Q___-_-_________________________
(2) R3garding Ovaraizad Holos:
Field Change Request (FCR)'ME9, dated 4/17/86, stated that the anchor bolts are out.of location by 1/8" and the bolt holes were to be reamed to allow for construction. In this case the bolt holes were changed to the over-sized condition (15/16" holes for 3/4" bolts). To assess the impact on safety, an evaluation was performed to establish the available bearing area (in the absence of washers). Upon review of the actual tensile forces in the anchors, it was concluded that the still available bearing area is more than the required bearing area and, accordingly, there is no safety impact because of the existing condition.
(3) Regarding Mounting Bolt Not Fully Engaged:
Adequate engagement was lacking by 1/2 to 3 threads. The applicable cal-culations were reviewed for the tensile loads in the mounting bolts and additional evaluations concluded that the existing thread engagement is more than adequate to transf r the design tensile loads to the mounting bolts.
. Corrective Action NCR's 13906 and 13917 and Startup NCR S-0490-M have been issued to docu-ment these discrepancies and dispositioned to rework.
Action to Prevent Recurrence In accordance with the Quality Control Program, a Corrective Action Report was issued against the Quality Control Engineer responsible for the per-formance of the inspection and acceptance of these units. Construction Mechanical Engineering has received formal training to reemphasize the requirement for detailed drawing / manual reviews.
I 1
l 4
l l'
1 MECHANICAL COMPONENTS /HVAC SYSTEM (LK-C-1902)
SECTION 4.3.2.1 Concern Three of four bolts attaching the horizontal brace between the fan suction bell and unit cooler 2Fv210 were loose. This unit cooler was part of the SWEC ICA. sample.
Response
Concur With Concern Cause The cited condition is attributable to the fact that the vendor manual /
drawing did not specify a tightening value or method for this bolted connection. As a result, these bolts were installed to a snug-tight condition.
Extent of Condition All safety related unit coolers have been reinspected; one additional installation was found to have a similar condition.
Significance American Air Filter (AAF) drawing FF-14838 shows a horizontal brace member between the saddle type supports of the fan housing (one each side). A review of the existing fan support calculations concluded that the subject horizontal brace performed a secondary redundant function and that the primary carrying system in the direction of brace was through the fan housing and the saddle type supports. Additional evaluations concluded that the primary system is capable of carrying the total loads and if the horizontal brace were not present no safety impact would have existed.
Corrective Action NCR's 13906 and 13917 were issued to identify and correct the discrepancies noted (both NCR's are now closed). In addition, Field Change Request ME-1419 has been issued to clarify the tightening requirements for the C-channel brace and the angle clip to the fan / cooler housing.
Action to Prevent Recurrence In accordance with Quality Control program requirements, a Corrective Action Report was issued against the Quality Control Engineer who performed the inspection and acceptance of these unit coolers. Construction Mechanical Engineering has received forn.al training reemphasizing thorough drawing / manual review requirements.
l l
l 1
_ _ _ . . _ _ _ = - . - - - . - - _ _ _ _ _ _ _ _ _ _ _ _ _ _ - - - - - - - - _ _ _ _ _
MECHAN 2 CAL COMPONENTS /HVAC SYSTEM (LK-C-1902) i I
SECTION 4.3.2.1 Concern Washers shown on the foundation anchor bolt' design drawing were not installed.
on the fixed end of fuel pool cooling water heat exchanger 2A and on the sliding ends of fuel pool cooling water heat exchangers 2A, 2B and-2C. In another example, heat exchanger 2C bolt holes had been significantly enlarged and were visible around the perimeter of the nut. Jam nuts shown on Bechtel Equipment Foundation Design Drawing C-631 were not installed on the fixed ends of any of these heat exchangers. Balance of Plant Condition Report M3144 was written by the licensee to document this NRC finding.
Response
Concur With Concern The cited condition is addressed in our response to the generic concern later in this section regarding the fact that various pumps and unit coolers did not have jam nuts installed in accordance with Note 5 on Bechtel Design Drawing C-626, " Reactor Building Equipment Foundation",
c
MECHANICAL COMPONENTS /HVAC SYSTFM (LK-C-1902)
SECTION 4.3.2.1 Concern Jam nuts on Bechtel Foundation Design Drawing C-631 were not installed on any of the anchor bolts for the fuel pool service water booster pumps.
This condition was also recorded by the licensee on Balance of Plant Condition Report M3144.
Response
Concur with Concern The cited condition is addressed in our response to the generic concern later in this section regarding the fact that various pumps and unit coolers did not have jam nuts installed in accordance with Note 5 on Bechtel Design Drawing C-626, " Reactor Building Equipment Foundation".
l
{' MECHANICAL COMPONENTS /HVAC SYSTEM (LK-C-1902)-
SECTION 4.3.2.1 Concern ,
various pumps and unit coolers did not have jam nuts installed on the foundation bolts. General Note 5 on Bechtel. Design Drawing C-626,
" Reactor Building Equipment Foundations", specified that two nuts were required for anchor bolts securing vibrating equipment.
l
RESPONSE
Concur With Concern Cause The cited-condition is attributable to the fact that equipment foundation details are not explicit or consistent as to when heavy jam nuts are required, although specified in general Note 5 of Drawing C-626. Some details call for a double nut specifically while others do not. For exampic Detail 6 on Drawing C-631 requires double nuts for the 2A-CP221 pumps, but Detail 11 on the same drawing does not require double nuts for the 2A-BP223 pumps. In some cases, the definition of vibrating equipment was misinterpreted oy Construction Engineering. In addition, an adequate drawing review was not performed in all cases prior to final acceptance of equipment installations.
Extent of Condition A subsequent inspection was performed of all mechanical equipment in the Reactor Enclosure and Diesel Generator Building for conformance to the applicable civil equipment foundation details. Of a total of 103 pieces of equipment inspected, an additional 6 were found to be missing required washers and 13 were missing required jam nuts.
Significance The safety significance of this issue is being evaluated by Project Engineering and will be addressed in our response to PECo Quality Assurance Finding Report 2r 38 which has been issued to track and close the NRC Region i Unresolved Item related to this concern. In addition, the final disposition of the above referenced deficiency documents will likewise be reconciled as part of that response.
Corrective Action The noted deficiencies, including those identified by the NRC inspection
-team, have been documented as follows:
EQUIPMENT DOCUMSWT 2AP210 NCR-S-0488-C 2BP210 NCR-S-0488-C 2AE201 NCR-S-0487-C 2BE201 NCR-S-0487-C 20T221 NCR-S-0496-C 20T204 NCR-S-0484-C
EQUIPMENT < DOCUMENT
- 2AE208 NCR-S-0412-M 2BE208 NCR-S-0412-M 20E207 NCR-S-0412-M 2AP256 NCR-S-0471-C 2BP256 NCR-S-0471-C 2AV209 NCR-S-0482-C 2BV209 NCR-S-0482-C 2EV211 NCR-S-0481-C 2FV211 NCR-S-0481-C 2GV211 NCR-S-0481-C 2HV211' NCR-S-0481-C 20T217 NCR-S-0495-C 2AP226 SWO-269D-020 2BP226 SWO-269D-020 2AP208 NCR-13877 2BP208 NCR-13877 2CP208 NCR-13877 2AP212 BOPCR-M3144 2BP212 BOPCR-M3144 i
2CP212 BOPCR-M3144 2AE202 BOPCR-M3144-2BE202 BOPCR-M3144 2CE202 BOPCR-M3144 l
Action to Prevent Recurrence Construction Mechanical Engineering and Quality Control have received additional training to reemphasize the requirements for a thorough drawing / manual review prior to the equipment installation and acceptance process.
I
^!
i l
(
1 l
_ _---__ _ _ _ _ - - - _ - - - - _ _ - - - - - - - _ - - - - - - - - - - - - - - - - - 1
MECHANICAL COMPONENTS /HVAC SYSTEM (LK-C-1902)
-SECTION 4.3.2.1 Concern The NRC inspection team reviewed WP's for the Unit 2 RHR pumps and heat exchangers. During this limited review, the team identified several discrepancies in the documentation and some apparent weaknesses in the control and execution of_ work-packages. These observations may point out contributory causes of the hardware discrepancies previously detailed.
The computer statusing report was not accurate. The status report' showed WPA, Revision 2, as the latest work package for RHR pump 2A.
However, the engineering file did not contain the original WP, or Revisions 1 or 2 to it, but did contain a completed Revision 4.
These missing file documents were later replaced by copies from the QC department information file. The status report' listed an MRP3,. Revision 1, for RHR heat exchanger 2A that had been voided.
This same voided WP was referenced on QC Inspection Report 2AE205-48-1.
The engineering files for the work packages were not well organized.
The files did not contain an index of contents or a listing of issued packages, which in some cases numbered 26 or more WP's.
The Quality Control Inspection Reports (QCIRs) in the vault were also extremely complex, with numero"s cross references to previous documentation (24 notes on one QCIR for heat exchanger 2A), and did a poor job of defining the scope of t? 'nspection effort and the source of the requirements for the inspection.
The stated scope of the type "A" WPs did not change with revisions, and the source of the requirements for the revision such as an NCR, FDDR, FCR, or engineering needs, was not indicated. Often the specific actions required by the work package were unclear; that is, in some cases the change to the WP only listed or revised a list of open FCRs or FDDRs. In general, the NRC inspector did not consider that changes to these documents were controlled or formatted adequately considerlag the importance of the equipment involved.
Work Packege "A" for RHR pump 2A was illustrative of the discrepancies noted. The original issue was signed off as craft complete, but the Installation Engineer and QC Representative completion signatures did not appear. Revision 1 and 2 in the file had no signatures for issue or completion. The inspector noted that the file copies of these documents had been taken from the QC information files. Revision 3 had been signed off as completed by craft and the Installation Engineer in March of 1987, but did not have the QC completion signature.
' *' Documents referenced in the WPs, such as unique construction aid documents, were not always located in the WP file. Active and. completed WPs were not' located in the master file but had to be retrieved from various locations in the engineering office.
This' inspection primarily focused on the hardware installation. However, the NRC inspection team was concerned about the condition.of mechanical equipment work packages and the separate and confusing QC inspection documents. The licensee should review the existing documentation to assure.that all aspects of installation and modification of mechanical equipment at Limerick have been completed and that the required documentation exists.
The NRC team alsc concluded that the policy of considering the work packages as nonpermanent QA records creates a potential information gap in the machinery history of the plant's mechanical equipment. The lack of a clear and detailed machinery history could hamper long term corrective efforts and does not provide an auditable record.
Response
Do Not Concur With Concern
.The work package program is the method used to identify to construction (i.e. the superintendent and craftsmen) the scope of work to be performed, special techniques to be used and the forms required for its completion.
The work package itself is a tool utilized oy Construction Engineering to get information to the craft in an expeditious manner to aid in organizing, planning and implementation of their tasks.
The work package is transmitted to Quality Control prior to the start of work for preparation of QCIR's and assignment of QC inspection witness and/
or hold points. Quality Control does not perform or accept inspections based upon the content of work packages; QCIR's are the objective evidence of QC inspections, surveillance, witness and hold point verifications.
The cited condition is not indicative of a lack of quality documentation for equipment, nor is it a contributory cause of any specific hardware problems.
Final acceptance of installations, modifications or rework is based upon approved design documents, not the work package, and is documented separately from the work package on Quality Control Inspection Records (QCIR's).
Construction Work Packages are reviewed by Quality Control to ascertain the type and extent of work to be performed on safety-related and ASME items.
An active open work package may contain design documents, written work instructions, material withdrawal requests, safety tag stubs, access per-mits, burning permits, data sheets, copies of NCR's, etc. When work is completed, any quality documents that have been accumulated in the work l package are removed by QC for incorporation into the QCIR. Those parti-cular work package forms that become part of the permanent record are defined as Quality Verification Documents (QVD's) in accordance with l Construction Procedure CP-M-1.
_ ~ _ _ _ _ _ _ _ - . _ . _ _ ___ -- _ _ _ _ _ _ _ _ - _ - _ . - - _ - - _ _ _ - - .
t The work packcga itself, including the signed cover chset, is not a QVD.
Therefore, tha existence of minor administrative errors, or even ths absence of the entire work package file, does not have any bearing on the hardware installations or the quality of the permanent plant records.
l The work package program was not instituted until early 1986. Much of the mechanical equipment in Unit 2 and all of the equipment in Unit 1 was installed prior to that time. In addition, work packages will not be retained or archived when Unit 2 construction is complete.
The fact that work packages are not classified as permanant QA records does not create an information gap in the machinery history for the reasons l provided above. This history exists in the many other documents that I detail the events over the years, including the time prior to implementation ,
of the work package program. Examples of these documents are design drawings, j FDI's, FDDR's, NCR's, IPRN's, ESR's, EPR's, FCR's, DCP's, SWO's and SCR's.
Although some complete QCIR packages are bulky and may appear somewhat confusing due to their size, all the required data and attributes are accounted for and accurately reflect the hardware installations. The Quality Control Inspection Record Program has been in place since the mid 1970's and has undergone periodic audits by Bechtel Quality Engineering, Bechtel Quality Assurance, PECo Quality Assurance, the NRC and INPO and has been determined to be adequate.
MECHANICAL COMPONENTS /HVAC SYSTEM (LK-C-1902)
SECTION 4.3.2.2 Concern Two bolts joining the companion angles of duct segments 73A and 80 were loose.
Response
Concur With Concern Cause The cited condition is attributea to an unauthorized alteration after final QC inspection.
. Extent of Condition An additional 92 companion flanges were inspacted for loose bolts. No discrepancies were founds.therefore, this is not a prevalent condition.
Significance Engineering's evaluation of this condition concludes that the stress on the bolts is. low and the structural integrity of the connection remains adequate and within design requirement. Any leakage as a result of the loose bolts would be within the room and would neither impact system performance nor room temperature.
Corrective Action NCR-13912 was issued to document this condition and has been dispositioned "Use-As-Is" by Project Engineering.
Action to Prevent Recurrence As part of the facility turnover process, the Facility Engineer and a Quality Control Inspector review installed commodities for deficiencies or damage. In addition, Start-up Engineers inspect the entire system before initial operation. Start-up Technical Program TT1.7, Section 7.2, Step 5 and 6, specifically addresses that the flanged connections be reinspected for compliance to design requirements.
l
MECHANICAL COMPONENTS /HVAC SYSTEM (LK-C-1902)
SECTION 4.3.2.2 Concern A splitter damper was installed at segment 73, but no damper was shown installed at that location on the Bechtel Design Drawing or the Schneider Construction Drawing.
Response
Concur with Concern Cause The cited condition is attributable to a drafting error. g Extent of Condition A review was performed on 46 other section views for similar conditions; no discrepancies were noted.
Significance Bechtel Drawing M-1172 was the only drawing that did not show the splitter l damper. Schneider Drawing M128-1201(1), submittal 6 (issued 4/29/87) to current submittal 8, always showed the splitter damper in the location installed in the field. In addition, Bechtel Drawing C-1352-2 shows a standard branch detail with splitter damper which is typical for supply air branch lines of similar configuration. Since field installations are performed in accordance with the Schneider drawings, the minor drafting error on the Bechtel drawing is insignificant.
Corrective Action The condition was documented on Bechtel NCR-13912 and Design Change Notice DCN #3 has been issued to incorporate an "SD" designator on Section 'A' of Drawing M-1172.
Action to Prevent Recurrence All HVAC installations are complete and, therefore, no further action is warranted.
I L
l f-l
~ ~ ~ ^ ~
o____ _ _ . _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ . _ _ _ . _ _ _ _ . _ _ . _ . . _ _ _ _ _ _ . _ _ _ _ _
l l
MECFANICAL COMPONENTS /HVAC SYSTEM (LK-C-1902)
SECTION 4.3.2.1 l
l Concern Tapered washers were not installed between the fasteners and the channel iron frame of a splitter damper, segment 73, as specified in the AISC Manual of Steel Construction. Paragraph 3.2.1 of HVAC Specification 8031-M-68A-2 stated that equipment and components shall conform to this standard.
Response
Do Not Concur With Concern A review of the cited condition concludes that it conforms to Specification M-68A-2 requirements. Per the AISC Manual of Steel Construction, 7th Edition, Page 5-196 requirement, the use of bevelled washers is to compensate for the lack of parallelism between mating surfaces. The subject bolt heads and nuts have been verified as achieving sufficient bearing of mating surfaces, therefore eliminating the need for bevelled washers.
This condition was documented on NCR-13912 because it was considered to be indeterminate at the time it was identified. The NCR has since been '
dispositioned to "Use-As-Is" for the reason stated above.
l l
i i
MECHANICAL COMPONENTS /HVAC SYSTEM (LK-C-1902)
SECTION 4.3.2.2 Concern Three tap holes used for checking pressures and flows during system I balancing were not properly sealed after the balancing work. Plugs consisting of bolts with gasket washers were not fully engaged. This situation provided potential leakage paths and the possibility that the bolts could back out completely due to vibration during system operation.
Response
Do Not Cor. cur With Concern The cited condition is attributable to the fact that balancing of system 2-34 had not started at the time of this inspection. The inspector misunderstood the Start-up Engineer to say all testing had been completed. Startup Technical Test TT1.10, Appendix A, verifies by notation that the static and traverse belts are wrench tight upon completion of final balancing. Final balancing of this system is scheduled to complete in March, 1989.
i l
CIVIL / STRUCTURAL JLK-C-1903)
SECTION 4.4.2.1 Concern
.The NRC team had one concern relating to sister splices made for cadweld testing.- Contrary to the requirements of Section 3.8.6.2.2.2 of the FSAR, sister splices were not tested prior to the start of production splices made by cadweld operator 93 for splice numbers 38271 and 41851. A similar concern was also raised by the SWEC reviewer for a different cadweld crew.
Response
concur with Concern pap s_e The cited condition is attributable to an oversight by Construction Engineering and Quality Control personnal to ensure that sister splices were made prior to production splices when the three day limit was exceeded after the welder requalified.
Extent of Condition A review of 120 records was performed to inspect for similar conditions 10 additional discrepancies were noted. We feel that it is reasonable to assume that this ratio of error is representative of the entire records population.
Significance This issue has been evaluated and the conclusion is that it is not significant for the following reasons:
- All welders were prequalified and every 25th splice was pull-tested.
- All splices were visually examined.
Corrective Action The deficiencies were documented on NCR-14244 which has been dispositioned to "Use-As-Is".
Action to Prevent Recurrence All cadwelding operations are complete and, therefore, no further action is warranted.
1 1
CIVIL / STRUCTURAL (LK-C-1903)
SECTION 4.4.2.2 Concern on platform 55, the NRC team noted that one bolt might not have enough projection to exclude the threads from the shear plane, as required by Bechtel Design Drawing C-ll84 Revision 31. PECo should confirm whether this bolt meets the requirements of the above mentioned design drawing.
Response
Concur Mith Concern cause The cited condition is attributable to inattentiveness on the part of the Construction Engineers when accepting these connections.
Extent of Condition An additional 2500 bolts were inspected for similar conditions; 35 discrepancies were noted.
Significance Per the AISC Manual of Steel Construction, 7th Edition, heavy hex structural bolts have shorter thread lengths than other standard bolte.
By making the body length of the bolt the control dimension it has been possible to exclude the thread from all shear planes, except in the case of thin outside parts adjacent to the nut. Depending on the amount of bolt length added to adjust for incremental (bolt) stock lengths, the full thread may extend into the grip as much as 3/8 inch for 1/2 inch, 5/8 inch, 3/4 inch, 7/8 inch, 1-1/4 inch and 1-1/2 inch diameter bolts.
Inclusion of some of the thread run-out into the plane of shear is permissible. At the other extreme, care should be taken to provide !
sufficient thread for nut tightening to keep the nut from jamming into the thread run-out. When the thickness of an outside part adjacent to the nut is less than these values, it may be necessary to specify the next increment of bolt length together with a sufficient number of flat circular washers to insure full seating of the nut.
This extension of threads into the grip by as much as 3/8 inch is significant in that, when used in a connection involving thin parts
(<3/8"), particularly when the thin part is adjacent to the nut, the possibility exists of having threads in the shear plane. Table 6 is provided on page 5-202 of the AISC Manual to enable the designer / engineer to specify the bolt length which is appropriate to provide for full thread engagement, but the table does not ensure that threads are excluded from the plane of shear. The designer / engineer must evaluate this attribute on a case-by-case basis and, as stated above, specify the next incremental bolt length together with a sufficient number of flat washers to ensure that threads are excluded from the plane of shear.
1
~
In otructural conosctions which utilize high-strsngth bolta to join thin materials, the bearing capacity of tha joinsd parts is usually the limiting condition. For a 7/8 inch diameter A-325 bolt in a standard hole (15/16 inch), bearing will govern when the thinner of the joined parts is less than or equal to 3/8 inch in thickness. In all of the connections identified as being discrepant, at least one of the connected parts is less than 3/8 inches in thickness. Hence, for these connections the bearing capacity of the thin parts governs rather than the shear capacity of the A-325 bolt. Considering these facts, the inclusion of threads in the plane of shear of these connections does not constitute a structurally deficient condition.
Corrective Action The bolt identified during the NRC inspection was documented on NCR-13885 and replaced. The bolts identified by the subsequent inspection were documented on NCR-1413G which was dispositioned "Use-As-Is" by Project Engineering.
Action to Prevent Recurrence Formal training has been given to the Construction Engineers and Quality Control Engineers to reinforce the requirement that it is their respon-sibility to ensure that threads are excluded from the plane of shear.
l 1
CIVIL / STRUCTURAL (LK-C-190H SECTION 4.4.3 Concern The NRC & SWEC reviews showed that the structural steel platform samples selected. conformed to the design requirements shown on the appropriate drawings, except for the missing connection welds on platform 2AV-212.
However., the NRC team believed that PECo should perform additional field inspections to determine whether a generic problem exists regarding missing connection welds.on Limerick Unit 2 Structural Steel. connections.
Response
concur with concern The generic implications.of this condition have been addressed in our response to Stone & Webster Construction Observation Report COR-29.
l l
CXVIL/ STRUCTURAL (LK-C-1903) l SECTION 4.4.3 Concern The'SWEC ICA review for the block walls was comprehensive and adhered to their review plan LK-C-1903. The NRC team believed that from the indepen-
! dent review performed on block wall 63, construction was in accordance with the FSAR requirements and is consistent with SWEC findings. However, the concerns raised by SWEC regarding construction documentation should be resolved by PECo before a conclusion can be reached on the ad.equacy of all masonry construction at Limerick Unit 2.
Response
Do Not Concur With Concern The cited condition regarding construction documentation has been resolved in accordance with Stone & Webster Construction Action Item CAI-82, which states that all of the records for block wall 104 are in the QC vault and available for review.
ELECTRICAL (LK-C-1904)
SECTION 4.5.2.1 Concern vendor installed cabling in several Class 1E motor control centers exhibited deficiencies in the bend radius. The teams examination of motor control centers 2AC224 and 2BC224 (control panels for the unit hydrogen recombiners) disclosed wiring between the heater breaker and the motor starter that exhibited significantly less radius than that required by industry and site standards. The wires in question were number 4 AWG (American Wire Gage) in size and were installed with a radius of approximately 1 inch. The team also noted that the insulation of these wires had buckled and pulled away from the individual conductors because of their excessive bending. Similar deficiencies were observed in cubicle 08 of motor control center 20B212, in which vendor wiring was formed with a radius of approximately 1/2 inch.
In response to this observation, the licensee issued NCR-13910 to document and evaluate the impact of the deficiencies on component and system operation.
However, based upon the frequency of this observation from a rr,latively small sample, it appeared that the licensee should consider conducting a review of vendor wiring in Class 1E motor control centers to determine the scope and significance of bend radius deficiencies.
Response
The concern identifies two conditions that are unrelated and must be addressed individually. As a means of differentiation, they are henceforth referred to as concern #1 and concern #2.
Response to Concern #1 Concur With Concern Cause The cited condition relative to the bend radius deficiencies identified in hydrogen recombiner power cabinets 2AC224 and 2BC224 is attributable to lack of attention to detail during the fabrication process.
Extent of Condition Construction Engineering pe. med a 100% inspection of these cabinets and no other discrepancies ac noted. Since these two panels are the only pieces of equipment of thin configuration supplied by this particular vendor (Time-Trol, Inc.)~ there is no reasons to believe that this condition exists in th' dass 11 motor control centers (supplied by Cutler-Hammer).
Significance (Interim Response)
The significance of this condition is under evaluation with Engineering and the vendor and will be submitted by 4/4/89.
Corrective Action NCR-13910 was issued to document the noted deficiencies and dispositioned to replace the discrepant wires with Rockbestos #4/0 cable; all work is complete.
Action to Prevent Recurrence j Based upon the discussion under Extent of Condition, no further action is warranted.
I Response to Concern #2 Do Not Concur With Concern The cited condition relative to the bend radius deficiency identified in motor control center 20B212 is attributable to the as supplied condition of the vendor supplied equipment.
The wire in question was a #4 SIS wire with a bend radius of approximately 1/2 inch. The vendor (Cutler-Hammer) has advised us that their standard wire Londing policy is that the inside bend radius is to be no smaller than I
the outside diameter of the cable being bent. The #4 SIS wire has an out-side diameter of .360 inch and therefore, a .5 inch inside bend radius-is acceptable.
i s
1
(
i ELECTR2 CAL (LK-C-1904)
SECTION 4.5.2.1 Concern During the examination of motor control center 20B217, the NRC inspector noted that cable 2CB21703A in cubicle 03 was improperly terminated. This three conductor cable was the power feed for the 2A reactor enclosure cooling water. pump and was terminated by bolted connection to the breaker terminal bars. The inspector noted that the phase 'B' lug had been bent by construction activity resulting .in a 1/2 to 3/4 inch gap between the lug and the terminal bar. Although circuit continuity had been maintained, this configuration resulted in a contact " hot spot" which may lead to operational problems. The licensee should determinate this cable replacing or repairing the lug in question.
Response
Do not concur with Concern The cited condition has been evaluated as acceptable. With input from the Burndy Corporation, the manufacturer of the lug in question, Design Change Notice, DCN 870, to Drawing E-1412 has been issued to clarify the criteria for bending of lugs. It is permissible to bend the Burndy type YA lug up to 30'; the lug in question was observed to be bent approximately 10' and, therefore, is acceptable.
In addition, all Unit II motor control center vertical bus bars have recently been replaced (aluminum to copper) which required determination and retermi-nation of all cables with this type lug, thus assuring compliance to DCN #70 bending criteria.
I
l ELECTRICAL (LK-C-1904) 1 SECTION 4.5.2.1 l Concern i l i
- l. During the examination, the inspectors noted that several Class lE penetrations contained circuits that feed both engineered safety features (ESP) and reactor protection system (RPS) functions.
. Additionally, the inspectors noted that several penetrations contained both Class lE and Non Class lE circuits.
The Limerick FSAR provided requirements for the physical independence or redundant systems in Section 8.1.6.1.14. This section detailed specific physical separation between ESF, RPS, and Non-Cla9s lE plant system circuits. These requirements reflected those specified in Regulatory Guide 1.75 and Institute of Electrical and Electronics Engineers (IEEE) Standard 384-1974, with stated exceptions.
With regard to electrical penetrations, the FSAR made provision for the routing of Class 1E and Non-Class lE circuits through a common penetration. However, no provision was made for the routing of ESF and RPS circuits or multiple divisions of RPS circuits through a common penetration. Consequently, the routing of both ESF and RPS circuits in this manner did not appear to meet the intent of the regulatory guide or the FSAR for the station.
Additionally, Section 5.5 of IEEE Standard 384, to which the station is committed, required that Non-Class lE circuits routed in penetrations containing class 1E circuits be treated as associated circuits in accordance with the requirements of Section 4.5. Section 4.5 stated that associated circuits shall comply with one of the following:
- 1. They shall be uniquely identified as such and shall remain with, or be separated the same as, those Class lE 'rcuits with which they are associated.
- 2. They shall be in accordance with (1) above from the class lE equipment to and including the isolation device. Beyond the isolation device, a circuit is not subject to the requirements of this document provided it does not again become associated with a Class 1E system.
- 3. They shall be analyzed or tested to demonstrate that Class 1E circuits are not degraded below an acceptable level.
The Limerick FSAR stated that associated circuits were not uniquely identified as such and were treated and identified as Class 1E up to an isolation device. Contrary to these requirements, the Non-Class lE circuits routed with Class lE circuits through common penetrations have not been identified or routed as Class lE once they exit the penetration.
In response to these concerns, Bechtel Project Engineering stated that the existing design conformed to the requirements of the FSAR for the station. The NRC inspectors believed that additional evaluation, clarification, or analysis would be required before this issue could be closed.
Response
Do Not Concur with Concern Project Engineering has analyzed the cited conditions and concludes j the following: l l
I
' Penetration 2OJX100D This penetration contains RPS Channel Z cables only and, therefore, electrical separation is maintained from redundant RPS, Class lE or Non-Class lE cables which are routed through different penetration assemblies dispersed around the circumference of the containment. l
- Penetration 2OJX103B l I
This penetration contains: l l
- 1. Class lE Divisions B&D Instrumentation Cables, and
- 2. Non-Class lE Instrumentation Cables l I
The following conditions apply:
- a. Inside the Penetration Assembly ,
The redundant Class lE and Non-Class lE instrumentation cables are i not routed through common feed-throughs. The reed-through steel !
casing forms the separation barriers between redundant Class lE and Non-Class lE feed-throughs. The size of the cables is #16 AWG. ;
Electrical separation is maintained between redundant Class lE !
and Non-Class lE cables per E-1406, Section 2.1.2.3, " Enclosed Raceways and Cable Trays", paragraph a.l.
- b. Inside the Penetration Box The redundant Class 1E cables are enclosed in stainless steel >
flexible conduits from the point they leave the feed-throughs up to a connector inside the Conax isolation junction box; ,
therefore, these are treated as enclosed raceways. The Non-Class I lE cables are considered dropout cables inside the Conax isolation junction box per E-1406, Section 2.2. Furthermore, all the cables l i
in the isolation box are #16 AWG and, therefore, separation is maintained between redundant Class lE and Non-Class 1E cables per j E-1406, Section 2.2.3, Paragraph a, " Separation Between Dropout Cables and Enclosed Raceway".
l l
l 1
Although no RPS cables have been found mixed with ESF or Non-Class lE cables in the subject penetrations, please be informed that, per E-1406, Section 2.1, Raceway Separation paragraph 2.1.1.a, Class lE raceway is defined as ESF and RPS raceways. Therefore, RPS cables are treated as Class lE and, as such, are separated from Non-Clars lE and ESF cables per E-1406, Sections 2.1.2.3 and 2.2.3 which is consistent with the FSAR, Section 8.1.6.1.14.a.l.
In addition, a review of the rest of the penetration assemblies for separation has been performed and our findings are similar to those described above.
We conclude that there is a continuity between IEEE 384-1974, the Project FSAR Section 8.1.6.1.14, paragraph b.8 and E-1406 and proper electrical separation has been maintained.
l
I ELECTRICAL (LK-C-1904)
SECTION 4.5.2.1 i concern The haphazard routing of cables and conductors in the main control boards made the boards congested and did not reflect the workmanship observed in other plant control panels. This condition appeared to be the result of the large number of General Electcic design changes issued for these panels.
The NRC did not inspect the panels in detail, and therefore did not identify specific construction deficiencies. However, the inspectors believed that the condition of these panels merits additional attention by the lice:. .
In particular, the licensee should assure that cable installation require-ments such as bend radius, separation, fire stops, and absence of cable damage, have been maintained.
Response
Do Not Concur With Concern The cited condition is programmatically addressed as part of the final separation and color coding inspection of electrical equipment performed by Quality Control in accordance with the requirements of E-1412. These inspection attributes include, but are not limited to the following:
Replacement equipment is the correct type and installed in the proper location with no damage.
Components removed to provide access for modification work are properly reinstalled.
Debris removed.
Work activities have been performed without damage to adjacent components or material.
Seoaration of components is maintained.
Permanent wire I.D. is installed.
Cable / wire size, type and dressing is correct.
Correct lugs or pins are installed.
Conductors are terminated at correct terminal point and are tight, soldering has been properly performed.
Wires are properly color coded.
Separation of wires and raceways is correct.
This inspecticn has been satisfactorily completed on all Unit 2 Main Control Boards.
l ELECTRICAL (LK-C-1904) j SECTION 4.5.2.1 Concern Valve HV-52-2F001C contained vendor supplied motor leads that appeared to be undersized for the application. Leads T1, T2, and T3 were number 18 AWG wire supplied by Belden Wire Company and providad interface between the actuator motor and the field power cable. Additionally, the use of Belden J, Type AWM crosslink wire in an environmentally qualified (EQ) valve was questioned by the NRC team.
To ascertain whether the identified wire was of the type and size required for this application, the NRC inspectors requested copies of environmental reports for this valve. Discussions with PECo and Bechtel personnel indicated that EQ packages had not been completed for the Unit 2 valve operators. The inspectors then attempted to review EQ packages for Unit 1 valve operators on the basis of similarity of design and purchase requirements. However, although it was initially indicated that these documents were on site, it was determined that all of the EQ packages were kept in the Philadelphia Office and thus were not made available to the inspectors.
On the basis of this lack of information and the questionable type and size of installed wire, the NRC inspectors considered the quality and the environmental qualification of valve operator HV-52-F001C to be indeterminate.
Response
DO NOT CONCUR WITH CONCERN Limitorque has informed us that the Peerless AC Motor, (Serial Number UAS4184WM) and the Belden motor leads are qualified in accordance with their Qualification Report B0003.
I 1
l 1
l
)
l l
I 4
U____________
i ELECTRICAL (LK-C-1904) i SECTION 4.5.2.1 Concern Examination of valve operator HV-55-2F002 disclosed a discrepancy in the termination of cable 2DB22418B. This cable was the power feed for l
the actuator motor and was terminated in accordance with the Bechtel connection list. Functionally the termination was correct and reflected the conductor-to-terminal point configuration specified on the connection list. However, the phase identification did net match connection list details in that Phase 'A' and 'B' had been reversed.
Discussions with Bechtel personnel indicated that phase conductors may be reversed in accordance with Specification E-1412. Section 4.22 of the specification stated that "for 480V, 3 phase motors the field may reverse two of the 3 phase conductors of the motor feeder cable at the MCC in order to obtain proper rotation of the motor. Revision of the connection list for this change is not required."
Reversing feeder cable conductors to achieve proper motor rotation is a common industry practice. However, such actions should be reflected in design documents to ensure accurate representation of existing field conditiors. Without formal updating of the Bechtel connection lists and wiring diagrams future system and component modifications may be adversely affected. The licensee should ensure that all conductor phase rotations are noted and incorporated into plant design documents.
Response
Concur With Concern The cited condition has been evaluated and, as a result, Field Change Rcquest EL-2557-E has been issued against E-1412, Section 4.22, which states:
"When reversal is required, the motor leads at the MCC shall be color phased with colored electrical tape to maintain black, white and red phasing.
This will ensure that the as-installed condition is in agreement with the connection list. A field walkdown is being performed to verify compliance to the above requirement which will complete in April, 1989.
ELECTRICAL (LK-C-1904)
SECTION 4.5.2.2 Concern In general, the installation of Class 1E cables was found to be in accordance with specified criteria. The routing of examined cables accurately followed the path shown on engineering documents and Quality Control inspection records.
Attributes such as cable size, type, and identification were also as specified on design documents. However, several cable installations were identified that did not meet the Limerick FSAR requirements for electrical separation. The examples observed involved Class 1E cables that left design designated raceways and ran " free-air" through walls or into electrical equipment. The inspectors noted that in many instances these cables did not maintain the required physical separation from redundant Class 1E cables and raceways and had not been protected through the use of fire barriers or cable wrap as required.
This issue was discussed with Bechtel personnel to determine the status of the cables in question. Bechtel's responses indicated that electrical separation would be addressed at the time of facility turnover through the use of Quality Contiol Instruction (QCI) E-3.0, " Raceway Barriers and Seals". This instruc'clon provided for the inspection of cable and raceway installations to ensure com-pliance with all attributes of electrical separation. Specific guidance was provided for inspection of " free-air" cables and the application of approved fire barriers. While the content of this document appeared thorough, the inspectors noted that aggressive implementation would be tequired in order to assure that Class lE cable installations met FSAR separation criteria.
Response
Do Not Concur With Concern The cited condition is programmatically addressed as part of the final facility electrical separation walkdown which aggressively assures cable separation in I accordance with QCI E-3.0. These inspection attributes include, but are not limited to the following:
Verify the methods ured to meet minimum separation between redundant Class lE raceway, or Class lE and Non-Class 1E raceway are acceptable.
- Verify the methods used to meet minimum separation between exposed cables or exposed cable and raceway.
Verify that dropout cable wrapping has been properly installed where required by design drawings.
Ver 79 that &An proper type of raceway covers have been correctly installed at 'the required locations.
Verify covers of proper type are installed on conduit fittings, panels and boxes. Where gaskets are required, verify they have been provided.
This inspection has been satisfactorily completed in 72 of the 84 total facilities to date.
PIPING AND PXPE SUPPORTS (LK-C-1905)
SECTION 4.6.2 Concern In the case of pipe support DCA-418-H8, the NRC team measured less than a 1/2 inch gap in the east-west direction between the 3-bolt pipe clamp (part number 6), which moves with the pipe, and the east side flange of a W6x20 column (part number 8). The potential interference was previously identified by the licensee via a Potential Interference Notification (PIN) and Disposition Form (PIN No. 2328, dated November 21, 1987). However, the PIN cited a gap of 3 5/8 inches versus the 1/2 inch measured by the NRC team. Bechtel Project Engineering had reviewed the PIN and determined that the reported gap was acceptable in view of analyzed piping displacements being less than the existing gap. The NRC inspector's review of the as-analyzed pipe displacements, however, revealed that seismic displacements as well as thermal pipe displacements exceeded the gap observed during the NRC inspection.
The single finding related to potential piping and pipe support interferences suggested the need for greater emphasis on interference checks in the final stress walkdown inspections of piping systems.
Response
Do Not Concur With Concern The cited condition is attributable to the fact that Construction Engineering mistakenly transposed dimensions and supplied erroneous information to Project Engineering for evaluation on PIN Number 2328.
The existing gap between the pipe clamp and the east side flange of the W6x20 column exceeds 1/2 inch. This was documented on PIN Number 4331 which super-sedes PIN number 2328 and submitted to Project Engineering for evaluation.
Project Engineering has concluded that the 1/2 inch clearance satisfies design requirements.
This condition would have been identified in the course of the final stress walkdown. Specification P-403, Section 10.1, indicates that the purpose of this walkdown is to review, by means of a visual inspection of the instal-lations a) The capability of the piping system to perform its design function, and b) Interferences which would prevent free thermal expansjon of the piping system.
Inspection attributes of this walkdown include, but are not limited to the followings t
Piping layout configuration consistent with that used in the stress analysis.
Fr:3 tharmal cxp nsion of tha piping systsm not blockad by intsrfortncas with oth2r commodities, i.o HVAC, c2ble trays, other pipe supporta, floor or wall penetrations, grouted-in penetrations, etc.
Existing branch lines have been considered in the stress analysis and are shown on the walkdown ISO's.
Pipe support type, line of action or location correct with respect to the stress analysis.
Addition or missing pipe support (s) with respect to the stress analysis.
Emphasis has been placed on the interference checks and formal training sessions have been conducted to assure program compliance.
f l
L____________
F PROCUREMENT, RECEIVING, AND STORAGE (LK-C-1905) 1 SECTION 4.7.2.1 Concern The team conducted a review of selected Long Term Maintenance (LTM) requirements as established under Bechtel Procedure CP-G-3,'" Storage /
Maintenance / Lubrication Requirements for Long Term Storage of Installed l Equipment Prior to Turnover," Revision 8, dated August 23, 1988. Tnis procedure required that, as a minimum, the vendor recommended maintenance would be performed at the required interval. Deviations from the vendor
- recommendations were allowed, if supporting documentation was provided-
.to justify the change.
CP-G-3 required, in part, that deviations noted during the performance of LTM activities be documented on either an In-procesr, Rework. Notice (IPRN) or if of a more significant nature, a Nonconformance Report (NCR).
This procedure also required that steps be taken to correct the deficiency and that action be taken to prevent its recurrence.
The team reviewed the list of outstanding preventative maintenance' actions for the-week of September- 23, 1988. One safety-related maintenance action had not been performed within the required time interval. In reviewing the background of this delinquent maintenance action, the team noted that fhe pressure in the Hydraulic Control Unit (HCU) accumulators was required to be checked, on a sampling basis, once a month. The total number of HCUs installed in the plant was 185, and the LTM program required checking 4 HCU accumulators each month. During the review of the maintenance records, the team noted the following facts:
DATE CONDITION ACTION TAKEN 6/10/88 HCUs 30-19 and 38-47 nitrogen Accumulators recharged pressure below minimum acceptable to 25 psig value 7/7/88 HCUs 30-19 and 38-47 nitrogen Accumulators recharged pressure below minimum acceptable to 25 psig value 8/2/88 HCUs 30-19 and 38-47 nitrogen Accumulators recharged pressure below minimum acceptable to 25 psig. IPRN 2417 value issued to investigate leakage problem
l DATE CONDITION ACTION TAKEN 8/19/88 HCUs 30-19 and 38-47 nitrogen Accumulators recharged
- pressure below minimum acceptable to 25 psig. HCU 38-47 l value leak checked with liquid soap solution 8/24/88 HCUs 30-19 and 38-47 inspected by IPRN 2417 closed out QC. Pressure was within the acceptable range (10-25 psig) 9/29/88 HCUs 30-19 and 38-47 checked in NCR 13908 issued retponse to an NRC request.
Pressure in the accumulator units was 8.2 psig and 0.4 osig respectively Vendor Manual Cll-D001-K001 required maintaining a nitrogen purge pressure between 10-25 psig to prevent corrosion of the HCU accumulator internals.
The NRC team noted that no record of the as-found or the as-left pressure was recorded on the Maintenance Action Card (MAC) for the accumulators in question, with the exception of the pressure check performed on 8/19/88.
Additionally, the corrective action taken to correct the leakage problem and prevent its recurrence was not documented when IPRN 2417 was closed out.
Vendor Manual C11-D001-K001, Section 4-19, authorized the use of only two approved leak detection liquids. The vendor manual also required rinsing the fittings with demineralized water after the leak check was completed to prevent corrosion of accumulator metals. However, because no work instructions were uted to conduct the leak check, it could not be determined if either of the approved leak detection fluids was used.
As documented in NCR 13908 a demineralized water rinse was not performed following the leak check made on August 19, 1988.
Vendor Manual C11-D001-K001, Section 3-27.F, specified a torque value of 150 to 200 inch pounds for the nitrogen charging connector cap. However, the MAC card, which was routinely used to check the accumulator pressure and recharge it accordingly, did not specify any torque values. Additionally, the NRC inspector noted that a torque wrench was not used in reinstalling the connector cap on September 29, 1988.
The team determined that the Bechtel Quality Control department failed to identify the HCU leakage problem as a nonconforming condition. The Dechtel Quality Control Department closed out the lower tiered correction document, IPRN 2417, without verifying that the root cause of the problem had been corrected and without taking appropriate measures to prevent its recurrence. It was not until the NRC inspector requested Bechtel to recheck the pressure in the HCUs in question that nn NCR was issued to resolve the nitrogen leakage problem.
l l
The NRC team reviered the results of the SWEC ICA effort in the LTM and Procurement areas. SWEC found some problems in the maintenance records for major RHR components. On the basis of a review of the LTM program for HCUs and the findings of the SWEC ICA review, the team was concerned about Bechtel's attention to detail in maintaining the material condition l of safety-related components. The attention given the maintenance of these components was not consistent with the operational safety importance of the HCUs.
Response
Concur With Concern Cause The cited condition is attributable, in part, to a lack of communication between PECo I&C and Bechtel LTM Engineering. PECo Field Engineering released the HCUs to PECo I&C on July 27, 1988 to be tested in acccrdance with Procedure TL11-388. LTM Engineering had no knowledge that the testing required the removal of the hydrogen purge at the same time that they were performing regularly scheduled maintenance inspections of the HCUs in accordance with CP-G-3, paragraph 5.5.5. In addition, Procedure TLll-388 does not require the establishment of a nitrogen purge on the HCU's following completion of testing.
The documentation discrepancies noted are attributable to the fact that LTM-QC did not document the loss of pressure on June 6, 1988 and July 7, 1988 but, rather, took corrective action to recharge the deficient HCUs without notifying Engineering or obtaining Engineering concurrence.
Since the inspector felt that no deficiency existed, an IPRN was not initiated.
Extent of Condition LTM records on the verification of the nitrogen purge indicate that the inspections had acceptable readings prior to the HCUs being released for blue tag testing. Subsequent to the NRC inspection, the nitrogen purge on the remaining 183 HCUs was measured and recorded; 143 of these units were found to be out of the 10-25 psig range.
Significance As stated above, long term maintenance records indicate that the HCUs had been properly purged per tse requirements of CP-G-3. During blue tag testing, instrument checks and calibrations of the level and pressure switches were performed por i 11-386 which resulted in the nitrogen pressures out of tolerance and water in the instrument blocks. The requirements for a 10-25 psig purge are not part of Vendor Manual Cll-D001-K001.5. This manual is an operational manual and is not intended for use during installation of the HCUs. The 10-25 psig purge .as an additional preventive measure taken by Bechtel with G.E. concurrence; the need for a nitrogen purge is not a G.E. requirement.
1 J
1
Corrective Action NCR-13908 was generated to document the discrepancies noted during the
~
NRC inspection and the subsequent Bechtel inspection. The.NCR has been dispositioned to rework HCUs 30-19 and 38-47; the remaining 183 L
units have been dispositioned to "Use-As-Is". The torque value for the '
nitrogen charging connector cap has been addressed by G.E. FDDR-HH2-8803 which states; " Torquing of the nitrogen fill cap is not required prior j to startup. Hand tightening and snugging with a wrench is acceptable during maintenance and storage. The 150-200 inch-pounds is an' operational requirement only." '
The use of snoop during the leak check made on August 19, 1988 was an Lisolated instance and dispositioned in accordance with NCR-13908. LTM.
does not use leak' detection fluids in performance of their normal activities.
Action to Prevent Recurrence Following a thorough and complete review of the facts and circumstances concerning the HCU problem, management directed that LTM establish first lina' inspection responsibility for the maintenance activities performed under CP-G-3. In addition, to preclude any future recurrence of this type of problem, management prescribed additional training for QC inspectors.
This training reemphasized the requirements for documenting inspection activities which do not meet the acceptance criteria of Long Term Maintenance.
QUALITY ASSURANCE AND QUALITY CONTROL (LK-C-1907)
SECTION 4.8.2.4 Pressure Testing of Piping and Components a
Concern The NRC inspector reviewed and compared a sample of hydrostatic and pneumatic pressure tests to the requirements of the pressure testing procedure CP-M-2 and the ASME code requirements. Documentation for hydrostatic tests 2M-49A-12, 2M-45A-16, 2M-44B-10, 2M-45A-59, 2M-45A-10, and 2M-52A-06 was reviewed. Pneumatic test 2M-83-75 for HCC-234-E24 and HCC-234-E25 at 240 psi was observed in progress. During these reviews and observation, two potential problems were identified with the pneumatic test process. First, although the pneumatic test was conducted in accordance with the CP-M-2 procedure and Section III of the ASME Code, the leak detection fluid was applied to the pipe welds just before the examination for leaks rather than during the examination. This practice reduced the sensitivity for leak detection such that there was a slight possibility that a leak, if present, might not be observed. Second, the leak detection fluid (Snoop) was not removed from the stainless steel piping of HCC-234-E24 after completion of the examination. This omission was in conflict with page 1.4-1 of BWR Operations Manual NEDE 20583A, dated November 1978 which recommends complete removal of Snoop from stainless steel after use.
Response
Do Not Concur with Concern The inspections using Snoop are conducted in accordance with procedures for pneumatic testing which meet the requirements of ASME III NB/NC/ND-6000.
Snoop solution is applied to the joints just prior to or at the time of inspection. The inspectors raapply the fluid if they feel it is warranted.
Snoop is comprised primarily of deionized water containing a surfacant of the alkyl aromatic class. Snoop has less than (1) ppm total halogens.
- The total content of the low melting point metals mercury and lead is less than (2) ppm. Snoop contains none of the low melting point metals bismuth, antimony, tin or zinc. The vendor has concurred that Snoop is not deleterious to stainless steel.
f
INSTRUMENTATION (LK-C-1908)
SECTION 4.9.2.1 Concern RHR pump differential pressure transmitter PDT-51-2N058B:
The team found one of four Unistrut nuts securing the transmitter to the Unistrut channel in rack 20CO27 was not fully engaged with the Unistrut track. The bolt was snug and appeared to have been tightened with the Unistrut nut misaligned. Bechtel prepared a work package to correct the discrepancy during this inspection.
The team identified 11 instruments with at least one loose or misaligned Unistrut mounting nut. These QC accepted installations did not meet the requirements of Bechtel Specification M-830-G000, paragraph 8.7.
On the basis of this NRC finding, Bechtel instrument engineers performed a walkdown inspection of all instrument racks in Unit 2 to determine the extent of the deficiency and to identify additional misaligned Unistrut spring nuts. Bechtel wrote Startup Nonconformance Reports 283C-385/S-132-J and 251A-673/S-133-J to document and resolve the discrepancies. The NRC inspector reviewed the Startup Nonconformance Reports and determined that they adequately resolved the technical issue. However, the team believes that the root cause determination for such a large number of deficiencies in installations that were inspected and accepted by Quality Control should be evaluated, and measures should be taken to prevent recurrence.
Response
Concur With concern cause The cited condition is attributable to the as-receiveu condition of the vendor supplied instrument racks and lack of attention to detail during the fabrication process.
Extent of Condition A subsequent inspection was performed of all C.E. supplied instrument racks for similar conditions. Of a total of 964 spring nuts inspected, 209 were found to be deficient.
l significance Project Engineering has evaluated this condition found on other commodities to determine the safety significance of the deficiencies. These conclusions are contained in our response to Stone and Webster Construction Observation Reports COR-09 and COR-28.
E ________. _ _ _ _ _ _ . _ _ _ _ _ _ _ _ . _ . _ _ _ _ _ . . _ _ _ _ _ _ _ _ _ _ . . . _ _ _ _ _ _ _ _
Correctiv9 Action 1 The discrepant conditions were documented on NCR's 13915 and 139443 and Startup NCR's.S-132-J, S-133-J, S-135-J,'S-140-J, S-141-J, S-142-J, S-150-J and S-153-J. FDDR-HH2-6633 was issued by G.E. authorizing rework of all )
209 spring nuts; all rework is complete.
Action to Prevent Recurrence All instrument racks are installed and instrumentation installations complete; therefore, no further action is warranted.
1
~ - - - - - - - - _ _ _ _ , _ _ _ _ _ _ . _ _ .
INSTRUMENTATION (LK-C-1908)
SECTION 4.9.2.1 Concern Emergency service water pump pressure switch PSL-12-001D:
The team found two anchor bolts for the baseplate of hanger H1 were missing flat washers under the' nuts. Washers were recommended by Section 3.3 of Specification M-8031-C-64 and were present on other installations. The washers were also recommended by vendor manuals.
The panel nuts securing the two drain valves to their respective valve brackets were loose, preventing the brackets from holding the valves securely.
The external cover protecting the pressure switch adjustment mechanism was loose.
This level instrument was QC accepted and in service supporting LGS Unit 1 operation. The licensee issued Apparent Discrepancy Notification (ADN) From L-72 to identify and document the discrepancies of this instrument installation.
Response
The concern identifies two conditions that are unrelated and must be addressed individually. As a means of differentiation, they are hence-forth referred to as Concern #1 and Concern #2.
Response to Concern #1 Do Not Concur With Concern Specification M-8031-C-64, paragraph 3.3, states that washers may be omitted when... hole size is no larger than the bolt diameter plus 3/16 inch, and paragraph 3.5, which requires bolt holes to be nominal anchor /
bolt diameter plus 1/8 inch.
Response to Concern #2 Concur With Concern ,
cause The cited condition is attributable to an unauthorized alteration after final Q.C. acceptance.
i
l I
Extent of Condition A subsequent inspection has been performed of an additional 30 Unit 2 instruments involving 71 valves for loose dragon drain valve panel nuts j at bracket attachments and pressure switch adjustment mechanism external I covers; one isolation valve panel nut was found loose.
Significance This condition is not considered to be significant in that the valve bracket-to-support is rigidly attached and the connecting tubing on the inlet and outlet of the valve is supported such that any movement of the valve is prevented.
Corrective Action The deficiency was documented on Startup NCR-S-0350-M and dispositioned to rework in order to expedite ongoing testing.
Action to Prevent Recurrence Several recent issues of the Quality Counts Bulletin which is published to the job site weekly have contained information to heighten awareness to the fact that it is everyone's responsibility to report damage that may be caused in the course of performing a task and that alterations to installed commodities must be accomplished in accordance with approved procedures.
l INSTRUMENTATION (LK-C-1908)
SECTION 4.9.2.1 Concern RHR 'B' loop flow transmitter FT-51-2N015B: The team found one of three support bolts holding the Dragon 3 valve manifold to the instrument support was loose. Section 8.19.9 of Construction Specification M-830-G000, i Sheet 14, required bolt torque to be in accordance with the manufacturer's specifications. In the absence of specific vendor recommendations, M-830-G000 provided general specifications for various bolt sizes. Bechtel instrument engineers prepared Startup Nonconformance Report 249A-2065/S-129M which docu-mented the discrepancy and approved the corrective action of tightening the bolt to design requirements.
Response
Concur with Cited Condition cause The cited condition is attributable to an unauthorized alteration after final Q.C. acceptance.
Extent of Condition A subsequent inspection has been performed of an additional 24 instrument installations with valve manifolds installed (a total of 72 bolts) for similar conditionst no discrepancies were found.
Significance This condition is not considered to be significant in that two of the three bolts were found to be fully torqued, rendering the manifold installation immovable. In addition, this has been demonstrated not to be a prevalent condition.
Corrective Action The bolt found loose during the NRC inspection was documented on Startup NCR S-129-M which was dispositioned to rework; all work is complete.
Action to Prevent Recurrence Several recent issues of the Quality Counts Bulletin which is published to the job site weekly have contained information to heighten awareness to the fact that it is everyone's responsibility to report damage that may be caused in the course of performing a task and that alterations to installed commodities must be accomplished in accordance with approved procedures.
INSTRUMENTATION (LK-C-1908)
SECTION 4.9.2.1 Concern TE-76-223B for unit cooler 2BV210: The field lead terminal lugs at the temperature element had insulation damage on the terminal lug barrels caused by inadequate clearance between the inside of the threaded cover and the wiring and terminal lugs. Three other air cooler temperature detectors had similar or more severe damage on the field wires, detector wire, or terminal lugs and are discussed in Section 4.9.2.2.
Response
Concur With Concern This issue is addressed in our response to.the first concern discussed in Section 4.9.2.2.
i i
L-___-_-_____--_____-___-__-__-__-______ ._-_ _ .
INSTRUMENTATION (LK-C-1908)
Section 4.9.2.1
~ Concern Local control panel for unit cooler 2BV210: The white conductor (wire 388) on point 10 of terminal board TB-1 was bent excessively. The bend radius of the wire was only 1.5 to 2 times the wire's outside diameter. The NRC inspector was informed by Bechtel and PECo Staff that the Limerick' Station did not have a construction requirement for the bend radius of single conductors. .The inspector was. concerned that without a standard, no assurance could be provided that vendor recommendations and requirements-for safety-related wire bend radii ~would be maintained in future modifi-cations and installations..
Response
Do Not Concur With Concern The cited condition has been evaluated by Project Engineering for acceptability.
The criteria for minimum bending radius (4 x 0.D. of conductor) is contained in,ICEA Standard S19-81 under Ozone Resistance Test. This is a generic criteria and it stipulates the use of conductors under worst case _ service conditions. Based upon specific usage, the following observations have been made:
The ICEA minimum bending radius criteria is based on cables carrying rated current while the conductors in question are for instrumentation circuits and carry only a small fraction of rated current.
The ozone resistance properties of the conductor insulation used on Limerick (Ethylene Propylene Rubber) reduces the probability of insulation degradation as does the inherent conductor protection provided by the equipment enclosure.
For these reasons, the installed radius of the subject wires are considered acceptable. The wire manufacturer (Fluorocarbon) has also given their concurrence in this matter.
Although formal bend radius criteria is not a requirement, wires are always trained in a craftsman like manner and do not have extreme bends which would cause sharp points.
i
---,----n---_ _ _ - - ---_-_- - - - - - - _ - . - - - . - -
INSTRUMENTAL 20M (LK-C-1908)
SECTION 4.9.2.2 Concern The tcam found wire damage in all four Unit 2 compartment air cooler resistance temperature detectors (RTDs): TE-76-223F, 223H, 224F, and 224B. The damage was due to abrasion from the RTD threaded cover that was supposed to protect the wiring and terminal board. The configuration of the connections caused the lugs and. wires to protrude and come in contact with the cover whenever the cover was installed or removed.
Although the team found wires or terminal lugs showing evidence of abrasion from the cover in all four RTDs, the most severe damage was noted in TE-76-223F. The insulation of field wire number 2 in this RTD I was worn to the point of exposing bare conductor. A bare wire in contact with the terminal board screw and the cover could short circuit the RTD and make the air cooler fan control circuit inoperable.
The NRC team identified the problem to PECo and Bechtel staff, along with the concern that the problem may be generic to all air coolers in both Limerick units and RTDs of that type and manufacture. Bechtel issued NCR-13911 to document and correct the air cooler RTDs with the damaged cables in Unit 2, and ADN Form D-50 to evaluate the air coolers in Unit
- 1. The NRC inspector reviewed the nonconformance report and the apparent discrepancy notification. The NCR, although addressing the specific RTDs, did not address the probable generic nature of the problem for all similar installations in Unit 2. 'The generic nature of this issue should be addressed for all safety-related RTDs of this type.
Response
Concur with concern Cause The cited condition is attributable to the fact that the cover for the MINCO temperature elements is cutting the wire insulation whenever the cover is installed as a result of the cover not being a polished surface and the small amount of space available for making the field cable connections.
Extent of Condition This condition potentially exists in all MINCO temperature elements, Series S9322. .
Significance Two independent control circuito operate the air cooler:
- 1) Temperature Detector (RTD)
- 2) ECCS Pump Operation
1
- 1. RTD Control Circuit The control circuit of the RTD has been reviewed to evaluate the impact of a short circuit. The conclusion is that a short circuit in the RTD 1 would result in an indeterminate resistance / voltage signal to the tem-perature transmitter (TIT) which is used to start the air cooler fan when a high room air temperature is detected. Assuming the short circuit causes the RTD/ TIT to fail and in turn render the air cooler incper:ble, the room temperature would increase above the unit cooler temperature set point but not exceed design temperature since the normal non safety-related ventialtion system maintains normal room conditions while the ECCS pumps are not in operation.
Therefore, failure of the RTD control circuit does not affect the capability of the HVAC system to maintain the room within its design temperature values during normal plant operation.
The EQ is not a concern during normal plant operation since the EQ tempera-tures are not exceeded. The design relies on the normal ventilation system to maintain environmental conditions.
- 2. ECCS Pump Operation Control Circuit The second path of the control circuit that starts the air cooler is ECCS pump operation. This control circuit bypasses the RTD circuit completely. Any failure, a short or otherwise, in the RTD circuit does not render the air cooler inoperable during ECCS pump operation.
There are two 100% air coolers aligned to each safety related pump.
Each air cooler is designed to maintain room temperature whenever the L pump operates. The second air cooler would start only if the first air cooler either could not maintain the room temperature below the set point or fails to start.
In the first condition, inability to maintain room temperature, the RTD would attempt to start the second cooler. However, due to a short in the control circuit, the second air cooler would not start. Since each cooler can handle 100% of the pump load, the operation of the second air cooler is not required to mitigate the consequences of an accident.
In the second condition, failure of the first fan to start would automatically start the second air cooler fan, whether the RTD control circuit is operable or not, thus, one air cooler remains available to perform its safety function.
In conclusion, the safety function of the air coolers to maintain the pump compartment temperature at design conditions during ECCS pump operation is not compromised due to bare RTD wire and short circuit in the control circuit.
At least one air cooler would start automatically when the associated pump starts without relying on the RTD control circuit.
Corrsctiva Action Startup Field Raport 266C-002 was ganaratsd for Projact Engineering to evaluate this condition. As a result, it was concluded that the field connections into the terminal block of MINCO temperature elements, series S9322 only, could not be terminated back-to-back. A new method of' term-ination has been developed which requires the connection of the lugs under one screw to be side by side. Design Change. Notice, DCN #68, has been issued against Drawing E-1412 to document this change. Due to the rearrangement of the terminations, Startup Change Request, SCR 2ES-0396, was issued to incorporate these changes in Drawing TT-9513, E-705 sheets 33 and 14, and E-707.
Per Engineering instructions in the disposition of SFR-266C-002, if scuffed insulation was noted during the course of implementing this change, the j unacceptable portion of the wire was cut, recrimped and reterminated per DCN #68. In addition, when reterminating in the temperature element, the saddle was to be eliminated for better connection. No wire was to be looped higher than the lug to preclude future damage to the insulation.
Action to Prevent Recurrence The design change as described above has been incorporated into all MINCO temperature elements, series S9322, in accordance with Startup Work Orders 234B-021, 260A-014, 260C-040, 266C-009, 267A-003, and 267B-004.
INSTRUMENTATION (LK-C-1908)
SEf. TION 4.P.2.2 Concern The team found all four mounting bolts for pressure transmitter PT-ll-003B loose. The transmitter measured the unit common emergency service water B loop discharge header pressure. The bolts did not appear to have been tightened.
The NRC inspector reviewed ADM Form L-73 which the licensee prepared in response to the teams finding. The team noted that the form identified f and documented the discrepancy but had not yet been evaluated for appro-priate action by PECo QA. The NRC team recommended that the root cause of the discrepancy be determined and that measures be taken to prevent its recurrence.
Response
Concur With Concern The cited condition was on a Unit 1 instrument. Construction Engineering performed a subsequent inspection of 27 locally mounted instruments in Unit 2 for similar conditions; r.o transmitter to bracket mounting bolts were found to be loose (a total of 108 bolts were inspected). The root cause of the discrepancy will be addressed in the Unit 1 report.
--_----_-____-._.____m__.____ _ . _ . _ _ _ _ _ _ . _ . _ _ _
1 INSTRUMENTATION (LK-C-1908)
SECTION 4.9.2.2 Concern The NRC inspection identified loose flexible conduit fittings in the installations of seven instruments. Bechtel Instrument Engineers pre-pared a FDDR that documented the deficiencies identified by the team.
The NRC inspector reviewed the FDDR and concluded that it adequately addressed the problem.
Response
Concur with Concern cause The cited condition is attributable to field adjustment of *:ie vendor supplied connection to facilitate installation of the flex conduit.
Extent of Condition This condition potentially could exist at any of the vendor supplied 90* flex connectors.
Significance G.E. has evaluated this condition and considers it to be insignificant since the flex connector cannot unscrew from the device with conduit attached to it. There is no impact to plant safety or system function as a result of this condition nor does it affect the qualification of the instrument rack.
Corrective Action The discrepancies noted by the NRC inspector were documented on NCR-13odl which was dispositioned to rework in accordance with the response to G.E.
FDDR HH2-6631; all work is complete. FDDR HH2-6631 was dispositioned to add lockwashers at the request of Construction Electrical Engineering although the as-found installations would not have impacted plant safety or system function.
Actie.1 to Prevent Recurrence All installations are complete and, based upon significance, no further action is warranted.
INSTRUMENTATION (LK-C-1908) 1 l
l SECTION 4.9.2.2 Concern The team found that the instrument tubing supports for drywell pressure transmitters PT-42-2N091B, PT-42-2N094F, and PT-42-2N050B in rack 20CO27 exceeded the maximum allowable span. Construction Specification M-830-G038, Note 1, specified a maximum span of five feet (60 inches) for 3/8 inch diameter tubing. The team found that the tubing for these instruments had an unsupported span of about 80 inches. Bechtel instru- l ment engineers confirmed the overspan condition with the NRC inspector.
Since the instruments were in a rack supplied by G.E., Bechtel staff told the inspector that the installation would be evaluated in conjunction with G.E.
Response
Concur With Concern Cause The cited condition is attributable to the fact that Construction Engineering did not take into consideration,the unsupported G.E. supplied bu.khead connection and adjacent tube length as part of the total field run tubing span requirement as specified in 8031-M-830.
Extent of Condition G.E.'s rack bulkhead connections are, in most instances, supplied as a rigid attachment. All 30 G.E. supplied instrument racks (approximately 160 bulkhead connections) have subsequently been reinspected for additional overspans by Construction Engineering; this non-rigid attachment has been confirmed to be limited to four racks. Three of these racks had an addi-tional support added at the time of the tubing installation to the racks, the fourth being the subject of this concern.
Significance This condition has been evaluated by Project Engineering for safety signi-ficance and is addressed in our response to Stone and Webster Design Action Item DAI-365, which evaluated a span length of 8'2". This configuration was modeled in the ME101 piping computer program. The calculated results, using worst case enveloped response spectra for the entire Reactor Building (and containment wall) for the seismic and hydrodynamic inertia loadings, showed that the maximum primary stress level was less than 63% cf its allowable, and that the maximum tubing support reaction loads were 17 pounds in the horizontal direction and 10 pounds in the vertical direction. Both of these calculated maximum loads are well below the design reaction loads of the tubing supports. This analysis demonstrates the inherent conservatism with the support span spacing requirement criteria.
Corrective Action Construction Engineering issued work package JFR-42-07-350-D and an addi-tional support has been installed.
Action to Prevent Recurrence Construction Instrument Engineers have been reinstructed to consider unsupported vendor connections whenever span requirements are being evaluated.
INSTRUMENTATION (LK-C-1908)
SECTION 4.9.2.2 Concern The NRC inspector found a loose support strap on the instrument tubing bulk-head connector for pressure transmitter PT-40-2N055 in rack 20-C073. The inspector also noted that the instrument tubing contacted safety-related conduit 2DIO59. Construction Specification.M-830-G000, Sheet 14, required torquing of clamp bolts and specified the torque requirements. Section 5.11 of M-830-G000, sheet 7, required instrument lines be protected from mechanical loads and wear.
Bechtel prepared FDDR J-196 in response to the NRC inspector's finding. The FDDR identified the discrepancy and proposed torquing the clamp to the required 72 inch pounds. Securing the clamp properly should eliminate the contact between the tubing and conduit. The inspector reviewed the FDDR and concluded that the proposed action was adequate to address the technical issue.
Response
Concur With Concern Cause !
The cited condition is attributable to an unauthorized alteration after i final Q.C. acceptance.
Extent of Condition Construction Engineering performed a subsequent inspection of all G.E. supplied instrument racks for similar conditions. Of a total of 124 bulkhead connections i
inspected, 14 were found to be loose.
l Significance j The impact of force effects of the conduit onto the tubing, which can be conservatively estimated by multiplying the mass of the conduit with the response spectra acceleration and by a factor identified as an equiva-lent dynamic lead factor (which is normally taken as 2.0 for an impulse type loading), will not result in cross sectional deformations in the tubing or excessive loadings on the rack which would affect system design I
or function.
Corrective Action The noted deficiencies were documented on Startup NCR's S-369-E, S-395-J, S-382-J, S-373-M, S-372-J and SWO 264A-125 and dispositioned to rework in accordance with FDDR HH2-6625; all work is complete.
Action to Prevent Recurrence Several recent issues of the Quality Counts Bulletin which is published to the job site weekly have contained information to heighten awareness to the fact that it is everyone's responsibility to report damage that may be caused in the course of performing a task and that alterations to installed commodities must be accomplished in accordance with approved procedures.
~ ~~ ~
I l
INSTRUMENTATION (LK-C-1908)
)
SECTION 4.9.2.2 Concern During ff. eld inspections of instrument tubing and supports, the NRC inspector
-noted that galvanized materials welded or cut in the field were left untreated.
Construction specification M-830-G038, Sheet 1, Note 8 required that " field cut or welded surfaces of galvanized materials shall be treated with a gal-vanizing spray to protect the surfaces". l l
Bechtel responded to the NRC inspector's finding by initiating Balance of j
Plant Condition Report J400. The condition report identified and documented {
the discrepant condition for future correction. 1 l
L l
Response
Concur With Concern Cause The cited condition is attributable to a misunderstanding on the part of Construction Engineering of the requirements in M-830-G038 to coat the ends of cut-off galvanized material. The general notes of M-830, Paragraph 8.4, states that for painting and welding, refer to project requirements.
Construction ' Engineering understood this to mean that general project painting requirements should apply to cut or welded Unistrut.
Extent of Condition Not applicable based upon action taken to prevent recurrence.
Significance This condition has been evaluate 1 and it has been determined that non-treated field cut or welded surfaces of galvanized materials are not detrimental to the plant.
Corrective Action Bal[nce of Plant Condition Report J-400 was issued and has been disposi-tioned by Project Engineering to "Use-As-Is".
Action to Prevent Recurrence Field Change Request, FCR MI-758-J, has been approved to remove this requirement from Specification M-830-G038.
INSTRUMENTATION (LK-C-1908) 1 SECTION 4.9 2.2 l
Concern }
The NRC's inspection also identified a number of fastener discrepancies that consisted of missing flat washers on base plate anchor bolts and missing lock washers on instrument attachment bolts.
Hanger H2 for emergency se.vice water pressure switch PSL-12-001B tubing: missing flat washer on one baseplate anchor bolt attachment.
Emergency service water pressure transmitter PSH-12-004B: missing lock washer on one support bracket bolt.
Emergency service water pressure transmitter PSH-12-004D: missing l lock washers on two support bracket bolts.
Conduit hanger CI-1485: missing flat washers under all four anchor boIt nuts for the hanger baseplate.
Bechtel prepared ADN Form L-72 to document the NRC observations for the trans-mitters and tubing support. The NRC inspectors discussed baseplate ar ior flat washer discrepancies with the Bechtel Lead and Assistant Electrical Lead Engineers !
and was presented with information from Specification 8031-C-64 for anchor instal-lations. The specification recommended the use of washers for all expansion anchor installations and that washers could be omitted when the plate hole size j is no greater than the bolt diameter plus 3/16 inch. It further stated that all bearing surfaces shall be plane surfaces to ensure that tightening of the nut has the desired seating effect on the anchor. The NRC inspectors had the following concerns regarding this portion of the specification:
" Since omission of flat washers for expansion anchors was allowable, then the expansion bolt hole size for those baseplates should be a
! quality attribute that should be QC inspected and recorded during assembly.
The specification did not provide criteria to determine if surfaces f were plane and, therefore, that determination was left to the judge-ment of the installation craftsmen, i
The specification's exception for the use of washers was independent j
of the anchor's use and expected load. The exception was also incon-61 stent with vendor recommendations, drawings, and qualification testing for specific anchors.
The NRC inspectors noted that most baseplace anchor bolts had flat washers installed and it appeared that a few isolated electrical conduit supports were missing washers. The team also noted that the intent in constructing the plant was to install washers under nuts on all anchor bolts. The NRP team recommended that PECo and Bechtel evaluate the appropriatences of Sp.cification C-64 requirements for expansion anchors.
1 1
Response
Do Not Concur With Concern The cited condition has been evaluated and we conclude the following:
Expansion anchor allowable ' loads contained in Specification C-64 are independent of the anchors' use. Omission of washers, as permitted by the specification, is similarly independent of the anchors' use and expected load. Washer omission at anchors installed to achieve specified torques has no significant affect on anchor capacities.
This has also'been confirmed by the anchor manufacturer.
l Anchor qualification testing and vendor recommendations do not identify washers as a significant parameter. Drawings for commodity support connections using expansion anchors do not specifically require washers. Expansion anchor hole sizes in baseplates are inspected as required by Section 3.6 of Project Quality Control Instruction (PQCI) 18240/C-1.50. Criteria to determine if bearing surfaces are plane is given in Section 3.3 of Specification C-64 by the requirement that the minimum allowable angle between the axis of the anchor and the bearing surface shall be 65 degrees.
This criteria is addressed in Section 2.1.B.l.(b) of the PQCI.
The PQCI, in Section 3.5, also requires that proper nut contact with +!.e bearing surface be verified in accordance with Section 3.3 of Specification C-64.
l
L" INSTRUMENTATION (LK-C-1908)
SECTION 4.9.2.2 Concern The'NRC inspector noted an apparent' interaction violation between an un-insulated steam pipe and a safeguards' junction box and safety-related conduit 2AJ170. The aninsulated pipe, EBD-205-E3 with a design surface temperature of 550 degrees'Fahrenke4.t, was routed within 2 inches of the face of'the junction box. Construction Specification M-830-E-1406, Section 2.9 required that uninsulated pipe with-surface temperatures above 200 degrees Fahrenheit be evaluated by Project Engineering for adequate clearance from electrical raceways.
The NRC inspector discussed the installation with Bechtel staff who provided NCR 13913 and PIN 369 dated August 11, 1987. .The inspector had the following concerns after reviewing the NCR and PIN.
PIN 369 stated that conduit 2AJ170 was reworked to achieve 3-1/4' inches separation between the bare pipe and the junction box. The PIN also required 2 inches of insulation on the pipe. The NRC field inspection of the installation found the pipe uninsulated and.only'2 inches of' clearance between the pipe and the junction box.
The PIN did not indicate any QC or QA action or acceptance.
The PIN was signed off as being " closed out" with "N/A" indicated in the'" Action Completed By" block.
The NRC inspectors concluded that the installation was not in accordance with requirements or the PIN, and had not been reworked as stated in the PIN.
Additional review and evaluation of corrective action for this deficiency were warrantew.
Response
. Concur with Concern Cause The cited condition is attributable to the following:
PIN-369, Revision 0, was written on 6/4/87 to document the fact that 2AJ170 did not meet the hot pipe separation requirements of E-1406 (1-1/4 inch clearance existed). The condition was evaluated and found unacceptable by Project Engineering and the PIN was dispositioned to rework the junction box to achieve 2" minimum separation from the insulation of the pipe.
_ _ _ _ _ _ _ ___. _ . . . _ _ _ _ - _ _ _ _ _ _ _ _ _ _ . _ - _ . . . _ - . - _ . _ . - - ~ - - _ . _ _ _ - - _ . _ - . _
- - The rework to the junction box was performed and PIN-369, Revision
-l', was written on 8/11/87 to furnish the new as-built dimension to Project Engineering for evaluation (a clearance of 3-1/4 inches now existed). This condition.was found acceptable by Project-Engineering ~and,that is why "N/A" appears in-the " Actions' Completed.
By" portion'of PIN-369, Revision 1.
.What wasn't realized at'the time was'the' fact that the pipe (SP-EBD-205-E3)-that 2AJ170 was in violation with was supported with temporary hangers. Subsequent installation of the permanent supports moved the line,.thus.creat'.Lg the condition found during this inspection.
Extent of Condition Not applicable based upon action taken to prevent recurrence.
Significance Not applicable _ based upon action taken to prevent recurrence.
Corrective Action NCR-13913 was written to document this condition and dispositioned in accordance with PIN-369, Revision 2, to add 2" thick MIN-K insulation blanket to cover the entire exposed section of pipe SP-EBD-205-E3; all work is complete.
Action to Prevent Recurrence This deficiency _would have been identified in the course of the final stress reconciliation walkdown performed by Project Engineering'as previously discussed and, therefore, no further action is warranted.
INSTRUMENTATION =(LK-C-1908)
SECTION 4.9.3-Concern l
l The' inspector gained additional insight on the. involvement of site QC inspection personnel in the construction process during review of completed work packages and quality control inapection records (QCIRs).
l' The NRC inspector noted that QC personnel were involved in the work i
activities from work planning tc final inspection-of the activities.
However, the deficiencies identified by the NRC team in the mechanical, HVAC, and instrument 4 tion areas, particularly the equipment mounting
( . discrepancies, indicated that QC activities were not completely effective in all cases.
The number-of discrepancies found during NRC field inspections of in-struments both within and outside the SWEC sample gensrall/ indicated that the quality of work by craftsmen was above average. The type of discrepancies found on QC-accepted installations would suggest weaknesses in attention to detail in ecceptance of completed installations or lank of detailed checklists for final verification.
Response
Do Not Concur with Concern We have reviewed the cited conditions and conclude that the vast majority of the valid concerns were noted to be relatively isolated cases which were assessed by Engineering to be minor in nature and can be categorized as follows:
- Deficiencies associated with Q.C. acceptance of commodities
- Deficiencies associated with the degradation of commodities after Q.C. acceptance
- Specification interpretation The only discrepancies attributable to the first category are those identified in the area of mechanical equipment mounting. As discussed in the response to each of the specific hardware concerns related to this subject, conflicting / confusing design requirements and lack of attention to detail during drawing review were the root causes of the problem. It has been determined that this condition is isolated to this issue as has been demonstrated by the Stone and Webster IDCA and this Inspection Report.
Fe provide a graduated program of assurances that a level of quality is maintained on safety related equipment sufficient to assure its function.
Each program provides controls to ensure that the work receives the appro-priate inspection. The controls for Bechtel Construction are provided
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ . _ . . - i
1 l^
in the Construction Procedures (CP's). These controls ensure the work package is routed through QC for the preparation of their ir.spection documentation. When blue tag testing is in progress, the controls are established in PECo Electrical Engineering Procedures (EE's). These procedures require the rework notice to be routed through QC for inspec-tion documentation. SWO's are used by the Startup organization to authorize all work. Their controls are noted in the Startup Administra-tive Procedures (AD's) and they include provisions for QC inspection.
These control functions eneure that all work on systems and facilities are carefully controlled under our quality program.
The discrepancies attributable to the second category are also discussed in the responses to each of the specific concerns. In addition, our quality program recognizes the possibility of inadvertent damage after i QC inspection. Two programs have been developed to ensure that the damage does not go ur. detected.
The first is a system turnover program. Prior to system turnover to Startup, a walkdown is performed by a multidiscipline team of Construction Engineers, Quality Control Engineers and Startup personnel. This walk-down is done to ensure that all construction is complete or noted as an exception and that any damage is documented. In conjunction with the walkdown, a QC records review is performed to ensure all inspections are completed and all open items are entered on the punchlist. Once turned over to Startup, the system is tagged to clearly identify that the system is under Startup control.
The second system developed is the facility turnover process. Again, a walkdown is performed by a multidiscipline team. This team ensures that the facility and its components are damage free and note that all con-struction is complete or documented as an exception. These areas are then locked-down to the extent possible to prevent unauthorized entry.
The two walkdown programs are described in Bechtel Construction Procedures CP-T-1 and CP-T-2.
Assurance that the design is maintained after post acceptance testing is recognized as an important element of our AD's. These documents assure that all design changes are properly controlled. In addition, these controls include provision for QC inspection and retest after work completion.
With only one exception, the concerns associated with specification inter-pretation have been demonstrated to be justifiable.
- - _ - - - _ _ . _ _ _ _ _ _ _ _ _