ML20234E348
| ML20234E348 | |
| Person / Time | |
|---|---|
| Site: | Comanche Peak  |
| Issue date: | 12/17/1987 |
| From: | TEXAS UTILITIES ELECTRIC CO. (TU ELECTRIC) |
| To: | |
| Shared Package | |
| ML20234E089 | List: |
| References | |
| TAC-R00284, TAC-R284, NUDOCS 8801080040 | |
| Download: ML20234E348 (288) | |
Text
- - - - - - - - -
gOMANCHE PEAK STEAM ELECTRIC STATION UNITS 1 AND 2 COMANCHE PEAK RESPONSE TEAM RESULTS REPORT FOR ISAP - Vll.C OCONSTRUCTION REINSPECTION / DOCUMENTATION REVIEW VOLUME IV-C MECHANigL POPULATION REPORTS TEXAS UTILITIES GENERATING COMPANY A DIVISION OF O TEXAS UTILITIES ELECTRIC COMPANY
[BR "388R Bussie PDR
I R2 Vision: 1 i Pego 1 of 41 l l
/T RESULTS REPORT l \_-) !
I ISAP VII.c (Cont'd)
Appendix 8 Large-Bore Piping Configuration i 1.0 REVIEW PROGRAM IMPLEMENTATION 1.1 Construction Work Category Description The construction work category of large-bore piping configuration is comprised of all safety-related large-bore piping (2-1/2 inch d1ameter and larger) and includes all in-line components; instrumentation control and indicating devices; branch connections; and vent, drain, and instrumentation piping up to and including the first root isolation valve.
1.2 Population Size and Sample Selection For this construction work category, a population of I t% approximately 1,050 items was identified as QC-accepted as of the date of this reinspection effort. The piping and is_) components on each safety-related QC-accepted piping isometric drawing were designated as an item in this population. Of the total of 100 accessible items reinspected 81 first sample items were randomly selected from the population to ensure that all attributes in the first random sample were ;
reinspected a minimum of 60 times except as noted below. An additional 15 second sample items were randomly selected to ensure that at least 60 safe-shutdown hardware items were reinspected.
The expansion joint installation attribute was reinspected three times; however, it was not necessary to expand the sample to reinspect 60 expansion joints because a construction deficiency was identified in one of the three sampled expansion joints, resulting in a recommendation to perform a reinspection of the rest of the population of safety-related expansion joints.
1.3 Attributes Selected Sample items were reinspected for the following attributes:
Attribute 1 - Piping and component identification Attribute 2 - Piping and component sequence and (n)
~
location
Revision: 1 Page 2 of 41 RESULTS REPORT ISAP VII.c (Cont'd)
Appendix 8 (Cont'd) 1.0 REVIEW PROGRAM IMPLEMENTATION (Cont'd)
Attribute 3 - Orientation of piping and components Attribute 4 - Piping clearances Attribute 5 - Expansion joint installation A documentation review was not performed as all attributes of piping configuration could be reinspected.
2.0 DISCUSSION OF RESULTS 2.1 Summary of Results For reinspection 52 Deviation Reports were issued describing 120 deviations. Approximately 6,000 inspection points were encountered in performing the reinspection. See Table 8-1 for results of the reinspection.
In the five attributes of this population, one special case, one unclassified trend and one construction deficiency were identified, as follows:
- 1) Attribute 3 - Orientation of piping and components:
Special case for flow element flow direction. See Section 2.2.3 for discussion of this attribute.
- 2) Attribute 4 - Piping clearances: Unclassified trend for piping clearances. See Section 2.2.4 for discussion of this attribute.
- 3) Attribute 5 - Expansion joint installation:
Construction deficiency for expansion joint tie rod installation. See Section 2.2.5 for discussion of this attribute.
2.2 Analysis of Reinspection Results This section provides, by attribute, a discussion of the O re7crted deviations, an analysis of the effect of the deviations on the function of the piping, and an analysis for the presence of trends. Piping configuration is important in order to maintain the structural integrity and flow capability of the piping system and to maintain system performance.
I Rsvision: 1 Pogs 3 of 41 f]
b RESULTS REPORT ISAP VII.c (Cont'd)
Appendix 8 (Cont'd) 2.0 DISCUSSION OF RESULTS (Cont'd) 2.2.1 Attribute 1 - Piping and Component Identification !
l Piping and component identification includes reinspection to verify compliance with the piping isometric drawing for nominal pipe diameter; the i presence of branch connections, screwed joint sealant and elbows; and identification of valves, orifices,
. flow elements, expansion joints, strainers, and moment restraints.
In 1,135 inspection points, the piping and component identifications were verified to be as required by the piping isometric drawing. No deviations were reported.
2.2.2 Attribute 2 - Piping and Component Sequence and 1 L Location Piping and component sequence and location includes reinspection to verify compliance with the as-built piping isomeltcic drawing for pipe centerline elevation; pipe location; pipe lengths; valve centerline elevation; mechanical joint in-line location; flow element and orifice in-line location; and component sequence. No deviations were reported for the reinspection for flow element and orifice in-line location and component sequence.
From 3,607 inspection points to verify piping and component sequence and location, 36 deviations were reported concerning 17 of the 100 sample items. None of the deviations was evaluated to be a construction deficiency.
The deviations involved conditions where the pipe centerline elevation, pipe location, pipe lengths, valve centerline elevation or mechanical joint in-line location were not in accordance with the as-built piping isometric drawing and the specified tolerance. .
I 1
- p. The specified tolerance for all of the above {
dimensions, except spray norzle orientation, is 2 i
\' inches. Of the 36 deviations, 16 exceeded the l tolerance by less than 1 inch; 11 exceeded the l t
I
Ravision: 1 Page 4 of 41 RESULTS REPORT ISAP VII.c ,
(Cont'd) l l
Appendix 8 )
(Cont'd) 2.0 DISCUSSION OF RESULTS (Cont'd) tolerance between 1 and 2 inches; and five deviations exceeded the tolerance by 2 inches up to a maximum deviation of 4-1/4 inches. The remaining four deviations involved two pipe dimensions that were 17 feet 9-1/8 inches and 2 feet 10-1/4 inches out-of-tolerance due to drawing errors and two spray nozzles that were 4.5 degrees out-of-tolerance.
An evaluation of each of the 32 deviations that ranged. '
up to 4-1/4 inches out-of-tolerance was performed to determine the change in the pipe stress and pipe support loads due to any change in pipe length or weight shift caused by the deviation. All pipe length or weight shift changes had a negligible effect on the pipe stress and pipe support loads. !
l In addition, a review was made to determine what effect each of the 32 deviations would have on system performance and flow capability. The changes in system pressure drops were negligible because 1) all pipe length changes resulting from these deviations were negligible in comparison to the overall pipe length in each system, 2) no nominal pipe diameter deviations were reported, 3) no additional fittings or components were found to be installed, and 4) there were no significant changes in elevations. Therefore, these deviations had a negligible effect on system performance and flow capability. Based on the above evaluations, the 32 deviations were determined to be insignificant.
The largest location deviation reported was a piping location that was 17 feet 9-1/8 inches out-of-tolerance from the location specified on the piping isometric drawing. A review of this condition determined that the piping had been installed ccrrectly and that the piping isometric was in error. As a result, a design change (DCA-31,009) has been issued by TU Electric to revise the piping isometric and a memo has also been
[". forwarded to the Design Adequacy Program (DAP). This
%J deviation was det. ermined to be insignificant. The other larger location deviation was the sum total of
Revision: 1 Page 5 of 41 3
,m
( ,)
RESULTS REPORT ISAP VII.c (Cont'd)
Appendix 8 (Conc'd) 2.0 DISCUSSION OF RESULTS (Cont'd) several individual pipe length dimensions that was 1 out-of-tolerance by 2 feet 10-1/4 inches when compared l with the overall dimension specified on the piping isometric drawing. A review of this condition determined that the piping had been installed correctly '
and that the overall dimension on the piping isometric was in error. TU Electric has been informed of this drawing error and a memo has also been forwarded to DAP. This deviation was determined to be insignificant.
The out-of-tolerance dimensions in the range found during the reinspection have a negligible effect on the gx-existing pipe stress, pipe support loads and system performance. In addition, due to the similarities in
(' ') the analytical methods used for ASME III Code Class 1, 2, and 3 piping and the insignificant impact resulting from the range of the out-of-colerance dimensions found, the effect would not be greater for any class of ASME III piping. Therefore, if deviations of a similar type were to occur in the uninspected portion of the population, it is expected that they too would have an insignificant effect on the functional capability of the piping configuration. Therefore, no adverse trend
.was identified.
The as-built piping isometric drawings, which are the piping drawings of record, were used as the basis for dimensional reinspection. Although the deviations were all evaluated to be insignificant, the large number of deviations that was found is more than would be expected following a comprehensive as-built p ogram.
As a result, a recommendation for improvement, which is applicable to both the Large and Small Bore Piping Configuration populations, is provided in Section 4.2 of Appendix 9.
The two spray nozzles with deviations exceeding the tolerance are part of the containment spray system,
()
which functions to remove heat from the post-accident
(/ containment environment. An evaluation of the two i
spray nozzle deviations was performed to determine the effect that the spray nozzle's orientation would have on the spray pattern and coverage inside the containment. The nozzles are located adjacent to the
Revision: 1 Page 6 of 41
() RESULTS REPORT ISAP VII.c (Cont'd)
Appendix 8 (Cont'd) 2.0 DISCUSSION OF RESULTS (Cont'd) shield wall and directed down and away from the center of the containment. Due to the increased nozzle angle, the spray will project further from the center of the containment resulting in increased water impingement on the containment wall. The evaluation determined that the spray nozzle coverage volume would decrease by approximately 3 percent, and that the volume ratio of the sprayed region to the total region would decrease by 4 percent. Because these deviations result in a negligible ef fect on the spray volume and volume ratio, these two deviations were determined to be insignificant.
A bounding analysis was performed to determine what O
effect a similar deviation occurring at the same frequency on the uninspected nozzles in the containment spray system would have on the functional performance of the system. The evaluation indicated that the overall effect on system performance would not be significant. Therefore, no adverse trend was id+atified.
2.2.3 Attribute 3 - Orientation of Piping and Components Orientation of piping and components includes reinspection to verify compliance with the piping isometric drawing for pipe slope, valve stem angle, valve flow direction, orifice flange tap orientation, flov element flow direction, strainer flow direction, and eccentric reducer orientation. No deviations were reported for the reinspection for valve flow direction, orifice flange tap orientation, strainer flow direction, and eccentric reducer orientation.
From 309 inspection points to verify orientation of piping and components, three deviations were reported:
one for flow element flow direction, one for pipe slope, and one for valve stem angle. Each of the deviations was evaluated not to be a c:nstruction deficiency.
1 I
Rsvision: 1 Pegs 7 of 41 l
1 f RESULTS REPORT ]
k.-)/ ISAP VII.c
\
i l
(Cont'd) l Appendix 8 (Cont'd) l 2.0 DISCUSSION OF RESULTS (Cont'd)
The first deviation was a flow element located in a 4-inch diameter component cooling water 31ne that was installed in the flow direction opposite to that required by tre piping isometric drawing. An evaluation of the reversed flow element was performed to determine the effect of this installation on the system functional capability.
The flow element functions to monitor fluid flow in the component cooling water discharge line from control ,
room air conditioning condensors and, upon low-flow conditions, causes a signal to be sent to a low-flow alarm in the main control room. The flow-indicating transmitter and the low-flow alarm associated with this flow element are both claasified as non-safety-related.
f- g A separate flow control loop associated with the
() control of flow (via flow control valves) at the air conditioning condensor outlets is safety-related and is not affected by the inaccuracy associated with this reversed flow element.
The flow through the flow element is measured by correlating the pressure drop across the flow element to a flow rate. By reversing the orientation of the flow element, less pressure drop occurs for the same flow rate. As a result, the indicated flow will always be less than the actual flow rate passing through the line. .
The evaluation determined that the actual cooling water flow rate through the reversed flow element will not decrease; however, an inaccutate (lower) flow measurement will result. This deviation was considered to be notable. Because this flow element does not affect a safety-related function, the deviation will not have any adverse safety-related effect on the piping system's performance of its intended function.
Although the deviation identified did not result in a construction deficiency, if a component that was I, associated with a safety-related function was reversed
's-/
1 in the uninspected portion of the population, a construction deficiency could result, though
Revision: 1 Page 8 of 41 RESULTS REPORT ISAP VII.c (Cont'd)
Appendix 8 (Cont'd) 2.0 DISCUSSION OF RESULTS (Cont'd) significant errors would likely be detected during testing. Based on the low potential for a construction deficiency and the low frequency of occurrence, no adverse trend was identified. However> because proper orientation may be important to system operation, it is prudent to correct any similar deviations, if they exist. Therefore, this orientation deviation was identified as a special case. A root cause and generic implications analysis is provided in Section 3.1. -
The second deviation was a steam pipe slope that did not meet the minimum slope specified on the piping isometric drawing. A review of the actual slope revealed that the pipe sloped in the correct direction; O however, it was slightly less than the specified angle of slope. As a result, the deviation would have a negligible effect on the drain capability of the steam pipe. A review was performed to determine what effect a similar deviation would have on steam piping in the uninspected portion of the population. The results of the review revealed that no other piping isometrics specified pipe slope angles and vertical offsets for steam lines in a similar manner to that contained on this isometric. Had the alternative form of specification been applicable to the subject case, there would have been no deviation, given the applicable tolerances. Therefore, this type of deviation will not occur in the uninspected portion of the population and no adverse trend can result.
The remaining deviation was a valve stem orientation on a 4-inch manual plug valve that did not meet the standard industry practice of installing 2-1/2 inch and larger valves in horizontal piping with the valve stem in a vertical upright position unless otherwise indicated on a piping drawing. The valve stem was oriented in a horizontal direction instead of a vertical upright direction.
The valve is located on the suction line to the prelube pump of a standby diesel generator. The valve and associated piping are supplied and designed by the diesel generator manufacturer. Installation is
Revision: 1 Page 9 of 41 RESULTS REPORT ISAP VII.c (Cont'd)
Appendix 8 (Cont'd) 2.0 DISCUSSION OF RESULTS (Cont'd) performed onsite, under the general direction of the manufacturer's representative. The manufacturer's piping drawing does not specify a valve stem orientation. A memo has been forwarded to the DAP regarding the lack of a valve stem orientation criterion on the part of the diesel generator manufacturer.
An evaluation was performed to determine what effect the horizontal valve stem orientation would have on the pipe stress and system flow performance.
Because the mass of the valve gear box and operating handle is small and the distance of the gear box and operating handle from the centerline of the pipe is O approximately 12 inches, the valve stem orientation will have a negligible effect on the pipe stress. In addition, this type of plug valve can operate properly in either the horizontal or vertical position. Its orientation vill have no ef fect on system flow parameters.
In addition, a review of the area in which this valve is installed revealed that the available valve stem positions in which the valve could be installed were limited due to hardware in the area. The current location of the valve ster provides good access for operation of the valve handwheel and the stem orientation for this valve is the same on all four diesel generators.
In accordance with CP-CPM-6.9E, where reorientation is required to avoid interferences, valve operators may be rotated in accordance with site engineering and/or manufacturers' instruction. No documentation of such an instruction was located, hcwever. Nevertheless, based on the circumstances of installation as discussed above the condition is acceptable as is and does not have any implications into the uninspected portion of the popul.ation. Therefore, no adverse trend was
() identified.
Ravision: 'l Pags- 10 of 41 RESULTS REPORT ISAP VII.c (Cont'd)
Appendix 8 (Cont'd) 2.0 DISCUSSION OF RESULTS (Cont'd) 2.2.4 Attribute 4 - Piping Clearances In 940 inspection points to verify piping and component clearances, 76 deviations affecting 24 of the 100 sample items were reported where the clearances Setween two pipes or between a pipt and other hardware (i.e.,
sleeves, supports, cable tray, and conduit, etc.) were less than that required by the piping erection specification. Verification of piping clearance is performed in order to assure that the mechanical ,
interaction between piping and hardware, under design loading conditions, is acceptable. The criteria for piping clearances varied depending on the temperature of the piping and the item with which clearance was to be maintained.
A summary of the reported deviations for piping clearance is contained in Table 8-2. All but eight of the deviations occurred where at least one item was insulated. Of the 76 deviations, 75 were determined to be insignificant and one was determined'to be notable.
The evaluation of the deviations consisted of determining 1) the heat transfer effect on the subject pipe and hardware as a result of the reduction in clearance and 2) whether the combined movements under design conditions were less than the available clearance, thereby avoiding hard contact.,
For all 76 deviations the heat transfer effect on the subject pipe and hardware as a result of the reduction in clearance was reviewed. The clearance deviations were localized in nature. This being the case, the l heat transfer effects due to a reduction in clearance were determined to be negligible.
In addition, an evaluation of each of the deviations was performed to determine if the combined movements were less than the available clearance, including allowances for compressibility of the insulation. If
(.-s the combined movements were less than the available clearance, the deviation was determined to be insignificant. If the combined movements resulted in
1 1
Revision: 1 Page 11 of 41 RESULTS REPORT
(-)-
ISAP VII.c (Cont'd)
Appendix 8 (Cont'd) 2.0 DISCUSSION OF RESULTS (Cont'd) hard contact between pipes or between a pipe and other hardware, further evaluations were performed to determine the effects on the pipe stress and pipe supports, and the effects on the impacted hardware.
Of the 76 deviations, 68 concerned piping and other piping or hardware where at least one item was insulated. The results of the evaluations for the combined movement indicated that 37 of the deviations-would not result in contact. The remaining 31 deviations would result in the insulation touching the ,
adjacent item. In all but one of these 31 cases, the insulation was sufficiently compressible to accommodate the movement with no measurable increase in the pipe
/ s) strecs or pipe support loads, or to have any ef fect on
\/
~ the hardware. Therefore, 67 of the devigtions were determined to be insignificant.
In the last of the 68 deviations, no clearance existed between an uninsulated 3-inch diameter pipe and a 4-inch diameter pipe insulated with reflective insulation. In this case, the reflective insulation was determined not to compress sufficiently to permit the predicted pipe movement. As 4 result, an evaluation was made to determine the additional pipe stress and support loads resulting from the interaction of the insulated and uninsulated pipe.
The analysis indicated that under design conditions, the impact load would cause the existing pipe stress in the 3-inch pipe to increase by a factor of seven and the existing stress in the 4-inch pipe to increase by 30 percent. A comparison of the new pipe stress levels for each pipe to the ASME-Code allowable stresses indicates that these stress levels are a maximum of 60 percent of the allowable values. Therefore, the new stresses are still within Code-allowable strasses and are acceptable. An additional evaluation was performed to determine if any local pipe wall deformation would
(~') occur on either pipe as a result of contact. The
(,,/ results of the evaluation indicated that the localized pipe stresses from contact fall within ASME .
Code-allowable stresses, and therefore no local pipe l wall deformation will occur. l
i Rsvision: 1 Page 12 of 41 j rs
- (,,) RESULTS REPORT I
ISAP VII.c (Cont'd)
Appendix 8 (Cont'd) 2.0 DISCUSSION OF RESULTS (Cont'd)
An evaluation of the effect of impact on the adjacent pipe supports for each line was also performed. The original design load for each of the supports was small relative to the capacity of each support. In the worst case loading condition, which includes the impact load, the new support load increased by approximately 63%.
However, since the new support loads are still a maximum of only 56% of the support design capacities, the existing pipe supports are acceptable.
As a result of.the increase in pipe stress resulting from the impact load caused by contact, this one deviation was determined to be notable.
s The eight remaining clearance violations involved
() uninsulated pipe and uninsulated pipe or hardware. In all eight cases the existing clearances were sufficient for the pipe movements involved, and no contact between pipes or hardware would occur under design conditions.
Therefore, these eight deviations were determined to be insignificant.
In summary, none of the clearance deviations affected the functional capability of the related piping, pipe supports or adjacent hardware. All of the deviations are insignificant with the exception of the one notable deviation, where the pipe stress design margin is reduced, but the pipe stresses are still within code-allowable stresses.
Based on the number of pipe clearance deviations found, a trend was identified. Although only one deviation was determined to be notable, it is difficult to determine whether more significant clearance deviations are likely to exist with pipes located in the uninspected portion of the population. This determination would require that representative piping with stress levels close to ASME-Code allowables be identified and evaluated for the effects of the worst l
case clearance. Based on the scope of this effort and I
\> the potential for additional clearance deviations dee to the large quantity of pipe and hardware installed
Revision: 1 Pego 13 of 41 RESULTS REPORT (v j ISAP VII.c (Cont'd)
J Appendix 8 (Cont'd) 2.0 DISCUSSION OF RESULTS (Cont'd) onsite, an. unclassified trend for piping clearances was identified. A root cause and generic implications analysis is provided in Section 3.2.
2.2.5 Attribute 5 - Expansion Joint Installation Expansion joint installation was inspected to verify that each expansion joint was installed in accordance with the expansion joint manufacturer's drawing, installation instructions and/or design drawings. In addition, an inspection was performed to verify that the installation of the connecting pipe had not resulted in the expansion joint being out of alignment.
In the 18 inspection points for expansion joint
(/
\_-
)
installation, five deviations were reported on one expansion joint that connects a staticn service water pump to its discharge piping.
Four deviations were reported for tie rod installation where only one of four tie rod nuts on the expansion joint was hand-tight, and all four tie rods had only one nut on each end instead of the two required by the design drawing.
The function of the tie rods is to restrain the axial movement due to thrust imposed on the expansion joint.
To perform this function, the tie rods ath required to be uniformly tight and both nuts are required to be installed so that the initial setting of the maximum extension of the expansion joint is maintained. In this application there are no other axial restraints on the pipeline to prevent axial movement of the expansion joint when the tie rods are loose.
An evaluation was performed to determine the effect that the reported condition would have on the integrity of the expansion joint. The analyais was performed assuming that only one tie rod was engaged. The analysis indicates that under design loading j
l
()
(_,) conditions, the load on a single acting tie rod would exceed the safeload of this tie rod by a factor of 5.4.
Similarly, assuming all tie rods were tight and shared 1
_ -- _ -__ x _
'I Rsvision: 1 Pass 14 of 41 RESULTS REPORT ISAP VII.c I (Cont'd) f1 Appendix 8 (Cont'd) 2.0 DISCUSSION OF RESULTS (Cont'd) the load equally, the load on each tie rod would exceed the safeload by a factor of 1.3. Because the safeload of.each tie rod is exceeded under design loading conditions, a memo has been forwarded to the DAP.
Under normal operating conditions, the load on a single acting tie rod would exceed the safeload by a factor of 3.7. If the load were equally distributed over all four tie rods, the load on each tie rod would be within the recommended safeload.
The omission of a second nut (jam nut) on each tie rod was also evaluated. Without jam nuts, the potential exists that the single nuts may migrate due to normal
(h. system vibration, resulting in expansion joint
\m / extension beyond the specified design length, with a potential for overstress of the expansion joint.
Based on the above analyses, it was concluded that the tie rods and the expansion joint may not be able to perform their safety-related functions over the lifetime of the plant. While failure may not occur, these four deviations of improper tie rod installation were determined to be a construction deficiency because the calculated design loads are also above the ;
safeload. A root cause and generic implications analysis is provided in Section 3.3.
The remaining deviation was reported where the as-installed end-to-end dimension of the expansion joint exceeded that required by the expansion joint manufacturer's drawing. The tolerance for installation is within 1/8 inch of the designated length. The actual measured length exceeded this tolerance by 3/8 inch. An evaluation was performed by the expansion joint manufacturer to determine the effect of the extension of the expansion joint on its ability to operate. The evaluation indicated that the expansion joint will operate satisfactorily in its extended position. Therefore, this deviation was determined to be insignificant and no adverse trend was identified.
Rsvision: 1 Pago 15 of 41
(')
v RESULTS REPORT ISAP VII.c (Cont'd)
Appendix 8 (Cont'd) 2.0 DISCUSSION OF RESULTS (Cont'd) 2.3 Analysis of Documentation Review Results A documentation review was not performed as all attributes of piping configuration could be reinspected.
3.0 ROOT CAUSE AND GENERIC IMPLICATIONS This section provides a root cause and generic implication analysis for the following:
- 1) A special case for a flow element flow direction.
- 2) An unclassified trend for piping clearances.
/
\_,)\ 3) A construction deficiency for expansion joint tie rod installation.
3.1 Flow Element Flow Direction Root Cause The following is a summary of the relevant information regarding the flow element flow direction deviation.
Prior to system turnover by Construction, Startup personnel are required to conduct a walkdown of each system in accordance with procedure CP-SAP-20, Guidelines for System Walkdown Inspections. This procedure requires, in part, that flow elements are verified to be properly oriented in the direction of flow. In the event that any discrepancies are noted during the system walkdown, the procedure requires that they be entered on the master data base or noted on the System Turnover Package. A review of the turnover package for this system revealed that this flow element was not installed at l the time of the walkdown.
The flow element was installed by the piping craft after the I
system was turned over to Startup. Construction retained rT primary responsibility for this punchlist item performed after
( ,) turnover, but the work was performed tnder the cognizance of Startup via a Startup Work Authorization (SWA) form. The inlet side of the flow element was clearly marked on the l
L__---- ;
R3 vision: 1 Page 16 of 41 Q
l ! RESULTS REPORT V
ISAP VII.c (Cont'd)
Appendix 8 (Cont'd) 3.0 ROOT CAUSE AND GENERIC IMPLICATIONS (Cont'd) handle of the flow element and the flow direction required in the piping was marked on the applicable piping isometric drawing. There is no procedural requirement after system turnover that Startup conduct a walkdown of flow elements installed by the piping craft, working under Startup cognizance, to verify flow direction. Startup is required to sign the SWA in the " work verified complete" block. The SWA was signed by Startup. Based on these circumstances, it is concluded that construction personnel made an installation error, which was not detected by Startup.
The Pipe Fabrication and Installation Procedure, CP-CPM-6.9E, which directs the activities of construction piping personnel,
,s does not specifically state that flow elements are required to
- be installed in a specific orientation. The procedure does N ') specifically mention the orientation requirements for pipe slope, globe valves, and valve stems.
Based on the above, a less-than-adequate construction procedure, which did not address the requirement to install flow elements in a particular flow direction, is the primary root cause for the incorrect orientation of the flow element.
In addition, less-than-adequate construction supervision or workmanship and less-than-adequate Startup overview are contributing causes.
Quality Control (QC) is responsible, in accordange with the ASME Pipe Fabrication and Installation Inspection Procedure QI-QAP-il.1-26, for the inspection of the piping installed by Construction to ensure that the requirements of the piping isometric drawing are met. There is, however, no specific requirement in this procedure to verify installation of flow elements in accordance with the required flow direction.
Correct valve orientation is the only component orientation specifically required by the procedure to be verified. A less-than-adequate QC inspection procedure is the secondary root cause. Sufficient information was available, however, for the QC inspector to have identified the disorientated flow element; less-than-adequate QC inspection is therefore, identified as a contributing cause.
(G)
l
Rcvision: 1 Page 17 of 41 l}
( ,,/ RESULTS REPORT ISAP VII.c (Cont'd)
Appendix 8 (Lont'd) 3.0 ROOT CAUSE AND GENERIC IMPLICATIONS (Cont'd)
TU Electric Startup tested the piping system with the flow element reversed. During normal operation, flow through this line is controlled by a valve that modulates flow based on the condensor refrigerant pressure. This valve operates independently of the flow element. For testing, this valve is manually opened and a minimum flow of 250 gpm is required to be present in the pipeline. The startup test results for this system indicated sufficient flow was available. Therefore, since Startup did not detect that the flow element was incorrectly oriented after installation by the piping craft, and due to the negligible effect that this condition had on the system performance, this incorrectly oriented flow element was able to remain undetected during startup testing.
(f" )j TU Electric Operations accepts systems from Startup, in accordance with their Final Acceptance of Station Systems and Equipment Procedure No. STA-802. This procedure requires each system to be walked down for operability, maintainability and identification of any damage. It does not serve as a mechanism for providing or verifying final QC inspections of construction activities. As a result TU Electric Operations would not normally be expected to detect a condition of this type unless it affected system performance.
Review of TU Electric Operations documentation (PR-85-352) for this flow element determined that a problem was identified where the associated low flow alarm in the control room was constantly alarming. The flow recorded was 102 gpm as compared to the low flow alarm set point of 192 gpm. TU Electric Operations determined that this problem was caused by a low heat load in the control room due to deenergized Unit 2 control room equipment and low cooling water temperature. The ;
TU Electric-recommended corrective action was to add a I parallel circuit from the condensor refrigerant pressure !
switch to the low flow alarm. Therefore, the low flow condition would not activate the alarm unless the condensor refrigerant pressure was also high. TU Electric Operations resolution to this problem did not include an inspection of the flow element orientation. The incorrect flow element (s
orientation, which was likely a contributing cause to the x constant alarming, remained undetected. Because the resolution to the constantly alarming low flow alarm in the control room did not include reorientation of the flow element, a memo has been transmitted to TU Electric to request reverification of their recommended corrective action.
Revision: 1
- Page 18 of 41 RESULTS REPORT ISAP VII.c (Cont'd)
Appendix 8 (Cont'd) 3.0 ROOT CAUSE AND GENERIC IMPLICATIONS (Cont'd)
A similar deviation for a reversed component for this attribute was reported during the reinspection of the Small-Bore Piping Configuration construction work category that was identified as a special case. The following supplemental review is provided to determine whether a common root cause may exist when the reversed flow element is evaluated in combination with the small-bore piping configuration special case.
The deviation reported in the small-bore piping configuration construction work category was a 3/8-inch globe valve with a metal diaphragm that was installed in the reversed flow direction. The factors regarding the installation of the valve are summarized as follows: 1) the flow direction was marked on the valve body; however, no flow direction was O indicated on the piping isometric drawing used for the installation; 2) the construction and QC inspection procedures provide specific instructions regarding installation and inspection, respectively, of globe valves; 3) the valve was installed by construction personnel under construction supervision prior to turnover to Start-up.
The primary root cause of the reversed valve installation was less-than-adequate engineering requirements since the piping isometric drawing lacked a flow arrow. A secondary root cause was a less-than-adequate QC inspection checklist in that it lacked a requirement for verification of the correct valve orientation.
A comparison of the causes for the reversed flow element and the reversed globe valve reveals that no common primary root causes were identified. A similarity between the secondary root causes of the reversed valve and of the reversed flow element does exist, however, in the' area of QC inspection procedures.
Generic Implications The generic implications of the above primary and secendary root causes for the reversed flow element in this attribute O are that other piping parts and components, if they are not addressed in the same piping construction and QC procedures and on the piping isometric drawing, may also be miseriented.
As a result, a review of the other piping parts and componeats, i.e. , valves, straineis, pipe slope, valve stem
i Rsvision: 1 Page 19 of 41
\' RESULTS REPORT ISAP VII.c i (Cont'd)
Appendix 8 (Cont'd) l 3.0 ROOT CAUSE AND GENFRIC IMPLICATIONS (Cont'd) angle, orifice flange taps and eccentric reducers, all of which have a preferred orientation, was made to determine i whether sufficient installation instructions were provided in the piping construction procedure or isometric drawings and whether sufficient QC inspection criteria were provided for installation of these components.
The piping construction procedure (CP-CPM-6.9E) does contain criteria for properly orienting globe valves, pipe slope, and valve stem angle. In addition, each piping isometric drawing specifies the orientation of eccentric reducers, pipe slope and orifice flange taps. Therefore, generic implicaticas concerning these items do not exist. The remaining components T that require installation in a specific orientation and that J//
\~- are not specifically addressed on the piping isometric drawing or in the const.ruction procedure.are strainers and valves, excluding globe valves.
In general,Lany disoriented valves, strainers or other flow sensitive piping components in a piping system either would not be expected to have a safety-significant impact on the pressure integrity, flow capability and performance of the piping system or the condition would h6ve a significant impact that would be expected to be detected by Startup or Operations during normal startup and preoperational testing. However, in order to confirm this determination, a complete review of all flow-sensitive components and the'startup test methods used to test them, would be required to be performed. Based on the scope of this effort, and since two conditions of disoriented components have been detected, i.e., one each in the large bore and small bore pipe configuration populations, a generic implication does exist for all piping components, e.g.,
valves, strainers, etc.
I' The QC inspection procedures (QI-QAP-11.1-26 and CP-QAP-12.1, Inspection Criteria and Documentation Requirements Prior to l.
System N-5 Certification) does require that inspection be performed to verify that the piping configuration be in accordance with the piping isometric drawing. Such an
("'
\ inspection would be expected to cover orientation of pipe slope, orifice flange taps and eccentric reducers. In addition, the QC inspection procedure (QI-QAP-11.1-26) specifically addresses the requirement to verify valvt etem 1
i Rsvision: 1 Peg 2 20 of 41 m
\ RESULTS REPORT I d
ISAP VII.c (Cont'd) i Appendix 8 (Cont'd) 3.0 ROOT CAUSE AND GENERIC IMPLICATIONS (Cont'd) orientation. Therefore, there are no generic implications of a less-than-adequate QC inspection procedure concerning all of the above items. The remaining items that require inspere?on l' for installation in a specific orientation are direction-sensitive piping components, i.e., valves, strainers, etc. Since a generic implication has already been i identified, as discussed alove, for these items, no further l
/
review is required.
In summary, there is one valid generic implication concerning disorientation of piping components, e.g., valves, strainers, etc., within this attribute due to less-than-adequate construction and QC procedures.
In addition, a review was performed to determine whether the
("'}
(_ - primary and secondary root causes of this special case would have any generic implications for any of the other attributes in this population. Implications of these root causes of less-than-adequate construction and QC procedures do exist for Attribute 4, piping clearances, and Attribute 5, expansion joint installation. However, as discussed in this Appendix, separate findings have been identified for each of these attributes, and the generic implications are addressed by the resulting recommended corrective actions. A similar generic implication exists but has already been addressed for the same Attribute 4 in the Small-Bore Piping Configuration Population (App. 9).
There are no generic implications for Attribute 1, piping and component identification, and Attribute 2, piping and component sequence and location, as the piping isometric drawings do identify the installation requirements and the QC inspection procedure does verify identification and sequence and location of piping and components for these attributes.
One orientation deviation, that of a 3/8-inch globe valve installed in the reversed direction, as discussed above, was l~ found in the small-bore piping configuration population.
Since the secondary and contributing causes of these two deviations are both associated with QC inspection, and since j))
(, similar installation drawings and procedures are used, a 1
i
Rsvision: 1 Page 21 of 41
/'*\
) RESULTS REPORT A_/ .
ISAP VII.c (Cont'd)
Appendix 8 (Cont'd) 3.0 ROOT CAUSE AND GENERIC IMPLICATIONS (Cont'd) generic implication has been confirmed to exist in the Small-Bore Pipe Configuration Population for installation of piping parts and components that require specific orientation.
This implication has been addressed, however, in the recommended corrective action for the globe-valve finding.
The generic implications identified for the special case of a reversed flow element are limited to flow elements and other direction-sensitive piping components, e.g., valves, strainers, etc., in the large-bore and small-bore piping configuration populations. Other generic implications exist, but were separately identified based on other findings.
Summary
-- The reversed flow element, which was declared to be a special case, is a result of multiple causes as identified below:
- 1) Less-than-adequate construction procedure, which did not specifically address the direction-of-flow installation criterion for flow elements, was determined to be the primary root cause.
- 2) Less-than-adequate QC procedure, which did not address inspection of flow element flow direction, was a secondary root cause.
- 3) Less-than-adequate construction supervision or workmanship and less-than-adequate QC inspection to install and verify, respectively, a clearly marked flow elewent in the correct flow direction were determined to be contributing causes.
- 4) Less-than-adequate overview by Startup, which did not l detect that thw' flow element installed under their ,
cognizance was reversed, was determined to be a contributing cause. j l
The generic implication identified for this special case is {
that other safety-related flow elements and other 1 (V)
Revision: 1 Page 22 of 41
() RESULTS REPORT ISAP VII.c (Cont'd)
Appendix 8 (Cont'd) 3.0 R00.' CAUSE AND GENERIC IMPLICATIONS (Cont'd) direction-sensitive piping components, e.g., valves, strainers, etc., may also be incorrectly installed in the uninspected portion of this population and the small-bore piping configuration population.
Recommendations for corrective action are provided in Section 4.1.
3.2 Piping Clearances Root Cause The following provides a summary of a. review of Engineering specifications and Construction and QC procedures to determine the causes for the piping clearance deviations.
Engineering Clearance criteria were not consistently or completely presented in the Piping Erection Specification'2323-MS-100, and other component erection specifications during the time of major construction activities. Inconsistent clearance criteria existed between the Piping Erection Specification, and the Electrical Erection Specification, 2323-ES-100, up until April 1982. For example, the Electrical Erection Specification did not provide clearance criteria between ASME Code Class 1 and 2 piping, apart from the weld areas, and I
conduit, conduit supports, cable trays, and cabib tray supports. The Piping. Erection Specification required pipe with operating temperatures greater than 200*F, predominantly Class 1 and 2, to maintain a one inch clearance with conduit, conduit supports, cable trays, and cable tray supports.
Miscellaneous other inconsistencies or lack of criteria similar to this example also' existed. Subsequent to April 1982, the clearance criteria presented in both the electrical and mechanical specifications evolved to become similar.
The Piping Erection Specification provides installation requirements for both piping and pipe supports. Clearance criteria between pipe and pipe supports are only specified in
l Rsvision: 1 Page 23 of 41
) RESULTS REPORT ISAP VII.c (Cont'd)
Appendix 8 (Cont'd) 3.0 ROOT CAUSE AND GENERIC IMPLICATIONS (Cont'd) the piping installation section (Section 4) of this specification, and no criteria are contained in the pipe support section (Section 7) for maintaining clearances with piping.
In summary, the inconsistencies and incompleteness of specifications are likely to have contributed to some of the reported clearance violations and is a primary root cause. A memo has been sent to the Design Adequacy Program regarding the inconsistencies in the clearance criteria provided in various specifications.
The Insulation Specification (for non-reflective insulation, fmg 2323-MS-30) that was in effect through March 1983 provides
() minimum insulation thicknesses based on maximum material conductivity values. This specification allowed the thicknesses to exceed those contained in the specification without engineering approval if a material with a higher conductivity was supplied. Thus, insulation with thicknesses greater than the minimum specified in the insulation specification could have been provided and installed without engineering knowledge. The insulation specification was revised to reflect the actual thicknesses of purchased and installed insulation and to require increases in insulation thickness to be brought to the attention of the engineers.
However, as the original minimum thicknesses were used by construction and QC to estimate clearances, clearance deviations resulted. Therefore, the potential for clearance deviations exists in the installed piping that is insulated, due to a less-than-adequate engineering interface with construction and QC. This is also a primary root cause.
Construction Piping was installed in accordance with construction procedure CP-CPM-6.9E. This procedure provides instructions to maintain clearances, including insulation, when installing piping.
This procedurre only specifies criteria for pipe-to-pipe clearances. Clearances and clearance criteria with other
( plant hardware are not included in this construction
\~ procedure, although these criteria do exist in the piping
l I
i Rsvision: 1
]
Page 24 of 41 l
j G
RESULTS REPORT
)
ISAP VII.c (Cont'd)
Appendix 8 (Cont'd) 3.0 ROOT CAUSE AND GENERIC IMPLICATIONS (Cont'd) erection specification, 2323-MS-100. This less-than-adequate construction procedure was identified as a primary root cause.
In addition, although the construction procedure states that clearances include insulation, there is no reference to insulation thicknesses to be used to establish clearances.
This was identified as a contributing cause.
In addition, construction procedure CP-CPM-6.9E requires that when the location specified on the drawing results in clearances less than those specified, Engineering is to be contacted for resolution. Based on the number of clearance deviations reported, for at least some cases construction either may not have complied with this requirement or was e- unable to estimate correctly the insulation thickness to be installed as described above.
In November 1984, construction procedure CP-CPM-6.2, Summary of Clearance Criteria, was issued to summarize clearance criteria for all hardware types. This procedure also provides ,
insulation thicknesses, which are the same as those contained I in the insulation specification, and states that notching of insulation may be necessary to maintain required clearances.
Because much of the piping and insulation (for Unit 1) was installed prior to the issuance of this procedure and because no provision was made to backfit the new requirements, the potential for clearance violations exists.
Once insulation was installed, notching of insul'ation was permitted by Engineering to maintain clearances. Based on the number of clearance violations, it is not evident that insulation was notched or notched sufficiently to maintain clearances. There is evidence of notching in the plant; however, it appears to be to avoid interferences and not to maintain clearance. The piping erection specification, 2323-MS-100, specifies that notching of insulation to meet clearance criteria is acceptable. There is no evidence that construction was cognizant of the criteria because, prior to the issuance of CP-CPM-6.2 in November 1984, none of the
- construction procedures contained any instructions relative to notching or reference to other documents containing these criteria. Therefore, less-than-adequate construction instructions prior to November 1984 relative to notching of insulation was a primary root cause.
% m _ _ _ _ _ - _ _ _ _ .
Revision: 1 Page 25 of 41 RESULTS REPORT ISAP VII.c (Cont'd)
Appendix 8 (Cont'd) 3.0 ROOT CAUSE AND GENERIC IMPLICATIONS (Cont'd)
Quality Control Final QC inspections to verify piping clearances were i performed in accordance with CP-QAP-12.1, Inspection Criteria and Documentation Requirements Prior to System N-5 Certification. Often insulation was not installed at the time of the walkdown, and anticipated clearances after installation of insulation were determined by subtracting the insulation I thickness from the measured clearance. Insulation thicknesses I
were obtained from the pipe line list and the insulation l specification.
This method of checking insulation by QC is considered adequate for a preliminary check. However, as insulation thicknesses varied from those contained in the insulation O specification, the lack of a walkdown to verify clearances after insulation is installed is considered a secondary root cause fcr clearance deviations.
Generic Implications _
A review was made to determine whether the primary and secondary root causes identified for this attribute would have any implication for any of the other attributes in this population. The primary root cause of a less-than-adequate construction procedure does have implications for Attribute 3, orientation of piping and components, and Attribute 5, expansion joint installation. However, as discubsed in this Appendix, separate findings have been identified for each of these attributes. There are no generic implications of this primary root cause with Attribute 1, piping and component identification, and Attribute 2, piping and component sequence and location, since the piping isometric drawings (instead of the construction procedure) do identify the installation requirements for these attributes; as further confirmation, no significant deviations have been determined to exist.
For the primary root cause of a less-than-adequate engineering interface with construction, a generic implication does exist
Revision: 1 Page 26 of 41 r
( RESULTS REPORT N.
ISAP VII.c (Cont'd)
Appendix 8 (Cont'd) 3.0 ROOT CAUSE AND GENERIC IMPLICATIONS (Cont'd) discussed in this Appendix, a separate finding has been identified for this attribute. For the remainder of the attributes in this population, the engineering interface with construction relating to these attributes is adequate, and no generic implications exist.
For the primary root cause of a less-than-adequate engineering specification, generic implications do not exist for any of the other attributes in this population.
For the secondary root cause of a less-than-adequate QC procedure, generic implications were considered for Attribute 1, piping and component identification, Attribute 2, piping and component sequence and location, Attribute 3, orientation N of piping and components, and Attribute 5, expansion joint
('
\s- l installation. The secondary root cause of less-than-adequate QC procedures does have implications for Attributes 3 and 5; however, as discussed in this Appendix, separate findings have been identified for each of these attributes. For Attributes 1 and 2, the QC procedure does adequately address the inspection of these attributes; therefore, no generic implication exists.
In other populations, any electrical, mechanical, or structural component whose functional capability is affected by the clearance from piping would be implicated by the causes previously described for the piping clearance deviations.
Based on this review, affected components include conduit (App. 1), cable tray (App. 2), and small-bore piping -
configuration (App. 9), as a minimum.
Summary l
l The piping clearance deviations, which were identified to be l
an unclassified trend, are a result of multiple causes as l identified below:
- a. Inconsistent and incomplete clearance criteria between various compcnent erection specifications was a primary root cause of clearance deviations.
Ravision: 1 Page 27 of 41 RESULTS REPORT ISAP VII.c (Cont'd)
Appendix 8 (Cont'd) 3.0. ROOT CAUSE AND GENERIC IMPLICATIONS (Cor.t'd) ,
- b. Specification of minimum insulation thicknest u with generic approval to excsed those thicknesses without Engineering knowledge (prior to March 1983) if a material with a higher conductivity is supplied, provided a less-than-adequate engineering interface with construction and QC. This was a primary root cause of clearance deviations.
- c. The lack of criteria in the construction piping procedure for clearances with components other than piping and lack of instructions relative to notching.of insulation were primary root causes of clearance deviations for piping installed prior to November 1984.
- d. The lack of a QC requirement to verify final piping clearances subsequent to the installation of insulation was a secondary root cause of clearance deviations.
- e. The lack of criteria in the construction procedure for insulation thicknesses to be used when estimating clearances was a contributing cause of clearance deviations.
The generic implication from these causes is that the large and small-bore piping configuration, conduit, and cable tray populations, as a minimum, are also likely to have piping clearance deviations.
- Recommendations for corrective action are provided in Section 4.2.
3.3 Expansion Joint Tie Rod Installation koot Cause The following is a summary of the relevant information regarding the Unit I service water pump expansion joint tie rod assembly installation.
Based on numerous historical documents and various sources, the service water expansion joint cie rod assemblies were intended to be a permanent design item as early as February 1980. The service water system was in startup testing when l
I Rsvision: 1 Page 28 of 41
\s / RESULTS REPORT ISAP VII.c (Cont'd)
Appendix 8 (Cont'd) 3.0 ROOT CAUSE AND GENERIC E"VLICATIONf. (Cont'd)
Startup engineering requested installation of the tie rod assemblies on the expansion joints to provide the axial restraint desired during the testing of the adjacent service water pumps. Gibbs & Hill provided Startup with a preliminary design, and with the concurrence of both Gibbs & Hill and TU ,
Electric, temporcry (nonpermanent) tie rod assemblies were l installed while waiting for the permanent design to be issued.
In February 1981, construction supervision initiated construction operation travelers MP-81-2373C-0400 and MP-81-2374C-0400. These two travelers were issued solely to identify specific operations for the installation of the temporary tie rod assembly hardware, and replacement with the permanent design and hardware. Per traveler NT-81-2373C-0400,
('N y,,) Operation 6.1, the temporary installation was completed and verified by QC on March 5, 1981. It should be noted that Operation 6.1 had a handwritten clarification that states,
" Threads may be burred on the temporary tie rod installation in lieu of double nutting. Double nuts will be used on the permanent tie rod installation." After that point, both travelers were lost in the field and were never completed (the travelers have been located recently and are now in the Permanent Plant Records Vault. Neither traveler shows any verification of operational steps after the verification of March 5, 1981).
Subsequently, in October 1981, construction supervision issued revision MP-81-2373D-0400 and MP-81-2374D-0400 to document the disassembly and closure of the flanges connecting both ends of the expansion joint. The D revisions to the travelers also state that each supersedes the previous revision of the travelers (revision C). The D traveler revisions contained no instruction to the pipefitters regarding the installation of the temporary tie rod assembly, or the removal and replacement with the permanent tie rod assembly hardware. Construction supervision's failure to duplicate the operational steps when superseding the lost travelers resulted in the failure to properly document and track the verification of the installation and removal of the temporary tie rod assembly.
/- s Therefore, less-than-adequate preparation and review of
(
'- operation travelers is determined to be a primary root cause.
Ravision: 1 Page 29 of 41 1 s- - RESULTS REPORT ISAP VII.c (Cont'd)
Appendix 8 (Cont'd) 3.0 ROOT CAUSE AND GENERIC IMPLICATIONS (Cont'd)
The construction deficiency identified is for single-nut construction in lieu of double-nut construction. Per Operation 6.1 of traveler MP-81-2373C-0400, the temporary tie rod assembly installation had utilized single-nut construction and had the threads burred to prevent the nuts from backing off. This suggests that the hardware configuration used for the temporary tie rod assembly was never altered or replaced, despite the issuance of the permanent design drawing and despite the expansion joint flanges being reworked thereaf ter.
This implies that af ter the initial installation of the temporary tie rod assembly, the contractor either did not recognize.that the tie rod assembly was a temporary installation and required replacement, or did not recogni?"
that a permanent tie rod assembly was ever required.
The tie rod assembly was installed prior to issuance of a final design drawing; therefore, it was installed as a temporary item (considered as a nonpermanent item substituted for a permanent item). In 1981, the installation and control of a temporary item was to be performed in accordance with construction procedure CPM 9.1, Rev. O, Control of Nonpermanent Materials. This procedure calls for all such material to be color coded red to provide visible identification (e.g. , painting, tagging, etc.). The procedure also stipulates that the installation of nonpermanent items should be via specific documentation (such as a construction operation traveler) and submitted to the Permanent Plant Records Vault where it would be identified for tracking and closure.
If the tie rod assembly had been properly identified and documented as a temporary installation by the craft performing the work, then the tie rod assembly hardware would have been clearly flagged for construction and quality inspection activities when the final design drawing was issued or during any subsequent rework. It appears that the craft and supervision failed to follow this construction procedure.
Therefore, less-than-adequate construction supervision or y construction workmanship associated with controlling the installation of nonpermanent materials in accordance with
.( established site procedures is determined to be a secondary root cause.
Rsvision: 1
]
I Page 30 of 41 q A 1 RESULTS REPORT l
%.)
ISAP VII.c i (Cont'd)
Appendix 8 (Cont'd) 3.0 ROOT CAUSE AND GENERIC IMPLICATIONS (Cont'd)
Support drawing SW-1-031-TIE-J33R was issued in August, 1981, which established the tie rod assembly as an ASME Code Section III Class 3 permanent pipe support. Subsequently, in June 1984, Engineering reissued the drawing as SW-1-031-TIE-J35R.
The drawing was reissued to change the classification of the tie rods to non-ASME III (class 5), but r.here wt.s no change made to the design configuration of the tie tod assembly.
A review was perf ormed to determine that auheequent activities could have led to removal of the temporary tJe rod assembly and replacement with the permanent assechly. The only reason for this work being initiated would be if the tie rods required rework or if the design drawing was issued or ;
reissued. Based on the review of various installation documentation, the only rework activities that took place in
[U the vicinity of the expansion joint cie rod assembly were activities associated with the disassembly and reassembly of the flanged connections on the expansion joint. During this subsequent rework or reassembly of the flanged connections on the expansion joint, the tie rod assembly can be removed and replaced without disturbing the tie rod nuts that are the subject of the identified construction deficiency. By removing the studs and nuts bolting the tie rod assembly lugs to the expansion joint flange connections, the tie rod assembly can be removed in sections. Therefore, the rework of the expansion joint flange connections would not necessarily initiate rework of the tie rods and the subject nuts.
Construction procedure DCP-3, CPSES Document Control Program, is the procedure that addresses the activities and responsibilities for the control and distribution of construction and design documents generated at the site or received by the Document Control Center (DCC). Revision 11 was in effect in August 1981 and Revision 18 was in effect in June 1984. Under the guidelines of both revisions to the procedure, a design drawing, or revision thereof, must be reviewed by responsible personnel in applicable disciplines to determine if the drawing affects that discipline's work activities. For this reason, controlled documents are issued to the construction supervision of those disciplines
{$
( determined to be applicable.
l Rsvision: 1 Page 31 of 41
-'s . f RESULTS REPORT ISAP VII.c (Cont'd)
Appendix 8 (Cont'd) 3.0 ROOT CAUSE AND GENERIC IMPLICATIONS (Cont'd)
Per discussion with personnel in the Hanger Task Force, support drawing SW-1-031-TIE-J33R was isaued to Welding Engineering for review in 1981. In 1984, one controlled copy of drawing SW-1-031-TIE-J35R vas issued to the Hanger Task Force. In both instances, there is no informatica available regarding the further issuance of the drawing to eaccific disciplines for review or the generation of any tr.velers concerning the expansion joint tie rod assembly.
Regardless of what groups had the opportunity to review the design drawing for implementation, no further installation documentation was generated. Even if visual inspection caused them to believe that the permanent installation was already in
/~N place, then the lack of quality installation documentation and
(,,) the difference between the as-built hardware configuration and the drawing should have been an indication to the contrary.
Therefore, the lack of configuration control is determined to be a primary root cause.
During the root cause analysis, various specification, construction procedures, and inspection procedures were reviewed due to the possibility that the expansion joint tie rod assembly could have been considered as being of a number of different hardware types (piping component, piping support, equipment support, etc.). The following is a brief discussion of the results of this review.
In accordance with the Piping Erection Specification MS-100 (Section 4), the installation of expansion joints and associated hardware is the responsibility of the piping installer. If the tie rod assembly was considered as
" associated hardware", then piping department pipefitters would perform the work per Pipe Fabrication and Installation Procedure CP-CPM-6.9E, which establishes the construction requirements for ASME III and non-ASME III piping installation. The current revision of this procedure specifically addresses expansion joint installation requirements; however, the installation of the associated tie s rods is not addressed.
The QC procedure that provides the requirements for inspection of ASME Code Section III pipe installations is QI-QAP-11.1-26, ASME Pipe Fabrication and Installation Inspections. A review
Revision: 1 Page 32 of 41 ]
n x- RESULTS REPORT ]
l ISAP VII.c !
(Cont'd) ]
Appendix 8 (Cont'd) 3.0 ROOT CAUSE AND GENERIC IMPLICATIONS (Cont'd) l of the current revision to this procedure determined that no specific inspection requirements are provided for expansion joints and the associated hardware.
The possibility that the tie rod assembly was considered an ASME III pipe support was investigated. The applicable !
installation specification is 2323-MS-100 (Section 7). The hanger department ironworkers would perform the work to the pipe support installation procedures, which are CP-CPM-7.3, General Fabrication Procedure; CP-CPM-7.3B, Fabrication of ASME Component Supports; CP-CPM-9.10, Component Support Installation; and CP-CPM-9.10A, Installation of Vendor Supplied Component Support Catalog Items. A review of these
(~'} procedures revealed that no specific installation requirements
\__/ for expansion joint tie rods were either included or incorporated by reference to vendor instructions.
The QC procedure that provides the requirements for the inspection of safety-related pipe supports is QI-QAP-11.1-28.
A review of this, procedure determined that no inspection requirements are provided specifically for expansion joint tie rods.
A review of the installation documentation for the similar Unit 2 service water pump expansion joint tie rod assemblies was performed to determine which procedures were followed by construction in that case. The review indicated'that the other service water expansion joint tie rods were considered an equipment support and were installed and inspected in accordance with construction procedure MCP-1, General Installation of Mechanical Equipment and QI-QAP-11.1-39, Mechanical Equipment Installation Inspection Procedure, respectively. A review of these procedures and the Mechanical Equipment Erection Specification, MS-101, revealed that neither document specifically addresses installation of expansion joint tie rods.
Because of the lack of information within the construction and QC inspection procedures concerning expansion joint tie rods, 7'~' it is ambiguous as to what procedures were applicable and as
( to what construction discipline was responsible for the l permanent installation of the expansion joint tie rod assembly. This ambiguity may have contributed to the
Ravision: 1 Page 33 of 41
. /O . RESULTS REPORT
(_/
ISAP VII.c (Cont'd)
Appendix 8 (Cont'd) 3.0 ROOT CAUSE AND GENERIC IMPLICATIONS (Cont'd) breakdown in configuration control. Therefore, less-than-adequate construction and QC procedures are.also determined to be a contributing root cause.
The original issue of the tie rod assembly drawing clearly indicated that the tie rods were an ASME III Code Class 3 support.. Subsequently, Revision 1 of the tie rod assembly ,
drawing was issued to change the classification to l non-ASME III; however, no notation was added to the drawing to indicate that the tie rods were still safety-related. As a result, if QC inspection used Revision 1 of the drawing to j determine the scope of inspection, the QC inspectors may not have known that the tie rod assembly hardware was f-s safety-related, and therefore, may not have inspected it.
( Therefore, less-than-adequate identification by engineering of the safety classification of equipment is determined to be a contributing cause.
Generic Implications The root cause analysis identified a number of circumstances that only collectively resulted in a construction deficiency of expansion joint installation involving tie rod assemblies.
This unique set of circumstances includes:
- a temporary installation that was not properly identified in the field or in the documentation; a lost traveler with detailed instructions that served both as the documentation of the temporary installation and the document designating the need for a permanent installation, and a replacement traveler that failed to recreate any of these details; I
- a new hardware item, with ambiguous construction discipline installation responsibilities, was issued on a drawing that was not directly associated with the piping isometric or the construction discipline that installed the original temporary installation, and
() -
resulted in a breakdown in configuration control; and an initial design was implemented after system turnover to Startup.
Rsvision: 1 Page 34 of 41 RESULTS REPORT
, ISAP VII.c (Cont'd)
Appendix 8 i (Cont'd) 3.0 ROOT CAUSE AND GENERIC IMPLICATIONS (Cont'd)
The primary and secondary root causes along with these circumstances have not been encountered, either separately or collectively, in the other attributes of this population. The fact that there were no other significant deviations related to these root causes or circumstances reported during reinspection of the other attributes provides direct evidence that a generic implication does not exist. This unique set of circumstances that collectively resulted in a construction deficiency is unlikely to occur for other attributes; therefore, this deficiency is considered isolated to Attribute 5, expansion joint installation.
The generic implication resulting from these root causes is p that any safety-related expansion joint installed on piping in the small-bore or large-bore piping configuration population
, ( may be subject to the same deviation. (To date, in accordance with Expansion Joint Specification MS-71A, no small-bore expansion joints have been specified f or purchase.)
Other possible generic implications stemming from the root cause analysis are as follows:
- a. Other piping hardware e.g., Dresser couplings with their associated tie rods, not specifically addressed in the piping and/or pipe support erection procedures and the corresponding QC procedures also may be subject to the same deviations. A complete reviey is required to determine whether installation requirements are ;
omitted for other hardware in these procedures.
- b. Any safety-related piping hardware that is classified as non-ASME III and does not have a clear safety-related identification on the installation 1 drawing may be subject to a similar deviation. I
- c. Other temporary (nonpermanent) hardware installations in piping systems may not be properly identified and documented, which could possibly result in the failure of the Project to install the permanent hardware prior to final acceptance of the system.
Rsvision: 1 Page -35 ' of 41 RESULTS REPORT ISAP VII.c (Cont'd)
Appendix 8 (Cont'd) 3.0 ROOT CAUSE AND GENERIC IMPLICATIONS (Cont'd)
Summary The tie rod installation deviations, which were determined to be a construction deficiency, have the following causes:
- a. Less-than-adequate preparation and review of operation travelers is a primary root cause.
- b. Lack of configuration control is also determined to be a primary root cause.
- c. Less-than-adequate construction supervision or workmanship associated with controlling the installation of nonpermanent materials in accordance with established site procedures is the secondary root
- cause.
- d. Insufficient identification of the classification of the tie rods on the installation drawing, by engineering, is a contributing cause.
- e. The Piping Erection Procedure, Pipe Support Installation Procedure, or the Mechanical Equipment Installation Procedure, whichever is applicable, did not specifically address the responsibility and requirements for the installation of expansion joint tie rods. The Piping, Pipe Support, and gechanical Equipment QC inspection procedures, whichever is applicable, did not provide specific criterion for the inspection of expansion joint tie rods. Therefore, less-than-adequate construction and QC procedures are also contributing root causes.
The generic implications resulting from the culmination of these causes is that other expansion joints installed in the large-bore and small-bore populations may be subject to the same deviation (if used).
Other possible generic implications stemming from the root cause analysis are as follows:
_ - _ _ - _ _ _ _ _ - - _ _ _ _ . _ _ . ~ .
Ravision: 1 Page 36 of 41 j RESULTS REPORT ISAP VII.c (Cont'd)
Appendix 8-(Cont'd) 4.0 RECOMMENDATIONS (Cont'd)
- a. Other piping hardware not specifically addressed in the piping and pipe support erection procedures may be subject to the same deviation. One example of hardware not addressed is Dresser couplings and their associated !
tie rods.
- b. Any safety-related hardware that is classified as non-ASME III and does not have a clear safety-related identification on the construction drawing may be subject to the same deviation.
- c. Other temporary (nonpermanent) hardware installations in piping systems may not be properly identified and documented, which could possible result in the failure of the Project to install the permanent hardware prior O- to final acceptance of the system.
Recommendations for corrective action are provided in Section 4.3.
4.0 RECOMMENDATIONS Recommendations for corrective action are:
4.1 Flow Element Flow Direction
- 1) It is recommended that TU Electric reinspect the flow I elements and other direction-sensitive piping components, e.g., valves, strainers, etc., performing a safety-related function in both the large- and small-bore piping to verify that they are oriented in the proper direction and correct as necessary. l
- 2) It is recommended that TU Electric review the appropriate piping construction procedures to verify j that instructions are provided to ensure that flow elements and any other direction-sensitive components are to be installed in accordance with the direction of flow indicated on the piping israetric drawing. l
-^ - - - _ _ _ _ _ _ _ . . . _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
Revision: 1 Page 37 of 41 l
V RESULTS REPORT ISAP VII.c (Cont'd)
Appendix 8 (Cont'd) 4.0 RECOMMENDATIONS (Ccot'd)
- 3) It is recommended that TU Electric review the appropriate QC inspection procedures to verify that inspections are provided to ensure that flow elements and any other flow sensitive components are installed in accordance with the direction of flow indicated on the piping isometrit drawing. !!
- 4) It is recommended that TU Electric review the appropriate Startup procedures to verify that instructions are pr(vided to ensure that flow elements and any other direction-sensitive components installed under Startup cognizance after system turnover be verified by Startup for proper orientation.
4.2 Piping Clearances
- 1) It is recommended that TU Electric verify the existing pipe clearance criteria in applicable specifications and procedures to ensure that consistent and complete criteria among disciplines are provided and that all types of plant hardware requiring clearance from piping are addressed.
- 2) It is recommended that the installers training program reemphasize to the piping craf t that construction procedure CP-CPM-6.9E requires that Engineering be notified for resolution when the location specified on the piping drawing results in clearances less than those specified.
- 3) It is recommended that the following specifications be reviewed / revised:
a) The Insulation Specification, 2323-MS-30, should include instructions relative to notching of insulation and should specify the maximum insulation thicknesses for types of insulation used at CPSES.
b) Section 7 of the Piping Erection O Specification, 2323-MS-100, should include pipe-to-pipe support clearance criteria or should reference the pipsvg section of the specification for piping clearance criteria.
Revision: 1 Page 38 of 41 i
()
(_ ,/ RESULTS REPORT f
4 ISAP VII.c (Cont'd)
Appendix 8 (Cont'd) 4.0 RECOMMENDATIONS (Cont'd)
- 4) It is recommended that TU Electric revise the appropriate QC procedure to require verification of final piping clearances subsequent to insulation installation.
- 5) It is recommended that TU Electric perform a walkdown of piping to verify that sufficient clearance exists to accommodate the total out-of-phase movements under design conditions between pipes and other pipes, and between pipes and other plant items and correct as necessary. This walkdown should be done with pipe and other hardware installed in their final location and with insulation installed.
4.3 Expansion Joint Tie Rod Installation
-~)
L) 1) It is recommended that TU Electric reinspect the saf ety-related piping system expension joints for correct installation in accordance with the manufacturer's drawings and instructions or Gibbs &
Hill / Westinghouse /TU Electric drawings and instructions, when applicable, and correct as necessary.
- 2) It is recommended that TU Electric review and supplement, as required, the appropriate construction and QC inspection procedures to include the installation requirements and responsibility for j expansion joints and associated hardware, and any other hardware not addressed by procedure. One example of an item that may be required to be added to the piping procedure is Dresser couplings and their associated tie rods. As an alternative to identifying each type of specialty hardware explicitly, appropriate generic provisions may be made in the procedure, i It is also recommended that TU Electric reinspect, as j applicable, any additional piping hardware items j identified above to verify that they have been properly 1 i f-implemented in the field.
ks i
j
Rsvision: ,l' i Page 39 of 41 l
RESULTS REPORT ISAP VII.c (Cont'd)
Appendix 8 (Cont'd) 4.0 RECOMMENDATIONS (Cont'd)
- 3) It is recommended that TU Electric verify that safety-related piping, piping components,'and-associated hardware are clearly identified as safety-related on the construction drawings. Any safety-related items that are not clearly identified as ;
such on the construction drawing should be reviewed to- l ensure that appropriate installation and inspection procedures have been followed, and correct as necescary.
- 4) It is recommended that TU Electric ensure, as applicable, that all previously installed temporary.
piping and associated hardware installations have been replaced with permanent design installations. It is also recommended that TU Electric' evaluate, develop and 7 'gu implement, as required, a method of tracking the installation and removal of'all temporary piping-hardware to verify that temporary installations will be
~
replaced by the permanent design installation.
5.0 CONCLUSION
S Based on the findings of the reinspection and satisfactory 1 implementation of the above recommendations, there will be reasonable assurance that the hardware in this construction work category is adequately installed in accordance with the design.
O
1 r
Revision: 1 Page 40 of 41 RESULTS REPORT' 1
ISAP VII.c (Cont'd)
Appendix 8 (Cont'd)
Table 8-1
- Summary of Reinspection Results Large-Bore Piping Configuration Deviation Classification Number of .
Inspection Number of Insigni- Construction Attribute Points Deviations ficant Notable Deficiency
- 1. Piping and Component Identification 1,135 'O O O O
- 2. Piping and Component Sequence and Location 3,607 36 36 0 0
. 3. Orientation of Piping and Components 309 3 2 1 0
- 4. Piping Clearances 940 76 75 1 0
- 5. Expansion Joint Installation 18 5 1 0
- 1 TOTALS 6,009 120 114 2
- 1 (Approximately) (6,000)
- Four deviations occurred on one expansion joint.
a a h t l 0 T ou N s n s n al 1 s I h a%
1 4 e Tu0 q5 L
f sE 2 0 o e s o
- r e rt n1 u L o o4 t i a s r i e e vg p ea m %
RP e r0 T e5 t
g an 4 1 n ea i d rh t e GT a t r a 1 e l p u O s n s n al n e I h a%
o p T u0 i i q5 0 1 t P sE 1 1 a s o l e rt L o
'e D
e T c R n %
O 8 2 a F r0 P c. ) ) - r a
0 e5 t
E Id xd 8 R I i' e 0 an 0 3 S
V 'tn d t nn l e l C
2 ea rh
\ T P o eo b n a
d e
GT c L AC pC a g U S( p( T n h # t S I A i T a 1 E p t l R i r ou n P e N s al t n h a%
f a I T u0 o e q5 r sE 1 0 y G s o r e rt a e L o m r m u u t S a r
e %
p r0 m e5 e t T an 7 3 ea g d rh n e GT i t t a a l n r u al 5 e s h a%
p n T u0 O I q5 sE e s o 5 1 p ert 1 1 i L o P
E R V#
S E E E#
A P H E I S TS LS E E FPN F0N FSN GC DC O - O O - O O - O NN EN OI OI OI I AFRA RTT RTT RTT TROI R E - A E - A E - A SA UA BEI BEI BEI I E%QE MPV MPV MPV XL EL UI E UI E UI E ECARC NPD NPD NPD
Revision: 1 Page 1 of 20 RESULTS REPORT (JT' ISAP VII.c (Cont'd)
Appendix 16 HVAC Equipment. Installation 1.0 REVIEW PROGRAM IMPLEMENTATION 1.1 Construction Work Category Description The construction work category of. heating, ventilating and air conditioning (HVAC) equipment installation is comprised of the installation of all safety-related HVAC equipment in the duct systems by the HVAC installation contractor. Included is the setting of the equipment and connection of in-line equipment such as vane axial fans, propeller fans, fire dampers, modulating dampers, isolation dampers, gravity dampers and air flow monitoring stations to the duct work. Flanges, bolts and gaskets are also included in this work category.
1.2 Population Size and Sample Selection 9
For this construction work category, a population of approximately 580 items was identified as QC-accepted as of the date of this reinspection effort. Of the total of 180 accessible items reinspected, 96 first sample items were randomly selected from the population to ensure that at least ;
60 reinspection of each attribute were performed except for i the configuration attribute. The configuration attribute is l proportionally sampled in combination with the HVAC Ducts and Plenums population, (Appendix 15) to obtain 60 sample items. l The proportional sampling results for configuration are discussed in Appendix 15. An additional eight second sample items were randomly selected to ensure that at least 60 safe-shutdown hardware items were reinspected. '
1.3 Attributes Selected Sample items were reinspected for the following attributes:
Attribute 1 - Identification Attribute 2 - Location Attribute 3 - Orientation Attribute 4 - Configuration O' Attribute 5 - Flange bolt holes 1
Revision: 1 Page 2 of 20 RESULTS REPORT ISAP VII.c (Cont'd)
Appendix 16 (Cont'd) 1.0 REVIEW PROGRAM IMPLEMENTATION (Cent' 1)
Attribute 6 - Duct connections - bolt tension Attribute 7 - Duct connections - gasket installation Attribute 8 - Duct connections - bolt thread engagement Attribute 7 i. cessory connections A documentation review was not performed as all attributes of HVAC Equipment Installation could be reinspected.
2.0 DISCUSSION OF RESULTS 2.1 Summary of Results For reinspection, 171 Deviation Reports were issued describing approximately 1,656 deviations. Approximately 50,500 inspection points were encountered in performing these reinspection. See Table 16-1 for the results of the reinspection.
In the nine reinspection attributes of this population, there were no construction deficiencies and no adverse trends. In Attribute 9, accessory connections, there were five deviations that were notable.
2.2 Analysis of Reinspection Results This section provides, by attribute, a discussion of the reported deviations, an analysis of the effect of the deviations on the functional capability of the HVAC equipment installation, and an analysis for the presence of trends. The function of HVAC equipment is to maintain the structural integrity, the operability, and the flow and pressure-retaining capabilities of the duct system in which it is located.
2.2.1 Attribute 1 - Identification In 180 inspection points to confirm the installation of the proper component, five deviations were reported where the identification number was missing. It was verified that, although the item identification number
Rsvision: 1 Pega 3 of 20 RESULTS REPORT
(N LJ . ISAP VII.c (Cont'd)
Appendix 16 (Cont'd) 2.0 DISCUSSION OF RESULTS (Cont'd) was missing, the correct equipment item was installed.
This verification was accomplished by visual comparison of the installed item to drawings and by review of the installation contractor's QC inspection reports, which documented that the correct item was installed in accordance with the applicable drawing.
Therefore this deviation type was determined to be insignificant. No adverse trend was identified.
2.2.2 Attribute 2 - Location In 122 inspection points for location, two deviaticas were reported for equipment located outside the 4-inch tolerance stated in the installation procedures. The fs
() two deviations were air flow monitors that are f rom 1 foot to 1 foot 9 inches downstream of the design location.
An evaluation concluded that air flow monitor accuracy is not affected as long as it is located at least one duct diameter from an abrupt change in air flow direction. In both cases the air flow monitors were not located as shown on the design drawings because of interference problems. They were at least one duct diameter from any change in air flow direction. As no impact on the design function occurred, these deviations were determined to be insignificant. Based on the nature of the deviations, it is unlikely that an air flow monitor would be improperly located, within one duct diameter of an abrupt change in air flow, in the remainder of the population. Compared to air flow monitors, deviations in location for other HVAC equipment would usually have less effect on the item's function. Therefore, if this type of deviation was to occur in the remainder of the population, it la not likely that it would result in a construction deficiency. No adverse trend was identified.
2.2.3 Attribute 3 - Orientation N/ In all 103 inspection points, orientation was verified to be as required by the design drawing, and no deviations were reported.
l 1
i
Revisi nr '1 i
'Page 4 of 20' I RESULTS REPORT-ISAP VII.c (Cont'd).
1
- Appendix 16 (Cont'd)
'j 2.0 DISCUSSION OF RESULTS (Cont'd) 2.2.4 ' Attribute 4 - Configuration Proportional sampling was used- in combination with the :
- HVAC Ducts and Plenums Construction Work Category for this attribute.- The following is a discussion of the deviations reported for this attribute in this population'only. The trend analysis for the individual deviation types for this attribute in this population
'is reported in this appendix. Refer to HVAC Ducts and Plenums (Appendix'15) for the combined trend analysis results for this attribute.
E- In 156 inspection points, 15 deviations were reported .
where configuration did not comply with the detail drawings. Deviations were reported for fire damper sleeves, fire damper bladu catches, modulating' dampers, and retaining angles. .No deviations were reported in ten inspection points ~ for flexible connections.
Fire Damper Sleeves In 57 inspection points, nine deviations were reported where the distance between a wall and a companion flange on a fire damper sleeve was not as shown on the design drawings.
An evaluation concluded that the 3-inch minimum and 6-inch maximum dimension between a wall apd companion
. flange was only specified for ease of construction and accessibility. The affected items were installed and I
remain accessible. Since this type of deviation would only affect ease of installation, which has been completed, this type deviation was determined to be insignificant. The functional capability of the fire damper sleeve would not be affected even if this type of deviation were to occur in the uninspected portion of the population. Therefore, no adverse trend was identified.
O l
[ _ _ _ _ _ ._ ._ _____-_ -
Ravision: 1 page 5 of 20 RESULTS REPORT
(}
ISAP VII.c (Cont'd) l Appendix 16 (Cont'd) 2.0 DISCUSSION OF RESULTS (Cont'd)
Fire Damper Blade Catches In 13 inspection points, three deviations were reported where fire damper blade catches were located below the i curtain travel path instead of above, as required by installation procedures. Only horizontal-mounted fire dampers have blade catches.
An evaluation of the blade catches below the curtain travel path' revealed that the fire daspers had been abandoned in place and are not required for fire protection. Thus, these three deviations were determined to be insignificant. Although the fire dampers were inoperable and not required for plant operation, they still formed part of the system flow O path-and were retained as valid sample items. In addition, the fire damper manufacturer reported that' the blade catches are non-directional and can be mounted above or below the curtain travel path. If a condition of this type should occur in an operable fire damper, it would not affect the fire damper's functional performance. Therefore, this deviation type was determined to be insignificant, and no adverse trend was identified.
Modulating Dampers In 36 inspection points, two deviations wbre reported where modulating dampers had fail-open springs that appeared to be mounted on the wrong side, compared with vendor installation instructions.
Further investigation showed that the springs were correctly mounted on all 36 dampers, but that the nameplates used by the vendor's installation instructions as an orientation guide were inverted on the two with deviations. Thus, the contractor had installed the fail open springs correctly. Therefore, these deviations were determined to be insignificant.
e No adverse trend was identified.
~
l i
i
_ _ _ . _ _ _ _ _ . _ . _ _ _ _ _ _ - . _ . . . ~ _
R2 vision: 1 Pago 6 of 20
/~N RESULTS REPORT ISAP VII.c (Cont'd)
Appendix 16 (Cont'd) 2.0 DISCUSSION OF RESULTS (Cont'd)
Retaining Angles In 40 inspection points for retaining angles, one deviation uns reported where retaining angles were not installed as required by the insta11er's detail drawing.
An evaluation showed that the retaining angles were not installed but a duct hanger was installed. Both retaining angles and duct hangers are designed to prevent axial movement of the duct. The duct hanger was shown on the contractor's duct support location drawing. Since a duct hanger and a retaining angle cannot be installed in the same location but serve the
/~' same purpose, this deviation type has been determined
(,, to be insignificant. If deviations of this type were to occur in the uninspected portion of the population, they would not affect the axial movement of the duct.
Therefore, no adverse trend was identified.
2.2.5 Attribute 5 - Flange Bolt Holes In 21,700 inspection points, 1,169 deviations were reported where flange bolt holes did not comply with the vendor or Gibbs & Hill drawings. Deviations were reported for these conditions: oversized holes, misalignment of mating flange bolt holes, improper flange bolt hole spacing and improper bolt-hole-centerline-to-edge-of-flange distance.
Oversized Holes In 5,425 inspection points on 230 flanges, 24 deviations on eight flanges, which included one deviation on an actuator mounting bracket, containing 120 inspection points, were reported where oversized holes resulted in washers not fully bearing on the flange surface. This type of deviation ranged from one to seven oversized holes on a flange, with seven of 21 lockwashers on one flange partially pulled into the holes being the most extreme case.
x.
- R2 vision: 1 Pa-gs 7 of 20 RESULTS REPORT V[ \
ISAP VII.c j
(Cont'd)
Appendix 16 (Cont'd) 1 l
2.0 DISCUSSION OF RESULTS (Cont'd) )
(
An evaluation, based on considering that the seven bolts with recessed washers were non-functional, showed that the resulting stress in the remaining effective bolts was still less than 25 percent of the allowable stress and did not affect the structural integrfty of the flange. Therefore, the bolting group's ability to compress the flange gasket and remain air tight was not affected. Because the remainder of the reported deviations were evaluated to be less revere than those considered in this evaluation, all of the deviations were determined to be insignificant. Because of the typically conservative design of all flanges in this population, the functional capability of the equipment flanges would not be affected if similar deviations t were to occur in the uninspected portion of the
( population. No adverse trend was identified.
's-Misalignment of Mating Flange Bolt Holes In 5,425 inspection points on 230 flanges, 347 deviations on 41 flanges, containing 1100 inspection points, were reported where bolt holes were misaligned.
This tesulted in bolts being cocked or bolt holes that were not used. A total of 218 of the deviations resulted in unused holes and 129 resulted in cocked bolts. The 218 unused holes are holes that were drilled in the duct flanges and did not match the equipment holes when installation took place. The !
unused holes are not requ1 red and have no effect on the structural integrity of the duct joints. Misaligned holes that resulted in cocked bolts ranged from one to 18 on a single flange. The most extreme case was where 18 of 36 bolt holes were misaligned on one flange.
An evaluation based on considering that the 18 misaligned bolts were non-functional, showed that the seismic loads used for design were conservative in comparison to the actual seismic loads at the flange.
The design loads are four times greater than the
/
seismic loads that would occur at the flange. The
( actual loads are based on the natural frequency of the rigidly supported equipeent at the flange. The
Revision: 1
- Page 8 of 20
. v)
(q RESULTS REPORT ISAP VII.c (Cont'd)
Appendix 16 (Cont'd) 1 2.0 DISCUSSION OF RESULTS (Cont'd) remaining effective bolts are structurally adequate.
Therefore, the bolting group's ability to compress the flange gasket so that the flange remains air tight is not affected. Because the remainder of the reported f deviations were evaluated to be less severe than this.
evaluated condition, all of the the deviations were determined to be insignificant. A bounding evaluation confirmed that for a similar deviation in a hypothesized worst case location, based on the conservative design of the joints afforded by the seismic design approach and inherent in the supplied equipment flanges, the functional capability of the ficage would not be affected. No adverse trend was f-identified.
d Flange Bolt Hole Spacing In 5,425 inspection points on 230 flanges, 397 deviations on 34 flanges, containing 800 inspection points, were reported where bolt holes were not within the spacing tolerance of 11/16-inch as shown on the vendor drawings. This deviation type occurred 213 times on the minus side, bolts being closer than required, and 184 times on the plus side, bolts being farther apart than required. The deviations ranged from 1/8 inch to 3-3/4 inches on the minus side and from 1/8 inch to 2 inches on the plus side. All 34 flanges had the required number of bolts Installed.
The most extreme case was on a flange where one bolt hole was 8 inches on center when 6 inches on center was required.
A bounding analysis was performed to determine the effect of the increased bolt spacing on the strength of the flanged joint. The 6-inch spacing causes an increased bending moment in the flange angles. The analysis showed, however, that the bolts and not the angles limit the strength of the joint. If the bolts were to be loaded to their code-allowable stress, the local angle bending stress with an 8 inch on center N~ bolt spacing would be 91 percent of the code-allowable bending stress. Since the bolts would reach their code-allowable stresses at lower loads, the increased I
R2 vision: 1 Page 9 of 20 RESULTS REPORT ISAP VII.c (Cont'd)
'AppendLa 16 (Cont'd) 2.0 DISCUSSION OF RESULTS (Cont'd) bending moment in the angles has no ef fect on the structural integrity of the joint. The design capacity of the bolts is not affected by this deviation, and there is no change in the design margin in the strength of the bolt. Because this analysis bounds all bolt spacing deviations, they were determined to be insignificant. Based on the insignificance of this deviation type, the functional capability of the flange would not be affected if similar deviations were to occur in the uninspected portion of the population. No adverse trend was identified.
Bolt-Hole-Centerline-to-Edge-of-Flange Distance In 5,425 inspection points on 230 flanges, 401 O deviations on 31 flanges, containing 1,000 inspection points, were reported where bolt-hole-centerline-to-edge distances on the flange did not comply with the vendor or architect / engineer's drawings.
A total of 398 of these deviations was for edge distances less than 3/4 inch but at least 3/8 inch. A calculation was developed to establish the minimum acceptable centerline-to-edge distance based on the flange material and the duct gauge thickness. The results of this analysis showed there was adequate material to prevert shear failure of the flange. In all 398 cases the molt would fail before the sheet metal or angle, and the less-than-specified-edge distance had no effect on the capacity of the joint.
Therefore, these deviations were determined to be insignificant.
The three remaining deviations were 3/16 inch to 1/4 inch from the edge on one 52-bolt flange. An evaluation based on considering the three bolts to be nonfunctional concluded that three of 52 (approximately six percent) represents an insignificant reduction in the capacity of the bolted flange. This evaluation showed that the remaining bolts on the flange were O
- structurally adequate, and the bolting group's ability to compress the flange gasket so that the flange i
1 remains air tight was not affected. Because the
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ - _ a
,T Rsvision: 1
]
Page 10 of 20
.O' RESULTS REPORT ISAP VII.c (Cont'd)
Appendix 16 (Cont'd)
I 2.0 DISCUSSION OF RESULTS (Cont'd) reported deviations were less severe than the evaluated condition, the deviations were determined to be insignificant. Based on the insignificance of this deviation type, the functional capability of the flange would not be affected if similar deviations were to occur in the uninspected portion of 'che population. No ,
adverse trend was identified. -l 1
An additional evaluation due to the number of.
deviations reported for this and other attributes is summarized in Section 2.2.10.
2.2.6 Attribute 6 - Duct Connections-Bolt Tension In 21,700 inspection points on 230 flanges, 387 deviations were reported where bolt tension did not comply with installation procedures. Deviations included no lockwashers installed, loose bolts, missing j bolts, and lockwashers not compressed.
I No Lockwashers Installed In 5,425 inspection points on 230 flanges, 347 deviations on 45 flanges, containing 600 inspection points, were reported where lockwashers were missing.
The Sheet Metal Air Conditioning Contractors National Association, Inc provides standards for duct fabrication. These standards do not specify the need for lockwashers. The installation contractor's ;
requirement for installing lockwashers was to provide a simple means of determining when the joint gasket was tight (i.e., tightness was achieved when the split lockwasher was compressed flat) . A specific evaluation of the deviations included a visual observation of the l affected joint to ensure that the gaskets were uniformly compressed, and that the bolts were at least hand tight. It was determined that the affected joints were sufficiently tight, even though lockwashers were O not always present to indicate that condition. Based on the insignificance of this deviation type, the functional capability of the affected equipment is not !
reduced. Therefore, no adverse trend was identified.
, Ravision: 1
, Page 11 of 20 f1
. RESULTS REPORT ISAP VII.c (Cont'd)
Appendix 16 (Cont'd) l 2.0 DISCUSSION OF RESULTS (Cont'd)
Loose Bolts, Missing Bolts, Lockwashers Not Compressed In 16,275 inspection points on 230 flanges, 40 deviations on 20 flanges, containing 1800 inspection points, were reported where 26 loose bolts, one missing bolt, and 13 lockwashers that were not compressed did not comply with the installation procedures. The 26 loose bolts were reported on 13 flanges, with six-loose bolts on a 38-bolt flange being the worst case. The one missing bolt was reported on another 38-bolt flange. The 13 lockwashers that were not compressed were reported on seven flanges, with three on a 24-bolt flange being the worst case.
() An evaluation, which considered the six loose bolts to be nonfunctional, showed the design seismic loads to be conservative in comparison to actual seismic loads at the flange. The 32 effective bolts are structurally adequate; therefore, the bolting group's ability to compress the flange gasket so that the flange remains air tight is not affected. Because the reported deviations, including the one missing bolt and 13 lockwashers not fully compressed, were evaluated to be less severe than this evaluated condition, all of the deviations were determined to be insignificant. Based on the insignificance of this deviation type, the functional capability of the flange would,not be affected if similar deviations were to occur in the uninspected portion of the population. No adverse trend was identified.
An additional evaluation due to the number of deviations reported for this and other attributes is summarized in Section 2.2.10.
2.2.7 Attribute 7 - Duct Connections-Gasket Installation In 230 inspection points, 41 deviations were reported where the gasket was not visible around the entire
() perimeter as required by insta11er's procedure. In 23 deviations for damaged gaskets there was a small section of gasket missing, i.e., 1/4 inch to 15 inches long and 1/4 inch to 1-3/4 inches deep (the maximum
Rsvision: 1 j Page 12 of 20 t ,j RESULTS REPORT I
ISAP VII.c (Cont'd)
Appendix 16 (Cont'd) l 2.0 DISCUSSION OF RESULTS (Cont'd) depth occurred on a gasket 2-1/2 inches wide). Another 18 deviationc occurred because parallel sheet metal flange surfaces were drawn together at their edges such that they caused a local visual obstruction for the confirmation of the presence of the gasket.
For 25 of the deviations a field evaluation conducted when the ventilation systems were operating determined that sufficient gasket material existed to provide an air-tight seal. For the deviations where this could not be confirmed, because systems were not operable, duct leakage was evaluated based on the size of the missing gasket. The flow area of each missing gasket section represents less than 2 percent of the joint's g--)t
( flow area, and the resulting air flow into the joint would have a negligible effect on the system's performance. Therefore, the deviations were determined to be insignificant. Five of these deviations were located in exhaust systems, s ,a cit operate under negative pressure relative to adjacent areas. All safety-related ventilation systems in this construction work category are designed potentially to collect air-borne radioactive contaminants and therefore they operate under negative pressure. Based on the insignificance of this deviation type, the functional capability of the flange would not be affected if similar deviations were to occur in the upinspected portion of the population. No adverse trend was identified.
2.2.8 Attribute 8 - Duct Connections-Bolt Thread Engagement In 5,425 inspection points, 15 deviations were reported regarding thread engagement. Eleven of these deviations involved duct-to-duct connections and four involved the attachment of a modulating damper bracket to the duct to duct connection.
Out of a total of 230 flanges, 11 deviations on six r
7-s flanges, containing 176 inspection points, were
' reported where bolts were not flush with, or did not project beyond, the face of the nut as required by the installer's procedures. All 11 deviating bolts were
Revision: 1 Page 13 of 20
()j f
RESULTS REPORT ISAP VII.c (Cont'd)
Appendix 16 (Cont'd) 2.0 DISCUSSION OF RESULTS (Cont'd) slightly below (one thread or less) the face of the nut. The mest extreme case' occurred where three bolts on one 8-bolt flange were slightly below the face of the nut.
Evaluations of each of these deviations determined that sufficient thread engagement existed to provide the full bolt design capacity. An analysis of bolt and nut capacity showed that 70 percent of nut thread engagement is required to obtain full bolt design capacity. Because the number of threads varies with nut size, the remaining 30 percent of nut threads not engaged is equivalent to approximately 1-3/4 to 2 threads. Therefore, these deviations were determined 7 ~g to be insignificant. Based on this analysis, should
(
similar deviations occur in the uninspected portion of the population, no reduction in design capacity would result, No adverse trend was identified.
The remaining four deviations involving a modulating damper bracket reported that the threads of four bolts were slightly below (less than two threads) the face of the nut. An analysis of bolt nut capacity showed that 70 percent of nut thread engagement is required to obtain full bolt design capacity. The number of-
- threads varies with nut size. The remaining 30 percent of nut threads not engaged is equivalent to approximately 1-3/4 to 2 threads. Based on this analysis using a 3/4-inch nut, approximately 68 percent of the nut thread was engaged resulting in less than a ;
3 percent reduction in design capacity. These four deviations were determined to be insignificant. Should similar deviations occur in the uninspected portion of q the population, it is unlikely that a construction a deficiency would occur. Therefore, no adverse trend {
1 was identified.
2.2.9 Attribute 9 - Accessory Connections In 777 inspection points for accessory connections, 22 deviations were reported where accessory connections did not comply with the installation procedures. The deviations were reported for modulating dampers, ]
Rsvision: 1 Page 14 of 20 f
f RESULTS REPORT ISAP VII.c (Cont'd)
Appendix 16 (Cont'd) 2.0 DISCUSSION OF RESULTS (Cont'd) gravity dampers, and fire damper sleeve attachments.
No deviations were reported in 10 inspection points for air flow monitor station instrument connections.
Modulating Dampers In 320 inspection points for accessory connections on modulating dampers, 5 deviations were reported on 40 actuator mounting brackets thst did nor comply with the installation procedure. The deviations were for three bolts missing, one loose bolt, ano one misdrilled hole.
For the deviation where two bolts were missing on a four-bolt actuator mounting bracket, further O- investigation determined that the damper had been locked open and the actuator, electrical and air connections were removed. The actuator bracket will not be utilized to support an actuator. Therefore, this deviation has been determined to be insignificant. The actuator mounting bracket was mounted offset due to an interference problem, making the installation of four bolts impossible without drilling two new holes. Based on the nature of the deviation and low severity, it is unlikely that an actuator mounting bracket would be improperly located in the uninspected portion of the population such that a construction deficiency would result.
For the cases of one missing bolt, one loose bolt, and one misdrilled hole, an evaluation that considered the bolts to be non-functional showed the design seismic loads to be conservative when compared to the actual seismic loads at the flange. The remaining effective bolts are structurally adequate; therefore, the bolting group's ability to compress the gasket was not l affected. Because the reported deviations were less severe than the evaluated condition, they were 1 determined to be insignificant. Based on the r~s insignificance of these deviation types and the f
conservatism afforded throughout the plant in the design approach for all of the HVAC equipment connections, the functional capability of the equipment would not be affected if similar deviations were to )
occur in the uninspected portion of the population. No I adverse trend was identified.
I
Revision: 1 Page 15 of 20 RESULTS REPORT ISAP VII.c (Cont'd)
Appendix 16 (Cont'd) 2.0 DISCUSSION OF RESULTS (Cont'd)
Gravity Dampers In 147 inspection points for 49 gravity damper accessory connections, 13 deviations on 10 dampers were reported where the damper counterweights and the counterweight arms were not installed according to vendor requirements. All of the gravity dampers in this population were inspected. The deviations were for five loose counterweight arms, three counterweights that were loose, four counterweight arms that stuck open at various positions, and one arm that was taped in the open position.
A field evaluation of the five counterweight arms that O were reported to be loose revealed that all the arms were fully threaded into their bushings and were threaded tight, but the locking nuts that are provided co secure the arms in place were loose. Because the arms were threaded tight and the dampers did perform their intended function, it was determined that the deviations were insignificant.
The three counterweights that were loose vary in size and weight from 10 to 20 pounds and are mounted on 3/4-inch diameter counterbalance arms. Each counterweight was attached to the counterbalance arm by a set screw. A field inspection revealed that the counterweights could be rotated about the axis of the counterbalance arms manually, but did not move up or down on the counterbalance arms more than 1/4 inch. At that point, resistance was encountered. After the resistance was encountered, it required a great deal of force to rotate or move the counterweights along the counterbalance arms. Because the gravity dampers did perform their intended function and because it was very unlikely that the counterweights would move more than 1/4 inch up or down the counterbalance arms, these deviations were determined to be insignificant.
() In the case of the four counterbalance arms that stuck open et various positions, each case was evaluated separately. Each evaluation for the damper, or the system in which it is located, showed that the deviation would not prevent the safety-related function from being performed when required.
Rsvicion: 1 Page 16 of 20
,3
) RESULTS REPORT l ISAP VII.c (Cont'd)
Appendix 16 (Cont'd) 2.0 DISCUSSION OF RESULTS (Cont'd)
The first case was a sleeve that was installed on the damper shaft with two set screws. The sleeve could be rotated around the axis of the damper shaft. The second case is one where a gravity damper stuck halfway open. This is attributed to the counterweight not being adjusted properly. Consequently, in both cases
.l the counterweights may not close the dampers when air flow is off. However, a field evaluation showed that in both cases the damper blades will tend to close due to their own weight and will stop at a slightly open position. When the emergency fan coil units that are serviced by these gravity dampers start, the pressure and developed flow in the opposite direction will force p the damper blades to close tight. Since the purpose of l I a gravity damper is to prevent the back f1ciw of air and this has been achieved, the dampers will perform their intended safety-related function.
The third case is where the day tank room exhaust fan that is serviced by this gravity damper operates at plant normal and emergency conditions. During t.ormal operations this fan is exhausting to atmosphere and the four diesel room fans are shut down. Therefore, a backflow of air is not possible. Even in an emergency condition, where this fan was out of service and removed, it is unlikely that any back flow of air could occur because the diesel exhaust fans are,also exhausting to atmosphere (no significant back-pressure).
The fourth case is where the emergency fan coil unit that is serviced by the gravity damper is installed in parallel with another emergency f an coil unit. These two emergency fin coil units are provided to maintain the design conditions in the electrical area during loss of off-site power or a LOCA condition. Both fans are energized simultaneously for an emergency condition. Back flow of air would exist in normal p
operations where flow could leak from the main air supply through the discharge of the fan, causing the (V) I
- - _ _ _ _ __ J
Ravision: 1 Page 17 of 20
("')
's/ RESULTS REPORT ISAP VII.c (Cont'd)
Appendix 16 (Cont'd) 2.0 DISCUSSION OF RESULTS (Cont'd) fan to free wheel backwards. An analysis showed that starting the fan while free wheeling backwards could increase run-up time by 18 percent, but this would not prevent the fan from obtaining full flow condition within an acceptable amount of time, and no damage to the fan motor would occur.
An evaluation of the gravity damper that has the counterbalance arm taped in the open position determined that the damper was installed to provide a; positive means of preventing the back flow of contaminated air into an area requiring personnel access during an emergency. This area is continuously ventilated by the Primary Plant Ventilation Exhaust T System. Since there is no forced air supply, air is l
) supplied by infiltration. An analysis showed that a negative pressure is maintained. Maximum pressure downstream of the subject damper is 0.017 inches of water less than the area pressure. Therefore, back flow of air through the taped-open gravity damper cannot occur. This deviation has been determined not to be a construction deficiency.
The four deviations with the counterweight arms that stick open at various positions, and the deviation with the arm taped open have been determined to be notable.
Even though the dampers will not affect the safety J function of the systems in which they are installed, they do not meet the requirements of the installation procedures. A recommendation for improvement is discussed in Section 4.0.
Because all of the gravity dampers in this population were reinspected, and no construction deficiencies were identified, no adverse trend exists. j Fire Damper Sleeve Attachments In 300 inspection points for 57 fire damper sleeve attachments, four deviations were reported where fire
( damper sleeve attachments did not comply with the installation procedures. Three of the deviations were for bolt spacing, and one deviation occurred where the bolt heads were located inside the sleeve instead of the outside. j l
l l
Revision: 1 Page 18 of 20 0 RESULTS REPORT ISAP Vll.c (Cont'd)
Appendix 16 (Cont'd) 2.0 DISCUSSION OF RESULTS (Cont'd)
An evaluation that compared the seismic loading on the dampers to the existing bolting capacity showed that the deviations do not affect the structural or functional integrity of the fire damper. The seismic loading of the dampers range from 20 to 275 pounds compared to the 1,300 to 9,500 pounds as-installed bolting capacity. Therefore, the three deviations were determined to be insignificant.
In the one case of the bolt heads being on the inside of the sleeve, the bolt heads were specified on the I outside to ensure that the bolts nuts could be tightened. A specific evaluation of the deviation included a field investigation, which ensured that the bolts were at least hand tight. This deviation was determined to be insignificant.
An evaluation concludeda th't it is unlikely thac, if a similar deviation exists in the uninspected portion of the population, it would affect the functional capabilities of the damper. No adverse trend was identified.
2.2.10 Summary Evaluation of Deviations The number of deviations identified during reinspection was judged to warrant additional evaluation.
The installation of HVAC equipment required a large number of bolts and associated hardware to connect the flanges of the two items. The tolerances on some of these requirements were unnecessarily restrictive, given the relative insensitivity of variations in these parameters. For example, bolts were to be spaced every 4 inches along a flange, with a plus-or-minus 1/16 inch tolerance. Many bolts were spaced closer than the minimum required distance, resulting in many deviations where the installed hardware was actually stronger than desi ned.
E Even those deviations where requirements were exceeded in an adverse direction uere I
O found to have minimal effects on the integrity of the ductwork. This is a result of the large design margin typical of HVAC systems, which in turn is due to the relatively low loading conditions.
Rsvision: 1 Page 19 of 20
- - RESULTS REPORT ISAP VII.c (Cont'd)
Appendix 16 (Cont'd) 2.0 DISCUSSION OF RESULTS (Cont'd)
The large number of deviations in the HVAC equipment installation of bolts and associated hardware was not unexpected, given the nature of the fabrication and installation procedures and the nature of the hardware. 1 The lack of any findings in this population was also not unexpected, given the insensitivity of the overall functioning of the hardware to deviations on individual parts of an item, and given the typically large design margin for HVAC systems.
l 2.3 Analysis of Documentation Review Results As all attributes were available for reinspection, a documentation review was not required.
3.0 ROOT CAUSE AND GENERIC IMPLICATION A root cause and generic implications analysis is not required.
4.0 RECOMMENDATIONS This section provides a recommendation for improvement based on the reinspection results for Attribute 9, accessory connections.
Gravity Dampers A recommendation for improvement is made to tack weld both the counterweights and locking nuts to the counterbalance arm and to initiate the appropriate procedural controls to implement this in i the future installations and on existing dampers.
5.0 CONCLUSION
S Based on the findings of the reinspection, and the results of the proportional sample trend analysis for the configuration attribute l reported in Appendix 15, HVAC Ducts and Plenums, there is )
reasonable assurance that the hardware in this construction work !
O. category is adequately installed in conformance with the design.-
i
Rsvision: 1 Page 20 of 20
. RESULTS REPORT-ISAP VII.c (Cont'd)
Appendix 16 (Cont'd) l Table 16-1 Summary of Reinspection Results HVAC Equipment Installation-Deviation Classification Number of Inspection Number of- Insigni- Construction Attribute Points Deviations ficant Notable Deficiency
- 1. Identification 180 5 5 0 0-
- 2. Location 122 2 2 0 0-
- 3. Orientation 103 0 0 0 0
- 4. # Configuration 156 15 15 0 0
- 5. Flange Bolt Holes 21,700 1,169 1,169 0 0
- 6. Duct Connections -
Bolt Tension 21,700 387 387 0 0
- 7. Duct Connections Gasket Installation 230 41 41 0 0
- 8. Duct Connections Bolt Thread Engagement 5,425 15 15 'O O
- 9. Accessory Connections 777 22 17 5 0 TOTALS 50,393 1,656 1,651 5 0 (Approximately 50,500)
- This attribute is proportionally sampled in combination with the HVAC Ducts and Plenums population. The numbers listed here only reflect that portion of the inspection points and deviations associated with the HVAC Equipment Installation fO population.
Revision: 1 Page 1 of 42 RESULTS REPORT ISAP VII.c (Cont'd)
Appendix 17 Mechanical Equipment Installation 1.0 REVIEW PROGRAM IMPLEMENTATION 1.1 Construction Work Category Description The construction work category of mechanical equipment installation is comprised of the installation, repair, and modification of safety-related mechanical equipment such as pumps, heat exchangers, shop-fabricated tanks, compressors, water chillers, filters, fuel transfer tubes, and large HVAC-equipment. All mechanical equipment was installed, repaired, or modified using similar techniques, procedures, and personnel, i
1.2 Population Size and Sample Selection l For this construction work category, a total of 466 items was identified, with a population of approximately 427 items accepted by QC. Of the total of 170 accessible items reinspected, 139 first sample items were randomly selected from the population. All attributes in the first random sample were reinspected a minimum of 60 times (or 100 percent of accessible samples) except as noted in Section 1.3. An additional three second sample items were randomly selected to ensure that at least 60 safe-shutdown hardware items were reinspected.
1.3 Attributes Selected Sample items were reinspected for the following attributes:
Attribute 1 - Equipment identification Attribute 2 - Orientation Attribute 3 - Levelness Attribute 4 - Configuration Attribute 5 - Grout Attribute 6 - Anchor bolt location Attribute 7 - Anchor bolt hardware, engagement, and tightness
R2 vision: 1 Page 2 of 42
,9 .
i
}
(/ RESULTS REPORT ISAP VII.c (Cont'd)
Appendix 17 (Cont'd) 1.0 REVIW PROGRAM IMPLEMENTATION (Cont'd)
Attribute 8 - Rotating equipment alignment Attribute 9 - Temporary coatings and preservatives The following attribute did not pertain to every sample item, but was reinspected whenever it applied:
Attribute 10 - Rust (stainless steel only)
This attribute was only reinspected to collect information on the possible existence of carbon steel contamination of stainless steel. Contamination had been found on field-fabricated liners but was not O expected on manufactured mechanical components.
h Because reinspection of this attribute was for information collection purposes only and a problem was not expected, this attribute was not required to meet minimum sample size requirements nor were statistical conclusions drawn for this attribute.
The reinspection results for the following attribute were combined with those from other populations on a proportional sampling basis to achieve the required sample size:
Attribute 11 - Concrete expansion anchors Documentation review was not required because alt safety-related attributes of mechanical equipment installation could be reinspected.
2.0 DISCUSSION OF RESULTS l 1 2.1 Summary of Results For reinspection, 217 Deviation Reports were issued describing l 891 deviations. Approximately 15,400 inspection points were
- encountered in performing the reinspection. See Table 17-1 1 l for results of the reinspection.
l In the 11 reinspection attributes of this population there was one configuration deviation regarding broken bolts that was j i
evaluated to be a construction deficiency. A special case
Ravision: 1 Page' 3 of 42 1 (~ .
RESULTS REPORT ISAP VII.c (Cont'd) l Appendix 17 (Cont'd) l 2.0 DISCUSSION OF RESULTS (Cont'd) regarding configuration deviations on manway covers.was also identified. See Section 2.2.4 for a discussion of the deviations associated with this deficiency and special case.
No adverse or unclassified trends were identified.
2.2 Analysis of Reinspection Results This section provides, by attribute, a discussion of the l
reinspection deviations, an analysis of the effect of the i deviations on the functional capability of the equipment, and an analysis for the presence of trends. The function of mechanical equipment varies with equipment type but, in general, is to maintain structural integrity and satisfy r"' operational requirements.
2.2.1 Attribute 1 - Equipment Identification In all but five of 170 inspection points, the equipment nameplate contained the equipment identification number (tag or vendor identification number) shown cnt the installation drawings. This attribute was inspected to ensure that the correct equipment was installed and to ensure that no confusion would arise during plant operations.
Two of the deviations involved reactor coolant pump nameplates that carried equipment identification numbers that were different from those used in the plant equipment list and in the manufacturer's instruction manual. The expected equipment identification numbers were TBK-RCPCPA-02 and TBX-RCPCPA-04, where the "RCPCPA" portion identifies the complete pump assembly. However, the nameplates contained the designation "RCPCPI", which identifies only the pump internals. The motor on each pump was clearly marked with the correct pump name, and the doorways of the associated pump cubicles were also clearly marked. An evaluation concluded that there was ample and obvious identification for each pump. The
(, deviations were determined to be insignificant.
Ravision: 1 Page 4 of 42
-(]
\ RESULTS REPORT ISAP VII.c (Cont'd)
. Appendix 17 (Cont'd) 2.0 DISCUSSION OF RESULTS (Cont'd)
The remaining three deviations dealt with equipment or portions of equipment that had been interchanged among redundant installations or between Units 1 and 2. In two of these cases, it was readily determined from other records and through physical inspection of the equipment (e.g. size, manufacturer, configuration, other markings, etc.) that the physical equipment interchange was authorized and had been correctly made; it was only the associated marking requirements that had not been implemented. Both of these cases involve equipment located in cubicles that'are clearly marked with the correct equipment identification number and/or equipment description. It was therefore concluded that there was no chance of later confusion regarding O equipment identification during plant operations.
These two deviations were determined to be insignificant.
In the last case of interchanged equipment, one of three identical pumps was found to have two different nameplates, one on the motor and another on the pump itself. The two nameplates each carried the equipment identification number of one of the other of the three pumps, and neither nameplate matched the equipment installation drawings for the pump being inspected. It was determined that, in the course of modifications and repairs, the motors and pumps had been interchanged as required to keep at least one pump in service. Because all three pumps and motors are identical, the equipment's ability to perform its intended function is not affected. Therefore, this deviation was determined to be insignificant. The misleading nameplates currently installed could result in maintenance or operating confusion, however, and a recommendation for improvement was made as described below.
In one of the above-described cases of interchanged equipment, the interchange was authorized in a construction operation traveler that required the O
A nameplates to be changed to reflect the transfer. This nameplate change was not properly made. In the other two cases, the as-found condition had been authorized {
by means of " Permanent Equipment Transfer" documents. l The authorizations required that metal tags be attached !
l l
Ravision: 1 Page 5 of 42 RESULTS REPORT ISAP VII.c (Cont'd)
Appendix 17 (Cont'd) 2.0 DISCUSSION OF LESULTS (Cont'd) to the equipment documenting the transfer. No such tags were installed. The fact that the equipment marking requirements specified in Permanent Equipment Transfer documents have been overlooked at least.twice suggests that all such documents and the associated equipment should be reviewed. A recommendation for improvement in Section 4.2.1 is made so that other occurrences of this problem are identified and corrected.
These five deviations constitute a trend. However, because it is considered highly unlikely that deviations of this type could result in a construction deficiency elsewhere in the construction work category, no adverse trend was identified.
2.2.2 Attribute 2 - Orientation In 533 inspection points, the equipment was oriented as shown in the applicable design drawings. The l reinspection for orientation also included, as applicable:
- Flow direction arrow pointing in the correct direction
- Nozzles oriented as shown on thg installation drawings
- Piping connected to correct nozzles No deviations were reported.
2.2.3 Attribute 3 - Levelness l In 337 inspection points, 16 deviations were reported in which the equipment was not level within the tolerances specified in the manufacturer's instructions or the equipment erection specification, as applicable.
l In four cases involving HVAC equipment, the reported i deviations were very small. During field observations i of the operating equipment it was noted that the I
_-- ___-_________ a
Rsvision: 1 Page 6 of 42 O
l 1( ,/ RESULTS REPORT l ISAP VII.c (Cont'd) !
'~
Appendix 17 (Cont'd) i 2.0 DISCUSSION OF RESULTS (Cont'd) equipment did not exhibit excessive vibration and appeared to be operating satisfactorily. The acceptability of a slight out-of-level condition was also confirmed with the equipment manufacturers. It was concluded that such deviations would not affect the safety-related function of the equipment. When the a HVAC equipment had attached piping, the effects of the )
out-of-level condition on piping connections and pipe j l supports were evaluated to ensure that any effect on '
stress levels during normal and design operating conditions would be negligible. These four deviations were determined to be insignificant.
f
Two deviations were reported in which the plumbness of
( vertical air accumulator tanks exceeded the plumbness tolerance of 1/16 inch end-to-end by up to 1/2 inch.
These deviations were evaluated to determine the effect on tank structural stability and the effect of any increased loads on the anchor bolts and concrete foundation. The effect of the out-of-plumb condition on piping connections was determined to be negligible because these connections were all made after installation of the accumulators. The evaluation showed that the out-of-plumb condition did not affect accumulator operability or stability and produced insignificant increases in the loads on the anchor bolts and on the concrete foundation. These deviations were determined to be insignificant.
One deviation concerned the tracks inside a fuel transfer tube; one track was found to be higher than the other by approximately 3/16 inch. This deviation was evaluated using field measurements to determine if the out-of-level condition would prevent the fuel transfer car from passing through the fuel transfer tube. The evaluation concluded that, although clearance.s were reduced, sufficient clearance remained for proper operation of the fuel transfer system.
Therefore, this deviation was determined to be insignificant.
l t
Revision: I f Page 7 of 42 j l
gs RESULTS REPORT
()
ISAP VII.c /
(Cont'd)
Appendix 17 (Cont'd) 2.0 DISCUSSION OF RESULTS (Cont'd)
Three deviations involve service water pumps that have j been operating regularly with no apparent problems. !
Analysis of the levelness deviations on these pumps j included a review of vibration levels recorded during initial pump operation. The vibration levels were normal and therefore support a conclusion that the slight out-of-level conditions will not affect the satisfactory operation of these pumps. The service water pump discharge piping includes an expansion joint f and thrust-resisting pipe supports that will accommoda'te the minor pump levelness deviations. Therefore, these deviations were determined to be insignificant.
fss Five additional deviations involve the levelness of a centrifugal charging pump, a containment spray pump, Q and two seal coolers attached to the containment spray 1 pump. Because the proper operation of components such I as the seal coolers is relatively insensitive to minor variations in levelness, the evaluation concluded that the deviations on these coolers are insignificant.
With regard to the containment spray pump and the charging pump, it was also determined that the slight j levelness deviations are insignificant and that {
deviations of this type and severity would not affect J the satisfactory operation of the pumps. Furthermore, difficulties in making piping connections or in ;
achieving proper alignment and operating problems such I as noise or vibration would make levelness deviations {
readily apparent before the safety-related function of I these pumps, or any pumps with similar levelness deviations in the uninspected portion of the population, would be affected. ;
i The remaining deviation involved a compressor skid in I which the motor and compressor were out of level with I respect to each other. Evaluation of this deviation showed that the out-of-level condition could affect the ability to align the equipment within allowable tolerances, thereby resulting in appreciable wear.
7 While this condition would not prevent the equipment t
V) from performing its safety-related function, it could .
I result in the need for more frequent maintenance, thereby reducing equipment availability. For this reason, the deviation was determined to be notable.
R3 vision: 1 Page 8 of 42
\- RESULTS REPORT ISAP VII.c (Cont'd)
Appendix 17 l (Cont'd) 2.0 DISCUSSION OF RESULTS (Cont'd)
No adverse trend was identified for any of the insignificant deviations described above because the l deviations have no safety significance and, even if deviations of this type were found in the uninspected portion of the population, there would be no impact on the affected equipment. With regard to the notable deviation, this condition was observed only once out of 144 inspection points for equipment with motor drivers.
Even if a deviation of this type were to exist in the uninspected portion of the population, it would not be safety-significant because 1) at least one additional alignment will be performed for doweling prior to plant operations and any difficulties in achieving satisfactory alignment would be likely to be noted and f/)
x- corrected, and 2) other operating indications such as noise or vibration vould provide ample warning before the safety-related performance of the equipment was affected. Accordingly, no adverse trend was identified.
2.2.4 Attribute 4 - Configuration In 359 of the 6,027 inspection points, the equipment had not been configured as specified in the applicable design document or Construction Operation Traveler.
Configuration reinspection focused on determining the overall adequacy of mechanical work associated with the assembly, modification, or repair of the equipment in the field to the extent that such reinspection could be performed without extensive disassembly. Only those samples affected by such activities were reinspected for configuration. The goal of the configuration reinspection was to permit conclusions regarding the overall adequacy of assembly, repair, or modification work typically performed by millwrights or similarly trained and experienced personnel. This was achieved by combined analysis of the data collected on the individual types of such activities. Examples of work reinspected under this attribute are component assembly
(/)x ,_
or reassembly, positioning and attachment of accessories, structural welding, and tightening of bolted connections and threaded fittings. In the combined analysis, the deviations associated with this attribute were sorted into three general categories:
bolt tightness, welding, and miscellaneous.
1
Rsvision: 1 Page 9 of 42 0
i RESULTS REPORT ISAP VII.c (Cont'd)
Appendix 17 (Cont'd) 2.0 DISCUSSION OF RESULTS (Cont'd)
While an effort was made to reinspect all field-completed assembly, repairs, and modifications for each of the sample items, it was not always possible to identify the exact scope of work performed. ,
Whenever it was possible to identify the scope of work .
performed in sufficient detail to specify the required reinspection and/or to provide acceptance criteria, reinspection of that particular assembly, repair, or .
modification was performed. Sufficient examples of the types of work required in performing repairs, modifications, and initial equipment assembly were reinspected to support conclusions about the overall adequacy of this work activity.
Bolt Tightness (O)
There were 303 deviations in which nuts or bolts turned when a torque equal to 70 percent of the latest torque specified for that bolted connection was applied. The 70 percent value used in performing the reinspection is based on experience with bolted structural joints and was adopted as the criterion for mechanical assembly bolting torque verification.
Two deviations involved compressor hold-down bolts on a control room air conditioning unit. There are six such bolts, three on a side. Two on one side vere found to be sheared off. A calculation of the resulting seismic loading on the remaining bolts determined that the stress in the one remaining bolt on the affected side of the compressor would exceed by a factor of two the stress allowed by AISC code. Accordingly, these deviations could affect the ability of the compressor and, consequently, the air conditioning unit to perform its safety-related function. A construction deficiency was therefore identified for configuration - broken bolts deviations. Refer to Section 3.1 for a discussion of the root cause.and generic implications of this deficiency.
O A total of 81 deviations occurred in 11 gasketed joints on ten separate pieces of equipment. In a gasketed joint, some reduction in bolt or stud tension will naturally occur due to relaxation of the gasket. The
Rsvision: 1 Page 10 of 42 1
) (T l RESULTS REPORT t.s_-)
ISAP VII.c (Cont'd)
Appendix 17 (Cont'd) 2.0 DISCUSSION OF RESULTS (Cont'd) amount of such " tension-relieving" that occurs will vary with time and is also affe:ted by the type of gasket material and by the magnitude of the initially-applied torque. Initial torque values are l
typica?.ly specified to ensure sufficient gasket compression to achieve a leak-tight joint and normally assume some tension-relievint. In such cases, maintenance of the bolt tensten resulting from the specified torque value is not a rigid engineering requirement, provided the joint remains leak-tight.
Accordingly, the evaluation of these deviations consisted of visual inspection of the joints to determine if there was cry evidence of leakage, review of installation requirements and procedures to identify 7l
(
factors that might further explain a reduction in the
\- # initially-induced tension and, in two cases, contact with the equipment manufacturer to confirm that reduced bolt or stud tension is acceptable as long as there is no evidence of leakage. No evidence of leakage was observed at any of the affected joints. In two cases, a pump and a heat exchanger, it was noted that the equipment had not yet been operated. As in the other cases, Icakage would not be expected for the as-found joint conditions; it was also concluded that if any leakage occurred, it would be readily detected during preoperational testing.
Other factors that typically contribute to tension relieving in the affected joints or in individual bolts or studs are lockwasher relaxation, bolt spacing, and torquing sequence. Several of the joints were assembled using split-ring lockwashers on each bolt or stud. This type of lockwasher will eventually relax, thereby reducing bolt or stud tension. The evaluation also noted that bolt spacing and torquing sequence occasionally contribute to tension reduction in the first few bolts or studs in the sequence, particularly when the bolts or studs are close together, even if site procedures for torquing sequence are followed. i rS Thus, it was concluded that the above-described cases I
(_,) of less-than-expected torque values are attributable to I
I
Revision: 1-Page 11 of 42
,m,
%- RESULTS REPORT ISAP VII.c (Cont'd)
Appendix 17 (Cont'd) 2.0 DISCUSSION OF RESULTS (Cont'd) the normal tension-relieving mechanisms associated with gasketed joints and have no effect on equipment safety function. These 81 deviations were determined to be insignificant.
Another 140 deviations were associated with sheet metal covers on two reactor coolant pump shaft lift oil pumps. .These covers, which were installed onsite, are made of stainless steel sheet metal plates and are bolted together with rubber sponge gaskets at each bolted joint. The evaluation determined that the tension reduction experienced by these bolts was due to relaxation of the rubber gasket material, a phenomenon T common to all gasketed joints as explained above. The (s/
evaluation concluded that these deviations do not jeopardize the safety function of either the shaft lift oil pumps or the reactor coolant pumps and they were determined to be insignificant.
Another 53 deviations affected attachment bolting of such items as couplings, fan covers, and seal housings on ten pieces of equipment. All of these deviations occurred on rotating equipment; an evaluation concludsd that the lowered torque values were the result of tension-relieving that normally occurs in the bolted connections of such equipment after operation. The evaluation also found that deviation from. design or recommended torque values does not by itself lead to loss of function of the affected part. Loss of function would only follow if all indications of a faulty assembly were ignored as the unit is operated for an extended period of time. This is an unlikely situation given the normal monitoring and preventive maintenance performed on rotating equipment.
Accordingly, these deviations were determined to be insignificant.
O
Revision: 1 Page 12 of 42 g
RESULTS REPORT ISAP VII.c (Cont'd)
Appendix 17 (Cont'd) 2.0 DISCUSSION OF RESULTS (Cont'd)
There was one deviation involving a missing bolt on an air conditioning unit electrical panel. Electrical cables are connected to the center of this panel; the missing bolt is one of 12 that secure the panel to the housing of the air conditioning unit. A likely explanation for this missing bolt is that it was once used to attach, or was removed in order to attach, a cable clamp, ground strap, or similar fixture. Such dual uses for electrical panel mounting bolts are conmon. The evaluation considered the seismic loading imposed on the panel by the cables. It was determined that the rematr.ing 11 bolts were adequate to withstand both vertical and lateral seismic loading. This fs
('} deviation was determined to be insignificant. An occasional missing bolt of this type is not likely to lead to a construction deficiency.
l Nine deviations were reported affecting nine of the 20 I studs and nuts that secure a safety injection accumulator manway cover. In evaluating these deviations, calculations demonstrated that the remaining 11 nuts, as spaced on the manway, are adequate to ensure the leak-tightness of the manway cover under both operating and hydrostatic test pressures. Based on design calet!1ations performed by the manufacturer, the supplied manvay cover bolting provides 59 percent excess capacity in th& design operating condition and 48 percent excess capacity at hydrostatic test pressure. If, as the evaluation of these deviations assumed, the nine undertorqued nuts are ineffective, this margin is reduced to 14 percent for the operating pressure condition and to 3 percent for the hydrostatic test pressure condition. The assumption that they are ineffective is, however, a very conservative one as they were all found sufficiently tight not to loosen during operation and to contribute effectively to the leak tightness of the manway. These deviations msy result in a capacity
/^' reduction above 10 percent and have been determined to be notable.
1
Revision: 1 Page 13 of 42
() RESULTS REPORT ISAP VII.c (Cont'd)
Appendix 17 (Cont'd) 2.0 DISCUSSION OF RESULTS'(Cont'd)
There were an additional nine deviations among the 16 bolts that secure the pressurizer manway cover. In the evaluation of these deviations, it was_shown by.
calculation that the existing bolt tension was sufficient to ensure the leak-tightness of the manway cover against design pressures. T'.e evaluation relied upon the fact that the manway cover did not leak during the hot functional test already performed on the pressurizer to estimate bolt tension in the nine deviating bolts These estimates were used in the calculation fot esign loading conditions. These deviations were determined to be notable.
Apart from the two deviations known to represent a construction deficiency, the remainder of the
\- configuration-bolt tightness deviations are bolted connections that did not meet the minimum tightness criteria. Most of these deviations were found to be insignificant and are attributable to normal loosening mechanisms such as bolt and gasket relaxation and vibration. The missing electrical par.el attachment bolt appears to be a simple and insignificant oversight. It is reasonable to conclude that similar deviations would be found in the uninspected portion of the population and that those deviations would likely also be insignificant.
The notabic deviations on the safety injection accumulator and pressurizer manway covers are less easily attributable to normal loosening mechanisms.
Because the analysis of these deviations concluded that substantial margin reductions may have occurred in these bolted connections, configuration - manway covers has been identified as a special case. A discussion of the root cause and generic implications of this special case is presented in Section 3.2.
As the foregoing discussions illustrate, there are several reasons why bolt tightness deviations may occur and, concomitantly, why the occurrence of such k's,/~') deviations is not unusual. Because deviations of the type reported are so typical, a variety of testing and monitoring programs are commonly relied upon to reveal
Revision: 1 Page 14 of 42 RESULTS REPORT ISAP VII.c (Cont'd)
Appendix 17 (Cont'd) 2.0 DISCUSSION OF RESULTS (Cont'd) problems that may arise from loore bolts. For pressure-retaining components and systems, hydrostatic tests at pressures above those expected during normal or accident conditions are conducted, emong other reasons, to ensure the leak-tightness of bolted connections. If leaks occur, and they occasionally do, the leaks are repaired. For rotating equipment, trial runs and vibration monitoring programs are relied upon to confirm satisfactory operation or to reveal, among other things, the presence of loose connections or attachments. Accordingly, deviations of the type and severity encountered are not likely to result in a construction deficiency in the uninspected portion of l
I the population and therefore do not represent an adverse trend.
Welding A total of 28 of the configuration deviations involved structural welds made in the course of equipment repairs or modifications on five pieces of equipment.
Four of the deviations involved the configuration of welds used to attach a pressure transmitter support to each of four control room air conditioning units. Each support was fabricated by welding the end of a length of 2-inch by 2-inch angle iron to the 2-1Ech side of a 2-inch by 3-inch housing brace on the air conditioning unit. The design drawing called for the attachment to be made by mecns of a 3/16-inch fillet veld all around.
However, the specified orientation of the attachment made it impossible for a fillet veld to be used on one side of each support; a penetration weld was used instead. The evaluation concluded that the use of a penetration weld on one side of each support is acceptable because a penetration veld is stronger than a fillet weld. These deviations were therefore determined to be insignificant.
The remaining 24 deviations all involve welds on 12 filter top clamp assemblies in a ventilation exhaust filter. Each end bracket of each top clamp assembly (total of 24 end brackete) is required by the design
Ravision: 1 Page 15 of 42
/N
'(_) RESULTS REPORT ISAP VII.c (Cont'd)
Appendix 17 (Cont'd) 2.0 DISCUSSION OF RESULTS (Cont'd) r drawing to have a tack weld on top and a full length fillet veld on each side. All of the tack welds exhibited overlap and lack of fusion, and all of the fillet welds were only part length. The evaluation determined that the tack w ids did not have any structural value but only av:ved to hold the top clamp assemblies in place while the fillet welds were made.
The evaluation also determined that the part length fillet welds were sufficient to permit the top clamp assemblies to function as intended during all loading conditions. Because the part length fillet welds afforded capacity far above the applied loads, these 24 deviations were determined to be insignificant.
/N
() _
Because of the minor nature of these deviations, it was concluded that deviations of the same type and severity occurring in the uninspected portion of the population would also be insignificant. Therefore, no adverse trend was identified.
Miscellaneous An additional 36 of the 359 configuration deviations did not fall into either of the preceding categories.
These deviations occurred on ten pieces of equipment.
1 There were six deviations involving the use of shims 1 under the mounting bolt washers of a diesel generator l building ventilation exhaust fan when shims were not explicitly specified in the design documents that i provided mounting details. The evaluation determined i that the shims were used to distribute the mounting bolt bearing load evenly over the bearing surface.
- This is a common pract ce and is preferable to the reduced bearing surface contact that would exist without the shims. These deviations were therefore determined to be insignificant.
Two deviations were reported in which the gap between (h
N/
safety injection pump bearing housings and the associated deflector rings was less than specified by the manufacturer. An evaluation determined that the function of the deflector rings is simply to keep dust
Rcvicion: 1 Pega 16 of 42
(g RESULTS REPORT
. \, /
ISAP VII.c (Cont'd)
Appendix 17 (Cont'd) 2.0 DISCUSSION OF RESULTS (Cont'd) and dirt out of the bearing housings. The manufacturer confirmed that a gap less than specified would not affect the operation of the pump provided some gap is present. Because a small gap does exist between the deflector rings and the associated bearing housings, these deviations were determined to be insignificant.
One deviation involved the motor support for a fan coil unit. This support is constructed of angle iron and consists of two horizontal members and two 45 degree braces. The braces were originally welded to the outside of each horizontal member. In the course of modifications intended to raise the horizontal members to permit alignment of the motor with the fan, the 45 degree braces were moved and welded to the inside of the members and were therefore not consistent with the Os design document specifying the modification. The evaluation of this deviation concluded that the change in configuration simply amounted to rotation of the members and braces about their axes, and that the structural capability of the motor support was unaffected. This deviation was therefore determined to be insignificant.
There was one deviation in which a containment spray pump constant level oiler was found to be less than ,
hand tight. The oiler was attached to the side of a I pump bearing housing by a horizontal length of pipe.
The evaluation determined that it was not possible to turn the connecting pipe by hand and that the oiler / pipe assembly did not turn easily. The evaluation concluded that the oiler was sufficiently tight that neither operating nor seismic loads would l cause it to loosen further. The evaluation also l determined that thread sealant on the threads of both l the oiler and its connecting pipe has prevented leakage thus far and, considering the small magnitude of the forces that would be experienced by this oiler, it is likely that no loosening and therefore no leakage will occur in the future. Accordingly, this deviation was
() determined to be insignificant.
Rsvision: 1 Page 17 of 42 f3
\I
% RESULTS REPORT ISAP VII.c (Cont'd)
Appendix 17 (Cont'd) 2.0 DISCUSSION OF RESULTS (Cont'd)
Another deviation in this category involved minor leakage of hydraulic fluid from the hydraulic snubbers on a steam generator. In this case, evidence of leakage was noted at a hose clamp connection in the hydraulic supply line, just upstream of the snubber control valves. A check of hydraulic reservoir level was conducted over a period of 30 days and no measureable drop in reservoir level was detected. In addition, the snubber manufacturer was contacted to confirm that low reservoir level would have no effect on snubber performance as long as the snubbers themselves remained full up to the outlet of the control valves. Because the leakage was noted on the
/~ inlet side of these control valves, and because'the
\ leakage rate was undetectable in the reservoir after 30 days, this deviation was determined to be insignificant.
The remaining 25 deviations in this category involved bolting hardware and the engagement of bolts used to attach five pieces of equipment to structural steel supports. There were six cases of missing washers, six cases where washers were trimmed to fit in the steel channel used as a support, four cases where shims had been substituted for the specified washers, and nine cases where a small gap existed between the nut and the support. The evaluations determined that.the missing washers were not required for satisfactory tightening of the attachment, the trimmed washers and the substituted shims provided sufficient bearing area for the bolting, and the gaps would have no effect on the ability of the equipment to operate satisfactorily under seismic loading conditions. All 25 of these deviations were therefore determined to be insignificant. Further discussion of deviations sLailar to the 25 deviations described above is found in Section 2.2.7, Anchor Bolt Hardware, Engagement, and Tightness.
Because of the minor nature of the 36 above-described deviations, it is expected that similar deviations )
occurring in the uninspected portion of the population would also be insignificant. No adverse trend was identified.
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ . _ _ - . _ _ _ _ )
Rsvision: 1 Page 18 of 42 i
l RESULTS REPORT ISAP VII.c-(Cont'd)
Appendix 17 (Cont'd) 2.0 DISCUSSION OF RESULTS (Cont'd)
Overall Trend Analysis Except for the deviations resulting in the construction deficiency involving broken bolts and the special case involving manway covers, all of the configuration deviations are insignificant. Such deviations are common in construction. It is therefore likely that similar deviations would be encountered in the uninspected portion of the population, but considering the insignificant nature of the deviations found during the reinspection, it is also likely that such deviations would be insignificant. Therefore, apart from the above-mentioned construction deficiency and special case, no adverse trend was identified for the (q,)
configuration attribute.
2.2.5 Attribute 5 - Grout In eight of 254 inspection points, the grout viewed at the edges of equipment base plates or support legs was found to exhibit voids or deterioration extending under the baseplate. In seven of these cases, it was determined that the relatively small amount of missing or damaged grout would not affect the ability of the equipment to perform its safety-related function nor would it impair the performance or capacity of the equipment. These seven deviations were determined to be insignificant.
In one case, a substantial amount of grout was found to be missing under one of four equipment support legs.
This condition was conservatively evaluated by assuming that there was no support at all under the faulted leg.
Despite this assumption, it was determined that the equipment was still capable of performing its intended function, even under design seismic loads. Although the evaluation assumed no support at all, the affected leg actually has a considerable amount of support from the remaining grout and from the. shims used to
() establish the grout clearance. Therefore, this deviation was determined to be insignificant.
1
Revision: 1 Page 19 of 42 RESULTS REPORT I ISAP VII.c (Cont'd)
Appendix 17 (Cont'd) 2.0 DISCUSSION OF RESULTS (Cont'd)
A bounding evaluation was performed to extend the above evaluation to the worst case condition (higher elevation and hence higher seismic load) fcr similarly supported equipment. It was concluded that deviations similar to either of these two types occurring in the uninspected portion of the population would also be insignificant. Furthermore, it is unlikely that i missing grout would affect the safety-related l performance of equipment without ample warning from such operating problem indications as vibration or noise. Therefore, no adverse trend was identified.
2.2.6 Attribute 6 - Anchor Bolt Location In 1,389 inspection points, there were 12 deviations reported in which the anchor bolts in the sliding feet of a heat exchanger were not positioned in accordance with the installation requirements.
The evaluation of one of these deviations concluded that the as-found position of the anchor bolts would allow sufficient room for thermal growth of the heat exchanger and would have no effect on its proper operation. This deviation was determined to be insignificant.
An additional two deviations involved the' sliding feet on a heat exchanger that had been subjected to hot functional testing. Because the sliding feet will not necessarily return to their original position after such testing, the acceptance criterion in this case was to confirm that no signs of deformation existed in the heat exchanger feet or the concrete foundation.
Deformation in these areas would indicate that the bolts may not have been properly positioned within the slotted holes of the sliding feet when the heat exchanger was originally installed. In the subject heat exchangers visible gaps under the center support and under one of the sliding feet appeared to be signs G of deformation. The evaluation determined that, although the supports do not fully bear on the embedded bearing plates in the concrete foundation, the condition was due to placement of the bearing plates in
'Ravicion: 1
. Peg 2 20 of 42 RESULTS REPORT-ISAP VII.c (Cont'd)
' Appendix 17 (Cont'd) )
2.0 DISCUSSION OF RESULTS (Cont'd) )
l the concrete and was not deformation caused by thermal j expansion of the heat exchanger. It was also determined that there is ample bearing surface to support the heat exchanger during both normal operating l and seismic conditions. These deviations were determined to be insignificant.
Another nine deviations involved the anchor bolts in the sliding saddle of the Unit 2 containment spray system chemical addition tank. The anchor bolts in the sliding saddle of this horizontal, cylindrical tank are installed along a line that is not parallel to the anchor bolts in the fixed taddle at the other end of the tank, resulting in some anchor bolts being too close to one end of the slots in the sliding saddle V while others are too close to the other end. Analysis of these deviations concluded that the installed position of the anchor bolts would permit sufficient room for thermal growth of the tank, but two anchor bolts would interfere with contraction of the tank should the ambient temperature drop below 70 degrees.
In the worst of these two cases, there was no clearance at all between the anchor bolt and the end of the slot.
The analysis considered the resulting stresses in the tank shell, in the weld between the sliding saddle and the tank, in the sliding saddle, and in the affected anchor bolt. The analysis showed that the resulting stresses remained well below ASME allowables, except in the case where all stress is assumed to be absorbed in the affected anchor bolt. In this case, the anchor bolt stress is increased to essentially 100 percent of allowable. The analysis concluded that the noted deviations would not affect the safety-related function of the tank. However, because of the magnitude of the potential increase in anchor bolt stresses, these deviations have been determined to be notable.
In investigating the possibility that deviations of this type might occur in the uninspected portion of the population, the counterpart tank in Unit I was examined. The anchor bolts in the sliding saddle of v that tank exhibited a skew pattern identical to that found on the Unit 2 tank. Field measurements confirmed
l Rsvision: 1 Page- 21 of 42
(. j RESULTS REPORT ISAP VII.c (Cont'd)
Appendix 17 (Cont'd) 2.0 DISCUSSION OF RESULTS (Cont'd) that it is the embedded anchor bolts that are skewed, not the tank saddles. Further investigation revealed that.the most likely explanation for the skewed anchor bolts is a survey error in locating the anchor bolt line during placement of the Unit 1 foundation. The error was apparently propagated to the Unit 2 foundation because templates made from Unit I were often used to locate the anchor bolts for Unit 2.
It therefore seems reasonable to assume that deviations of this type and severity could occur in the uninspected portion of the population. However, it has been determined that deviations of this type do not constitute an adverse trend, for the following reasons.
O Three of the deviations encountered were determined to be insignificant and, considering the nature of these deviations, it is concluded that other, similar l' deviations in the uninspected portion of the population would likely also be insignificant. Regarding deviations of the type encountered on the containment spray system chemical addition tank, the analysis of these deviations was extremely conservative in that it was assumed for each of the stressed elements (i.e.,
l tank shell, saddle weld, saddle, and anchor bolt) that l all stress resulting from contraction of the tank would l be absorbed in that element. In addition, it was assumed that all stress would be absorbed at the sliding end of the tank. In fact, the fixed saddle and its anchor bolts would share in absorbing the stress.
In addition to being highly conservative, the analysis is also bounding because, given the cause of these deviations, the as-found anchor bolt arrangement is far more likely than all anchor bolts being at one end of the slotted holes. This is because the anchor bolt line is determined by survey relative to building lines. A survey error of the type likely to have resulted in these deviations could only mean that the O I
I R vision: 1 I
Page 22 of 42 1
j RESULTS REPORT ISAP VII.c (Cont'd)
Appendix 17 (Cont'd) 2.0 DISCUSSION OF RESULTS (Cont'd) survey device was mispositioned; the building lines are not moveable. Mispositioning the survey device would therefore always result in a skewed anchor bolt line and not an anchor bolt line that is uniformly too close to one end of the slotted holes. Skewed anchor bolts will always mean that only one, or possibly two, anchor l' bolts become highly stressed during thermal growth or contraction.
Even if it is assumed that these one or two anchor bolts fail, analyses of other deviations in Section 2.2.7 have shown that as many as half of the anchor bolts on a piece of equipment can be assumed to be O missing without affecting the safety-related function of equipment. The skew pattern also implies that not all of the anchor bolts on the sliding end can become overstressed. Therefore, even in the unlikely event that some of the anchor bolts should fail, at least half of the anchor bolts on the sliding end would remain to provide vertical restraint during a seismic event, thereby resisting any overturning or bending moment on the equipment.
Other factors also support the bounding nature of the analysis done for these deviations. The inspection included 11 out of 30, or nearly 37 percent, of the population items with sliding ends. Furthermore, aside from the affected tank and its Unit I counterpart, all other population items with a sliding end are heat exchangers, which are more rigid than the affected tank and, therefore, are better able to withstand increased stresses than a relatively thin-walled, atmospheric tank. In addition, if all anchor bolts were mispositioned at one end of the slotted holes, all bolts would share the applied load and the resulting stress in each bolt would be less than that considered in the analysis.
Although, for the reasons discussed above, the anchor bolt location deviations do not constitute an adverse (n] trend, the potential for reduced anchor bolt capacity and unnecessary stresses in equipment and supports
i Revision: 1 Page 23 of 42 lh RESULTS REPORT ISAP VII.c (Cont'd)
Appendix 17 (Cont'd) 2.0 DISCUSSION OF RESULTS (Cont'd) suggests that it would be prudent to identify and !
correct all deviations of this type. Therefore, a recommendation for improvement is given in Section 4.2.2.
2.2.7 Attribute 7 - Anchor Bolt Hardware, Engagement, and Tightness In 402 cases out of approximately 5,968 inspection j points, anchor bolt hardware, engagement, or tightness' was not as specified in the applicable foundation details or in the applicable installation procedures.
The acceptance criterion for hardware required that the nuts, washers, spacers, shims, and/or other hardware required by the applicable foundation detail be G installed on the equipment. For engagement, it was required that anchor bolt nuts be in contact with and flat (no greater than a 1-to-20 slope or 3 degrees angularity) against the baseplate, and that the anchor '
bolt should be no lower than flush with the top of the anchor bolt nut. For tightness, anchor bolts were checked against the tightness specification given in the applicable design document or Construction Operation Traveler.
In 214 cases, washers called for on the applicable foundation detail were found to be missing. All of these deviations were determined to be insignificant because none involved circumstances in which washers would be important, i.e., to prevent galling or brinnelling of high-strength parts, to permit accurate torque application, or to provide adequare bearing area for the design hole size.
In 82 cases, washers were either trimmed or a shim plate or beveled washer was substituted for the flat washer that had been specified. In all of these cases, the evaluation concluded that the deviations were not safety-significant because the remaining portions of trimmed washers or the substituted hardware were G adequate to provide a non-galling surface and to distribute anchor bolt load evenly. These deviations were determined to be insignificant.
l Rsvision: 1 Page 24 of 42 I
RESULTS REPORT ISAP VII.c 3 (Cont'd)
Appendix 17 )
(Cont'd) f 2.0 DISCUSSION OF RESULTS (Cont'd) i In 34 cases of missing locknuts on rotating equipment, f 49 cases of insufficient anchor bolt engagement, and 15 cases of insufficient anchor bolt tightness, it was shown either by direct calculation or by reference to desigr. calculations that the deviations would not affect the safety function of the equipment, even under seismic loads. Many of these deviations were evaluated by assuming that the affected anchor bolts provided no j vertical restraint during a seismic event. In the most (
conservative of these evaluations-(enveloping the maximum of five deviations found on a single item), six out of 12 anchor bolts were assumed ineffective, thereby doubling the seismic load in the remaining anchor bolts. Even in this case, the anchor bolt
) stresses remained well within allowable limits. The V evaluations also considered bending forces when anchor bolts were found to be installed as much as 6 degrees from vertical (twice the allowable bolt angularity).
It was concluded that bent anchor bolts would straighten during the first seismic loading cycle in the vertical direction, and would thereafter resist seismic loads as well as anchor bolts with no deviation. Gaps between the anchor bolt nut and the equipment baseplate were typically evaluated by assuming that the affected nut was missing altogether.
In cases where the straightening of bent anchor bolts under seismic loads was evaluated, it was determined that any small gaps left after straightening would not result in significant impact loads on the bolts. In those cases when small baseplate movements were postulated as a result of the assumption of missing anchor bolt nuts, the effect on piping stresses was evaluated and was determined to be insignificant.
Considering the conservatism in the aforementioned 3 analyses, these deviations were therefore determined to be insignificant.
In eight cases of missing locknuts on one piece of I; equipment, it was determined that the missing hardware l was not actually required. It was determined that the l
- foundation detail for the catalytic hydrogen recombiner l l
l was erroneously labeled " compressor", implying that the recombiner contains rotating equipment. A review of
_ - _ _ ___-___________ __-_ Q
R1 vision: 1 P933 25 of 42 1
RESULTS REPORT
()
ISAP VII.c (Cont'd) 1 Appendix 17 (Cont'd)
I I
2.0 DISCUSSION OF RESULTS (Cont'd) 1 the manufacturer's assembly drawings for this equipment revealed that there is no compressor or any other rotating equipment associated with this device, nor are anchor bolt locknuts a manufacturer's requirement. The l evaluation concluded that single anchor bolt nuts are adequate for the catalytic hydrogen recombiner.
Accordingly, these deviations are insignificant.
None of the above-described deviations was determined to be an adverse trend because it is highly unlikely that deviations of these types would result in a construction deficiency in the uninspected portion of the population. This determination is based upon a number of factors, as dis <assed below, including the
[~) insignificant nature of the deviations and the highly
\~/ conservative assumptions used in the evaluations.
With regard to missing, trimmed, or substituted washers, the evaluations demonstrated that washers are only important in certain limited situations. None of the noted deviations involved those situations despite the fact that nearly half of all safety-related equipment with anchor bolts was inspected. If a deviation of this type were to occur in a situation where washers are important, the result would be insufficient anchor bolt tightness or engagement (due to reduced bearing area), conditions that are addressed in the following paragraph.
As noted above, it was assumed in the evaluations that an affected anchor bolt would provide no vertical restraint during a seismic event. This assumption was used in evaluating many of the missing locknut, engagement, and tightness deviations to facilitate evaluation and permit a determination of the safety-significance of the deviations. However, this assumption takes no credit for the contribution to vertical restraint provided by the hardware that is actually in place. For example, even though a locknut may be missing, there is still an anchor bolt nut in
(/
x.
s place. Even though an anchor bolt may be bent or the nut may not be down tight and flat against the baseplate, the nut is still there and will provide some
Revision: 1 Page 26 of 42 RESULTS REPORT ISAP VII.c (Cont'd)
Appendix 17 (Cont'd) 2.0 DISCUSSION OF RESULTS (Cont'd) vertical restraint. When these considerations are taken into account, it is concluded that similar I deviations involving missing locknuts, insuf ficient engagement, and insufficient tightness are not likely to result in construction deficiencies if they were to exist in the uninspected portion of the population.
Ii Although none of the anchor bolt deviations described above were notable and no adverse trend was identified, the frequency of these deviations suggests the need for i more careful attention to anchor bolt details. A 1 recommendation for improvement is given in Section 4.2.3 in order to bring the actual installations into conformance with the design drawings.
2.2.8 Attribute 8 - Rotating Equipment Alignment Out of 221 inspection points, the parallel alignment of 45 pieces, the angular alignment of 13 pieces, and the coupling gap of 21 pieces of rotating equipment were found to be outside of the tolerances specified in the mechanical equipment erection specification, manufacturer's instructions, Construction Operation Traveler, or other applicable documents governing the performance of the most recent alignment work.
In 46 of these deviations, the alignment and/or coupling gap were found to be outside of the tolerances specified in the site instructions, but within the tolerances allowed by the manufacturer. Such deviations were determined to be insignificant. The high number of deviations clearly results from an attempt to maintain site-imposed tolerances that are highly conservative.
In the remaining 33 deviations, the alignment and/or coupling gap were outside of both the site tolerances and the tolerances allowed by the manufacturers.
Because these deviations could result in increased wear
l Ravision: 1 )
Page 27 of 42 j
('~)
~# - RESULTS REPORT ISAP VII.c (Cont'd)
Appendix 17 (Cont'd) 2.0 DISCUSSION OF RESULTS (Cont'd) and the need for more frequent maintenance, they were determined to be notable. However, these deviations are not unusual for rotating equipment that has undergone the periodic testing, operation, and maintenance associated with plant construction.
I Because of the expected frequency of such activities, a final alignment and the installation of dowels in the drivers of most rotating equipment to aid in maintaining alignment are not performed until just prior to plant operation. The planned performance of a final alignment just prior to doweling, coupled with the normal periodic checks of alignment conducted as part of the regular preventive maintenance and ASNI
(
Section XI inservice inspection programs, make it very
.k/ ) unlikely that misalignment deviations of the type and nature encountered could result in a construction deficiency in the uninspected portion of the population. Even when the above programs do not apply, the gradual increases in noise, vibration, and temperature resulting from misalignment would signal the need for maintenance long before the equipment would fail to perform its safety function.
Further, all of the QC-accepted and accessible safety-related rotating equipment was inspected; the sample therefore included over 60 percent of all safety-related rotating equipment in the plant. This further reduces the likelihood of a construction deficiency in the uninspected portion of the plant. It has therefore been determined that rotating equipment alignment deviations of the type encountered do not constitute an adverse trend.
2.2.9 Attribute 9 - Temporary Coatings and Preservatives In all 134 inspection points, it was observed that temporary coatings and preservatives had been removed from the installed mechanical equipment as required by
<~ the mechanical erection specification. No deviations k ,') were reported.
._ __________________a
R2 vision: 1 Page 28 of 42
() RESULTS REPORT ISAP VII.c (Cont'd)
Appendix 17 (Cont'd) 2.0 DISCUSSION OF RESULTS (Cont'd) 2.2.10 Attribute 10 - Rust (Stainless Steel Only)
In 47 inspection points involving stainless steel equipment, one deviation was reported in which a rust-colored corrosion product buildup was observed on a stainless steel surface. An evaluation was conducted to determine if the rust resulted from carbon steel contamination of the stainless steel surface. It was determined that the rust had been deposited on the surface as a result of run-off from lightly rusted carbon steel parts in the vicinity, and that no significant penetration of the stainless steel had occurred, nor was any significant penetration from this 7g source possible in the future. Therefore, this i deviation was determined to be insignificant.
t s_/
2.2.11 Attribute 11 - Concrete Expansion Anchors l Inspection of concrete expansion anchors was divided into several characteristics: size, length, type, embedment, spacing, angularity, nut engagement, nut bearing, and concrete damage.
Of the 362 inspection points, nine deviations were identified in the characteristics of nut engagement and bearing, spacing, and embedment. The evaluation of these deviations and analysis for adverse trends are included in the ISAP VII.b.4, "Hilti Anch'or Bolt Installation", Results keport.
2.3 Analysis of Documentation Review Results Documentation review was not required because all safety-related attributes cf mechanical equipment installation could be reinspected.
3.0 ROOT CAUSE AND GENERIC IMPLICATIONS This section provides a root cause and generic implications
(
%- analysis for 1) the construction deficiency for configuration -
i broken bolts and 2) the special case involving configuration -
manway covers.
Revision: 1 Page 29 of 42
( RESULTS REPORT ISAP VII.c (Cont'd)
Appendix 17 (Cont'd) 3.0 ROOT CAUSE AND GENERIC IMPLICATIONS (Cont'd) 3.1 Configuration - Broken Bolts This construction deficiency involves two broken compressor hold down bolts on a control room air conditioning unit.
There are a total of six compressor hold down bolts on the unit, three on a side. Both of the broken bolts are on one side of the compressor.
Root Cause The control room air conditioning units had a history of vibration problems from their original installation in 1979 until late 1984, when a number of modifications were made to stiffen the spring-isolated foundations. The modifications on O the unit in question occurred in October 1984. At this time, the compressor was removed from the unit while a number of stiffener plates and gussets were welded to the support frame.
l The compressor was then reinstalled and the origioal six 5/8 inch hold down bolts were tightened " snug tight". The operational traveler step for the installation of the hold I
down bolts was signed of f by the rigging department, and the subsequent step involving checking each hold down bolt for a
" soft foot" was signed off by both the millwright department and by QC. There is no evidence that this compressor has been removed subsequent to the above-described activity, nor is there any evidence that the hold down bolts have subsequently been retightened or checked for tightness. Nevertheless, the two bolts were either twisted off during the reinstallation of the compressor in 1984, or they failed in service subsequent to that time.
The " snug tight" criterion used for the hold down bolts is defAr.ed in Mechanical Equipment Erection Specification 2323-MS-101, Section 4.11.8, as "the tightness attained by a few impacts upon a hammer wrench or the full effort of a man using an ordinary spud wrench or standard wrench". This is a common definition videly used in the construction industry when the application does not demand a specified torque value.
It is not, however, very specific and it is reasonable to l
() assume that it could occasionally lead to overtightening small bolts. However, the bolts in question are 5/8 inch, A325 high-strength bolts. They are readily accessible and would l
Revision: 1 Page 30 of 42 O RESULTS REFORT ISAP VII.c (Cont'd)
Appendix 17 (Cont'd) 3.0 ROOT CAUSE AND GENERIC IMPLICATIONS (Cont'd) require no unusual wrench configuration for installation.
Although it cannot be positively confirmed, it seems highly probable that these bolts were installed using an ordinary combination wrench or a socket vrench with ratchet drive, neither of which would be much longer than 12 inches. It is very unlikely that the typical workman could apply enough force at the end of a 12 inch lever arm to twist off a bolt of this size and material. It is even more unlikely that he would do it twice and that both occurrences would be left uncorrected and go unnoticed by QC. Furthermore, bolts that have been twisted off exhibit a characteristic 45 degree angle fracture and a relatively smooth fracture surface that typically, except in very brittle materials, shows some
('
plastic deformation. The subject bolts are broken nearly straight across, with only a small central area showing signs of plastic deformation. The remainder of the fracture surface is rough and crystalline in appearance. It therefore seems unlikely that these bolts were twisted off either during installation or in a subsequent retightening.
The possibility that the bolts failed in service is supported by a number of factors. The affected control room air conditioning unit has been in operation on a more or less regular basis since 1982. From that time until late 1984 the unit experienced excessive vibration that was not corrected by repeated attempts to realign the equipment. After a number of modifications specified by engineering, including defeating the spring isolators and the above-mentioned stiffening of the equipment support frame, vibration was finally reduced to acceptable levels. Because the same bolts were used when the compressor was reinstalled in 1984, it is possible that those bolts had been weakened by the preceding two-year period of excessive vibration so that they were no longer able to withstand the normal operating vibration associated with rotating equipment. The history of excessive vibration experienced by this equipment, coupled with the physical appearance of the fracture surfaces of the bolts, support a conclusion that the probable failure mechanism of the
/T compressor hold down bolts is vibration-induced fatigue.
.U There is some evidence to suggest that the vibration problem with this air conditioning unit has only been masked, and not solved, by the modifications undertaken to date. At least one
Revision: 1 Page 31 of 42 RESULTS REPORT ISAP VII.c (Cont'd)
Appendix 17 (Cont'd) 3.0 ROOT CAUSE AND GENERIC IMPLICATIONS (Cont'd)
TU Electric work order subsequent to the 1984 modifications refers to excessive vibration of the air conditioning unit control cabinet. Startup engineers also reported an increase in control room noise level'following the modifications; this is consistent with a transfer of equipment vibration to the floor above the control room. These factors raise the possibility that the modifications undertaken on this equipment did not correct the vibration problem but only cured the apparent symptoms, i.e., the amplitude of the vibration.
If the cause of the vibration has not been corrected, making the equipment support frame more rigid would simply transfer the vibration to the floor. The vibrational stress experienced by the compressor hold down bolts would remain
') A
(\~
excessive and would contribute further to fatigue of the bolts. This possibility, along with the fact that the weakening of the hold down bolts should have been anticipated and bolt replacement should have been specified as part of the 1984 modifications, leads to a conclusion that the primary root cause of the bolt failure was less-than-adequate engineering evaluation.
The documentation associated with the air conditioning unit modification work supports a conclusion that the work was carried out and inspected in accordance with procedures. It is also concluded that the construction and inspection procedures themselves are adequate and did not contribute to failure of the compressor hold down bolts.
- Generic Implications Because all four of the control room air conditioning units experienced excessive vibration and were subsequently modified in the same manner, one generic implication of this deficiency is that all of the compressor hold down bolts on all four units have likely been weakened and could fail prematurely.
Another generic implication derives from the failure to replace the hold down bolts as part of the modifications.
Other equipment within the inspected part of the population
( also experienced problems with excessive vibration and were modified in a similar fashion, i.e., containment spray pump pedestal supports were stiffened without replacement of the pedestal bolts. It is therefore likely that other rotating equipment in the uninspected portion of the population may
Revision: 1 Page 32 of 42 x
~
RESULTS REPORT ISAP VII.c (Cont'd)
Appendix 17 (Cont'd) 3.0 ROOT CAUSE AND GENERIC IMPLICATIONS (Cont'd) have experienced vibration problems that-were subsequently corrected without taking into account the possibility of weakened parts.
Further evaluation is required to determine the overall adequacy of the engineering modifications undertaken to correct the vibration problem on the control room air conditioning units. However, such an evaluation is in the scope of the Design Adequacy Program (DAP). Therefore, this construction deficiency has been referred to DAP for further evaluation and the determination of any additional generic implications and corrective action.
, Summary The primary root cause of the fracture of two compressor hold down bolts on a control room air conditioning unit was less-than-adequate engineering evaluation in failing to anticipate the effect of excessive vibration and specify the replacement of these bolts and, possibly, failure to specify an adequate solution to the vibration problems experienced by the control room air conditioning units.
The recommended corrective actions are presented in Section 4.1.1.
3.2 Configuration - Manway Covers This special case involves the tightness of the manway cover bolts on a primary system pressurizer and on a safety j injection system accumulator. In the case of the pressurizer, nine of the 16 manway cover bolts failed to meet the minimum tightness criterion, i.e., the bolts moved before a torque I equal to approximately 70 percent of the specified torque value was applied. In the case of the safety injection system accumulator, the nuts on nine of 20 manway cover studs failed ;
to meet the minimum tightness criterion.
Root Cause The pressurizer and the safety injection system accumulator are both supplied by Westinghouse as part of the Nuclear Steam Supply System (NSSS). The operational travelers for
Rsvision: 1 Page 33 of 42 RESULTS REPORT ISAP VII.c (Cont'd)
Appendix 17 (Cont'd) 3.0 ROOT CAUSE AND GENERIC IMPLICATIONS (Cont'd) installation of the pressurizer and safety injection system accumulator manway covers indicate that the procedures specified by Westinghouse were fellowed and that appropriate QC checks were performed. Although initials and signatures on the travelers do not guarantee adequate performance of the work, a system hydrostatic test was performed on the safety inspection accumulator and a hot functional test was performed on the pressurizer subsequent to the installation of the manway covers. A review of the hot functional test report and subsequent work documents gave no indication that any manway
. cover leakage was experienced during the test. The possibility of poor workmanship cannot be ruled out altogether. Nevehaeless, the satisfactory performance of these manway covers during hydrostatic and hot functional I testing in conjunction with.the documentary assurance of U adequate procedural and quality controls provided by the operational travelers make it seem unlikely that these deviations resulted from poor workmanship, inadequate supervision, or inadequate quality control.
The effect of torque wrench calibration tolerances on the accuracy of both the torque applied during manway cover installation and the torque applied during the reinspection was considered. Depending on the torque wrench configuration used, such tolerances can amount to as much as nine percent.
However, it does not seem likely that a difference in torque wrench calibration between the time of manway cover installation and the time of reinspection could account for the noted deviations, because such a difference would have affected all bolts or studs on the manway covers, and not just some of them.
The effect of pressure and temperature cycling during hydrostatic and hot functional testing was also considered.
However, such factors were also ruled out because they would affect the entire manway cover and therefore cannot account for deviations on only some of the bolts or studs.
As mentioned earlier, both of the manway covers were installed using the procedures specified by Westinghouse. Although the L number of bolts / studs are different for each of these manways, the procedures employed the same basic methodology:
1
l l
l Revision: 1 Page 34 of 42 RESULTS REPORT ISAP VII.c (Cont'd)
Appendix 17 (Cont'd) 3.0 ROOT CAUSE AND GENERIC IMPLICATIONS (Cont'd)
- 1) Clean and lubricate all bolts / studs and the internal threads on the manway cover flanges;
- 2) Tighten all bolts / nuts to the specified intermediate torque value, following a specified sequence;
- 3) Tighten all bolts / nuts to the specified final torque value, following the specified sequence;
- 4) Remove each bolt / nut, one at a time and in the specified sequence, clean and relubricate the threads, and then retorque to the final specified torque value.
Continue torquing each bolt / nut, in sequence, until there is no more movement when the specified final O torque value is applied.
This general methodology is commonly employed in tightening manway covers, or any bolted connection with many bolts, and will usually result in reasonably uniform tightness of the connection. However, it is particularly difficult to achieve uniform tightness on such installations because the tightening is accomplished by applied torque. Even in an ideal connection, approximately 80 percent of the torque is taken up by friction in the threads and between the bolt head or nut and the manway cover. Only 20 percent of the torque contributes to bolt or stud tension. Dirty or damaged threads, galling in the threads or under the bolt head or nut, .
and any other problem that increases friction in the l connection further reduces the tension achieved for a given ,
level of applied torque. In addition, some elastic recovery I
in the bolt or stud occurs Lemediately upon removal of the wrench, and a large percentage (up to 90 percent) of the l normal tension-relieving experienced in a bolted connection occurs within twenty-four hours. Whr:n many bolts or studs must be tightened, the first few will have already experienced a significant amount of tension-relieving by the time the last is tightened. It is because of this " spring-back" phenomenon that large torques are typically applied in steps and a
" star-shaped" or "criss-cross" torquing sequence is followed so that the tightness of the bolted connection remains ll reasonably uniform. When gaskets are used, the bolted connection will also experience some tension-relieving due to
l l
Rsvision: 1 Page 35 of 42 O
( / RESULTS REPORT l
I ISAP VII.c (Cont'd)
Appendix 17 (Cont'd) 1 i
3.0 ROOT CAUSE AND GENERIC IMPLICATIONS (Cont'd) gasket relaxation; if the gasket is not compressed uniformly, gasket relaxation will not be uniform. Furthermore, gasket relaxation accelerates the " spring back" phenomenon and makes uniform tightness even more difficult to achieve.
The most likely explanation for the noted deviations on the pressurizer and safety injection system accumulator manway covers is that they are due to some combination of the above described factors, with galling under the bolt head or nut likely to be the principal abnormal factor. In this explanation, galling under the bolt head or nut is' thought to have caused indicated torque to reach the specified value-before the required bolt tension was established. Subsequent tension-relieving would then permit incremental rotation at a O torque below 70% of the specified value.
The specified procedure does not preclude the possibility of minor damage to the bolt or stud threads during manway cover installation. The pressurizer manway cover weighs approximately 650 pounds, but it is initially drawn loosely into place by only a few bolts. The lateral load on these few bolts during this first step could result in enough thread damage to cause galling. In fact, evidence of galling was noted on several bolts during a prior removal of the pressurizer manway cover. The safety injection accumulator manway cover is supported on a davit arm and is swung into place over the studs. If the manway cover just brushes any of the studs as it is being positioned, enough thread damage could be sustained to result in galling when the nuts are tightened. Another reason why galling is likely to be a major factor in these deviations is that washers are not used under the pressurizer manway cover bolt heads nor under the safety injection accumulator manway cover nuts. Although washers have been determined to be unnecessary in many mechanical equipment applications, the AISC code indicates that they are necessary to minimize irregularities.in the torque-tension ratio when bolts or nuts are tightened by the calibrated wrench method. The AISC code also points out that, without a hardened washer under the bolt head or nut, galling can be of such a magnitude that resistance.to turning will register a O' high torque reading before the required tension is achieved.
l Revision: 1 Page 36 of 42
!%)
RESULTS REPORT iSAP VII.c (Cont'd)
Appendix 17 (Cont'd) 3.0 ROOT CAUSE AND GENERIC IMPLICATIONS (Cont'd)
The primary root cause of the pressurizer and safety injection system accumulator manway cover deviations is therefore determined to be less-than-adequate vendor procedures because the procedures were not as exacting as required to ensure the uniform tightness of these manway covers. A contributing root cause is less-than-adequate vendor engineering evaluation because of the failure to specify the use of washers under the manway cover bolt heads or nuts.
Westinghouse technical bulletin 87-01 dated January 20, 1987 also suggests that achieving uniform tightness of manway covers may require more exacting procedures than previously thought. Although this bulletin was issued because a number
()
\s -
of sites have experienced bolt seizing on pressurizer and steam generator maavay covers, discussions with Westinghouse indicate that the problems that resulted in bolt seizing at other sites are essentially the same as those believed to have l
resulted in the manway cover Ceviations discussed above. This l.
technical bulletin provides considerably more guidance l
regarding manway cover installation than did the installation procedures in effect at the time tha pressurizer and safety injection system accumulator manway covers were installed.
Among the recommendations made in this technical bulletin are:
match-marking bolts to ensure that bolts are always used in the same holes; the use of guide pins, rather than actual fastener hardware, to align the covers; mock-up training; several passes through the torque sequence at each torque level, rather than the formerly specified single pass at each level below the last; the use of hardened washers under the bolt heads; and a lower final torque value. The technical bulletin also strongly urges utilities to consider the use of axial stud tensioning equipment to simultaneously tension all fasteners in order to reduce the likelihood of fastener seizure and achieve uniform tightness and gasket compression.
Generic Implications The nature of the root cause and the similarity of the manway r cover deviations on both the pressurizer and the safety i i (, injection system accumulator indicate that deviations of this type could affect all manway covers, inspection ports, l
handholes, and other bolted connections within the ;
________a
Revision: 1 Page 37 of 42 RESULTS REPORT ISAP VII.c (Cont'd)
Appendix 17 (Cont'd) 3.0 ROOT CAUSE AND GENERIC 1 IMPLICATIONS (Cont'd)
Westinghouse scope of supply where the installation procedure calls for tightening of the individual bolts or nuts by means l of a torque wrench. This generic implication would not extend to bolted connections such as the reactor vessel head or the reactor coolant pump flanges where stud tensioning devices or the bolt / stud elongation method of tightening is used because these other tightening methods do not rely on the application of torque to individual fasteners.
The generic implication also does not extend to similar bolted connections on other population items supplied by different ;
manufacturers. Even though a number of bolt tightness l deviations on gasketed connections with many bolts were e discovered elsewhere in the population, all of those deviations were of a lesser severity, they occurred on equipment in lower pressure applications, and they were attributable to normal tension-relieving mechanisms.
Deviations of the severity encountered on the pressurizer and safety injection system accumulator are much more likely to occur when large torque values are used to achieve the f astener tension required to contain high system pressures.
For applications outside of the NSSS, less exacting tightening methods will still result in reasonably uniform tightness and satisfactory, leak-tight connections.
The possibility that the manway cover deviations might have generic implications that extend to other attributes in this population was considered. The anchor bolt tightness attribute bears the closest resemblance to the manway cover tightness check that resulted in the noted deviations.
However, anchor bolts are not often torqued; more commonly they are simply installed " snug tignt". Even when anchor bolts are torqued, they are far less sensitive to variations in applied torque than a bolted connection that must be leak-tight. It is therefore not likely that less-than-adequate vendor procedures would have any generic implications extending to the anchor bolt tightness attribute.
A generic implication would exist for other attributes in this O population if they could be affected by less-than-adequate vendor procedures. However, as explained below, a review of
Revision: 1 Page 38 of 42
(
RESULTS REPORT ISAP VII.c (Cont'd)
Appendix 17 (Cont'd) 3.0 ROOT CAUSE AND GENERIC IMPLICATIONS (Cont'd) the evidence for such effects concluded that generic implications for other attributes do not exist. As noted in the root cause discussion above, the vendor procedure that was used in the installation of the manway covers would be adequate in most applications; it was only in somewhat limited application that the level of detail was determined.to be less-than-adequate. None of the other deviations recorded against any other attribute appear to have been due either entirely or in part to inadequate vendor procedures.
Furthermore, none of the other attributes in this population are likely to be as sensitive to the level of detail provided in vendor procedures as the tightening of high-pressure, gasketed, bolted connections. Rotating equipment alignment is
.f the only other attribute of similar complexity, and the analyses of deviations recorded against this attribute have shown that, although failure to achieve alignment within tolerances may result in increased wear of the equipment, operating and periodic monitoring program indications will provide ample warning of a misalignment problem long before the safety-related function of the. equipment is affected. It is therefore concluded that the generic implications of the manway cover deviations do not extend to other attributes within the mechanical equipment installation population.
The possibility that the manway cover deviations might have a generic implication extending into the field-fabricated tanks population was also considered. Although these tanks have manway covers, the tanks were not supplied by Westinghouse and they are not pressure vessels. The manway covers on the field-fabricated ~ tanks are therefore not likely to be as sensitive to the level of detail provided in the manway cover installation procedure as were the manway covers on the pressurizer and the safety injection system accumulator.
Summary The primary root cause of both the pressurizer and the safety injection system accumulator maavay cover deviations is a less-than-adequate vendor installation procedure that failed to provide sufficient detail regarding the steps necessary to l achieve uniform tightness of the manway covers. Because of the failure to specify the installation of hardened washers
_ _____________ __ w
R vision: 1 Page 39 of 42 n
_, RESULTS REPORT ISAP VII.c (Cont'd)
Appendix 17 (Cont'd) 3.0 ROOT CAUSE AND GENERIC IMPLICATIONS (Cont'd) under the manway cover bolts or nuts, a contributing cause of the devittions is less-than-adequate vendor engineering evaluation.
A recommendation for corrective action is presented in Section 4.1.2.
4.0 RECOMMENDATIONS 4.1 Recommendations for Corrective Action 4.1.1 This section provides recommendations for corrective action resulting from the configuration - broken bolts
(}
%_)
construction deficiency:
- 1) Compressor hold down bolts on the four control room air conditioning units should be replaced with new bolts of the correct size and material.
- 2) Rotating equipment that has experienced a vibration problem requiring engineering modifications should be evaluated to determine the effect of fatigue on the
- ability of component parts, including electrical connections in on-skid control cabinets, to fulfill their design function.
Upon completion of the evaluation, any necessary corrections should be implemented.
4.1.2 Configuration - Manway Covers This section provides recommendations for corrective action resulting form the configuration - manway covers special case:
- 1) Revise procedures for installing manway covers and similar closures on vessels and
, ~~s accumulators supplied by the NSSS vendor,
() particularly those that are part of the primary system pressure boundary, to
Ravision: 1 Page 40 of 42 RESULTS REPORT
- ISAP VII.c-(Cont'd)
Appendix 17 (Cont'd) 4.0 -RECOMMENDATIONS (Cont'd) incorporate, as appropriate, the guidance and recommendations in Westinghouse Technical Bulletin 87-01,
- 2) Inspect the manway covers and similar closures on vessels and accumulators supplied by the NSSS vendor, particularly those that are part of the primary system pressure boundary, to ensure that the manway cover ~ ,
flange internal threads and the bolt or stud and nut threads are clean and undamaged. The bolts and nuts should then be retightened, using the revised procedures, to the torque values specified in the applicable design v documents.
4.2 Recommendations for Improvement This section provides, by attribute or group of related attributes, recommendations for improvement:
4.2.1 Equipment Identification Each Permanent Equipment Transfer (PET) affecting i safety-related equipment should be reviewed to ensure that equipment marking requirements associated with the transfer have been implemented. -
4.2.2 Anchor Bolt Location i
All anchor bolts installed in the sliding end of equipment subject to thermal growth should be inspected to ensure that the bolts are positioned as specified in the applicable design documents.
4.2.3 Anchor Bolt Hardware, Engagement, and Tightness All mechanical equipment installations involving anchor ,
bolts should be reinspected to ensure that: 1
- 1) the anchor bolt hardware required on design drawings is installed;
Revision: 1
.Page 41 of 42 J
-!' "1 J
RESULTS REPORT ISAP VII.c !
(Cont'd).
Appendix 17 (Cont'd) 4.0 REC 019ENDATIONS (Cont'd)
- 2) all anchor bolts are tight and flat (no greater.than a 1-to-20 slope) against the baseplate or support;
- 3) the tightness criteria specified in design documents, manufacturer's instructions, and/or Construction Operation Traveler's, or other work documents, as applicable, have been met.
5.0 CONCLUSION
S Based on the findings of the reinspection and satisfactory implementation of the above recommendations, there will be k reasonable assurance that the construction work activities associated with mechanical equipment. installation have been adequately performed in accordance with the design. Conclusions for concrete expansion anchors are separately sta:ed in ISAP VII.b.4.
~
__ l
)
Revision: 1 Page 42 of 42 RESULTS REPORT ISAP VII.c (Cont'd)
Appendix 17 (Cont'd)
Table 17-1 Summary of Reinspection Results Mechanical Equipment Installation Deviation Classification Number of Inspection Number of Insigni- Construction Attribute Points Deviations ficant Notable Deficiency
- 1. Equipment Identification 170 5 5 0 0
- 2. Orientation 533 0 0 0 0
- 3. Levelness 337 16 15 1 0
- 4. Configuration 6,027 359 339 18 1*
- 5. Grout 254 8 8 0 0
- 6. Anchor Bolt Location 1,389 12 3 9 0
- 7. Anchor Bolt Hardware, Engagement, and Tightness 5,968 402 402 0 0
- 8. Rotating Equipment Alignment 221 79 46 . 33 0
- 9. Temporary Coatings and Preservatives 134 0 0 0 0
- 10. Rust (SS only) 47 1 1 0 0
- 11. Concrete Expansion Anchors 362 9 # # #
TOTALS 15,442 891 819 61 1 (approximately 15,400)
- This construction deficiency accounts for 2 deviations.
- Evaluation of Concrete Expansion Anchors deviations and an analyses for adverse trends is included in ISAP V11.b.4.
Esvision: 1 Pego 1 of 25 l
RESULTS REPORT N_- =
ISAP VII.c (Cont'd)
Appendix 9 Small-Bore Piping Configuration i
1.0 REVIEW FP.0 GRAM IMPLEMENTATION 1.1 Const'r,uction Work Category Description The construction work category of small-bore piping configuration is comprised of all safety-related small-bore piping (2 inch diameter and smaller) and includes all in-line components; instrumentation control and indicating devices; branch connections; and vent, drain, and instrumentation piping up to and including the first root isolation valve.
1.2 Population Size and Sample Selection For this construction work category, a population of g approximately 930 items was identified as QC-accepted as of g ,) the date of this reinspection effort. The piping and components on each safety-related QC-accepted piping isometric drawing were designated as an item in this population. Of the total of 103 accessible items reinspected, 77 first sample items were randomly selected from the population to ensure at least 60 reinspection of each attribute were performed. An additional 20 second sample items were randomly selected to ensure that at least 60 r<afe-shutdown hardware items were reinspected.
1.3 Attributes Selected Sample items were reinspected for the following attributes:
Attribute 1 - Piping and component identification Attribute 2 - Piping and component sequence and location Attribute 3 - Orientation of piping and components Attribute 4 - Piping clearances A documentation review was not performed as all attributes of piping configuration could be reinspected.
Q
.m _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
R2 vision: 1 Pega 2 of 25 rm
() RESULTS REPORT ISAP VII.c (Cont'd)
Appendix 9 (Cont'd) 2.0 DISCUSSION OF RESULTS 2.1 Summary of Results For reinspection, 83 Deviation Reports were issued describing 164 deviations. Approximately 3,750 inspection points were encountered in performing the reinspection. See Table 9-1 for results of the reinspection.
In the four attributes of this population, one special case l and one unclassified trend were identified as follows:
- 1. Attribute . - Orientation of piping and components:
Special case for valve flow direccion. See Section 2.2.3 for discussion of this attribute.
s'~m
- ) 2. Attribute 4 - Piping clearances: Unclassified trend
\' for piping clearances. See Section 2.2.4 for discussion of this attribute.
2.2 Analysis of Reinspection Results This section provides, by attribute, a discussion of the reported deviations, an analysis of the effect of the deviations on the function of the piping, and an analysis for the presence of trends. Piping configuration is important in order to maintain the structural integrity and flow capability of the piping system, and to maintain system performance.
2.2.1 Attribute 1 - Piping and Component Identification Piping and component identification includes reinspection to verify compliance with the piping isometric drawing for nominal pipe diameter; the presence of branch connections, screwed joint sealant and elbows; and identification of valves, orifices, flow elements, strainers, and moment restraints.
In all 371 inspection points, the piping and component identifications were verified to be as required by the piping isometric drawing. No deviations were reported.
O L)
Revision: 1 Page 3 of 25 k RESULTS REPORT ISAP VII.c i (Cont'd) i Appendix 9 (Cont'd) 2.0 DISCUSSION OF RESULTS (Cont'd) 2.2.2 Attribute 2 - Piping and Component Sequence and Location Piping and component sequence and location includes reinspection to verify cocpliance with the piping as-built isometric drawing for pipe centerline elevation, pipe location, pipe length, mechanical joint in-line location, valve centerline elevation, flow element and orifice in-line location, and component sequence. No deviations were reported for the reinspection for mechanical joint in-line location, flow element and orifice in-line location, and component sequence.
From 2,394 inspection points to verify piping and lf component sequence and location, 51 deviations were reported concerning 29 of the 103 sample items. None of the deviations was evaluated to be a construction deficiency.
The deviations involved conditions where the pipe centerline elevation, pipe location, pipe lengths, and valve centerline elevation were not in accordance with the piping isometric drawing and the specified tolerance.
The specified tolerance for all of the abpve dimensions is within 2 inches. Of the 51 deviations, 17 exceeded the tolerance by less than 1 inch; 14 exceeded the tolerance between 1 and 2 inches; 11 exceeded the tolerance by greater than 2 inches up to a maximum deviation of approximately 4 inches; and there were nine deviations that exceeded the tolerance by more than 4 inches up to a maximum of 1 foot 3 inches.
An evaluation of each of the deviations was performed to determine the change in the pipe stress and pipe support loads due to any change in pipe length or weight shift caused by the deviations. All pipe length or weight shift changes had a negligible effect on the G pipe stress and pipe support loads. The nine deviations that exceeded the tolerance by more than 4
- inches are described below.
- _ - _ - _ ~ . _ . _ _ _ . _ _ _ _ _ _ _ . _ _ _ _ _ _ _ _ _ . _ _ _ _ _ _ _ _ _ _ _ _ . , _ _ _ _ _ _ _ _ _ _ _
i i
Rsvision: 1 Pego 4 of 25
-[ RESULTS REPORT ISAP VII.c (Cont'd)
Appendix 9 l (Cont'd) 2.0 DISCUSSION OF RESULTS (Cont'd)
Two deviations consisted of couplings that were located incorrectly on straight runs of pipe. These deviations exceeded the allowable tolerance by 8 and 10 inches.
The deviations had a negligible effect on the piping ;
i stresses and no effect on the pipe support design capacity. 'These deviations were determined to be insignificant. Two of the deviations involved two vertical runs of pipe separated by a 45-degree offset where one pipe length was 1 foot 4-1/2 inches less than that required by the isometric drawing and the remaining pipe length was 1 foot 5 inches greater than the linear dimension on the isometric. Because the overall pipe length changed by only 1/2 inch, resulting
-s, in a negligible effect on the total weight of the pipe g
and vertical span length, the two deviations were determined to be insignificant. One deviation consisted of a piping location that was 9 inches beyond the allowable tolerance. This was evaluated to be a drawing error in the location shown on the isometric with respect to the reactor centerline. As the configuration remained unchanged there was no effect on pipe stresses and pipe support design capacity. The deviation was determined to be insignificant. One deviation consisted of a piping length between the tangent point of an elbow and the centerline of a branch cor.nection that was 7 inches shorter that the dimension specified including the allowable tolerance.
Because the acceleration of this piping section did not change, there is no effect on existing pipe stress or pipe support loads. This deviation was determined to be insignificant.
The last three deviations were determined to be a result of drafting errors on the piping isometric drawings. One deviation was reported for a pipe locating dimension to the tangent point of the first of two 45-degree elbows located in series that was approximately one foot beyond the installed position.
The isometric also specified a one foot dimension between the tangent point of the first and second
/ 45-degree elbows. It was determined that the locating dimension that was applied to the first 45-degree l
t c __-
Rsvision: 1 Pags 5 of 25 r
RESULTS REPORT ISAP VII.c I (Cont'd)
Appendix 9 (Cont'd) 2.0 DISCUSSION OF RESULTS (Cont'd) elbow should have actually been shown to the second elbow. This was confirmed by the locating dimensions on the adjoining isometric drawing. Based on the above review, the deviation was determined to be insignificant.
The second deviation involved a vertical length of pipe that was 8-3/4 inches longer than the linear dimension specified on the piping isometric. A review of the related piping elevations on the isometric indicated that the measured length was actually correct and that the linear dimension on the piping isometric was in error. In addition, a review of the existing stress A analysis revealed that the measured length was used as V input in the analysis with acceptable results. Based on the above review the deviation was determined to be insignificant. A memo documenting this specific discrepancy between the isometric and the analysis war, forwarded to the Design Adequacy Program.
The last deviation was a pipe length that measured 12-1/2 inches longer than the linear dimension specified on the isometric. Further review determined (
that the discrepancy was due to a drawing error, which j has been corrected. Based on the above, the deviation 1 was determined to be insignificant.
A separate memo documenting the generic problem of dimensional discrepancies on piping isometric drawings, as illustrated by all three of the deviations described )
above, was forwarded to the Design Adequacy Program.
In addition, a review was made to determine what effect each of the 51 deviations would have on system performance and flow capability. The resultant changes in system pressure drops were negligible because (1) all pipe length changes resulting from these deviations were negligible in comparison to the overall pipe c length in each system, (2) no nominal pipe diameter i deviations were reported, (3) no additional fittings or components were found to be installed, and (4) there were no significant changes in elevation (slope). l Therefore, these deviations had a negligible effect on i system performance and flow capability.
Rsvision: 1 Paga 6 of 25 RESULTS REPORT ISAP VII.c (Cont'd) "
Appendix 9
-(Cont'd) 2.0 DISCUSSION OF RESULTS (Cont'd)
Based on the above evaluation, the 51 deviations were determined to be insignificant.
The out-of-tolerance dimensions'in the range found during the reinspection have a negligible effect on the existing pipe stress, pipe support loads and system performance. In addition, due to the similarities in the analytical methods used for ASME III Code Class 1, 2, and 3 piping and the insignificant impact resulting from the range of the out-of-tolerance dimensions found, the effect would not be greater for any class of ASME III piping. Therefore, if similar type deviations were to occur in the uninspected portion of the population, it is expected that they would also have an
.h insignificant effect on the functional capability of
~D the piping configuration. Therefore, no adverse trend was identified.
I The as-built piping isonetric drawings, which are the piping drawings of record, were used as the basis for dimensional reinspection. Although the deviations were all evaluated to be insignificant, the large number of deviations that was found is more than would be expected following a comprehensive as-built program.
As a result, a recommendation for improvement is provided in Section 4.2.
2.2.3 Attribute 3 - Orientation of Piping and Components Orientation of piping and components includes J reinspection to verify compliance with the piping isometric drawing for valve stem angle, valve flow direction, orifice flange tap orientation, flow element flow direction, and strainer flow direction. No deviations were reported for the reinspection for valve stem angle, orifice flange tap orientation, flow element flow direction, and strainer flow direction.
O i
Rovision: 1 Pags 7 of 25 RESULTS REPORT ISAP VII.c (Cont'd)
Appendix 9 (Cont'd) 2.0 DISCUSSION 0F RESULTS (Cont'd)
Out of 132 inspection points to verify orientation of piping and components..one deviation was reported for valve flow direction. The deviation was evaluated not to be a construction deficiency.
The deviation involved a 3/8-inch globe valve with a stainless steel metal diaphragm that was installed in the flow direction opposite to the flow arrow noted on the valve. The valve is manually operated, and is located in a gaseous waste sample return line. Unused gaseous cample is returned from the Grab Sample Module via this valve to the Waste Gas Compressor. The flow requirements are low, and sufficient flow is permitted
-'s ' even though the valve is not installed in the correct flow direction. In addition, since the valve is rated for a maximum design pressure of 2500 psig at 650 F, and the actual design conditions are only 150 psig at 180 F valve reversal will not continuously expose the valve diaphragm and stem packing to high pressures that would increase the likelihpod of stem leakage.
Therefore, this deviation was determined to be insignificant.
The piping isometric drawing is the document used by construction to determine the flow direction in which the piping and associated components are to be installed. The specific piping isometric.used for installation of this valve did not indicate a flow direction. Therefore, this deviation was attributed to insufficient information on the piping isometric drawing.
Although only one deviation was reported, if this deviation type were to occur in the uninspected portion of the population, it could affect the operation of a piping system, but the pressure integrity would not be violated. While it is possible that other valves in the uninspe,cted portion of the population may be disoriented, it is unlikely that such a situation would
{s result in an undetected construction deficiency.
Therefore, no adverse trend is identified. However, because proper orientation is important to system operation, a special case was identified for valve flow direction.
Rsvision: 1 Page 8 of 25 i
RESULTS REPORT
(_ )
ISAP VII.c (Cont'd)
Appendix 9 (Cont'd) 2.0 DISCUSSION OF RESULTS (Cont'd)
A root cause and generic implications analysis is provided in Section 3.1.
2,2.4 Attribute 4 - Piping Clearances In approximately 834 inspection points to verify piping clearances, 112 deviations affecting 40 of the 103 sample items were reported where the clearances between pipe and other pipe, and pipe and other hardware, i.e.,
sleeves, supports, cable tray and conduit, etc. were less than that required by the piping erection specification. Verification of piping clearance is performed in order to ensure that the mechanical
,s interaction of piping and hardware under design loading
( conditions is acceptable. The criteria for piping
) clearances varied depending on the temperature of the piping and the hardware with which clearance was to be maintained.
A summary of the reported deviations for piping clearances in contained in Table 9-2. All but 17 deviations occurred where at least one item was insulated. Of the 112 deviations, 111 were determined to be insignificant and one was determined to be notable.
The evaluation of the deviations consisted of determining (1) the heat transfer effect on the subject pipe and hardware as a result of the reduction in clearance and (2) whether the combined movements under design conditions were less than the available clearance, thereby avoiding hard contact.
For all 112 deviations the heat transfer effect on the subject pipe and hardware as a result of the reduction in clearance was reviewed. The clearance deviations were localized in nature. This being the case, the heat transfer effects due to a reduction in clearance were determined to be negligible.
I v
l
Revision: 1 Page 9 of 25 RESULTS REPORT ISAP VII.c (Cont'd)
Appendix 9 (Cont'd) 2.0 DISCUSSION OF RESULTS (Cont'd)
In addition, an evaluation of each of the devictions was performed to determine if the combined movements we.re less than the available clearance, including a'Llowances for compressibility of the insulation. If the combined movements were less enan the available clearance, the deviation was determined to be insignificant. If the combined movements resulted in hard contact between pipes or between a pipe and other hardware, further evaluations were performed to determine the effects on the pipe stress and pipe supports and the effects on the impacted hardware.
Of the 112 deviations evaluated, 95 involved piping or other hardware where at least one item was insulated.
O
\ The results of the evaluations for the combined movement indicated that 41 of the 95 deviations would result in the insulation touching the adjacent hardware. In all of these 41 cases, the insulation was sufficiently compressible to accommodate the movement with no messurable increrse in the pipe stress or pipe support loads, and 5dth no effect on the hardware.
Therefore, 95 of the devia:icno were determined to be insignificant.
The 17 ressinirg clearance deviations involved uninsulated pipe and uninsulated hardware. In all but four cases, the ericting clearances were pufficient for the pipe movements involved, and no contact between pipe and hardware would occur under design conditions.
Two of the four cases involved contact between an uninsulated pipe and tube. An evaluation of the condition determined that based on the flexibility of the tubing in comparison to the stiffness of the piping and the high frequency of vibration of the equipment upon which this piping and tubing are mounted, the deviation will have an insignificant ef f ect on the stress levels of both the tubing and pipe. In addition, the effects due to abrasion were determined to be insignificant.
Revision: 1 Pago 10 of 25 O RESULTS REPORT V
ISAP VII.c (Cont'd)
Appendix 9 (Cont'd) 2.0 DISCUSSION OF RESULTS (Cont'd)
The third of the four cases involved contact between a pipe and support. The pipe movements were determined to be small with the only non-negligible movement resulting from thermal growth. Since the seismic movements were negligible, there will be no impact between ths pipe and support. The small thermal movements will not be restrained as the pipe will slide across the support. Given the small interaction forces where sliding is predicted between the pipe and support and given the similar materials involved, these thermally induced movements will not cause galling or increase the pipe stress. In addition, the pipe support loads will not be significantly affected.
1 The last of the four cases was a condition where 1/2-inch clearance existed between a 1-1/2 inch dianeter safety injection pipe and a steel support channel containing two 3/8-inch diameter instrument air tubes. One of the instrument air lines supported by the channel is required to operate a post-accident sample system containment isolation valve and one is a valve stem leakoff line that is not asfety related.
Because the safety injection line has an operating temperature greater than 200 F, the clearance required between the 1-1/2 inch diameter pipe and support channel is 1 inch. The evaluation of this devistion concluded that the maximum combined movement, which was virtually all contributed by the channel, was 1.95 inches. Therefore, hard contact would occur under design conditions. In addition, since the calcalated combined movement of 1.95 inch exceeded the required clearance of 1 inch, the original clearance criterion does not appear to have been sufficient. As a result, a concern regarding the adequacy of the clearance criterion was transmitted to the Design Adequacy Program for further evaluation.
The results of the analysis indicated that the impact q load causes the existing pipe stress in the 1-1/2 inch i pipe to increase by a factor of approximately 2.9. A comparison of the new pipe stress levels to the ASME Code-allowable stress indicates that the stress levels
-- --_______.__________._J
Rsvision: 1 Pags 11 of 25 I
l^
RESULTS REPORT ISAP VII.c (Cont'd)
~
Appendix 9 (Cont'd) 2.0 DISCUSSION OF RESULTS (Cont'd) are a maximum of 53 percent of the allowable values.
The stress level in the post accident sampling instrument air line was also calculated and found to increase by approximately 45 percent. The new stress level is approximately 73 percent of the Code-allowable stress. Therefore, the new stresses in both the pipe and tube are still within Code-allowable stresses and are acceptable.
An additional evaluation was performed to determine if damage would occur to the pipe wall of the safety injection line or the instrument air line due to localized stresses resulting from contact. The g evaluation for the pipe showed the stress levels to be within the material allowables at the design t
temperature. Thus, no local pipe wall deformation would occur and the minimum wall thickness requirements and subsequent pressure boundary integrity would be maintained. For the tubing, since 1) the tubing is firmly attached to the channel with steel clamps with q no gap between the cir.mp and tubing, and 2) the channel )
takes all the localized impact, the tubing wall is also )
unaffected by impact, j An enslysis was also performed to determine the effect f that the iepact load due to contact would have on the ]
pipe eupports and channel supports on either side of j the point of contact. The recults of the analysis j deternined tSat under design conditions the load due to !
contact causes the existing pipe support loads to l increase substantially (i.e., by a factor of 2.1 and 5) for each of the adjacent supports. However, the l original support loads were small and the new pipe !
support loads were determined not to exceed the capacity of the existing pipe support designs. For the channel supports, the increase in load due to contact was determined to ha insignificant and would not affect the capability of the supports.
( Based on the increase in pipe and tube stresses and pipe support loads, this one deviation was determined to be notable.
l
Rsvision: 1 Page 12 of 25 fy RESULTS REPORT ISAP VII.c (Cont'd)
Appendix 9 (Cont'd) 1 1
2.0 DISCUSSION OF RESULTS (Cont'd)
In summary, none of the clearance deviations affected the functional capability of the related piping, pipe supports or adjacent hardware. All of the deviations are insignificant, with the exception of the one notable deviation where the pipe stress and pipe support design margin is reduced, but the existing stresses are still within ASME Code-allowable stresses.
Based on the large number of pipe clearance deviations-found, a trend was identified. Although only one notable deviation was reported, it is dif ficult to determine whether more significant clearance deviations are likely to exist with pipes located in the rg uninspected portion of the population. This determination would require that all piping with stress
() levels close to ASME Code-allowables be identified and evaluated for the effects of the worst-case clearance.
Based on the scope of this effort and the potential for additional clearance deviations due to the largs quantity of pipe and hardvare installed onsite, an unclassified trend for p,iping_ clearances was identified. A root cause and generic implications analysis is previded in Section 3.2.
2.3 Analysis of Docunantacion Paview Results 1
A documentation review was not performed as all attributes of j piping configuration could be reinspected. j 3.0 ROOT CAUSE AND GENERIC IMPLICATIONS j i
This section provides a root cause and generic implications j analysis for the following: {
- 1) A special case for valve flow direction.
- 2) An unclassified trend for piping clearances.
( 3.1 Reversed Globe Valve Root Cause The following is a summary of the relevant information regarding the valve flow direction deviation:
l
l l
Rsvision: 1 Pegs 13 of 25-1 O
k~ / RESULTS REPORT ISAP VII.c I (Cont'd)' l Appendix 9 f
4 (Cont'd) 3.0 ROOT CAUSE AND GENERIC IMPLICATIONS (Cont'd)
The design document used by Construction for installing pipe is the piping isometric drawing. For this particular isometric, no flow direction arrow was indicated. Thus, there was insufficient information for Construction to be sure that the valve installation was correct. The primary root cause for the deviation was a less-than-adequate engineering drawing, which did not contain a flow direction arrow.
Although no flow direction was indicated on the isometric, Construction should have verified the flow direction prior to fit-up. The Construction Procedure CP-CPM-6.9E provides specific instructions to orient globe valves in the correct flow direction. The arrow on the valve body is likely to have
/ raised a question about orientation. Failure to resolve the (s,)T question correctly may be attributable to training, workmanship, supervision, or even an incorrect verification, verbally obtained. A specific cause could not be determined.
Quality Control (qC) is responsible, in accordance with the ASNE Pipe Fabric.ation anu Installation Inspection Procedure.
QI-QAP-11.1-26, for the inspection or the piping installed by Construction to ensure that the requirements of the piping isomet:ric drawing are uct. Ther:e arc specific instructions in this procedure to vv.rify correct valve orientation, but there is ne required sign-off. While othar factors may have ;
contributed to the failure to detect mizorientation during inspection, a secondary root cause for the deviation is ,
identified to be a less-than-adequate inspection checklist, {
sihich did not include a sign-of f for orientation.
Generic Implications The generic implications of the identified primary and secondary root causes for this attribute are that other piping parts or components that are required to be installed in a specific orientation may not have been installed in the correct flow direction due to the piping isometric drawing lacking a flow direction arrow or a less-than-adequate QC i
l gg checklist. No other conditions of this type have been found as a result of reinspection conducted in this population;
' .f
) however, a generic implication does exist for all other piping parts and components within this attribute.
- _ . - - - _ - _ _ _ - _ _ . _ _ , _ - _ - - _ _ - . _ _ , , , _ _ , - - - . , . _ . . _ - . - _ . - , , _ ~ " ' ' ~ ~ ~' ~ ' ' ' ' ' ' ' ' ' ' ' ~ ~ ~ ' - ' ~ ~ ~
- 0 Revision: 1 Pegs 14 of 25
) RESULTS REPORT ISAP VII.c (Cont'd)
Appendix 9 (Cont'd) 3.0 ROOT CAUSE AND GENERIC IMPLICATIONS (Cont'd)
In addition, a review was performed to determine whether the primary and. secondary root causes of this deviation would have eny generic implications with any of the other attributes in this population. The primary root cause of less-than-adequate engineering drawings (a flow arrow missing from the isometric) was assessed for each attribute. No drawing errors for Attribute 1, piping and component identification, were detected and therefore no generic implication is concluded to exist.
For Attribute 2, piping and component sequence and location, no omissions were detected, however, numerous dimensional errors on piping isometrics have been noted during the
- reinspection process. The trend evaluation for Attribute 2,
,' s however, concluded that these types of errors would not lead
~ to an undetected construction deficiency and, therefore, no generic implication exists. (A recommendation for improvement is made, however, in Section 4.2 addressing those errors for both the Large-Bore and Small-Bore Construction Work Categories). For Attribute 4, piping clearances, no significant requirement is typically specified on isometrico and therefore, no generic implication exists.
For the seconduy root cause of a less- than-adequate checklist, QC inspection of Attributes 1 and 2 is adequately ,
addressed by QI-QAP-11.1-26 and by the Mechanical Component Installation verification, U-5 certification progedure, CP-QAP-12.1. Inspection for Attribute 4 is addressed in Section 3.2. Thus no additional generic implications are identified. l The Large-Bore Piping Configuration Construction Work Category (Appendix 8) treats orientation, Attribute 3, in the same manner as the Small-Bore Piping Construction Work Category (i.e., flow arrows on isometric drawings and flow orientation not on the inspection checklist). A generic implication for the Large-Bore Construction Work Category is, therefore, identified. In fact, a special case for component flos i orientation is identified in Appendix 8 for a component with !
clear indication of flow direction on the isometric.
.()
Rsvision: 1 Pcgs 15 of 25 r"'N k_,) RESULTS REPORT l ISAP VII.c (Cont'd)
Appendix 9 (Cont'd) 3.0 ROOT CAUSE AND GINERIC IMPLICATIONS (Cont'd)
Summary The valve flow direction deviation, which was determined to be a special case, is a result of the causes identified below:
- 1) A less-than-adequate engineering drawing, which did not contain a flow direction arrow, was determined to be I
the primary root cause.
- 2) A less-than-adequate QC inspection checklist, without a .!
requirement for verification of correct valve orientation was determined to be a secondary root cause.
p)
(, The generic implications of this deviation are that certain other components that are required to be oriented in a specific flow direction may also be incorrectly installed in the uninspected portion of this Ccustruction Work Category or in the Large-bore Piping Configuration Construction Work Category. The potentially a.ffected components are those that ace subject to similar inspection (1.2., flow orienu> tion not on the checklist), whether the design isometric drawina provides a flow direction or not.
Recomur.ndations for corrective action are provided in Section 4.1.
' i 3.I Piping Clearances j l
Root Cause i
The following provides a summary of a review of erection I specifications and Construction and QC procedures to determine the causes for the piping clearances deviations. J l
Engineering Clearance criteria were not consistently or completely presented in the Piping Erection Specification 2323-MS-100;.
i 7- and other component erection specifications during the time of l
\- major construction activities. Inconsistent clearance l criteria existed between the Piping Erection Specification and the Electrical Erection Specification, 2323-ES-100, up until
Revision: 1 Pags 16 of 25 RESITLTS REPORT 1
ISAP VII.c (Cont'd)
Appendix 9 q (Cont'd) {
3.0 ROOT CAUSE AND GENERIC IMPLICATIONS (Cont'd)
April 1982. For example, the Electrical Erection Specification did not provide clearance criteria between ASME Code Class 1 and 2 piping, apart from the veld areas, and conduit, conduit supports, cable trays, and cable tray supports. The Piping Erection Specification required pipe with operating temperatures greater than 200 F, predominantly-Class 1 and 2, to maintain a one inch clearance with conduit, conduit supports, cable trays, and cable tray supports.
Miscellaneous other inconsistencies or lack of criteria similar to this example also existed. Subsequent to April 1982, the clearance criteria presented in both the electrical and mechanical specifications evolved to become similar.
eg The Piping Erection Specification provides installation s
requirements for both piping and pipe supports. Clearance criteria between pipe and pipe supports are only specified in the piping installation section (Section 4) of this specification, anc' no criteria are contained in the pipe support section (Section 7) for maintaining clearancee with piping.
In r,ummary, th6 inconsistencies and incompletacess of specifications are likely te have contributed to come of the reported clearance violatiens and are a primary root cause. A memo was sent to the Design Adequacy Program regarding the inconsistencies in the clearance crittria provided in various 1 specifications. ,
The Insulaticn Specification (for non-reflective it.sulatien, 2323-MS-70), that was in effect through March 1983 provides minimum insulation thicknesses based on maximum material conductivity values. Ihis specification allowed the thicknesses to exceed those contained in the specification without Engineering approval if a material with a higher conductivity was supplied. Thus, insulation with thicknesses greater than the minimum specified in the insulation specification was provided and installed without Engineering knowledge. The insulation specification was revised to reflect the actual thicknesses of purchased and installed 9 insulation and to require increases in insulation thickness to I
' (V be brought to the attention of Engineering. However, as the original minimum thicknesses were used by construction and QC to estimate clearances, clearance deviations resulted.
. s
Revision: ~1 1 Paga 17 of 25 )
1 RESULTS REPORT ISAP VII.c (Cont'd)
Appendix 9 (Cont'd) j 3.0 ROOT CAUSE AND GENERIC IMPLICATIONS (Cont'd)
Engineering did not require a backfit of the new specification l provisions. Therefore, the potential for clearance deviations exists in the installed piping that is insulated due to a less-than-adequate Engineering interface with construction and QC. This also is a primary root cause.
Construction Piping was installed in accordance with construction procedure-CP-CPM-6.9E. This procedure provides instructions to maintain clearances, including insulation, when installing piping.
This procedure only specifies criteria for pipe-to-pipe clearances. Clearances and clearance criteria with other f- g plant hardware are not included in this construction procedure although these criteria do exist in the piping erection
(") specification, 2323-MS-100. This less-than-adequate construction procedure was also identified as a primary root cause. In addition, although the constriction procedure states that clearances include insulation, there is no reference to insulation thicknesses to be used to establish clearances. This was identified as a contributing cause.
In addition, :nnstrtaction procedure CP-CPM-6.9E requiree that when the location specified en che drawing results in cleTrances less then these spectflad, Engineering is to be contacted for resciutien. 5esed on the number of clearance deviations reported, for at least sone cases construction either may not have complied with this requirement, or was I
unable to estimate correctly tne insulation thickness to be installed, as described above.
In November 1984, construction procedure CP-CPM-6.2, Summary of Clearance Criteria, was issued to summarize clearance criteria for all hardware types. This procedure also provides insulation thicknesses, which are the same as those contained in the insulation specification, and states that notching of insulation may be necessary to maintain required clearances.
Because much of the piping and insulation (for Unit 1) was installed prior to the issuance of this procedure, and because no provision was made to backfit the n'ew requirements, the d potential for clearance deviations exists.
I
.._.________________._____________A
Rsvision: 1 Pegs 18 of 25
,/~% .
(,) RESULTS REPORT ISAP VII.c (Cont'd)
Appendix 9 (Cont'd) i 3.0 ROOT CAUSE AND GENERIC IMPLICATIONS (Cont'd)
Once insulation was installed, notching of insulation was permitted by Engineering to maintain clearances. Based on the number of clearance violations, it is not evident that insulation was notched or notched sufficiently to maintain clearances. There is evidence of notching in the plant; however, it appears to be to avoid interferences and not to maintain clearance. The Piping Erection Specification, 2323-MS-100, specifies that notching of insulation to meet clearance criteria is acceptable. There is no evidence that Construction was cognizant of the criteria because. prior to the issuance of CP-CPM-6.2 in November 1984, none o" the construction procedures contained any instructions r elative to notching or reference to other documents containing ;hese e' criteria. Therefore, less-than-adequate instruction, prior to
( ,)s November 1984 relative to notching of insulation was a primary root Cause. i Quality Control Final QC inspections to verify piping clearances were perfctmed in accordance with CP-QAP-12.1, inspection Critoria and Lecumentation Requirements Prior to System N-$
Certification. Of ten insulction was not installed at the time of the walkdown, and anticipated clearances after installation of insulation were determined by subtracting tho insulation thickness from the measured clearance. Insulation thicknesses were obtained from the pipe line list and insuistion specification.
This method of checking insulation by QC is considered adequate for a preliminary check. However, as insulation thicknesses varied from those contained in the insulation specification, the lack of a walkdown to verify clearances after insulation is installed is considered a secondary root cause for clearance deviations.
Generic Implications gs A review was made to determine whether the primary and l secondary root causes identified for this attribute would have
\')
t any implication for any of the other attributes in this I
l Construction Work Category. The primary root cause of a I less-than-adequate construction procedure does have l
t
Revision: 1
'Paga 19 of 25 l.
i RESULTS REPORT ISAP VII.c l (Cont'd)
Appendix 9 )
(Cont'd) {
l I
3.0 ROOT CAUSE AND GENERIC IMPLICATIONS (Cont'd) implications for Attribute 3, orientation of piping and components; however, as discussed in this Appendix, separate findings have already been identified for this attribute.
There are no generic implications of this primary root cause for Attribute 1, piping and component identification, and Attribute 2, piping and component sequence and location, since the piping drawings (instead of the construction procedure) do identify the installation requirements for these attributes; and as further confirmation, no significant deviations have been determined to exist.
For the primary root causes of a less-than-adequate Engineering interface with construction and a m less-than-adequate erection specification, generic implications were considered for the remaining attributes of this population. However, since no significant deviations have been reported for these attributes and, nore importantly, since the Engineering interface with construction and the erection specifications relating to these attributes are adequate, generic implications do not exist.
For the secendary root cause of a less-thsu-adequate QC .
procudure, generic duplications were cor.siccred for the l remaining attributes in this Construction Work Catt. gory. For Attributes 1 and 2. since adequate QC procedures are available pertaining to these attributes, the generic implication does not exist. For Attribute 3, orientation of piping and components, a generic implication does exist with regard to the adequacy of the QC procedure to address all components which are required to be verified to be installed in the correct flow direction. A recommendation is provided for the Large-bore Piping Configuration Construction Work Category (Appendix 8) and this appendix for this generic implication.
With respect to other Construction Work Categories, any electrical, mechanical or structural component where the clearance from piping may have an effect on the functional capability of the component would be implicated by the causes previously described for the piping clearance deviations.
Based on this review, affected components include conduit (Appendix 1), cable tray (Appendix 2), and large-bore piping configuration (Appendix 8), as a minimum.
Ravision: 1 Paga 20 of 25 p)
( RESULTS REPORT wJ ISAP VII.c (Cont'd) )
Appendix 9 (Cont'd) 1 3.0 ROOT CAUSE AND GENERIC IMPLICATIONS (Cont'd)
Summary The piping clearance deviations, which were identified to be an unclassified trend, are a result of multiple causes as identified below:
- a. Inconsistent and incomplete clearance criteria between various component erection specifications was a primary root cause of clearance deviations.
- b. Specification of minimum insulation thicknesses with generic approval to exceed those thicknesses without engineering knowledge (prior to March 1983) if a material with a higher conductivity is supplied.
7.-
- provided a less-than-adequate Engineering interface
\- with conrecruction and QC. This was a primary root cause of clearance deviations.
- c. The lack of critoria in the construction piping procedure for citarances with items other than piping and lack of instructions relative r.o notching of insulation were primary root causes, of clearance deviations for piping installed prior to November 1984.
L The lack of a QC requiren nt to verify final piping cle.arances subsequent to the installation of insulation was a secondary root cause of clearance deviations.
I
- e. Thra lack of criteria in the construction procedure for insulation thicknesses to be used when estimating clearances was a contributing cause of clearance deviations.
The generic implication from these causes is that large-bore piping configuration, conduit, and cable tray Construction Work Categories, as a minimum, are likely to also have piping clearance deviations.
Recommendations for corrective action are provided in Section 4.1.
tO) v
i Ravision: 1 Pags 21 of 25 l
!, ,) RESULTS REPORT ISAP VII.c (Cont'd) ,
s Appendix 9 (Cont'd) i 4.0 RECOMMENDATIONS 4.1 Recommendations for corrective action:
Valve Flow Direction
- 1) It is recommended that TU Electric review all safety-related piping isometric drawings (large and small-bore) to identify any that do not contain the required flow direction arrows and correct as necessary.
- 2) It is recommended that TU Electric inspect the safety-related components (e.g., valves, strainers, etc.) whose flow orientation is essential to maintain system operability for safety-related functions in both
()
ex the large and small bore piping to verify proper orientation, and correct as necessary.
k
- 3) It is recommended that the QC inspection checklist be l revised to address orientation of flow ransitive components cad require a QC inspector signoff. In addition, those components which are requized te be f installed in a specific orientation (flow directiors) should be re-emphasized in trcining for both construction and QC personnel.
l Piging Clearances s
- 1) It is recommended that TU Electric verify'the existing pipe clearance criteria in applicable specifications and procedures to assure that consistent and complete criteria among disciplines are provided and that all types of plant hardware requiring clearance from piping are addressed. .
- 2) It is recommended that the installer's training program reemphasize to the piping craft that construction procedure CP-CPM-6.9E requires that Engineering be notified for resolution when the location specified on the piping drawing results in clearances less than I those specified.
U}
- 3) It is recommended that the following specifications be reviewed / revised:
Revision: 1 Pagn. 22 of 25 RESULTS REPORT i
ISAP VII.c (Cont'd)
Appendix 9 (Cont'd) 4.0 RECOMMENDATIONS (Cont'd) a) The Insulation Specification, 2323-MS-30, should include instructions relative to notching of insulation and should specify the ,
maximum insulation thicknesses for the types of insulation used'at CPSES.
b) Section 7 of the Piping Erection Specification, 2323-MS-100, should include pipe-to-pipe support clearance criteria or should reference the piping section of the specification for piping clearance criterion.
- 4) It is recommended that TU Electric revise the f.
appropriate QC procedure to require verification of
-; final piping clearance subsequent to insulation
\- installation.
- 5) It is recommended that TU Electric perform a walkdown of piping to verify that sufficient clearance exists to accommodate tha total out-of-phase movements under l desi6n conditions between pipes and other pipes, and between pipes and other plant hardware and corrset as necessary. This walkdort should be done with pipe and other hardware set in their final location and with ,,
insulation installed.
4.2 Recommendation for Improvement ,
Piping Isometric Drawing It is recommended that a general review of piping isometrics be performed for large and small-bore piping to verify the accuracy and consistency of the linear and locating dimensions specified, as a minimum, and correct as necessary. For example, verify that the sum total of sub-dimensions is equal to the overall dimension specified.
In addition, it is recommended that a review of the general program for as-built piping verification be performed to
/~'\ determine the scope of and the cause for the inaccuracy or !
( s/ inconsistency of piping location and dimensional measurements, and to ensure appropriate corrective action.
s
Revision: 'l Page. 23 of 25 RESULTS REPORT ISAP VII.c .1 (Cont'd) f l
Appendix 9 !
(Cont'd)' l 5.0 . CONCLUSIONS Based on the findings of the reinspection and satisfactory implementation of the above recommendations, there will be reasonable assurance that the hardware in this construction work category is adequately installed in accordance with the design.
O O
Revision: 1 Page 24 of 25 l
.g RESULTS REPORT ISAP VII.c (Cont'd)
Appendix 9
-(Cont'd) i Table 9-1 Summary of Reinspection Results -l Small-Bore Piping Configuration i
Deviation Classification j l
Number of Inspection Number of Insigni- Construction-Attribute Points Deviations' ficant N,o table - Deficiency
- 1. Piping and Component Identification 371 0 0 0 0
- 2. Piping and Component Sequence & Location 2,394~ 51 51 0 0 j
-3. Orientation of> Piping and Corponents 132 1 1 0 0
- 4. Piping and Component ~
Clearances 834 112 111 1 0 Totals 3,731 164 163 1 0 (Approximately 3,730) l O
o -_ -
rb' T s
- a t u l 0 s os n e N n al 5 L I h a%
l' 2 T u0 e q5 5 1 f r sE o u s o
- t e rt n5 a L o -
o2 r i e s p i a m vg e ea T %
RP r0 g e5 n t i an 1 0 t ea 1 a d rh r e CT e t p a O l 1 u
e s n s p n al n i I h a%
o P T u0 q5 8 i 5 t sE 4 a s o i ert v L o e
D e
T c R n F %
O 9 2 a r0 P
E
- c. ) xd
)
9
- r a
0 0
e5 t 1 1 I d R I i e 2 an S
V 'tn dnn't l e l C n ea rh T P o e o b a d GT L AC pC a g h e U S( p( T n T t 2
S I A i # a E p r l R i e t u n P t os al a N n h a%
f e I T u0 o r q5 1
G sE 2 y s o r e e rt a r L o m u m t u a S r e
p %
m r0 e e5 T t an 4 6 g ea n d rh i e GT t t a a r l n e u al 0 p s h a%
O n T u0 I q5 e sE 1 9 p s o 1 i ert P L o E
V# R S F. E# E A P E H I S LS TS E E FPN FS N F0N GC DC O - O O - O O - O NN EN OI OI OI
) I AFRA RTT RTT RTT TROI R E - A E - A E - A I'b SA I E%QE XL UA EL BEI MPV UI E BEI MPV UI E aEI FPV UI E ECARC NPD NPD NPD
R; vision: 1 Pegs 1 of 14
\
[d RESULTS REPORT ISAP VII.c ;
(Cont'd)
Appendix 10 Piping Bend Fabrication 1.0 REVIEW PROGRAM IMPLEMENTATION 1.1 Construction Work Category Description i The construction work category of piping bend fabrication is comprised of all safety-related, field-fabricated piping bends in small-bore piping (2 inches and smaller).
1.2 Population Size and Sample Selection For this construction work category, a population of approximately 270 piping isometries with pipe bends was identified as QC-accepted as of the date of this reinspection effort. Reinspection and documentation reviews were performed for a total of 94 items with accessible pipe bends. Of these, 60 first sample items were. randomly selected from the
/' 'T population to ensure that at least 60 reinspection or
'\_sl documentation reviews of each attribute were performed. An additional 19 second sample items were randomly selected to ensure that at least 60 safe-ohutdown hardware items were inspected or reviewed. Each sample item, i.e. , piping isometric, may have one or more bends. A total of 220 accessible bends on the 94 sample item isometrics were reinspected and associated documentation was reviewed.
1.3 Attributes Selected Sample iter.s were reinspected for the following attributes:
Attribute 1 - Radius Attribute 2 - Ovality Attribute 3 - Buckling and flattening Sample item documentation was reviewed by the Construction Evaluation Engineering Group (CEEG) for the following attributes:
Attribute 4 - Pipe bending machine qualification
/'~% All of the pipe bending machine qualification forms were reviewed because the forms applied
(_) to all pipe bends.
Attribute 5 - Minimum wall thickness verification
_ ._________________a
Revision: 1 Pega 2 of 14 RESULTS REPORT N --:
ISAP VII.c (Cont'd)
Appendix 10 (Cont'd) 2.0 DISCUSSION OF RESULTS 2.1 Summary of Results For reinspection, one Deviation Report was issued describing two deviations. There were 625 inspection points encountered in performing the reinspection. See Table 10-l'for results of the reinspection.
For documentation review, one Deviation Report was issued describing 32 deviations. There were 233 review points encountered in performing the documentation reviews. See Table 10-2 for results of the documentation reviews.
In the three reinspection attributes of this population, no deviations were evaluated to be construction deficiencies and no adverse trends were identified.
- (,j,,
' In the two documentation review attributes of this population, deviations reported for Attribute 5, minimum wall thickness verification, were determined to represent an unclassified trend for wall thickness verification. See Section 2.3 for discussion.
2.2 . Analysis of Reinspection Results This section provides, by attribute, a discussion of the l
reported daviations, an analysis of the effect of the deviations on the functional capability of the piping bends, and an analysis for the presence of trends. The function of a pipe bend is to maintain the structural integritt and flow capability of the piping in which each bend is located.
2.2.1 Attribute 1 - Radius 1
In all but two of the 201 inspection points, the bend radius was verified to meet or exceed the minimum bend radius as shown on the piping isometric.
Two deviations were reported where two SD (five times pipe diameter) bends were installed instead of the 6D (six times pipe diameter) bends required by the piping 1 isometric. An evaluation determi'ed n that the increase l (N in flow resistance was negligible. The overall !
increase in flexibility for each of the pipe sections {
l
Revision: 1 Page 3 of 14 RESULTS REPORT ISAP VII.c (Cont'd)
Appendix 10 (Cont'd) 2.0 DISCUSSION OF RESULTS (Cont'd) was less than two percent, and there was no change in the ASME Code-specified stress intensification factor. ]
In each case, this results in a negligible effect on the stress level for the piping system in which the bend is located. Both bends were evaluated and determined to meet the design requirements for pressure and allowable stresses for all design loadings. The deviations were determined to be insignificant. If this type of deviation were to occur in the uninspected portion of the population, a similar negligible effect on the stress level can be expected. Hence, no adverse trend was identified.
2.2.2 Attribute 2 - Ovality G In all 211 inspection points, the ovality of the bend was verified not to exceed a maximum of eight percent as required by the piping specification. No deviations )
were reported.
2.2.3 Attribute 3 - Puckling and Flattening In all 213 int,pection points, the absence of buckling, flattening, or any other defctmation of the pipe as verified in accordance with the piping specification.
No deviations were reported.
2.3 Analysis of Documentation Review Results This section provides, by attribute, a discussion of the documentation review deviations and an analysis of the effect of the deviations on the ability of the evidence collected to provide reasonable assurance that the piping bends are properly fabricated. The documentation review was performed directly by the CEEG engineer during the initial investigation performed to support the preparation of the quality instruction for documentation review. The results of this engineering review are presented below.
2.3.1 Attribute 4 - Pipe Bending Machine Qualification An engineering assessment was made of the qualification of the pipe bending machines and shoes by reviewing all of the qualification forms available in the records
-Ravision: 1 Paga 4 of 14
- ('\
RESULTS REPORT ISAP VII.c (Cont'd)
' Appendix 10 (Cont'd) 2.0 DISCUSSION OF RFSULTS (Cont'd) vault, resulting in 83 review points. Each qualification form addressed one combination of pipe bending machine, desired bend radius, nominal pipe diameter, nominal pipe vs11 thickness, and pipe material. The review found that the pipe banding machines and shoes (collectively referred to as the pipe bender) were all accepted as qualified in accordance with the construction procedure. Therefore, no deviations were identified.
Further engineering evaluation of the details of the pipe bender-qualification test results determined that the percentage of wall-thinning after bending ranged from_0.0 to 12.3 percent of the original wall
[ thickness.- It was also noted that one of the pipe bender qualification forms was accepted despite the fact that the recorded wall thickness after bending was less than the manufacturer's minimum wall thickness.
In addition, the wall thinning effect of the two qualified bending machines differs significantly for each combination of pipe diameter, schedule (wall' thickness), and material type. The Engineering specification MS-43B, " Shop Fabrication of Piping in the Field", allows the use of pipe for bending with a wall thickness at manufacturer's minimum wall thickness prior to bending. Therefore, the potential exists for a pipe wall to be thinned below manufacturer's minimum wall thickness after bending, if the pipe
- wall thickness prior to bending does not have enough margin to compensate for the wall thinning effect. Thus, it is necessary to know the actual wall thickness of the ;
pipe prior to bending and the pipe bender used for j bending to determine if the manufacturer's minimum wall #
1 thickness requirement after bending was always satisfied. See Attribute 5 below for further discussion.
2.3.2 Attribute 5 - Minimum Wall Thickness Verification The Manufacturing Record Sheet (MRS) for each piping i bend contained in the reinspection samples was reviewed. This resulted in 150 review points to determine whether the pipe wall thickness prior to
_____________________j
Revision: 1 Page 5 of 14 j RESULTS REPORT g
ISAP VII c (Cont'd)
Appendix 10 (Cont'd) 2.0 DISCUSSION OF RESULTS (Cont'd) bending had been recorded. A chronological review of revisions to the construction procedure CP-CPM 6.9E indicated that the construction personnel were always required to measure and verify a specified minimum pipe wall thickness prior to bending. The additional requirement to record the pipe wall thickness measurements prior to bending on the MRS was added to the construction procedare (CP-CPM 6.9E, Revision 6) for pipe bends fabricated after August 12, 1982.
Of the 150 Manufacturing Record Sheets reviewed, 46 were prepared af ter August 12, 1982. Only 14 of these 46 Manufacturing Record Sheets have the pipe wall thickness prior to bending recorded as required by the construction procedure. This resulted in 32 deviations 9 being reported for this attribute.
For the 104 Manufacturing Record Sheets prepared prior to August 13, 1982, six have the pipe vall thickness recorded prior to bending. The other 98 Manufacturing Record Sheets did not have notations indicating whether minimum vall thickness had been measured and verified by the construction personnel. No deviations were identified as the construction procedure in effect at the time did not require the recording of this information.
The 32 deviations for not recording the will thickness measurements prior to bending after August 12, 1982 were determined to be notable because no other documentation is available to permit determination of resultant pipe wall thickness after bending. These 32 deviations and the unavailability of pipe wall thickness data for pipe bends fabricated prior to August 13, 1982 result in a situation where wall thickness of pipe bends cannot be verified by available documentation. As noted in Section 2.3.1, Attribute 4, the potential exists for a pipe wall to be thinned below manufacturer's minimum wall thickness in the bending proc (ss for some combinations of pipe bending 9 machine and pipe wall thickness.
Revision: 1 Page 6 of 14
() RESULTS REPORT ISAP VII.c (Cont'd)
Appendix 10 (Cont'd) 2.0 DISCUSSION OF RESULTS (Cont'd)
To provide additional information regarding pipe bend wall thicknesses, the Project performed ultrasonic testing of the pipe bends included in the reinspection samples. A 10 percent overview inspection was also conducted by the QA/QC Review Team. Review of this overview inspection data indicated that five of the 23 l
pipe bends in the 10 percent overview inspection sample had wall thicknesses that were up to 8 mil below manufacturers minimum wall thickness. This review also-resulted in uncertainty about the wall thicknesses of pipe bends in the remainder of the reinspection sample items. This same uncertainty also exists for the uninspected portion of the population. These conditions were not evaluated for safety significance.
O Rather, an unclassified trend was declared for Attribute 5.
The lack of a record of wall thickness verification and of the bender used for bending, combined with the identification of pipe bends with less than manufacturer's minimum wall thickness, were declared an unclassified trend for wall thickness verification.
See Section 3.0 for root cause and generic implications analysis. A QA/QC Program Deviation Report (QA/QC PDR-073) documenting the failure of craft and QC to check wall thickness after pipe bending was also written as a result of the review for root cause and generic implications. This PDR was evaluated to be a QA/QC program deficiency. The corrective action recommendation made in Section 4.0 also covers the PDR.
3.0 ROOT CAUSE AND GENERIC IMPLICATIONS A root cause and generic implications analysis was performed for the unclassified trend of wall thickness verification for pipe bends.
Root Cause The pipe bending section of engineering specification MS-43B, " Shop Fabrication of Piping in the Field", requires that pipe meet a specified minimum wall thickness prior to bending. In addition,
Rsvision: 1 Pego 7 of 14
( ); RESULTS REPORT ISAP VII.c (Cont'd)
Appendix 10 (Cont'd) 3.0 ROOT CAUSE AND GENERIC IMPLICATIONS (Cont'd) the general standards section requires that the manufacturer's minimum wall thickness be met after fabrication. Site engineering evaluation and resolution are also required by the specification when the manufacturer's minimum wall thickness is not met due to fabrication processes. These requirements are sufficient to ensure that all production bends meet the design requirements. Because only the pipe wall thickness requirements prior to bending are specified in the pipe bending section, it is possible that the requirements for fabrication of bends given in other sections of the specification were overlooked when the related construction and inspection procedures were prepared. The requirements specified in the pipe bending section do not alone provide assurance that, considering the wall thinning effect of the bending operation, the manufacturer's minimum wall thickness will be met after bending C' even if the requirements prior to bending are satisfied. Thus, the
\m specification was determined to be less-than-adequate and was a contributing cause for the unclassified trend.
The construction procedure 35-1195-PCP-8, " Forming and Cold Bending l of 'Q' and Non 'Q', Stainless Steel and Carbon Steel Piping",
l Revision 0, dated November 17, 1977, required that the pipe bending l machine and corresponding shoe be qualified utilizing one bend sample for each nominal pipe diameter, pipe schedule (vall thickness), and material type combination. The construction procedure also required that the reduction in wall thickness be 6 percent or less for the qualification bend sample to be acceptable.
This demonstrated maximum wall thinning effect, together with a separate construction procedure requirement that the &inimum pipe I wall thickness prior to bending be 1.06 times the manufacturer's minimum wall, would have ensured that all the production bends met the specification requirement that the pipe wall thickness after bending be greater than or equal to the manufacturer's minimum wall. However, Interim Change Notice Number 2, dated January 5, 1978, revised the construction procedure to require only that the wall thickness after bending of a qualification specimen meet the manufacturer's minimum wall thickness while retaining the same requirement prior to production bending of 1.06 times manufacturer's minimum wall thickness. This revised requirement did not take into consideration the pipe wall thickness of the specimen used in the qualification which could have been ;
s significantly thicker than the required minimum. However, for bend !
fabrication, a pipe section with a wall thickness of only 6 percent !
above manufacturer's minimum wall thickness would not be enough to I
Rovision: 1 Pags 8 of 14 RESULTS REPORT
(}
ISAP VII.c (Cont'd)
Appendix 10 (Cont'd) 3.0 ROOT CAUSE AND GENERIC IMPLICATIONS (Cont'd) compensate for a higher than 6 percent thinning effect and still meet the manufacturer's minimum wall thickness after bending. The construction procedure CP-CPM-6.9, " Pipe Fabrication and Installation", Revision 0, subsequently superseded construction procedure 35-1195-PCP-8, Revision 0, and incorporated similar requirements as stated in the interim change notice.
The construction procedure allows the use of pipe for bending with a wall thickness of 1.06 times manufacturer's minimum wall. Based on a review of the results of all 83 pipe bending qualification tests performed and accepted by construction per the requirement of construction procedures CP-CPM 6.9E for the two bending machines, the wall thinning effect due to bending ranged from 0.0 to 12.3 percent of the original wall thickness. The wall thinning
(]
(m;/ .
percentages between the two bending machines for a bend with the same pipe diameter, schedule, and material type combination also differed significantly. Hence, pipe section with wall thickness less than 1.123 times manufacturer's minimum wall thickness prior to bending, with the potential of up to 12.3 percent wall thinning, would have a wall thickness less than the manufacturer's minimum wall after bending. The construction procedure permitted a qualification process that could result in failure to meet the specification requirements in the fabrication of production bends.
Therefore, the primary root cause of inadequate demonstration that the specification requirements for manufacturer's minimum wall thickness were met by all installed pipe bends was the less-than-adequate construction procedure for pipe bender qualification and for the selection of pipe for the fabrication of pipe bends.
A chronological review of the revisions to construction procedures CP-CPM-6.9E and 35-1195-PCP-8 determined that the requirement to measure and verify minimum pipe wall thickness prior to bending has always been in the procedure. However, the requirement to record the measured wall thickness on the MRS form was added to the !
construction procedure CP-CPM 6.9E in August 1982. The added requirement was clearly stated but did not specify where on the MRS form to record the wall thickness. Furthermore, the documentation i review determined that the form was not revised to provide a space j r- for the data and that construction personnel infrequently recorded !
- f-s the pipe wall thickness. Construction personnel interviewed were not f amiliar with the recording requirement. Therefore, a l contributing cause of inadequate demonstration that the (
Revision: 1 Pags 9 of 14
_RESULTS REPORT
[
ISAP VII.c (Cont'd)
Appendix 10 (Cont'd) 3.0 ROOT CAUSE AND GENERIC IMPLICATIONS (Cont'd) specification requirements for manufacturer's minimum wall thickness were met by all installed pipe bends was the less-than-adequate construction supervision and/or craft training l
for implementing procedure changes. Another contributing cause was the less-than-adequate construction procedure which failed to specify in sufficient detail where on the MRS form the pipe wall thickness data should have been recorded.
A comparison of the QC inspection procedure, QI-QAP-11.1-26 " Pipe Fabrication and Equipment Installation Inspections", and the construction procedure determined that there was almost a one-for-one correlation between the construction procedure pipe bend fabrication process steps and the QC inspection procedure.
However, the QC inspection procedure did not require the QC inspector to verify that construction personnel measured and 5 recorded the pipe wall thickness prior to bending, as' required by the construction procedure. Therefore, the secondary root cause of inadequate demonstration that the specification requirements for manuf acturer's minimum wall thickness were met by all installed pipe' bends, was a less-than-adequate QC procedure that did not require QC to verify that construction personnel measured and recorded pipe wall thickness prior to bending on the MRS form.
Generic Implications The primary root cause of inadequate demonstration that all installed pipe bends meet the specification requirements of manufacturer's minimum wall thickness was the less-than-adequate construction procedure for pipe bender qualification and for the selection of pipe for fabrication of pipe bends. One of the possible generic implications is that the construction procedure CP-CPM-6.9E is less-than-adequate in the qualification of other machines or tools requiring pre-qualification. Further review of the construction procedure indicated no other machines or tools require special pre-qualification. Hence, the implication does not exist.
Another possible generic implication is that the construction procedure CP-CPM-6.9E is less-than-adequate for ensuring that pipe fabrication of any sort meets the specification requirements for O minimum wall thickness. Counter-boring for veld-end preparation of butt welds is t e only other fabrication work done on pipe that is also covered by construction procedure CP-CPM-6.9E.
All revisions
l Rsvision: 1 Pcg2 10 of 14
[]
'O RESULTS REPORT ISAP VII.c (Cont'd)
Appendix 10 (Cont'd) 3.0 ROOT CAUSE AND GENERIC IMPLICATIONS (Cont'd) of CP-CPM-6.9E required the construction personnel to check the pipe end wall thickness after counter-boring to ensure that manufacturer's minimum wall thickness had been met. No requirement for recording wall thickness by the construction personnel was specified in the procedure. However, any pipe not conforming to manufacturer's minimum wall thickness was required to be dispositioned with a nonconformance report. The quality control e inspection procedure QI-QAP-11.1-26, Revision 17, requires the QC inspector to measure and record the pipe wall thickness on the veld.
data card after counter-boring. Recording wall thickness, however, was not a requirement until Revision 15 of the inspection procedure which was issued on April 18, 1984. To ensure manufacturer's minimum wall thickness was maintained at the treld ends, the Project measured (via ultrasonic testing) and recorded the wall thickness for all butt welds completed prior to this date. This effort was V(7 documented by Nonconformance Report No. M-857 Therefore, between the requirements of the construction and inspection procedures and the referenced nonconformance report, there is indication that all manufacturer's minimum wall thickness deviations have been formally evaluated for all pipe fabrications other than pipe bending.
Hence, the generic implication does not exist.
The less-than-adequate construction procedure that did not specify in sufficient detail the appropriate use of the Manufacturing Record Sheet form by the construction personnel for the recording of pipe wall thickness data prior to bending was a contributing cause. A possible generic implication is that the procedure does not provide sufficient detail regarding the use of the ,
Manufacturing Record Sheet for the recording of other pertinent !
data for the fabrication of pipe bends. A review of both the procedure and the Manufacturing Record Sheet forms revealed that no other data related to the bending process are required to be recorded on the Manufacturing Record Sheet. Hence, the implication 1
does not exist.
The less-than-adequate QC inspection procedure was a secondary root ,
cause involving the failure to verify that construction personnel l had measured and recorded pipe wall thickness prior te bending. A ,
possible generic implication is that the QC inspection procedure (
was less-than-adequate to verify that construction personnel have q y~~
followed other construction procedure requirements for bend j fabrication. A comparison of the QC inspection procedure and the i
_ _ _ . _ _ _ )
Rsvision: 1 Pcge 11 of 14
/~' RESULTS REPORT C) ISAP VII.c (Cont'd) j Appendix 10 (Cont'd) 3.0 ROOT CAUSE AND GENERIC IMPLICATIONS (Cont'd) construction procedure revealed that there was a one-for-one correlation for each attribute of the pipe bend fabrication process except in the verification and recording of pipe wall thickness prior to bending. Hence, the generic implication does not exist.
Based on the review of the primary and secondary root causes and contributing cause discussed previously, a possible generic implication is that the construction procedure and the QC procedure were also less-than-adequate for the other attributes (i.e.,
radius, ovality, and buckling and flattening) of the pipe bend fabrication. A review of the construction procedure and the QC procedure determined that all the requirements for the other attributes were clearly stated. In addition, the low frequency of deviations (only two deviations were reported for the 625
(T inspection points) for these three attributes tends to confirm the (s,) conclusion that the generic implication does not exist.
Summary The primary root cause for the unclassified trend of wall thickness verification was the less-than-adequate construction procedure for the qualification of pipe benders and for the selection of pipe used for the fabrication of pipe bends.
The secondary root cause was the less-than-adequate QC inspection procecure that did not require QC to verify construction personnel have measured and recorded the pipe wall thickness prior*
to bending on the MRS form.
The contributing causes include: 1) the less-than-adequate Engineering Specification that did not provide clear direction to ensure the manufacturer's minimum wall thickness would be met after bending; 2) the less-than-adequate construction procedure that did not provide sufficient detail regarding the use of the MRS form for recording the pipe wall thickness data prior to bending; and 3) the less-than-adequtte construction supervision and/or craft training for implementing procedure changes.
There are no generic implications that have not been adequately f'N addressed by existing procedures. Refer to Section 4.0 for
( ,) recommended corrective action. I 1
i 1
- - - - - - - - - -- _ A
Revision: 1 Page 12 of 14 RESULTS REPORT ll 1 ISAP VII.c (Cont'd)
Appendix 10 (Cont'd) 4.0 RECOMNINDATIONS It is recommended that the following corrective actions be performed to provide reasonable assurance that this construction work category is adequately installed.
4.1 Verify the Adequacy of Installed Pipe Bends Verify by ultrasonic testing (UT) that the wall thickness of the installed site fabricated pipe bends on 2 inch and smaller piping meets the manufacturer's minimum wall thickness requirement.
4.2 Provide Assurance of Adequate Future Construction Work Revise Construction Procedure CP-CPM-6.9E to add the e requirement that manufacturer's minimum wall thickness be met after bending and to provide an appropriate space on the MRS form for the recording of pipe wall thickness prior to bending.
Retrain construction personnel regarding the construction procedure :equirement to record the pipe wall thickness prior to bending on the MRS form.
Revise QC Inspection Procedure QI-QAP-11.1-26 to require QC verification, by ultrasonic testing, that the pipe wall thickness after bending meets manufacturer's minimum wall thickness, and that the measured wall thickness be recorded.
Revise engineering specification MS-43B to include, in the pipe bending section, a reference to the applicable requirements for wall thickness after fabrication.
5.0 CONCLUSION
S Based on the findings of the reinspection and documentation reviews and satisfactory implementation of the above recommendations, there will be reasonable assurance that the hardware in this construction work category is adequately constructed in conformance with the design.
Rzvision: 1I Page. 13 of 14 4
RESULTS REPORT ISAP VII.c (Cont'd)
Appendix 10 (Cont'd)
. Table 10-1 Summary of Reinspection Results Piping Bend Fabrication Deviation Classification Number of' Inspection Number of Insigni- Construction Attribute Points _ . Deviations ficant Notable Deficiency i
- 1. Radius 201 2' 2 0 0
- 2. Ovality 211 0 0 0 0-
- 3. Buckling and Flattening 213 0 0 0 0 O
TOTALS 625 2 2 0- 0 1
i O i
____-.__.__.________-__._______________a
Revision: 1 Pcgs 14 of 14 (T RESULTS REPORT
..)
ISAP VII.c (Cont'd)
Appendix 10 (Cont'd) !
Table 10-2 Summary of Documentation Review Results Piping Bend Fabrication l
Deviation Classification i Number of ;
Review Number of Insigni-Attribute Points Deviations ficant Notable I
- 4. Pipe Bending Machine Qualification 83 0 0 0
- 5. Minimum Wall .
Thickness Verification 150 32 0 32 O
LJ TOTALS 233 32 0 32 l
)
i l
1 1
l 1
i
Revision:- 1 Page 1 of.17
() RESULTS REPORT ISAP VII.c (Cont'd)
Appendix 11 Piping System Bolted Joints 1.0 REVIEW PROGRAM IMPLEMENTATION 1.1 Construction Work Category Description The construction work category of piping system bolted joints is comprised of all safety-related pressure-retaining bolted joints in piping systems (large and small-bore piping).
Flanges, bolts, studs, cap screws, nuts, and gaskets are I
included in this construction work category.
1.2 Population Size and Sample Selection For this construction work category, a population of approximately 2,600 items was identified as QC-accepted as of the date of this reinspection effort. Reinspection and I\ documentation reviews were performed for a total of 108
\-- accessible items. Of these, 91 first sample items were randomly selected from the population to ensure that at least 60 reinspection or documentation reviews of each attribute were performed. An additional seven second sample items were randomly selected to ensure that at least 60 safe-shutdown hardware items were reinspected or reviewed.
1.3 Attributes Selected Sample items were reinspected for the following attributes:
Attribute 1 - Bolt, stud, or cap screw diameter Attribute 2 - Nut height Attribute 3 - Thread engagement Attribute 4 - Bolt, stud, cap screw, and nut material, and gasket type Attribute 5 - Flange pressure rating and flange type Attribute 6 - Bolted joint completeness Attribute 7 - Bolt tightness l
Attribute 8 - Gasket coverage over the flange sealing surface
Revision: 1
.Page 2 of 17 RESULTS REPORT ISAP VII.c (Cont'd)
' Appendix 11
['
l' (Cont'd)
I 1.0 REVIEW PROGRAM IMPLEMENTATION (Cont'd)
Sample item documentation was reviewed, when applicable, for the following attributes:
Attribute 9 - Bolt, stud, cap screw, and' nut material ~!
traceability Attribute 10 - Alignment of pipe flanges prior to bolt-up Attribute 11 - QC inspector certification 2.0 DISCUSSION OF RESULTS 2.1 Summary of Results For reinspection, ten Deviation Reports were issued describing 18 deviations. Approximately 7,700 inspection points were encountered in performing these reinspection. See Table 11 l for results of the reinspection.
For documentation review, 11 Deviation Reports were issued describing 11 deviations. Approximately 400 review points were encountered in performing the documentation reviews. See Table 11-2 for results of the documentation reviews.
Documentation review deviations related to inspector certification have not been included in the aforementioned totals. As discussed in Section 2.3.3, it has bien determined that these deviations do not necessarily relate directly to actual inspector qualifications. These deviations were considered in the ISAP I.d.1 evaluation together with supplemental information as necessary to determine actual inspector qualifications. The ISAP I.d.1 conclusions are summarized in Section 2.3.3.
In the eight reinspection attributes of this population, no deviations were evaluated to be a construction deficiency and no adverse trends were identified.
I In the two documentation review attributes of this population T (excluding QC inspector certification), all deviations were determined to be insignificant.
. l
Ravision: 1 Paga 3 of 17 RESULTS REPORT ISAP VII.c (Cont'd)
Appendix 11 (Cont'd) 2.0 DISCUSSION OF RESULTS (Cont'd) 2.2 Analysis of Reinspection Results This section provides, by attribute, a discussion of the reinspection deviations, an analysis of the effect of the deviations on the functional capability of the bolted joint, and an analysis for the presence of trends. The function of a bolted joint is to join adjacent sections of pipe, piping components, and equipment, while maintaining the= structural integrity and flow capability of the piping in which it is located.
2.2.1 Attribute 1 - Bolt, Stud, or Cap Screw Diameter In 846 inspection points, the diameter was verified to A. .be in accordance with the piping isometric drawing, and
- ( ,) no deviations were reported.
2.2.2 Attribute 2 - Nut Height In 1,472 inspection points to check this attribute, one deviation was reported where the nut height was not in accordance with the piping specification. This deviation identified one nut on a four-bolt flange which was approximately 1/32-inch thinner than required by the specification. The difference in nut height and width indicated a standard hex nut was installed in lieu of a heavy hex nut. An evaluation of the reported condition determined that installation of'a standard hex nut would not result in a reduction in the bolt design capacity. Because no reduction in design capacity occurred, this deviation was determined to be insignificant.
Furthermore, a bounding analysis, which encompasses the use of a standard hex nut in lieu of a heavy hex nut, was prepared for the loose bolt deviations described in Section 2.2.7. This analysis determined the effect of one loose bolt en any four-bolt flange in the population. The analysis was based on the conservative assumption that the loose bolt is ineffective, which could also be assumed in analyzing the use of a
()~
'- standard hex nut in lieu of a heavy hex nut. This bounding analysis determined that, should a similar deviation occur in the uninspected portion of the
_ __ ______-____-__A
Rsvision: 1 Pagt 4 of 17
- [')l
\~-
RESULTS REPORT ISAP VII.c (Cont'd)
Appendix 11 (Cont'd) l 2.0 DISCUSSION OF RESULTS (Cont'd) population, it would also be insignificant, as no reduction in design capacity would occur. Therefore, based on the insignificance and low frequency of this deviation type, no adverse trend was identified.
2.2.3 Attribute 3 - Thread Engagement For this attribute, samples were reinspected to verify that a minimum of one thread protruded beyond the nut-face or tapped hole, if accessible, or that thread engagement equivalent to a minimum of one nut existed for bolts or. cap screws installed in blind tapped holes. In 1,608 inspection points for thread engagement, 11 deviations were reported on three
. ,Q flanges where there existed less than one full thread
\/ protrusion beyond the nut face, as required by the piping erection specifications. The requirement to have one full thread protrude beyond the nut face is conservative, and facilitated the original inspection.
Nine deviations were reported on two flanges where the stud extended less than one full thread beyond the face of the nut, and two deviations were reported on one flange where the studs were slightly below (nearly flush with) the nut face. An evaluation of the reported conditions determined that, in all cases, j sufficient thread engagement exists to provide the full i bolt design capacity. Because no reducti8n in the design capacity occurred, these deviations were determined to be insignificant.
An analysis of bolt and nut capacity for bolts 1/2-inch to 1 1/2-inches diameter showed that a minimum of 72 percent of nut thread engagement is required to obtain full bolt design capacity. Because the number of threads varies with nut size, the remaining 28 percent I
of nut threads not engaged is equivalent to approximately 1-3/4 to 2 threads. Based on this analysis, it was concluded that should similar i deviations occur in the uninspected portion of the O population, no reduction from design capacity would o . _ _ _ _ _ _ _ - _ - _ -
Rtvision: 1 Pago. 5 of 17
()
LJ RESULTS REPORT ISAP VII.c (Cont'd)
Appendix 11 (Cont'd) 2.0 DISCUSSION OF RESULTS (Cont'd) result, and these deviations would be' insignificant.
In the worst cases encountered, the studs were almost flush with the face of the nuts, and the functional capability of each joint was maintained. Therefore, based on the insignificance and low frequency of the reported deviations, no adverse trend was identified.
2.2.4 Attribute 4 - Bolt, Stud, Cap Screw, Nut Material, and Gasket Type In all but one of 2,270 inspection points, the bolt,-
stud, cap screw, and nut materials were verified to be in accordance with the piping isometric drawing.
j] One deviation was reported where cap screws were
(_) installed instead of studs as required by the piping isometric drawing. The cap screws were of the same diameter and material as required for the studs. An evaluation determined that the design actually required cap screws to be installed; however, the bill of materials on the piping isometric drawing erroneously specified studs. An investigation later identified that both studs and cap screws were listed in the bill of materials, and the revision of the drawing that was current at the time the flange was installed correctly specified cap screws for this flange. Subsequent to the installation of the flange, however, the drawing was revised, and the marker for this flange was incorrectly relocated next to the entry for studs. As a result of the identification of this drawing error, the isometric drawing has been revised to specify cap screws for this flange.
Therefore, because the use of cap screws was in accordance with the design requirements, this deviation was determined to be insignificant. Since the design requirement is met, no adverse trend was identified.
In 102 inspection points, the gasket type was verified to be in accordance with the piping isometric drawing, r'~ and no deviations were reported.
( )g
Revision: 1 j Page 6 of 17
]
1 RESULTS REPORT q
) ;
LJ ISAP VII.c )
I (Cont'd)
Appendix 11 (Cont'd) 2.0 DISCUSSION OF RESULTS (Cont'd) 2.2.5 Attribute 5 - Flange Pressure Rating and Flange Type In 326 inspection points, the flange pressure rating and the flange type were verified to be in accordance with the piping isometric drawing. No deviations were reported.
2.2.6 Attribute 6 - Bolted Joint Completeness In 107 inspection points to check the quantity of bolts, studs, cap screws, nuts, and gaskets, all the joints were verified to be in accordance with the piping isometric drawing. No deviations were reported.
O 2.2.7 Attribute 7 - Bolt Tightness O The bolt tightness requirements were established to verify the minimum structural integrity of each bolted joint. Where the Construction Operation Traveler required a bolted joint to be tightened to specific torque values, the joint was reinspected by retorquing to verify that the bolt was tight when a minimum of 70 percent of the original specified torque value was applied. For those bolted Joints where no specific torque values had been specified, the bolts were reinspected to verify hand-tightness as a minimum.
In 866 inspection points to check this attribute, five cases were identified where bolt tightness was found to be less than hand-tight.
Five cases were reported where one nut in each bolted joint was not hand-tight. Four of these cases occurred in four-bolt flanges and one occurred in a 12-bolt flange. Each deviation was evaluated to determine if a leak could result from one loose nut and if the additional stress applied to the remaining studs would be such that they could be stressed beyond the ASME III Code allowable. The analysis of the deviations was based on design loading conditions and the assumption Q that the less-than-hand-tight bolts are ineffective.
V 1
I n _ . _ . _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
Rsvision: 1 Page 7 of 17 RESULTS REPORT f}-
U ISAP VII.c (Cont'd)
' Appendix 11 (Cont'd) 2.0 DISCUSSION OF RESULTS (Cont'd)
One loose bolt on the 12-bolt flange would result in an increase in stress on the remaining 11 bolts; however, j this would not result in a loss of structural integrity of the joint based upon the analysis described below for one bolt being loose out of total of four flange bolts. In addition, an analysis was made to determine the effect that one loose bolt would have on the leak-tightness of the bolted joint. This analysis was based on the location of the flange gasket relative to the bolting location. It was determined that if any one bolt was completely removed, the entire gasket would still be contained within the area encompassed by the remaining eleven bolts. Because the gasket is completely restrained by the remaining bolts, it'will s be totally compressed between the flange faces.
( Therefore, due to the close proximity of the bolts to one another, the analysis determined that no significant leakage would result from one loose bolt on a twelve-bolt flange. Based on the ability of this flange to maintain leak tightness and structural integrity of the piping system, this deviation was determined to be insignificant.
The analysis of the four-bolt flange deviations evaluated the effect of load redistribution among the tightened bolts, based on the assumption that the bolted connection is subjected to a bending moment t' 't would cause the flange face to pivot about the bolt opposite the loose bolt, causing the other two bolts to assume the entire load. This analysis determined that, in the worst case encountered, the stress on the remaining bolts increased from 13 to 26 percent of the ASME III Code allowable. For all four cases, the evaluation determined that the piping system integrity and system flow capability would be maintained; howevca, minor leakago might occur. The effect of leakage on the system performance, and on the local plant area, would be insignificant. Based on the increase in stress and frequency of occurrence, these four cases were determined to be notable and were identified as a trend. The following is an evaluation
>O of the potential severity of this trend.
Revision: 1 Paga 8 of 17 -
./s RESULTS REPORT ISAP'VII.c (Cont'd)
Appendix 11 .:
(Cont'd) l 2.0 DISCUSSION OF RESULTS (Cont'd)
For the bolted joints where a loose nut was observed, an engineering review of the installation documentation revealed that all required construction practices were satisfied. Furthermore, review of the Hydro-pneumatic Test Records indicates that the required leakage criteria were also satisfied. This leads to the conclusion that the loose nuts may not have been caused by a construction activity.
It is not uncommon for a joint that was initially tight to loosen slightly, due to a slow relaxatien of the gasket. This may be caused by a change from the ambient to the operating temperature of tha piping.
Therefore, a loose bolt deviation may occur on other joints in the remainder of the population. Asla i result, a bounding analysis was performed on four-bolt flanges to determine if a similar deviation could result in bolt stress greater than that allowed by the ASHE III Code. A bounding analysis was also performed
~
to determine if under a similar situation leakage could occur that might adversely affect the operation of the
~
plant. These analyses considered four-bolt flanges only, as the loosening of one bolt on a four-bolt flange would result in a greater increase in stress on the remaining bolts than if a similar deviation were to occur on a flange with more than four bolts.
Calculations were performed to determine the maximum allowable bending moment for each size, material and ,
schedule of pipe for which the use of a four-bolt I flange would be appropriate. This bending moment was then applied to each type of flange containing only three bolts, and the resulting stress was compared to that allowed by the code for the appropriate bolting material. Based on the results of this analysis, it was determined that for all four-bolt flanges,'the loss of one bolt would not result in a stress on the remaining bolts greater than that allowed by the ASME III Code.
O
Rsvision: 1 Pags 9 of 17
/"'I RESULTS REPORT O ISAP VII.c (Cont'd)
Appendix 11 (Cont'd) 2.0 DISCUSSION OF RESULTS (Cont'd)
The bounding analysis regarding leakage concluded that four-bolt flanged joints in ASME III piping systems with one loose bolt would not leak, provided that they were initially properly tightened.
Should a loose bolt result in leakage during hydrostatic or pr2 operational testing, common practice is to merely retight,a the bolts. Once the piping system is turned over from TU Electric Startup to Operations, the importance of minimizing potential leakage in an operating system becomes more critical.
Thus, TU Electric Operations has implemented procedural requirements to maintain-bolted joint tightness and to minimize leakage. Leakage caused by loose bolts in the uninspected portion of the population would be limited or detected through implementation of the following existing procedures:
'(
- 1) If it becomes necessary to disassemble a piping bolted joint, the joint is reassembled to specific torque values established by TU Electric engineering and verified by QC.
This is a more conservative practice than that used during construction and startup.
This is a requirement of Flange Bolting Procedure MMI-203, Revision 0,
- 2) Periodic walkdowns are performed on radioactive systems located outside the containment for evidence of piping bolted joint leakage. This is called the Radioactive System Leakage Inspection Program (procedure STA-705, Revision 1). This program includes provisions for identifying the af fected bolted joint and tracking and verifying corrective action.
- 3) Systems located inside the containment are monitored for leakage by checking the reactor containment sump water inventory and <
discharge rate. This is monitored every 12 I O hours during various plant opereting modes.
This is a requirement of the Reactor Coolant System Water Inventory Procedure OPT-303A, Revision 1.
R:; vision: 1 Pegs 10 of 17 RESULTS REPORT
()3 ISAP VII.c (Cont'd)
Appendix 11 (Cont'd) 2.0 DISCUSSION OF RESULTS (Cont'd)
- 4) Walkdowns are performed on safety-related support systems located outside the containment for leakage and for abnormal operation of piping and equipment. This walkdown is performed by each auxiliary operator twice per shift. This is a requirement of the Operations Department Logkeeping and Equipment Inspections Procedure OWI-104, Revision 4.
- 5) Periodic pressure testing is performed on safety-related systems in accordance with the inservice inspection requirements of ASME Code,Section XI. This is a requirement of the Inservice Inspection Program (procedure STA-705, Revision 1) .
g]
Q) In addition to the TU Electric Operations procedures described above, several features are incorporated into CPSES to detect leakage before the leak becomes critical. Major systems that can detect unidentified leakage in the plant are the Process and Effluent Radiological Monitoring Systems, Sampling Systems, and ,
the Vents and Drains Systems. Additionally, )
I containment temperature and pressure monitors can detect a leak in a high-energy line in the Containment Building. These features are described in greater detail in the Final Safety Analysis Report (FSAR),
Volumes VI and IX through XI. -
Therefore, should a leak result from a loose bolt occurring in the uninspected portion of the population, I it is expected that it would be detected by the plant l systems identified above, or through the implementation of the TU Electric Operations procedures, and corrected.
Based on the lack of severity of those deviations i identified, and the evidence that the piping system !
leakage criteria were met, it has been determined that a construction deficiency is unlikely to occur. .
Therefore, no adverse trend was identified. i
(\ws'j
Revision: 1 Paga 11 of 17 !
] RESULTS REPORT (V ISAP VII.c
(
(Cont'd) l Appendix 11 i (Cont'd) 2.0 DISCUSSION OF RESULTS (Cont'd) 2.2.8 Attribute 8 - Gasket Coverage Over the Flange Sealing Surface In 103 inspection points to check gaskets, it was verified that the gaskets cover the entire. flange sealing surface, and no deviations were reported.
2.3 Analysis of Documentation Review Results This section provides, by attribute, a discussion of the documentation review deviations and an analysis of the effect of the deviations on the ability of the evidence collected to provide reasonable assurance that the bolted joints are properly assembled.
p 2.3.1 Attribute 9 - Bolt, Stud, Cap Screw, and Nut Material Q Traceability Bolted joint assembly is controlled using a Construction Operational Traveler form, which requires recording of material identification information for l the bolts, studs, or cap screws, and nuts. For those items larger than 1-inch diameter, the information required is the heat number. For those items with diameters of 1-inch and smaller, the information required is the color code marking of the item.
However, in the early stages of construction work, for 1-inch and smaller diameter studs and nuts, heat number and color code recording were required, bat the primary means of material traceability was by color code. In addition, the Construction Operational Traveler form requires recording of the Material Requisition form number, providing a redundant means of material traceability.
In order to verify material traceability, the documentation for each sample item was reviewed for the I following:
- 1. Verification that the Construction*
Operational Traveler contains a rceord of the ig) color code or heat number of the material V installed with a Brown & Root QC signoff.
l 1
- _____-_ - 1
Rcvisien: 1 Paga 12 of 17 RESULTS REPORT f] '
.(/ ISAP VII.c (Cont'd)
' Appendix 11 .
~
(Cont'd) 2.0 DISCUSSION OF RESULTS (Cont'd)
- 2. Verification that the color code or heat number recorded during reinspection matches that recorded on the Construction Operational Traveler.
- 3. Verification that the color code or heat number of the material recorded on the Construction Operational Traveler and the corresponding Certified Material Test Report is for the same material as required by the piping isometric drawing.
In all but 11 of 305 review points to check the bolt, !
stud, or cap screw and nut material, the material was l A verified to be traceabin.
U One deviation was reported where the heat numbers of the 5/8-inch diameter studs and associated nuts on one joint were different from the heat numbers recorded on the Construction Operational Traveler. However, the color code recorded on the Construction Operational Traveler indicated that the material was correct. The heat number on the studs and nuts was also traceable to a Certified Material Test Report, further confirming the use of the correct materials. It is likely that an error was made in recording the heat numbers on the Construction Operational Traveler. Therefore, this deviation was determined to be insignificant.
A second deviation was reported where the heat number recorded on the Construction Operational Traveler corresponded to a 1/2-inch diameter nut, instead of 7/8-inen diameter as required by the piping isometric.
Further investigation into the related documentation, i.e., the Receiving Inspection Report and the Material Requisition form referenced on the Construction Operational Traveler, revealed that the heat number I corresponding to the 7/8-inch diameter nuts was recorded on these documents. It is apparent that an error was made in recording the heat number on the l
R0 vision: 1 Peg 2 13 of 17 j RESULTS REPORT ISAP VII.c (Cont'd)
Appendix 11 (Cont'd) 7.0 DISCUSSION OF RESULTS (Cont'd)
Construction Operational Traveler, and that traceability was maintained by the heat number recorded on the Material Requisition form and the Receiving Inspection Report. Therefore, this deviation was determined to be insignificant.
Nine deviations were found for 1-inch and smaller studs and nuts where the Construction Operational Traveler contained neither the color code nor the heat number..
In each of these cases, the Material Requisition form number was present and provided the necessary record that the appropriate material was used. Therefore, these nine recording omissions had no effect on material traceability and were determined to be
/ ) insignificant.
V A possible cause for these nine deviations is the interpretation of the QC Inspection Procedure, which directs the QC Inspector to verify that the heat number or color code of the bolting hardware is recorded on the Construction Operational Traveler. This may have been interpreted by Brown & Root QC Inspection to mean that the intent of the procedure is satisfied by recording, on the Construction Operational Traveler, the identification number of the Material Requisition form containing the bolting hardware heat number or color code. As a result, the color code and heat number are not always recorded on the Construction Operational Traveler. A recommendation for improvement of this procedure is provided in Section 4.0.
2.3.2 Attribute 10 - Alignment of Pipe Flanges Prior to Bolt-up In 107 review points to check documentation of alignment, it was confirmed that the Construction Operational Traveler included a QC signature indicating proper alignment of flanges prior to bolt-up. No deviations were reported.
O
Ksvision: 1 Paga 14 of 17
[ RESULTS REPORT
. w} .
ISAP VII.c (Cont'd)
Appendix 11 (Cont'd) 2.0 DISCUSSION OF RESULTS (Cont'd) 2.3.3 Attribute 11 - QC Inspector Certification A total of 108 reviews were performed to determine whether the QC inspectors who signed or initialed and dated the Construction Operational Traveler forms were appropriately certified to Part I, Supplement D of the Brown & Root Training Qualification and Certification Manual, which covers visual examination of mechanical installations.
Deviations were identified whenever evidence of certification to the established procedural 7tquirements at the time of inspection could not be found. However, as a result of the ISAP I.d.1
(
i evaluations of inspector certifications, it has been determined that such deviations do not necessarily relate directly to actual inspector qualifications.
Consequently, the identified deviations were not relied upon (though they were considered) in determining inspector capability. Inspector qualifications were assessed for all inspectors (those with and without reported deviations) using the ISAP I.d.1 methodology.
The ISAP I.d.1 evaluation of inspectors whose work affected this population concluded that all inspectors either were certifiable to applicable criteria at the time of inspections, were found to be capable of performing satisfactory inspections (inclnding those l with substantial positive evidence), or were otherwise shown to be of no further concern. Evaluation details are discussed further in the ISAP I.d.1 Results Report.
3.0 ROOT CAUSE AND GENERIC IMPLICATIONS No adverse trends or construction deficiencies were identified; a root cause and generic implications analysis is not required.
- .. J
1 Ravision: 1-Pega 15 of 17 p RESULTS REPORT ISAP VII.c (Cont'd)
Appendix 11 (Cont'd) 4.0 RECOMMENDATIONS Based on the results of.the reinspection and documentation-reviews, no corrective action recommendations are required.- J Furthermore, the effective implementation of the.TU Electric Operations procedures and programs will assure bolt tightness and minimize leakage.
As a result of the nine deviations discussed in Section 2.3.1, where the Construction Operational Traveler did not contain the color code or heat number required for 1-inch and smaller bolting .
hardware, a recommendation for.the improvement of procedure QI-QAP-11.1-27, Revision 18, is made. In order to reduce the potential.for future misinterpretation, it is recommended that the procedure be revised to state clearly that an acceptable alternate is to list the Material Requisition form number on the Construction Operational Traveler in lieu of the color code or heat-number.- It
~b.
V is also recommended that, in the future, Material Requisition forms be maintained in the Permanent Plant Vault.
5.0 CONCLUSION
S I Based on the findings of the reinspection and documentation reviews, and the applicable recommendations in the ISAP~I.d.1 g results report, there is reasonable assurance that the hardware in q this construction work category is adequately installed in' ]
conformance with the design.
I i
)
. 1 O l
Rsvision: 1 Pag 3 16 of 17' J l
N' RESULTS REPORT
()T ISAP VII.c (Cont'd)
Appendix 11 1 (Cont'd) l Table 11-1 Summary of Reinspection Results Piping System Bolted Joints ,
1 l
Deviation Classification Number of
, Inspection Number of Insigni- Construction i L Attribute Points Deviations ficant Notable Deficiency
- 1. Bolt, Stud, or Cap Screw Diameter 846 0 0 0 0
- 2. Nut Height 1,472 1 1 0 0 T
(/
\, . 3. Thread Engagement 1,608 11 11 0 0 l
- 4. Bolt, Stud, Cap Screw, and Nut Material, and Gasket type 2,372 1 1 0 0 l .
I 5. Flange Pressure Rating and Flange Type 326 0 0 0 0
-6. Bolted Joint Completeness 107 0 0 0 0 l
- 7. Bolt Tightness 866 5 1 4 0
I Surface 103 0 0 0 0 TOTALS 7,700 '18 14 4 0 (Approximately 7,700) e il
Revision: 1 Page 17 of 17 i
RESULTS REPORT ISAP VII.c (Cont'd)
Appendix 11 (Cont'd)
Table 11-2 Summary of Documentation Review Results Piping System Bolted Joints Deviation Classification Number of Review Number of Insit;ni-Attribute Points Daviations ficant Notable
- 9. Bolt, Stud, Cap Screw, l and Nut Material
! Traceabil?ty 305 11 11 0
- 10. Alignment of Pipe Flanges Prior to Bolt-up 107 0 0 0 TOTALS I) 412 11 11 0 (Approximately 400)
() QC Inspector Certification review points and deviations are excluded from totals and the deviations are not classified because the results of ISAP I.d.1 indicated that these deviations do not relate directly to actual inspector qualifications.
O
Revision: 1 Page 1 of 29
() RESULTS REPORT ISAP VII.c (Cont'd)
Appendix 12 l Pipe Welds and Material l
1.0 REVIEW PROGRAM IMPLEMENTATION 1.1 Construction Work Category Description The construction work category of pipe welds and material is comprised of site-made safety-related pipe welds. The site-made welds, weld materials, and base materials for the l
welding of piping to other piping, fittings, and components
' are included in this construction work category.
1.2 Population Size and Sample Selection For this construction work category, a population of approximately 59,700 items (pipe welds) was identified as having QC acceptance of the final weld as or the date of this reinspection effort. Reinspection and documentation reviews O were performed for a total of 187 accessible items. Of these, 98 first sample items vere randomly selected from the population to ensure that at least 60 reinspection or documentation reviews of each attribute were performed. An additional four second sample items were randomly selected to ensure that at least 60 safe-shutdown hardware items were reinspected and reviewed. Furthermore, there were 79 more sample items randomly selected to ensure that at least 60 items from each of the following groups were reinspected and reviewed: carbon steel welds, stainless steel welds, gas tungsten acc welding process, and shielded metal arc welding process. ,
1.3 Attributes Selected Sample items were reinspected for the following attributes:
Attribute 1 - Weld identification and location Attribute 2 - Weld configuration Attribute 3 - Weld size Attribute 4 - Radial weld shrinkage O
- -- - - _- _ -- -.______._ ?______ _______ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ ______ _____ _
Revision 1 Pags 2 of 29
') RESULTS REPORT l
\J l ISAP VII.c (Cont'd)
Appendix 12 (Cont'd) ;
I 1.0 REVIEW PROGRAM IMPLEMENTATION (Cont'd)
Attribute 5 - Weld undercut Attribute 6 - Surface of weld Attribute 7 - Weld cracks end lack of fusion Attribute 8 - Weld rust (stainless steel only)
Attribute 9 - Base material defects Attribute 10 - Base material rust (stainless steel only)
Sample item documentation was reviewed for the following attributes:
/3
- l. ,) Attribute 11 - Base material traceability Attribute 12 - Weld material traceability Attribute 13 - Weld procedure application This attribute also encompasses the review of any applicable post-weld heat treatment, which is a non-recreatable activity associated with this construction work category.
Attribute 14 - Weld procedure qualification Attribute 15 - Welder qualification Attribute 16 - QC acceptance at hold points This attribute encompasses the review of a number of non-recreatable activities associated with this construction work category, such as the following:
fitup, preheat, purge, NDE, repairs, cleanliness, end prep, offset (ficup), etc.
Attribute 17 - Pressure test completion
/^
Attribute 18 - QC inspector certification
(
R2 vision: 1 Page 3 of 29 RESULTS REPORT ISAP VII.c (Cont'd)
Appendix 12 (Cont'd) 2.0 DISCUSSION OF RESULTS 2.1 Summary of Results For reinspection, seven Deviation Reports were issued describing seven deviations. Approximately 3,000 inspection points were encountered in performing the reinspection. See Table 12-1 for results of the reinspection.
For documentation review, five Deviation Reports were issued !
describing five deviations. Approximately 2,520 review points '
were encountered in performing the documentation reviews. S e's Table 12-2 for results of the documentation reviews.
Documentation review deviations related to inspector certification have not been included in the aforementioned totals. As discussed in Section 2.3.8, it has been determined that these deviations do not necessarily relate directly to actual inspection qualifications. These deviations were considered in the ISAP I.d.1 evaluation together with supplemental information as necessary to determine actual inspector qualifications. The ISAP I.d.1 conclusions are summarized in Section 2.3.8.
In the 10 reinspection attributes of this population, there were no reinspection deviations that were evaluated to be a construction deficiency, and no adverse trends were identified. One special case was identified in Attribute 4, radial weld shrinkage, regarding radial weld shr_inkage at butt- welded stainless steel piping joints. See Section 2.2.4 for a discussion of this attribute. One unclassified trend was identified in Attribute 9, base material defects, regarding base material reduction at transition type weld joints. See Section 2.2.9 for a discussion of this attribute.
In the seven documentation review attributes of this population (excluding QC inspector certification), all five documentation review deviations were determined to be insignificant.
With the exception of the above special case, the reinspection and documentation review deviations identified no trends as O, being unique within any of the following groups: carbon steel welds, stainless steel welds, gas tungsten arc welding process, or shielded metal arc welding process.
Rsvision: 1 Page 4 of 29
(N RESULTS REPORT ISAP VII.c (Cont'd)-
Appendix 12 (Cont'd) 2.0 DISCUSSION OF RESULTS (Cont'd) 2.2 Analysis of Reinspection Results This section provides, by attribute, a discussion of the reinspection deviations, an analysis of the effect of the deviations on the functional capability of the weld and base material, and an analysis for the presence of trends. The function of the weld and the base material is to maintain the structural integrity and flow capability of the piping in which it is located.
2.2.1 Attribute 1 - Weld Identification and Location In all 186 inspectica points, the weld was verified, by weld identification and/or location, to be in accordance with the isometric drawing. No deviations f- were reported.
(
2.2.2 Attribute 2 - Weld Configuration In all 186 inspection points, the veld configuration (e.g., butt, fillet, or tack) was verified to be in accordance with the site piping fabrication and erection specifications and/or the isometric drawing.
No deviations were reported.
2.2.3 Attribute 3 - Weld Size In all but four of 257 inspection points,'the weld size was verified to be in accordance with the site piping fabrication and erection specifications. The four deviations concerned two undersize socket welds and two underfilled branch connection welds that were not welded flush as required. There were approximately 85 socket welds and 22 branch connection welds included in the sample.
Socket Welds Two deviations were reported where socket-welded joints had undersized fillet welds. One of these deviations
() was a weld with a throat that was undersized by 0.012 inch (a 7 percent reduction in size) for the pipe i
o __
Revision: 1 Page 5 of 29 RESULTS REPORT ISAP VII.c (Cont'd)
Appendix 12 (Cont'd) 2.0 DISCUSSION OF RESULTS (Cont'd) schedule used. The throat of a fillet weld is the distance f rom the root of the joint to the f ace of the weld. This condition occurred on a 3/4-inch diameter Schedule 160 socket weld for lengths of 23 percent and 11 percent of the circumferential. weld length. The second deviation was a 2-inch diameter Schedule 160-fitting-to-fitting socket weld that was undersized in two places. At one location, the weld throat was undersized by no more than 1/64-inch, (a 6 percent reduction in size) for approximately 13 percent of the circumferential weld length. At the other location, the weld throat was undersized by 1/32-inch (a 12 percent reduction in size) for approximately 10 percent of the circumferential weld length.
( Even though the deviations encompassed only a small percentage of each weld circumference, the deviations were conservatively evaluated by assuming that the undersize weld condition existed for 100 percent of the circumferential weld length.
The evaluation of the first deviation determined that under the worst case loading condition, the reduction in weld size resulted in stresses increasing from 44 percent to 50 percent of the ASME Section III Code-allowable stresses. This constitutes a stress increase of approximately 14 percent, with a 50 percent margin from the Code-allowable stresses still available. The evaluation utilized conservative assumptions to permit simplified calculations. More sophisticated. calculations, for which these conservative assumptions would not be required, would indicate a stress increase of less than 10 percent.
Therefore, the deviation was determined to be insignificant.
The evaluation of the second deviation determined that under the worst case loading condition, the reduction in weld size resulted in stresses increasing from 56 percent to 64 percent of the Code allowable stresses.
O This constitutes a stress increase of approximately 14
Revision: 1 Page 6 of 29 RESULTS REPORT ISAP VII.c (Cont'd)
Appendix 12 (Cont'd) 2.0 DISCUSSION OF RESULTS (Cont'd) percent with a 36 percent margin from the Code-allowable stresses still available. The evaluation utilized conservative assumptions to permit simplified calculations. More sophisticated calculations, for which these conservative assumptions would not be required, would indicate a stress increase of less than 10 percent. Therefore, the deviation was determined to be insignificant.
Both of these deviations were near the threshold of ,
detectability. More significant weld undersize conditions were more likely subject to detection by the original quality control inspection; therefore, this l deviation type is limited to the undersize weld condition found. Based on the insignificant nature of these deviations, it is unlikely that a similar 4 deviation, if it were to occur elsewhere in the uninspected portion of the population, would result in a construction deficiency. Therefore, no adverse trend was identified.
Branch Connection Welds The two remaining deviations occurred at two separate weld joints where a small-bore sockolet fitting was welded to a large-bore pipe. Both deviations were reported where a full penetration branch connection weld joining a 3/4-inch diameter sockolet,to a 4-inch diameter pipe failed to meet the requirement to be welded flush with the wall of the sockolet. One weld was intermittently underfilled by 3/64 inch (an 11 percent reduction in wall thickness), and the other weld was underfilled at one location by no mere than 1/64 inch (a 6 percent reduction in vall thickness).
Both deviations occurred at the top (or start) of the sockolet bevel where the greatest amount of wall thickness exists, and had smooth transitions without notches or sharp contours. Evaluations determined that the remaining wall thickness of the sockolet at the underfilled branch connection weld location was at G least twice that of the socket weld joining the sockolet to the small-bore branch pipe. The stress-limiting point of this type of piping
Ravision: 1 Pega 7 of 29 RESULTS REPORT
()
(_/
ISAP VII.c (Cont'd)
Appendix 12 (Cont'd) 2.0 DISCUSSION OF RESULTS (Cont'd) configuration is the socket weld, and there was no reduction of design capacity at the branch connection weld. Therefore, the deviations were determined to be insignificant.
Due to the possibility for a similar condition to occur on other branch welds, a bounding analysis was performed on 3/4-inch diameter sockolet branch welds.
This analysis was performed to determine the maximum underfilled weld depth that would cause the -
stress-limiting point to shift from the socket weld to the underfilled branch connection weld location. This evaluation determined that more than 44 percent reduction in wall thickness at the branch connection e weld must occur to reduce the strength of the sockolet
(,)s to that of the fillet weld of the small-bore pipe. In trying to meet the requirement of filling a branch weld out to the wall of a sockolet, it is highly unlikely that a branch weld could be underfilled enough to !
i result in a wall thickness reduction of up to 44 percent. This conclusion would also be applicable for other size sockolet branch welds typical of this population. j I
Based on the results of the conservative evaluations i performed and the low probability that an underfilled j weld would exceed the limits determined in the bounding l
analysis, it is unlikely that this type of deviation, 1 if it were to occur in the uninspected portion of the l population, would result in a construction deficiency. i Therefore, no adverse trend was identified.
2.2.4 Attribute 4 - Radial Weld Shrinkage i
In all but one of 256 inspection points, the amount of l j' radial weld shrinkage in the heat-affected zone of the j weld joint was verified to be in accordance with the site piping fabrication and erection specifications. I Although both butt welds and socket welds were
()
s reinspected for radial weld shrinkage, this is primarily a concern on stainless steel piping joined by a full penetration butt weld, particularly when j extensive repair or rework nas been performed. ;
Rtvision: 1 Pags 8 of 29
[ ) RESULTS REPORT LJ ISAP VII.c (Cont'd)
Appendix 12 (Cont'd) 2.0 DISCUSSION OF RESULTS (Cont'd)
One deviation was reported where there was radial weld shrinkage 1/8 (0.125) inch deep on a 3-inch diameter schedule 40 stainless steel butt-welded joint that has an allowable radial weld shrinkage of 0.108 inch (one-half the nominal wall thickness). Even though the deviation only encompassed approximately 7 percent of the circumferential wald length, the deviation was conservatively evaluated by assuming that the radial weld shrinkage extended around 100 percent of the ,
circumference of the pipe.
The evaluation of the deviation determined that under worst case loading conditions, the radial weld shrinkage resulted in stresses increasing from 45
[~ ) percent to 50 percent of the ASME Section III
\ _/
~ Code-allowable stresses. This constitutes a stress increase of approximately 11 percent, with a 50 percent margin from the Code-allowable stresses still available. The evaluation utilized very conservative assumptions to permit simplified calculations. More sophisticated calculations, for which these conservative assumptions would not be required, would indicate a stress increase of less than 10 percent.
Therefore, the deviation was determined to be insignificant.
This deviation type was found only once during the reinspection of approximately 90 butt welds, of which 26 were located on schedule 80 or thinner stainless steel pipe that is susceptible to radial weld shrinkage. Subsequent investigation of this deviation identified a concern regarding weld shrinkage inspection requirements. Prior to December 1982, QC I inspection did not perform radial weld shrinkage l inspections; therefore, there could potentially be more !
instances of radial weld shrinkage in the uninspected I portion of the population. It is difficult to determine if such radial weld shrinkage would have an adverse effect on piping in the uninspected portion of the population. This determination would require that f}
(m,/ all Schedule 80 or thinner stainless steel butt-welded pipe with stress levels close to ASME Section III Code-allowable stresses be identified and evaluated i j
I
R vision: 1 Peg 2 9 of 29
,m i i RESULTS REPORT G
ISAP VII.c (Cont'd)
Appendix 12 (Cont'd) 2.0 DISCUSSION OF RESULTS (Cont'd) for the effects of radial weld shrinkage. This concern was of a programmatic nature; however, it was deemed to be desirable to perform a root cause analysis and to develop recommendations for corrective action.
Therefore, a special case for radial weld shrinkage was declared. A root cause and generic implications analysis is provided in Section 3.0.
2.2.5 Attribute 5 - Weld Undercut In all 372 inspection points, the base material at the toe of the weld was verified to be free of undercut that exceeded 1/32 inch or 12-1/2 percent of the nominal wall thickness, whichever depth is less, as
(
') required by the site piping fabrication and erection
's -
/ specifications. No deviations were reported.
2.2.6 Attribute 6 - Surface of Weld In all 558 inspection points, the surface of the weld was verified to be free of irregularities (such as overlap, abrupt ridges, ripples, etc.) that would have interfered with the performance and proper interpretation of radiographic and other required non-destructive examinations, as required by the site piping fabrication and erection specifications. No deviations were reported, ,
2.2.7 Attribute 7 - Weld Cracks and Lack of Fusion In all 558 inspection points, the weld was verified to be free of any visible cracks, lack of fusion, r.nd crater cracks, as required by the construction procedure and the site piping fabrication and erection specifications. No deviations were reported.
2.2.8 Attribute 8 - Weld Rust In all 83 inspection points, the stainless steel weld
/" surface was verified to be free of rust that could be k.,,T) attributed to improper use of materials or tools, as required by the site piping fabrication and erection specifications. No deviations were reported.
Rsvision: 1 Psg2 10 of 29 l
,-s RESULTS REPORT i
\- ') ISAP VII.c (Cont'd)
Appendix 12 l (Cont'd) 2.0 DISCUSSION OF RESULTS (Cont'd) 2.2.9 Attribute 9 - Base Material Defects In all but two of 372 inspection points, the base material on each side of the weld was verified to be free of welding-related defects that cxceeded 12-1/2 percent of the nominal wall thickness, in accordance with the site piping fabrication and erection specifications. An additional conservative requirement to limit a surface defect to a depth of 1/32 inch was imposed by this reinspection program. ,
The two base material defect inspection points involved an arcstrike and a base material reduction. The arcstrike was caused by touching a tungsten electrode to a 2-inch diameter schedule 40 pipe approximately 3
/ inches from the nearest weld. The arcstrike was
\--) reported to be 3/16 inch diameter with a 1/32 inch tungsten inclusion. The allowable defect depth for the 2-inch pipe was established at 0.019 inches. The base material reduction was caused by overgrinding while blending the reinforced branch connection weld of a 1 inch sockolet into the base material of a 4 inch diameter schedule 40 pipe. The overgrinding was reported to have removed up to 1/16 (0.063) inch from j the base materisl of the pipe and the condition was j reported to be around the entire weld. The allowable j defect depth for the 4 inch pipe was established at ]
1/32 inch. Subsequent ultrasonic measurements !
indicated that at the worst location the defect was j approximately 3/64 inch deep. 1 The arcstrike and the grinding cavity required to remove it are considered to be localized defects. For localized defects, the increase in mechanically-induced stresses (thermal, seismic, dead weight) were evaluated by conservatively calculating the change in cross-section modulus caused by the deviati.on. The cross-section modulus of the 2-inch diameter pipe was calculated to change less than 5 percent; therefore, the mechanically-induced stresses were evaluated to
/~N increase less than 10 percent. For pressure-induced l k,) stress, the evaluation determined that there was a l I
l l
Ravision: 1 Page 11 of 29 RESULTS REPORT
(}
ISAP VII.c (Cont'd)
Appendix 12 (Cont'd) 2.0 DISCUSSION OF RESULTS (Cont'd) negligible increase in stress, and that the. remaining wall thickness was in excess of the minimum wall thickness required by the ASME Section III Code for internal pressure design. Therefore, the deviation was determined to be insignificant.
The base material reduction is also considered a localized defect, and for localized defects such as this, the increase in mechanically-induced stresses were evaluated by conservatively calculating the change in cross-section modulus caused by the deviation. The.
cross-section modulus of the 4-inch diameter pipe was calculated to change less than 5 percent; therefore, the mechanically-induced stresses were evaluated to
[_ increase less than 10 percent. For pressure-induced
=
stress, the evaluation determined that the remaining wall thickness was over twice the minimum wall thicknese required by the ASME Section III Code for internal pressure design, but the resulting stress increase was approximately 25 percent. Therefore, the deviation was determined to be notable.
The two deviations are considered different types of base material defects, and a review was performed to determine whether or not these deviations should be trended separately. The following are the differences identified between the deviations: .
- The overgrind is the apparent result of welder, grinder, or helper in the piping department who worked the metal trying to blend the transition weld into the base material. The arcstike can be declared as accidental " damage" that could be associated with any construction discipline performing welding with tungsten electrodes, and not neceae.stly the piping department. .
i
- The time of occurrence of the overgrind can be determined to be after welding was O' complete and prior to the final visual acceptance of the weld by QC inspection. The
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ ._ -_ i
1 Rsvision: 1 l Pega 12 of 29 l l
h RESULTS REPORT
's-l \
ISAP VII.c (Cont'd)
Appendix 12 (Cont'd)
~
2.0 DISCUSSION OF RESULTS (Cont'd) arcstrike time of occurrence cannot be determined since it cannot be associated with any one particular work activity within the area.
- The overgrind condition is an item that would have been examined during the visual weld inspection and is a smooth transition that would quite likely be passed over during subsequent walkdown inspections. The arcstrike is an item that would not be directly inspected during a weld inspection, but would only be identified during one of the many walkdowns that inspect specifically
(}
LJ for arcstrikes.
Based on these differences identified above, it was determined that these two deviations should be trended separately.
Tne arcstrike deviation was the only reported damage encountered while reinspecting 372 pieces of piping and associated connections. The arcstrike encountered was in an inconspicuous place on a Class 3 pipeline and was only 3/16 inch diameter with no signs of cracking. In accordance with present project acceptance criteria (as incorporated into revision 9 of pipe fabrication and I
erection specification 2323-MS-100) an arascrike of such characteristics is acceptable, provided it is not on a Class 1 pipeline. No deviations were found on Class 1 pipelines. In addition, there are a number of walkdowns performed on pipelines, with specific intent to identify and resolvz damage such as arcstrikes.
Walkdowns are performed prior to pressure testing, prior to the installation of insulation, prior to N-5 certification, and prior to Operations accepting the system, with particular attention given to Class 1 pipelines.
~
(Vh
_ - _ ~ _ _ _ _ _ _ _ _ _ _ _ _ _
Rsvision: 1 Pega 13 of 29 A(,,J - RESULTS REPORT ISAP VII.c (Cont'd)
Appendix 12 (Cont'd) 2.0 DISCUSSION OF RESULTS (Cont'd)
Based on the insignificance of the found deviation, the low frequency of occurrence, the recognition that an arcstrike of a more severe magnitude is not likely to go undetected, and the recognition that piping is not typically stressed near Code-allowable levels, it is unlikely that a deviation of this type, if it were to occur in the uninspected portion of the population, could result in a construction deficiency. Therefore, no adverse trend was identified. ,
The base material reduction was identified during the reinspection of 187 weld joints and associated base-materials. In order to determine the impact that a
- deviation of this type and magnitude would have on the g uninspected portion of the population, a bounding analysis was performed. The analysis conservatively assumed all pipe to already be at manufacturer's minimum wall thickness, assumed the deviation to extend around the encire pipe circumference, and assumed that i this deviation type could occur anywhere. The results of the analysis identified a subset of approximately 540 pipelines in which a deviation of this magnitude could violate the ASME III Code minimum wall thickness.
Due to the quantity of pipes identified with potential code wall violations, an unclassified trend for base material reduction was identified. A root cause and generic implications analysis is provided,in Section 3.2.
2.2.10 Attribute 10 - Base Material Rust In all 164 inspection points, the stainless steel base material surface on each side of the weld was verified to be free of rust that could be attributed to improper use of material or tools, as required by the piping erection specification. No deviations were reported.
2.3 Analysis Of Documentation Review Results
\
- This section provides, by attribute, a discussion of the
[bd documentation review deviations and an analysis of the effect of the deviations on the ability of the collected evidence to provide reasonable assurance that the hardware is properly installed.
-______________-_____-____ _ a
Revision: 1 Pags 14 of 29 l
,7
\/
) RESULTS REPORT l
l ISAP VII.c (Cont'd)
Appendix 12 l (Cont'd) 2.0 DISCUSSION OF RESULTS (Cont'd) 2.3.1 Attribute 11 - Base Material Traceability l In all but three of 366 review points, it was verified that traceability of the base material was maintained through applicable documentation, as required by the site piping fabrication and erection specifications.
In addition, if legible heat or code numbers were l available on the pipe, these numbers were verified to be the same as the heat or code numbers in the weld ,
documentation.
One deviation was reported where an installed elbow was marked with the heat code "FBCW"; however, all applicable documentation indicated that the elbow
(}
\_ / should have a heat code "FBCN". Evaluations determined that the vendor certification document lists both heat codes under the same material description, and shows that both heat codes meet the requirements of the material specification and the applicable requirements of the ASME Section III Code; therefore, the deviation was determined to be insignificant.
A second deviation was reported where a material designation on the drawing bill of materials indicated that a stainless steel valve had been installed; however, the valve-type designation on the same bill of materials and all other applicable documentation indicated that the valve was carbon steel. Evaluations determined that the incorrect material designation was recorded on the drawing bill of material, but the correct valve was installed per the weld documentation and site piping f abrication specification; therefore, the deviation was determined to be insignificant.
A third deviation was reported where a containment spray nipple piece number designated on the isometric drawing did not match the piece number recorded in the weld documentation. At the time the nipple was welded r' ' to the header, the nipple piece number recorded in the
( weld documentation was the same as the piece number designated on the isometric drawing; however, the alpple lengths did not match. The drawing was revised to show the as-built nipple length and to assign a new l
l
_--_____----------_________----_-__________J
Revision: 1~
Pegs 15 of 29 RESULTS REPORT ISAP VII.c (Cont'd)
Appendix 12 (Cont'd) 2.0 DISCUSSION OF RESULTS (Cont'd) piece number to the nipple; however, the weld documentation was not modified to reflect this new piece number. Evaluations determined that the material description and heat number recorded in the weld documentation was actually the correct information for the subject nipple. Because the material description and heat number were determined to be correct, the material was traceable to certification documentation; therefore, the incorrect notation of the piece number '
l in the weld documentation was determined to be-insignificant.
2.3.2 Attribute 12 - Weld Material Traceability In all but one of approximately 259 review points, it was verified that traceability of the weld filler materials and the consumable insert materials used in the production of the weld, was maintained through applicable documentation, as required by'the site piping fabrication and erection specifications.
One deviation was reported where the weld filler material heat number 5005-9 was not traceable to any document certifying the material. A TU Electric response to a Technical Information Request indicated that heat number 5005-9 was not received at CPSES. The number appears to be a minor recording error. The heat number 5005-9 should have been 50059-1, as was recorded on the two previous days on the same weld filler material log. The proper heat number, 50059-1, was traceable to documentation certifying.the material; therefore, the deviation was determined to be insignificant.
2.3.3 Attribute 13 - Weld Procedure Application In all 224 review points, the application of the weld procedure used in the production of the weld was verified for the type of joint being welded, as required by the site piping fabrication and erection specifications. No deviations were reported.
i L l- ,
Revision: 1 Page 16 of 29 RESULTS REPORT
(}
ISAP VII.c (Cont'd)
Appendix 12 (Cont'd) 2.0 DISCUSSION OF RESULTS (Cont'd) 2.3.4 Attribute 14 - Weld Procedure Qualification In all 224 review points, the weld procedure was verified to be approved for use by the architect / engineer or the utility prior to its use, as required by the site piping fabrication and erection specifications. No deviations _were reported.
2.3.5 Attribute 15 - Welder Qualification In all 320 review points, the welder was verified to be properly qualified to make the weld at the time the welding was performed, as required by the site piping fabrication and erection specifications. In addition, if legible welder symbols were available on the pipe, Os these symbols were verified with the welder symbols in the weld documentation. No deviations were reported.
2.3.6 Attribute 16 - QC Acceptance at Hold Points-In all but one of 1,008 review points, the proper hold points established by the piping erection specification and construction procedures were verified to be signed, or initialed, and dated as satisfactory by the QC inspector on the weld installation and repair documentation (when applicable), in accordance with the requirements of the site piping fabrication and erection specifications.
One deviation was reported where a 3/4-inch diameter ASME Class 3 socket weld, which required either a magnetic particle or liquid penetrant surface examination, per the construction procedure, did not receive either surface examination. Only a final visual examination was performed. The ASME Section III .
Code does not require a Class 3 socket weld to receive l a magnetic particle or liquid penetrant surface examination; the final visual examination is adequate to assure the quality of the weld surface. Upon fs further investigation, it was found that the
( )- requirement for magnetic particle or liquid penetrant examination of this weld was a typographical error placed on the inspection matrix, and has been removed from the construction procedure. Therefore, the deviation was determined to be insignificant.
Revision: 1 Page 17 of 29 l
lllI RESULTS REPORT ISAP VII.c (Cont'd)
Appendix 12 (Cont'd) 2.0 DISCUSSION OF RESULTS (Cont'd) 2.3.7 Attribute 17 - Pressure Test Completion In all 124 review points, the welds were verified to have been pressure tested, and documentation was signed, or initialed, and dated by the QC inspector, in accordance with the piping erection specification and the construction procedure requirements. No deviations were reported.
2.3.8 Attribute 18 - QC Inspector Certification A total of 1,497 reviews were performed to determine if the QC inspector who signed, or initialed, and dated the inspection results in the weld documentation was 9 appropriately certified to perform QC Level II inspections.
Deviations were identified whenever evidence of certification to the established requirements at the time of inspection could not be fcund. However, as a result of the ISAP I.d.1 evaluations of inspector certifications, it has been determined that such deviations do not necessarily relate directly to actual inspector qualifications. Consequently, the identified deviations were not relied upon (though they were considered) in determining inspector capability.
Inspector qualifications were assessed for all inspectors (those with and without reported deviations) using the ISAP I.d.1 methodology.
The ISAP I.d.1 evaluation of inspectors whose work affected this population concluded that all but one inspector either were certifiable to applicable criteria at the time of inspections, were found to be capable of performing satisfactory inspections (including those with substantial positive evidence),
or were otherwise shown to be of no further concern.
Evaluation details are discussed further in the ISAP I.d.1 Results Report.
1 R$ vision: 1 Page 18 of 29
( RESULTS REPORT ISAP VII.c (Cont'd)
Appendix 12 (Cont'd) 2.0 DISCUSSION OF RESULTS (Cont'd)
An analysis was performed on the inspection discrepancies noted during the ISAP I.d.1 evaluation that are attributable to the inspector not-found to be either certifiable or capable of performing satisfactory inspection by the ISAP 1.d.1 evaluation. ,
This analysis took into account the nature, scope, and 1 relevancy to this population of the reported discrepancies and considered the possibility that such discrepancies might result in a construction deficiency, adverse trend, or unclassified trend if ,
they were to occur on items in this population. This analysis determined that these discrepancies are all of a relatively minor nature and are similar to deviations (those that did not result in findings) noted during the reinspection conducted on items in the random
(' - sample for this population. Based on the analysis, it
! ( was concluded that inspection discrepancies of the type and severity noted during the ISAP I.d.1 Phase III evaluation, were they to occur on a reinspection or documentation review attribute for an item in this population, would not result in a construction deficiency, adverse trend, or unclassified trend.
Thus, the possibility of such discrepancies existing does not detract from the conclusions for this construction work category.
3.0 ROOT CAUSE AND GENERIC IMPLICATIONS .
This section provides a root cause and generic implications analysis for the special case of radial weld shrinkage identified in Section 2.2.4, and the unclassified trend of base material reduction identified in Section 2.2.9.
3.1 Radial Weld Shrinkage Root Cause The subject field weld is a replacement weld of a previously completed weld that was field-fabricated in March 1981 i utilizing Weld Procedure Specification 88021. Subsequently, I
- ( ) the replacement weld received a repair later in March 1981 and another repair in April 1981. Both repairs utilized Weld 1
i l
- - - - ~_ _ _ . _ _ . _ . _ _ _ _ _ .
o _ _ _
Revision: 1 Page 19 of 29 RESULTS REPORT ISAP VII.c (Cont'd) .
Appendix 12 (Cont'd) 3.0 ROOT CAUSE AND GENERIC IMPLICATIONS (Cont'd)
Procedure Specification 88025. The repairs involved through-the-wall excavation and rewelding to resolve unacceptable lack of fusion results determined from radiographic examination. Both repair areas corresponded to the limited location where the radial weld shrinkage deviation was reported.
Pipe velding was performed by qualified pipe welders in accordance with construction procedure CP-CPM-6.9D. This procedure provides instructions to the welder for pipe welding and welding-related processes. From 1981 through the present revision, this procedure provided no specific instruction or requirement associated with radial weld shrinkage, or necking, of the pipe in the heat affected zone. Construction procedure O CP-CPM-6.9D required qualified welders to perform welding to qualified and approved weld procedure specifications, which contain weld technique sheets for reference, to ensure that the proper welding paramaters were used. These weld parameters include preheat, travel speed, interpass temperature, amperage, etc.
The qualifications of each welder associated with the weld were reviewed, and each welder had the proper qualifications to work to the weld procedure specifications used during the fabrication and subsequent repair welding. The weld procedure I
specifications were reviewed in conjunction with the weld data l package documentation, and there is no evidence to suggest that the weld procedures or the welding parameter requirements were not followed during the fabrication or repair of the veld.
Although variations in the previously mentioned welding parameters can contribute to radial weld shrinkage, the shrinkage is usually the result of excessive cycles of high heat input from rewelding and reworking of the metal. Such a situation could result from a replacement weld or an area of a weld that has received extensive repair involving through-the-wall excavation and rewelding, particularly on schedule 80 or thinner wall stainless steel pipe.
l Rsvision: 1 Pags 20 of 29 A
f $
\s/ RESULTS REPORT ISAP VII.c (Cont'd)
Appendix 12 (Cont'd) 3.0 ROOT CAUSE AND GENERIC IMPLICATIONS (Cont'd)
The evidence reviewed indicates that this is the cause of the radial weld shrinkage. The radial weld shrinkage was identified on a replacement weld and only at a limited location that received two repairs. Both repairs involved through-the-wall excavation and rewelding.
Construction procedure CP-CPM-6.9D provided no limitation to the number of times that extensive repairs can be made to the ,
same area (other than to obtain owner approval of the third i repair to control sensitization of stainless steels);
therefore, radial weld shrinkage may have occurred during the replacement and/or extensive repair of schedule 80 or thinner stainless steel piping joined by full penetration butt welds.
The control of radial veld shrinkage was apparently left to
(/}
x- Quality Control inspection for' detection and to Engineering for resolution of excessive shrinkage.
Final visual acceptance of the weld was performed by certified Quality Control inspectors in April 1981, in accordance with Quality Control piping inspection procedure QI-QAP-11.1-26 Revision 4, which correlates with construction procedure CP-CPM-6.9D, and Quality Control inspection procedure QI-QAP-10.2-7, Revision 0. Neither inspection procedure QI-QAP-11.1-26, inspection procedure QI-QAP-10.2-7, nor any referenced inspection documents, provided any criteria or requirements to inspect for radial weld shrinkage of pipe. If excessive radial weld shrinkage was in existence on a weld during final visual examination in 1981, it is quite possible that it would not have been reported because of these less-than-adequate inspection procedures.
Criteria for the requirements in the inspection procedures were formulated from various source documents. One such document was an erection specification. In 1981, piping field welds were fabricated and installed to the requirements of Gibbs & Hill Piping Erection Specification 2323-MS-100, Revision 5. At that time, the specification contained no criteria for acceptable radial weld shrinkage. It only
/- g required that " welding parameters shall be controlled to avoid excessive heat input resulting in excessive necking down of
('~) butt welds in thin wall pipe". Even if Quality Control inspection were to take it upon themselves to inspect for radial weld shrinkage, acceptance would be purely judgmental
Rsvision: 1 Pags 21 of 29 RESULTS REPORT ISAP VII.c (Cont'd)
Appendix 12 (Cont'd) 3.0 ROOT CAUSE AND GENERIC IMPLICATIONS (Cont'd) as to what " excessive necking" meant because of the lack of quantitative criteria in MS-100. This erection specification is considered less-than-adequate for its failure to provide quantitative criteria for the inspection of excessive radial weld shrinkage (necking) and for not addressing shrinkage that can occur due to replacement and repair of a weld.*
Radial weld shrinkage and lack of quantitative criteria were addressed on May 18, 1982, when Design Change Authorization ,
(DCA) 13,335 was issued to MS-100, Section 4.27.8, to specifically provide quantitative acceptance criteria for radial weld shrinkage. No such design change was ever issued to MS-43B. This criteria was not incorporated into Quality Control piping inspection procedure QI-QAP-11.1-26, Section 3.6.4, until December 16, 1982 (Revision 7 of QI-QAP-11.1-26 4O had superseded procedure QI-QAP-10.2-7 in its entirety). No evidence was found that would suggest that any effort was made to retrofit or justify formally not retrofitting inspection to the new quantitative criteria for butt welds on stainless steel pipe inspected prior to December 1982.
In summary, regardless of how well weld parameters are controlled during fabrication of a weld, radial weld shrinkage is still expected to occur because of construction activities such as weld replacement-and/or extensive weld repair of butt welds on schedule 80 or thinner stainless steel pipe. It then becomes Quality Control inspection's responsibility to detect and verify acceptability of excessive radial weld shrinkage in accordance with the criteria established by Engineering.
Based on the above review, the existence of excessive radial weld shrinkage can be traced to the following causes:
- Gibbs & Hill Piping Erection Specification 2323-MS-100 is the engineering specification for the erection of all fabricated piping. The site shop-f abricated piping is f abricated in accordance with Gibbs & Hill Piping-Nuclear (Shop Fabrication of Piping in the Fiald) Specification 2323-MS-43B. Specification 2323-MS-43B also failed to provide quantitative criteria for the O inspection of radial weld shrinkage, and presently, still provides no quantitative acceptance criteria.
1 l
Ravicion: 1 Paga 22 of 29 I RESULTS REPORT U)
ISAP VII.c (Cont'd)
Appendix 12 (Cont'd) l l
l 3.0 ROOT CAUSE AND GENERIC IMPLICATIONS (Cont'd)
- The lack of quantitative acceptance criteria for radial weld shrinkage, resulting from a less-than-adequate l
piping erection specification, was the primary root I cause for the existence of excessive radial weld shrinkage. This resulted in the less-than-adequate construction procedure and the less-than-adequate QC inspection procedure prior to December 1982.
l - The lack of a QC requirement to detect and report I excessive radial weld shrinkage, resulting from a ,
less-than-adequate piping inspection procedure, was the secondary root cause for the existence of excessive radial weld shrinkage.
/ N - Engineering's failure to require a retrofit to inspect
\m- ) the welds completed prior to December 1982, was a contributing cause for the existence of excessive radial weld shrinkage.
Generic Implications The requirement to inspect for radial weld shrinkage using quantitative criteria has been in effect since December 1982; however, a retrofit was not performed to reinspect welds made prior to that date. On piping inspected prior to December 1982, there could be stainless steel butt welds in schedule 80 or thinner pipe with excessive radial weld shrinkage that have gone undetected. In addition, other welding and material attributes may have been affected by the failure of engineering to retrofit changes to the acceptance criteria in piping erection specification 2323-MS-100.
This implication is also applicable to site pipin; fabrication specification 2323-MS-43B, which is the specification for the shop fabrication of piping in the field.
Summary The primary root cause for the existence of excessive radial r 's weld shrinkage is a less-than-adequate piping erection
() specification 2323-MS-100. The secondary root cause is a less-than-adequate piping inspection procedure QI-QAP-11.1-26.
The contributing cause is the failure of engineering to m - .___-.________________o
L l Rsvision: 1:
Page 23 of 29 O RESULTS REPORT ISAP VII.c (Cont'd)
Appendix 12 l
(Cont'd) 1 3.0 ROOT CAUSE AND GENERIC IMPLICATIONS (Cont'd) require retrofitting, in accordance with the acceptance criteria added in December 1982. Refer to Section 4.0 for recommendations for corrective action.
3.2 Base Material Reduction Root Cause The base material reduction occurred where grinding removed up to 0.044 (approximately 3/64) inch from the base material of a 4 inch diameter Schedule 40 pipe around the entire reinforced branch connection weld of a 1 inch sockolet.
Construction procedure CP-CPM-6.9D, Rev. 7, " Welding and Related Processes", which provides the weld reinforcement and configuration requirement for sockolets and the wall thickness requirements for piping, specifies that the weld surface shall make a smooth transition into the surface of the pipe. In addition, the final surface shall be free of surface irregularities (cracks, laps, spatter, etc.) and in no case shall the surface condition interfere with the interpretation of NDE, when required. The procedure also stipulates that at no time shall there be a violation of the manufacturer's minimum wall thickness. The construction procedure does not provide any instruction or limitation on in-process grinding activities, but does provide criteria for the quality results required after grinding is complete. Therefore,'the grinding that occurred, although excessive, was not due to an inadequate construction procedure.
Grinding may be performed by welder, fitters, or helpers.
Fitters and helpers are not identified on the weld data card; therefore, the individual responsible for the overgrinding condition could not be identified. The grinding was likely done by an individual trying to blend the weld into the base ,
material. Therefore, less-than-adequate workmanship by the craft is the primary root cause.
The QC inspection procedure QI-QAP-11.1-26, "ASME Pipe Fabrication and Installation Inspections", which providas the weld inspection requirements, specifies that the QC inspector shall verify that no base metal defects (laps, seams, minimum wall violation, etc.) and no arc strikes are present. A
Revision: 1 Page '24.of 29 p
RESULTS REPORT J
ISAP VII.c (Cont'd) l Appendix 12 (Cont'd) l 3.0 ROOT CAUSE AND GENERIC IMPLICATIONS (Cont'd) review of the applicable weld data card for the weld of concern determined that the "VT" (visual inspection) block was initialed by a certified QC inspector as being satisfactory, and the weld data card had been completed by that inspector in compliance with the above procedure.
In addition _to the original weld inspection completed under QI-QAP-11.1-26, a Pre-pressure Test /N-5 walkdown, in accordance with CP-QAP-12.1, was also performed, which included an inspection of the' pressure boundary for base metal defects.- A review of the QC Pre-pressure Test /N-5 checklist L for this pipeline indicates that the walkdown of the pipeline where the overgrind occurred was accepted by yet another QC
, inspection.
With regard to QC inspection, at least two different QC inspectors inspected the overground area, directly or indirectly, and determined that the condition was acceptable.
This indicates that the ground area was in a difficult location or position, or that it was blended so smoothly into the base material that a thickness reduction was difficult to detect. It appears that certified QC inspectors closely followed adequate QC inspection procedures; nevertheless, the deviation still remained undetected after direct inspection.
Therefore, the root cause associated with QC inspection is considered indeterminate.
Generic Implications The root cause of less-than-adequate workmanship does have valid generic implication with other pipe grinding activities on-site; however, certain configurations and pipe schedules are more susceptible to an overgrind condition that could violate the ASME Section III Code minimum design wall thickness. Some welds are merely lightly ground (polished) to enhance the weld surface for NDE, while others, such as the reinforce branch connection welds identified, stanchion welds, and welds joining uneven thicknesses of material receive more extensive and difficult grinding to blend the weld into the base material to make a smooth transition.
(
Rcvision: 1 Pega 25 of 29 g ~3 x_- RESULTS REPORT ISAP VII.c (Cont'd)
Appendix 12 (Cont'd) 3.0 ROOT CAUSE AND GENERIC IMPLICATIONS (Cont'd)
In addition, thin wall piping is obviously more susceptible to a violation of code minimum wall thickness, since there is less margin between the code minimum wall and the manufacturer's minimum wall. The one deviation that was identified was a reinforced branch connection weld that occurred on a pipe with a schedule 40 wall thickness.
Based on the above, the generic implication of this root cause is that other pipelines that contain a reinforced branch connection weld, stanchion weld, and/or welds joining uneven thicknesses of material, etc., which require a higher level of workmanship, are susceptible to a similar deviation.
r'N With regard to the remaining attributes of this construction
(_,) work category, each of these attributes indirectly verifies the adequacy of the craft workmanship in the performance of each weld. Since no significant deviations were identified for these attributes, less-than-adequate workmanship does not have any valid generic implications with any of the other at tribute s .
Summary The primary root cause of the existence of the base material reduction is less-than-adequate construction workmanship. The generic implication of this root cause is that other piping that contains reinforced branch connection welds, stanchion welds, and/or welds joining uneven thicknesses of material, are most susceptible to a similar deviation, in particular, thin vall pipelines. Refer to Section 4.0 for recommendations for corrective action.
4.0 RECOMMENDATIONS j The following recommendations for corrective action are made.
4.1 Radial Weld Shrinkage g)
'~
- 1) Ic is recommended that TU Electric reinspect butt welds in schedule 80 and thinner stainless steel piping that are replacement welds and/or have received extensive I
i I
l l
I
o- Revision: 1 i
Page 26 of 29
(%
\~~' RESULTS REPORT ISAP VII.c (Cont'd)
Appendix 12 (Cont'd) 4.0 RECOMMENDATIONS (Cont'd) repairs to verify that the welds meet the present criteria for radial weld shrinkage, and correct as necessary. The reinspection is limited to welds that were inspected prior to December 1982.
- 2) It is recommended that TU Electric evaluate the need for adding criteria to both piping erection specification 2323-MS-100 and site piping f abrication specification 2323-MS-43B to limit the number of i extensive weld repairs or replacements to avoid excessive radial weld shrinkage, and correct as necessary.
- 3) It is recommended that TU Electric review revisions and
[N.- design changes to both piping erection specification 2323-MS-100 and site piping fabrication specification 2323-MS-43B to ensure that proper retrofitting has been performed, and correct as necessary.
- 4) It is also recommended that TU Electric add quantitative criteria for acceptable radial weld shrinkage to site piping fabrication specification 2323-MS-43B.
4.2 Base Material Reduction
- 1) It is recommended that TU Electric perform an evaluation and analysis based on actual piping data, when available, to determine those pipe lines that are likely to be adversely affected by as much as a 3/64 inch deep grind. This analysis should not only consider the current design temperatures, pressures, and corrosion allowances, but should also assess the likelihood of an excessive grind taking place based upon the weld configuration, inspection requirements, and grinding activities associated with the welds that comprise those lines. Subsequent to completion of the TU Electric analysis, it is recommended that TU
'O Electric perform a walkdown to determine whether or not base material reduction exists on those pipelines that
() were determined to be susceptible to an overgrind l
condition of this magnitude and that could violate the ASME Section III Code minimum design wall thickness.
Revision: 1 Page 27 of 29 RESULTS REPORT ISAP VII.c (Cont'd)
Appendix 12 (Cont'd) 4.0 RECOMMENDATIONS (Cont'd)
- 2) It is recommended that TU Electric emphasize to QC inspection, by retraining and/or modification to QC procedures QI-QAP-11.1-26 and CP-QAP-12.1, those pipe weld / joint configurations that are most susceptible to overgrinding.
5.0 CONCLUSION
S Based on the findings of the reinspection, documentation reviews, and satisfactory implementation of the above recommendations and the applicable recommendations in the ISAP I.d.1 Results Report, there is reasonable assurance that the hardware in this construction work category is adequately installed in accordance O with the design.
l I
O
Revision: 1 Page 28 of 29 RESULTS REPORT ISAP VII.c (Cont'd)
Appendix 12 (Cont'd)
Table 12-1 Summary of Reinspection Results Pipe Welds and Material Deviation Classification Number of Inspection Number of Insigni- Construction Attribute Points Deviations ficant Notable Deficiency
- 1. Weld Identification and Location 186 0 0 0 0
- 2. Weld Configuration 186 0 0 0 0
- 3. Weld Size 257 0) 4 4 0 0
- 4. Radial Weld 256 1 1 0 0 Shrinkage
- 5. Weld Undercut 372 0 0 0 0
- 6. Surface of Weld 558 0 0 0 0
- 7. Weld Cracks and Lack of Fusion 558 0 0 0 0
- 8. Weld Rust 83 0 0 0 0
- 9. Base Material Defects 372 2 1 1 0
- 10. Base Material Rust 164 0 0 0 0 TOTALS 2,992 7 6 1 0 (Approximately 3,000)
(') Includes the inspection points sssociated with the verification of butt weld reinforcement.
Revision: 1 Page 29 of 29 ,
'I RESULTS REPORT 1
ISAP VII'.c j (Cont'd) {
l Appendix 12. -l 5
(Cont'd) i Table 12-2 Summary of Documentation Review Re~sults Pipe Welds and Material ;
Deviation Classification Number of l Review- Number of Insigni- 3 Attribute Points Deviations ficant Notable !
- 11. Base Material Traceability 366 3 3 0
- 12. Weld Material Traceability- 259 1 1 0
.O
- 13. Weld Procedure 0 0 Application 224 0
- 14. Weld Procedure 224 0 0 0 Qualification 320 0 0 0
- 15. Welder Qualification
- 16. QC Acceptance at Hold Points 1008 1 1 0
- 17. Pressure Test Complet:1on 124 0 'O O 2,525 5 5 0 TOTALS (1)
(Approximately: 2,520)
II) QC Inspection Certification review points and deviations are excluded from the totals and the deviations are not classified because'the results of the ISAP O' I.d.1 Indicated that.these deviations do not relate directly to actual inspector qualifications.
r i Ravision: 1
'Page 1 of 17
- RESULTS REPORT
'G(^T' ISAP VII.c (Cont'd)
Appendix 13 Tubing Welds and Material 1.0 REVIEW PROGRAM IMPLEMENTATION 1.1 Construction Work Category Description The construction work category of tubing welds and material is comprised of all site-made safety-related tubing welds. The site-made welds, veld materials, and base materials for the welding of tubing to fittings and components are included in this construction work category.
1.2 Population Size and Sample Selection For this construction work category, a population of approximately 1,540 items (tubing welds) was identified as having QC acceptance of the final weld as of the date of this e-' reinspection effort. Reinspection and documentation reviews
( j) were performed for a total of 101 accessible items. Of these, 84 first sample items were randomly selected from the population to ensure that at least 60 reinspection or documentation reviews of each attribute were performed. An additional 17 second sample items were randomly selected to ensure that at least 60 safe-shutdown hardware items were reinspected or reviewed.
1.3 Attributes Selected Sample items were reinspected for the following attributes:
Attribute 1 - Weld identification and location Attribute 2 - Weld configuration Attribute 3 - Weld size Attribute 4 - Weld undercut Attribute 5 - Surface of weld (overlap)
Attribute 6 - Weld cracks and lack of fusion Attribute 7 - Weld rust (stainless steel only)
O
l Revision: 1 l Page 2 of 17.
RESULTS REPORT ISAP VII.c (Cont'd)
Appendix 13 (Cont'd) 1.0 REVIEW PROGRAM IMPLEMENTATION (Cont'd)
Attribute 8 - Base material defects Attribute 9 - Base material rust (stainless steel only)
Sample item documentation was reviewed for the following attributes:
Attribute 10 - Base material traceability Attribute 11 - Weld material traceability Attribute 12 - Weld procedure application Attribute 13 - Weld procedure qualification Attribute 14 - Welder qualification Attribute 15 - QC acceptance at hold points This attribute encompasses the review of a number of non-recreatable activities associated with this construction work category, such as: fit-up, preheat, purge, NDE, repairs, cleanliness, etc.
Attribute 16 - Pressure test completion Attribute 17 - QC inspector certification
'\.
2.0 DISCUSSION OF RESULTS 2.1 Summary of Results For reinspection, two Deviation Reports were issued describing two deviations. Approximately 1,500 inspection points were encountered in performing these reinspection. See Table 13-1 for the results of the reinspection.
) For documentation review, 13 Deviation Reports were issued l
describing 13 deviations. Approximately 1,180 review points were encountered in performing the documentation reviews. See Table 13-2 for results of the documentation reviews.
i
Revision: 1 Page 3 of 17
[)
%J RESULTS REPORT ISAP.VII.c (Cont' d) -
Appendix 13 (Cont'd) 2.0 DISCUSSION OF RESULTS (Cont'd)
Documentation Review deviations related to inspector certification have not been included in the aforementioned totals. As discussed in Section 2.3.8, it has been determined that these deviations do not necessarily relate directly to actual inspector qualifications. These deviations were considered in the ISAP I.d.1 evaluation together.with supplemental information as necessary to determine actual 4 inspector qualifications. The ISAP I.d.1 conclusions are summarized in Section 2.3.8.
In the nine reinspection attributes of.this' population, no reinspection deviations were evaluated to be construction deficiencies, and no adverse trends were identified.
In the seven documentation review attributes (excluding QC 1 inspector certification) of this population, four documentation review deviations were determined to be insignificant and nine were determined to be notable.
2.2 Analysis of Reinspection Results This section provides, by attribute, a discussion of the reinspection deviations, an analysis of the effect of the deviations on the functional capability of the weld and base material, and an analysis for the presence of trends. The function of a weld and associated base material is to maintain the pressure boundary of the system to which it is attached and to maintain the flow capability of the tubing in which it is located. The instrument tubing transmits a plant process variable such as flow, pressure, temperature, or level for use in monitoring, recording, or controlling plant processes.
2.2.1 Attribute 1 - Weld Identification and Location In all 101 inspection points, the weld was verified, by weld identification and/or location, to be in accordance with the isometric drawing. No deviations were reported.
O m - - - . _ _ . _ _ _ _ _ _ _ _ _ _ _ _ _ , _ ___ _
Rovision: 1 Page 4 of 17 RESULTS REPORT ISAP VII.c (Cont'd)
Appendix 13 (Cont'd) 2.0 DISCUSSION OF RESULTS (Cont'd) 2.2.2 ' Attribute 2 - Weld Configuration In all 101 inspection points, the weld configuration was verified to be a socket weld in accordance with the tubing erection specification and/or the isometric drawing. No deviations were reported.
2.2.3 Attribute 3 - Wald Size In all but one of 101 inspection points, the weld size was veri ~fied to be in accordance with the tubing erection specification.
Approximately seventy percent of all inspection points involved socket weld couplings on 1/2-inch diameter and f
- smaller tubing. The balance were socket welds on the incore detector system, which contains tubing larger than 1/2-inch diameter.
The one deviation identified was a socket weld on the incore detector system tubing with a throat that was undersized by 0.026 inch (approximately a 14 percent reduction) for the nominal tube wall thickness used. The throat of a fillet veld is the distance from the root of the joint to the face of the weld. This condition occurred on a 1-inch diameter tube-to fitting connection for approximately 16 percent of the circumferential weld length. -
Even though the deviation encompassed only a small percentage of the weld circumference, the deviation was conservatively evaluated by assuming that the undersize weld condition existed for_100 percent of the circumferential weld length.
The evaluation of the deviation detern.ined that under worst case loading conditions, the reduction in weld size resulted in stresses increasing from 38 percent to 41 percent of the ASME Section III Code-allowable stresses. '
This constitutes a stress increase of approximately 8 -
I Ravisicn: . 1 .!
Page- 5 of 17
~s RESULTS REPORT l- ISAP VII.c (Cont'd)
Appendix 13 l i
(Cont'd) 2.0 DISCUSSION OF RESULTS (Cont'd) percent, with a 59 percent margin from the Code-allowable stress still available; therefore, the deviation was determined to be insignificant.
The deviation occurred on one of the 58 (1-inch diameter) tubes that are used specifically for one system within each unit, the incore detector system. The configuration and supports for each tube run are very similar and the operating conditions (temperature and pressure) are identical. A bounding analysis determined that at the f highest stress point of the incore detector tubing system, a similar deviation around the entire pipe circumference would increase stresses from 83 percent to 89 percent of-the ASME Section III Code allowable stresses. This constitutes a stress increase of approximately 7 percent.
Based on the bounding analysis results and the
- ' insignificant nature of this deviation, it is unlikely that a similar deviation would result in a construction deficiency if it were to occur in the uninspected portion of the incore instrument tubing system. Therefore, no adverse trend was identified.
During the reinspection, a concern was raised regarding TU Electric's method for the determinscion of socket veld size on 1/2-inch diameter and smaller tubing. This method had also been utilized in the CPRT reinspection. The concern involved the inability to determine the edge of the base material by visual inspection after completion of the weld, thereby leaving weld size indeterminate.
The weld size for the type of weld used to join the tube and fittings in the 1/2-inch diameter and smaller tubes is determined by measuring the lengths of the two legs at the interface between the weld filler material and the base material. The minimum leg length is based on the nominal tube wall; the ASME Code requirement for thin walled pipe and tubing is 1.09 times the nominal tube wall thickness.
In order to determine this length, the edge of the base material must be located and the weld leg length measured at that point. For 1/2-inch diameter and smaller tubing,
() the hub or lip of the socket weld coupling is very close
Ravision: 1.
Page~ 6 of 17
- . RESULTS REPORT A ISAP VII.c (Cont'd)
Appendix 13 (Cont'd) 2.0 DISCUSSION OF RESULTS (Cont'd) to the required weld size. This means that any consumption of the coupling during the welding process would result in reduction in weld size and an indeterminate edge of the base material.
A different approach to the determination of weld size has been incorporated into TU Electric QC inspection procedure QI-QP-11.21-1, which includes the requirements for visual weld inspection. This inspection method uses scribe lines, placed on the coupling by warehouse or field parsonnel prior to fitup, as a reference point to deterniine the location of the original edge of the coupling. When the original edge of the coupling is located, any change to the fitting that occurred during the welding process can be evaluated and the size of the weld can be determined. Future inspections of tubing 7
socket welds in Units 1 and 2 will utilize the inspection procedures recently added to the requirements for visual weld inspection.
TU Electric has committed to resolve the issue concerning weld size for 1/2-inch diameter and smaller tubing (refer to the discussion in Section 4.0, addressing ongoing-activities). The CPRT reinspection have not been reperformed to the revised criteria. Thus, conclusions cannot be drawn regarding the strength characteristic of weld size for socket welds on 1/2-inch diameter and smaller tubing until the Project's dispositions of the NCRs documenting this concern have been implemented and the results of the reinspection have been evaluated.
2.2.4 Attribute 4 - Weld Undercut In all 202 inspection points, the base material at each ;
toe of the weld was verified to be free of undercut that exceeded the allowable defect depth defined by the tubing erection specification. No deviations were reported.
l O
l R vision: 1 Page 7 of 17 RESULTS REPORT l /N l 0 ISAP VII.c l.
(Cont'd)
Appendix 13 (Cont'd) 2.0 DISCUSSION OF RESULTS (Cont'd) 2.2.5 Attribute 5 - Surface of Weld (Overlap)
In all 202 inspection points, the wold surface at each toe of the weld we.s verified to be free of overlap as required by the tubing erection specification. No deviations were reported.
2.2.6 Attribute 6 - Weld Cracks and Lack of Fusion l
In all 303 inspection points, the weld was verified to be free of any visible cracks and lack of fusion as required by the tubing erection specification and the construction procedure. No deviations were reported.
,f g 2.2.7 Attribute 7 - Weld Rust kb In all 101 inspection points, the weld was verified to be free of rust that could be attributed to improper use of material or tools, as required by the tubing erection specification. No deviations were reported.
l 2.2.8 Attribute 8 - Base Material Defects In all but one of 202 inspection points, the base material on each side of the weld was verified to be free of welding-related defects that exceeded the allowable defect depth as defined by the tubing erection specification.
The deviation occurred on a 1-inch diameter tube (.240 _
.024 inch wall thickness), and consisted of an overgrind condition that measured 0.040 inch in depth, 3/8 inch in width (longitudinally), and 1 inch in circumferential length. The ground area had a smooth transition without notches or sharp contours. The allowable defect depth for ,
this tube size is 0.020 inch. Although the overgrind I condition occurred on only approximately one-third of the tube circumference, the deviation was conservatively evaluated by assuming the entire circumference of the tube was overground.
o _ _____
Ravision: 1 Page 8 of 17
[T RESULTS REPORT.
V ISAP VII.c (Cont'd)
Appendix 13 (Cont'd) l 2.0 DISCUSSION OF RESULTS (Cont'd)
The evaluation of the deviation determined that under the worst case loading condition, the decrease in section thickness resulted in stresses increasing from 54 percent
'to 69 percent of the ASME Section III Code-allowable ,
stresses. This constitutes a stress increase of l I
approximately 28 percent, with a 31 percent margin from the Code allowable stresses still available; therefore, the deviation was determined to be notable.
There are a total of 58 (1-inch diameter) tubes that are specifically used for one system within each unit, the incore detector system.- These-tubes are all exposed to identical pressure and temperature and the configuration and supports are very similar for each tube run. The
~N deviation occurred at one of the welds that joins each of these tubes to the corresponding reactor vessel nozzle.
The specific weld grinding requirements for the reactor vessel nozzle to incore detector tube weld call for the weld to be blended (ground) to a 1/4-inch radius. The 1/4-inch radius blending requirement, which is believed to be the cause for the excessive grinding, is not required for the smaller diameter tubes, and is not required elsewhere for the 1-inch diameter tubes. A welder performing normal surface grinding or polishing activities on other types of tubing welding (without the blending requirement) is unlikely to cause an excessive grind.
Therefore, this type of deviation was determined to be restricted in nature, with the possibility of occurring only on incore detector reactor vessel nozzle tubing welds.
As a result of this deviation a nonconformance report was initiated by the installer and the overgrind condition was reinspected for minimum wall thickness violation.
Measurements made by Brown & Root QC inspection utilizing an ultrasonic test (UT) determined that the tube wall thickness ranged from 0.220 to 0.228 inch at the overgrind area. Thesa measured values are above the manufacturer's required minimum wall of 0.215 inch. In this case, even though the allowable
- defect depth was exceeded, the minimum wall thickness was maintained.
I I
- -- n
,-_____-~ ,. -- -
1 i
i R2visio' n: 1 Page 9 of 17
<s -
RESULTS REPORT-l ISAP VII.c (Cont'd) j Appendix 13 .
(Cont' d) l 1
i 2.0 DISCUSSION.0F RESULTS (Cont'd)
Due to the possibility for a similar or greater overgrind to occur on other incore detector tubing welds, a bounding analysis was performed to determine the maximum overgrind depth that must take place to bring the stresses at the highest stress point of the incore detector tubing system to the code-allowable limits. This evaluation determined that an overgrind of 0.089-inch deep, around the' entire tube circumference, will result in a residual wall thickness of 0.151 inch and will increase the stresses up to the code-allowable limits. However, it is unlikely .
that a welder trying to meet the required 1/4-inch radius blending would cause an overgrind condition of 0.089-inch.
Based on the results of the conservative evaluations performed and the low probability that an overgrind will O enceed the bounding analysis determination, it is unlikely that this deviation type, if it were to exist in the uninspected portion of the population, would result in a construction deficiency. Therefore, no adverse trends were identified.
2.2.9 Attribute 9 - Base Material Rust In all 202 inspection points, the base material on each side of the weld was verified to be free of rust that could be attributed to improper use of materials or tools, as required by the tubing erece. ion specification. No deviations were reported. ,
2.2 Analysis of Documentation Review Results This section provides, by attribute, a discussion of the documentation review deviations and an analysis of the effect of the dariations on the ability of the collected evidence to provide reasonable assurance that the hardware is properly installed.
2.3.1 Attribute 10 - Base Material Traceability In all but 13 of the 200 review points, it was verified l that traceability of materials was maintained through !
.[,,}-
'- applicable documentation, as required by the tubing
R: vision: 1 Page 10 of 17
[ RESULTS REPORT C
ISAP VII.c (Cont'd)
Appendix 13 (Cont'd) 2.0 DISCUSSION OF RESULTS (Cont'd) erection specification. In addition, if legible heat or code numbers or color coding were available on the tube, these were verified to be the same as the heat or code numbers or color coding listed in the documentation.
Three deviations were reported where the color code found in the field for traceability of the tubing did not agree with the color code documented on the installation records for those welds. In all three instances, certified documentation was found that verified both color codes to have the same size and material specifications, and both were in compliance with the isometric drawing. Therefore, the deviations were determined to be insignificant.
/ One deviation was reported where a Kerotest valve serial
(]./ number as identified in the field did not match the serial number documented in the installation records. Evaluation of this deviation determined that an interchange had occurred between two identical, adjacent valves that were installed by the same welder on the same day. This deviation was determined to be insignificant.
Further investigations determined that all valves used for welded tubing applications (i.e., instrument isolation, vent, or drain) are safety-related and made by the same manufacturer (Kerotest). The body size and characteristics of these Kerotest valves are the same for all applications (i.e., the valves are identical),
regardless of the tube diameter to which they are connected. By comparison, the non-safety-related valves are approximately one-half the body size of the safety-related valves, are threaded (not welded), and are mad; by other manufacturers. It is very unlikely that a mixq, between a safety-related and a non safety-related valve could occur. Therefore, a similar deviation could not result in a construction deficiency, and no adverse trend was identified.
,Nine deviations concerned a material discrepancy for the A
- tube-ends of 1/2-inch diameter flexible metal hoses. The (j tube-este material, stainless steel grade 316, required by the installation drawing did not agree with the material,
Rzvisicn: 1 Pags 11 of 17 ,
RESULTS REPORT r
' .ISAP VII.c (Cont' d)
Appendix 13 j (Cont' d) {
2.0 DISCUSSION OF RESULTS (Cont'd) stainless steel grade 304L, in the material certification documentation. Since the documentation indicates acceptance of an out-of-specification condition, these deviations were determined to be notable.
The flexible metal hoses are vendor-supplied assemblies.
Each flexible metal hose is comprised of a bellows or hose-body assembly with short sections of tubing welded on either end. The bellows portion of a flexible metal hose has a lower allowable design load (i.e., is a weaker link) than the tube-ends. Stainless steel grade 304L has a lower allowable stress value than grade 316; however, evaluations determined that even with the substitution of 304L for the tube ends, the bellows portion of the metal hose is still the stress-limiting part and the design capacity of the hose had 'not been reduced. From the-O e'raluations performed, it was determined that this dtviation type does not affect the ability of the flexible metal hoses to perform their intended function. Based on this evaluation, it is concluded that the flexible metal hose assemblies are acceptable as-installed.
However, a concern exists in that the QC receipt inspector accepted components made from material not in accordance with specification requirements. Therefore, these deviations have been referred to ISAP VII.a 9, " Receipt and Storage of Purchased Material and Equipment Receiving Inspection", for incorporation into the ISAP VII.a.9 evaluation of the receipt inspection activfty.
l 2.3.2 Attribute 11 - Weld Material Traceability In all 100 review points, it was verified that traceability of the weld filler material used in the production of the weld was maintained through applicable documentation as required by the tubing erection specification. No deviations were reported.
O
l Revision: 1 Page 12 of 17
, <"N RESULTS REPORT j
(~) \
ISAP VII.c f (Cont' d)
Appendix 13 (Cont'd) 2.0 DISCUSSION OF RESULTS (Cont'd) 2.3.3 Attribute 12 - Weld Procedure Application In approximately 100 review points, the application of the weld procedure used in the production of the weld was verified to be in accordance with the tubing erection specification. No deviations were reported.
2.3.4 Attribute 13 - Weld Procedure Qualification l
In approximately 100 review points, the weld procedure was verified to be approved for use by the architect / engineer or the utility prior to its use, as required by the tubing erection specification. No deviations were reported.
2.3.5 Attribute 14 - Welder Qualification
( )
In approximately 100 review points, the welder was verified to be properly qualified to make the weld at the time the velding was performed, as required by the tubing erection specification. In addition, if legible welder symbols were available on the pipe, these symbols were verified with the welder symbols in the weld ,
documentation. No deviations were reported.
2.3.6 Attribute 15 - QC Acceptance at Hold Points In approximately 500 review points, the selected QC hold points on the weld installation and weld re* pair documentation (when applicable) were verified to be signed, or initialed, and dated as satisfactory by the QC inspector in accordance with the tubing erection specification requirements. No deviations were reported.
2.3.7 Attribute 16 - Pressure Test Completion In all 80 review points, the welds were verified to have been pressure-tested, and documentation was signed or initialed and dated as satisfactory by the QC inspector in accordance with the tubing erection specification and the construction procedure requirements. No deviations were
(N reported.
\~-
Revision: 1 Page. 13 of'17f r RESULTS REPORT
\
ISAP VII.c ;
(Cont'd)
Appendix 13 (Cont'd) 2.0 DISCUSSION OF RESULTS (Cont'd) 2.3.8 Attribute 17 - QC Inspector Certification A total of 780 reviews were performed to determine if the QC inspector who signed, or initialed, and dated the inspection results in the weld documentation was appropriately certified. For ASME weld inspections, a total of 265 reviews were performed to determine if the QC inspector was' appropriately certified to perform QC Level II inspections. For non-ASME weld inspections, a total of 515 reviews were performed to determine if the QC inspector was appropriately certified to all the applicable inspection procedures.
Deviations were identified whenever evidence of certification to the established procedural requirements t
\- at the time of inspection could not be found. However, as a result of the ISAP I.d.1 evaluations'of inspector certifications, it has been determined that such deviations do not necessarily relate directly to actual inspector qualifications. Consequently, the identified deviations were not relied upon (though they were considered) in determining inspector capability.
Inspecto qualifications were assessed for all inspectors (those with and without reported deviations) using the ISAP I.d.1 methodology.
The ISAP I.d.1 evaluation of inspectors whose work affected this population concluded that alf but one inspector either were certifiable to applicable criteria at the time of inspections, were found to be capable of performing satisfactory inspections (including those with substantial positive evidence), or were otherwise shown to be of no further concern. Evaluation details are discussed further in the ISAP I.d.1 Results Report.
An analysis was performed on the inspection discrepancies noted during the ISAP I.d.1 evaluation that are attributable to the inspector not found to be either certifiable or capable of performing satisfactory
/ inspection by the ISAP I.d.1 evaluation. This analysis ,
\_- took into account the nature, scope, and relevancy to this
-_l_____---
l Ravicion: 1 Page 14 of 17
/% RESULTS REPORT N~A i ISAP VII.c (Cont'd) )
l Appendix 13 (Cont'd) 2.0 DISCUSSION OF RESULTS (Cont'd) population of the reported discrepancies and considered the possibility that such discrepancies might result in a construction deficiency, adverse trend, or unclassified trend if they were to occur on items in this population.
This analysis determined that these discrepancies are all of a relatively minor nature and are similar to deviations (those that did not result in findings) noted during the reinspection conducted on items in the random sample for this population. Based on the-analysis, it was concluded that inspection discrepancies of the type and severity noted during the ISAP I.d.1 Phase III evaluation, were they to occur on a reinspection or documentation review attribute for an item in this population, would not result in a construction deficiency, adverse trend, or e'- unclassified trend. Thus, the possibility of such
( ,s) discrepancies existing does not detract from the conclusions for this construction work category, p 3.0 ROOT CAUSE AND GENERIC IMPLICATIONS No adverse trends or construction deficiencies were identified.
Therefore, a root cause and generic implication analysis is not required.
4.0 RECOMMENDATIONS AND ONGOING ACTIVITIES .
No recommended corrective actions are required.
To address the concern raised regarding indeterminate weld size on 1/2-inch diameter and smaller instrument tubing socket welds (refer to discussion in Section 2.2.3), TU Electric has generated Nonconformance Reports (NCR) M-86 103363 and CI-87-5621, covering all such tubing in Unit 1, Common, and Unit 2. Currently, this issue has not been fully resolved. The interim disposition to NCR M-86-103363 requires that QC l inspection reinspect all Unit 1 and Common safety-related 1/2-inch ,
tubing welds in accordance with the requirements of inspection procedure QI-QP-11.21-1 (latest revision). This reinspection is
<-s scheduled for completion on December 31, 1987. Upon completion of V
Rsvision: 1 Page 15 of 17 RESULTS REPORT
/^'sl ISAP VII.c (Cont'd)
Appendix 13 (Cont'd) 4.0 RECOMMENDATIONS AND ONGOING ACTIVITIES (Cont'd) i this reinspection, the results will be evaluated by engineering and corrective action will be provided, as required. No disposition has yet been provided for the remaining NCR, which deals with Unit 2 tubing welds.
5.0 CONCLUSION
Based on the findings of the reinspection, documentation reviews, and the applicable recommendations in the ISAP I d.1 results report, there is reasonable assurance that the hardware in this construction work category is adequately installed in accordance with the design.
Conclusions regarding socket weld size on 1/2-inch diameter and smaller tubing will be made by TU Electric as discussed in Section q 4.0.
v/
> \
/
Revision: 1 Page 16 of 17 RESULTS REPORT ISAP VII.c (Cont'd)
Appendix 13 (Cont'd)
Table 13-1 Summary of Reinspection Results Tubing Welds and Material Deviation Classification Number of Inspection Number of Insigni- Construction Attribute Points Deviations ficant Notable Deficiency
- 1. Weld Identification 0 0 0 0 and Location 101 0 0 0 0
- 2. Weld Configuration 101 101 1 1 0 0
- 3. Weld Size 202 0 0 0 0
- 4. Weld Undercut
- 5. Surface of Weld 202 0 0 0 0 (Overlap)
- 6. Weld Cracks and Lack 303 0 0 0 0 of Fusion 101 0 0 0 0
- 7. Weld Rust
- 8. Base Material 202 1 0 1 0 Defects .
- 9. Base Material 202 0 0 0 0 Rust 1,515 2 1 1 0 TOTALS (Approximately 1,500)
O
r Revision: 1 Page- 17 of 17
() RESULTS REPORT ISAP VII.c
.(Cont'd)
Appendix 13 (Cont'd)
Table 13-2 Summary of Documentation Review Results Tubing Welds and Material' Deviation Classification Number of Review Number of Insigni-Attribute Points Deviations ficant Notable
- 10. Base Material Traceability 200 13 4 '9
- 11. Weld Material-Traceability. 100 0 0 0
.f) 12. Weld Procedure
'\d Application' 100 0 0 0
- 13. Weld Procedure 0 0 0
-Qualification 100 0 0 0
- 14. Welder Qualification 100
- 15. QC. Acceptance at Hold Points 500 0 0 0
- 16. Pressure Test Completion 80 0 0 ,
O TOTALS (l) 1,180 13 4 9 (Approximately 1,180)
I (1) QC Inspector Certification review points and deviations are excluded from totals and the deviations are not classified because the results of ISAP I.d.1 indicate that these deviations do not relate directly to actual inspector qualifications.
1
Revision: 1 Page 1 of 24 RESULTS REPORT ISAP VII.c (Cont'd)
Appendix 14 Field-Fabricated Tanks 1.0 REVIEW PROGRAM IMPLEMENTATION 1.1 Construction Work Category Description The construction work category of field-fabricated tanks is comprised of all safety-related field-fabricated tanks.
All of the tanks in this category were designed and fabricated by the Chicago Bridge and Iron Company (CBI).
1.2 Population Size and Sample Selection For this construction work category, all field-fabricated tanks had QC acceptance, yielding a population of eight items.
These eight items were two boric acid tanks, two recycle holdup tanks, and four diesel fuel oil storage tanks.
As there are only eight items in the field-fabricated tank population, all accessible attributes applicable to each tank were inspected. As the four diesel fuel oil tanks were buried underground and contained fuel oil, it was not possible to reinspect them. When attributes were inaccessible or non-recreatable, documentation reviews were performed for all tanks.
1.3 Attributes Selected Sample items were reinspected for the following getributes:
Attribute 1 - Nameplates Attribute 2 - Shell top angles Attribute 3 - Anchor bolt rings, anchor bolt chairs, and anchor bolt nut tightness Attribute 4 - Weld seam offsets Attribute 5 - Nozzle type, size, and location Attribute 6 - Nozzle insert reinforcement plates
l Revision:- -1 Page 2 of 24 RESULTS REPORT I
ISAP VII.c (Cont'd)
Appendix 14 (Cont'd)'
1.0 REVIEW PROGRAM IMPLEMENTATION (Cont'd)
Attribute 7 - Nozzle and manhole reinforcing pads Attribute 8 - Seismic restraints Attribute 9' - Welding Attribute.10 - Base material defects
' Attribute 11'- Weld and base material rust Attribute 12 - Grout Sample item documentation was reviewed for the following attributes:
Attribute Dimensional verification Attribute 14 - Hydrostatic testing Attribute 15 - Record drawings Attribute 16 - Welding materials Attribute 17 - Procedure approval Attribute 18 - Welder and welding operator qualifications Attribute 19 - Nondestructive examinations (NbE)
Attribute 20 - NDE personnel certification Attribute Seismic restraint as-built dimensions Attribute 22 - Material verification 2.0 DISCUSSION OF RESULTS 2.1 Summary of Results For reinspection, 14 Deviation Reports were issued describing 403 deviations. Approximately 10,000 inspection points were encountered in performing the reinspection. See Table 14-1 for results of the reinspection.
Rsvision: 1 Pega 3 of 24 rm RESULTS REPORT
()
ISAP VII.c (Cont'd)
Appendix 14 (Cont'd) 2.0 DISCUSSION OF RESULTS (Cont'd)
For documentation review, 13 Deviation Reports were issued describing 23 deviations. Approximately 6,200 review points were encountered in performing the documentation reviews. See Table 14-2 for results of the documentation reviews.
Documentation review deviations related to inspector certification have not been included in the aforementioned totals. Ae discussed in Section 2.3.8, it has been determined that these deviations do not necessarily relate directly to actual inspector qualifications. These deviations were considered in the ISAP I.d.1 evaluation together with supplemental information as necessary to determine actual inspector qualifications. The ISAP I.d.1 conclusions are p- summarized in Section 2.3.8.
i'-) In this population, no reinspection deviations were evaluated to be construction deficiencies and no adverse trends were identified. All documentation review deviations were determined to bs insignificant.
2.2 Analysis of Reinspection Results This section provides, by attribute, a discussion of the reinspection deviations, an analysis of the effect of the deviations on the functional capability of the field-fabricated tanks, and an analysis for the presence of trends.
Field-fabricated tanks must maintain their struegural integrity and fluid retention capability under all required design loadings in order to perform their function.
2.2.1 Attribute 1 - Nameplates j Nameplates were examined to determine whether the "N" and " Class 3" stamps required by the ASME Code were affixed and whether the equipment number matched the number shown on the design drawing to indicate location. In all four inspection points, no deviations were reported. i l
?
U
L l
Revision: 1 Pago 4 of 24 i=(G L _,/ RESULTS REPORT ISAP VII.c (Cont'd)
Appendix 14 (Cont'd) 2.0 DISCUSSION OF RESULTS (Cont'd) 2.2.2 Attribute 2 - Shell Top Angles The shell top angle is a length of angle iron that has been shaped to_ fit the circumference of the tank. It is attached to the outer perimeter of the top of the tank wall and is used to support the tank roof.
A visual inspection was made to determine whether shell top angles were. installed in accordance with the nominal dimensions shown on the CBI drawings.' In all four inspection points, no deviations were reported.
2.2.3 Attribute 3 - Anchor Bolt Rings, Anchor Bolt Chairs, and Anchor Bolt Nut Tightness
-O The anchor bolt ring is a horizontal steel plate that is welded to the outside of the tank wall.just above the bottom of the tank. The bottom of the tank extends beyond the tank wall and the anchor bolts pass through both the anchor bolt ring and the tank bottom extension. An anchor bolt chair consists of two vertical pieces'of-steel plate welded between the anchor bolt ring and the tank bottom extension on either side of each anchor bolt. The purpose is to transmit load from the tank to the anchor bolt.
Anchor bolt rings and anchor bolt chairs yere visually inspected to determine whether they were installed in accordance with the nominal dimensions shown on the CBI drawings. In addition, each anchor bolt nut was determined to be tight by an attempt to further turn the nut by hand. In all 180 inspection points on the four above-ground tanks, no deviations were reported.
2.2.4 Attribute 4 - Weld Seam Offsets The above-ground field-fabricated tanks were constructed from steel plates that were welded together 3 J
at the site. In order to minimize the length of continuous weld the' plates were arranged in such a way that the vertical weld seams of two adjacent courses of
Ravicion: 1 Pags 5 of 24
) RESULTS REPORT ISAP VII.c (Cont'd)
Appendix 14 (Cont'd) 2.0 DISCUSSION OF RESULTS (Cont'd) plate did not coincide. The horizontal distance between the vertical weld-seams of two adjacent plate courses is known as the vertical weld seam offset.
The floors of the tants are made by joining overlapping plates with fillet welds. The plates are laid out in such a fashion that there is a minimum separation distance:between the points where three plates come together. This distance is known as the lap joint spacing.
For all four above-ground tanks, all vertical weld seams were inspected to determine whether they were, offset in accordance with ASME Code. In addition, the
() floor plate weld seams of the boric acid tanks were inspected to determine whether the lap joint spacings were in accordance with ASME' Code. The floor plate weld seams in the recycle holdup tanks could not be inspected, because these tanks contained liquid. .In all-22 inspection points on the four tanks inspected, no deviations were reported.
2.2.5 Attribute 5 - Nozzle Type, Size, and Location Nozzle type and size were inspected to determine whether the nozzle sizes and types were as shown on the CBI drawings. In addition, nozzle locatign was inspected to determine whether the vertical placement of the principal nozzles was within 1 inch of the dimension shown on the CBI drawings. In the approximately 125 inspection points, no deviations were reported.
2.2.6 Attribute 6 - Nozzle Insert Reinforcement Plates An inspection was made to determine whether nozzle insert reinforcement plates installed in the boric acid tanks agreed with the nominal dimensions as shown on the CBI drawings. The recycle holdup tanks were not inspected because they contained liquid. In all 28 O inspection points, no deviations were reported.
I
- - - - - - _ _ _ _ _ _ _ _ _ _ _ _ _ ________j
Revision: 1 Page 6 of 24 RESULTS REPORT ISAP VII.c (Cont'd) l Appendix 14 (Cont'd) 2.0 DISCUSSION OF RESULTS (Cont'd) 2.2.7 Attribute 7 - Nozzle and Manhole Reinforcing Pads An inspection was made to determine whether nozzle and manhole reinforcing pads were installed in accordance with CBI drawings. In all 50 inspection points, no deviations were reported.
2.2.8 Attribute 8 - Seismic Restraints An inspection was made to determine whether the seismic restraints had been installed and welded in accordance with the CBI drawing. In all 72 inspection points, no deviations were reported.
lh 2.2.9 Attribute 9 - Welding As a result of a review of the CBI design drawings it was determined that there were an estimated 1,100 welds to be reinspected on the boric acid and recycle holdup tanks. Each veld was subjected to several different types of inspection. Defining an inspection point as a specific type of inspection on one weld there were approximately 6,700 welding inspection points applicable to the 1,100 welds.
A total of 395 deviations were reported where the welds were not in compliance with manufacturer'y or code requirements. Deviations were reported for 11 pressure-retaining welds. The remaining 384 deviations for non-pressure-retaining welds occurred on items such as anchor bolt chairs and seismic supports. Of these 395 deviations 78 were determined to be notable and 317 vere determined to be insignificant. These 395 deviations are described in the following paragraphs.
Weld Configuration, Size, and Profile - Welds were inspected to determine whether the weld configuration and location were as required by the CBI drawings. In
, addition, weld size and length were measured to determine if they met or exceeded the minimum specified lf on the CBI drawings.
-]
Revision:- 1 Page 7 of 24 A
Q RESULTS REPORT ISAP VII.c (Cont'd)
Appendix 14 (Cont'd) 2.0 DISCUSSION OF RESULTS (Cont'd)
Out of the approximately 1,100 inspection points, 338 l deviations were reported. All but seven of the welds !
containing deviations were non-pressure-retaining welds, such as welds on anchor bolt chairs, seismic supports, and other support members. An evaluation of the seven deviations in pressure-retaining welds showed that the conditions were insignificant and did not affect the capability of the tanks to retain liquid or withstand seismic forces.
In two cases of pressure-retaining welds, a 1/4-inch fillet weld covering a full penetration veld attaching a 1-inch coupling to the tank shell and a 2-inch p coupling to a reinforcing pad were reported to be f 1/16-inch undersize. On the CBI design drawings, the couplings are shown to be welded to the tank shell and reinforcing pad with a full penetration veld for strength. The fillet weld is a cover veld added to prevent stress concentration and to provide an additional margin of safety for strength. No credit for the cover fillet was taken when calculating the stress in the tank shell under external loading.
Because there is no code requirement for reinforcing nozzles 2-inches in diameter or less, the additional weld metal is not required for reinforcement. In addition, the actual fillet welds are adequate to prevent stress concentration. Therefore, it was determined that the deviations had an insignificant impact on the structural integrity and fluid-retaining capability of the tank and were insignificant. If a similar deviation were to occur on the diesel fuel tanks it would have an insignificant effect on tank integrity.
In another case of a pressure-retaining weld, a 59-foot circumferential butt veld joining two 1/4-inch thick plates was not welded flush over a section 1/2 inch in length and 1/32 inch in depth. This deviation can be classified as a localized thin area. Localized thin areas that are totally reinforced by surrounding O metal are permitted by the ASME Code. General membrane stress in the vicinity of the localized thin area is unchanged from the stress in the surrounding full
Rsvision: 1 Pags 8 of 24 L
RESULTS REPORT l
ISAP VII.c (Cont'd) l Appendix 14 (Cont'd) 2.0 DISCUSSION OF RESULTS (Cont'd) i thickness metal. Stress concentrations due to irregularities in the metal surface at the thin area are of no concern because the operating conditions do not cause cyclical stresses. It was determined that the deviation had an insignificant impact on the structural integrity and fluid-retaining capability of I
the tank and was insignificant. If a similar deviation j
were to occur on the diesel fuel tanks it would have an E insignificant effect on tank integrity because the diesel fuel tanks have shells and heads that are significantly thicker than those required by the ASME Code.
In two other cases of pressure-retaining welds, the manufacturer's drawing specified a 3/8-inch fillet veld s
(~ attaching a 1/4-inch manway reinforcing pad to the tank shell. The existing fillet weld is 1/4-inch. A L
3/8-inch fillet weld is not possible on a 1/4-inch l
reinforcing pad. It has been determined that the manuf acturer's drawing is in error and should have specified a 1/4-inch fillet veld. Therefore these deviations were determined to be insignificant. These two deviations were sent to the Design Adequacy Program for information.
The remaining cases for pressure-retaining welds concerned two deviations for weld size on,two manway-reinforcing pads. In these cases, a 3/8-inch fillet weld attaching a manway neck to a reinforcing pad was found to have a throat 3/32-inch undersized and in the other case 1/8-inch undersize, both for approximately 25 percent of the circumference. The fillet weld is a cover veld added to prevent stress concentration and as an additional margin of safety for strength. No credit for the cover fillet was taken when calculating the stress in the tank shell under external loading. It was determined that the deviations had an insignificant impact on the structural integrity and fluid-retaining capability of the tank and were insignificant. If a similar O- deviation were to occur on the diesel fuel tanks it would have an insignificant effect on tank integrity.
Revision: 1 Page 9 of 24 RESULTS RE' ORT ISAP VII.c (Cont'd)
Appendix 14 (Cont'd) 2.0 DISCUSSION OF RESULTS (Cont'd)
In addition to the pressure-retaining weld deviations mentioned above, 254 deviations that were determined to be insignificant were reported for structural and attachment welds. The deviations reported for these velds were found on anchor bolt chairs, davit brackets, and the anchor bolt chair ring. The welds ranged in size from 1/4 inch to 5/16 inch, and the deviations were either for undersized legs or throats. In the worst case, a 28-inch-long weld on an anchor bolt chair was specified to be a .625 inch fillet weld. The reported deviation consisted of a .537 inch weld for 14 inches of length and a .581 inch veld for 4-3/4 inches of length. An evaluation showed that the stress e increased by 7-1/2 percent, and still remained below the Code-allowable stress. Therefore, it was determined that the 254 deviations had an insignificant impact on the structural integrity and fluid-retaining capability of the tanks. If a similar deviation were to occur on the diesel fuel tanks it would have an insignificant effect on tank integrity.
The 77 remaining deviations were reported for structural and attachment welds and were determined to be notable. The evaluations showed that the capacity of the welds decreased from 10 to 33 percent, but stresses still remained below Code-allowable stresses.
The worst case occurred where the weld capacity decreased 33 percent at a davit bracket-to-shell weld.
An analysis showed that the fillet weld was 1/16-inch undersized at various points around the davit bracket.
The Code-allowable stress is 12,000 psi and design stress with no deviation is 5143 psi. The weld size deviation increased the stress to 6847 psi, which does not exceed the Code-allowable stress. Therefore, it has been determined that this type of deviation did not have a detrimental impact on the structural integrity and fluid-retaining capability of the tanks because sufficient design margin remained. In considering the effect of similar deviations on the uninspected portion 9 of the population, it was determined that, based on the original design calculations, the most limiting case involved certain saddle support welds on the buried diesel generator fuel oil storage tanks in which the
Revision: 1 Page 10 of 24
) RESULTS REPORT ISAP VII.c (Cont'd)
Appendix 14 (Cont'd) 2.0 DISCUSSION OF RESULTS (Cont'd) stresses could exceed Code allowables if a 33 percent capacity reduction $s assumed. However, those calculations are extremely conservative in that the tanks are assumed to be free-standing for the purposes of calculating support veld stresses and a factor of 1.5 was applied in calculating the load resulting from seismic acceleration. No credit is taken for the resistance to bending and other stresses afforded by the surrounding earth. The structural welds in these buried tanks are therefore expected to retain adequate design margin. No adverse trend was identified.
Butt Weld Reinforcement - A small crown is normally added to the top surface of butt welds in order to strengthen or reinforce the veld. This practice is O recognized by the ASME Code, which contains dimensional standards for this butt veld reinforcement. The surface of the reinforcement of all butt-welded joints was inspected to determine whether the thickness was in compliance with design requirements. In approximately 50 inspection points, no deviations were reported.
Undercut - Welds were inspected to determine whether undercut at the toes of each weld exceeded 1/32 inch.
Out of approximately 2,200 inspection points, four deviations were reported where localized undercut occurred on one of the recycle holdup tanks. Except for one of the welds, all of these cases of undercut were on non-pressure-retaining structural and attachment welds.
An undercut of 3/64 inch deep by 1/16 inch long, occurring on a 59 foot circumferential seam weld at a point where the tank wall plates are 1/4 inch thick.
- was the only case of undercut found on a pressure-containing weld. This deviation can be classified as a localized thin area. Localized thin areas that are totally reinforced by surrounding metal are permitted by the ASME Code. General membrane stress in the vicinity of the localized thin area is unchanged from the stress in the surrounding full thickness metal.
O Stress concentrations due to irregularities in the metal surface at the thin area are of no concern because the operating conditions do not cause cyclical stresses. This deviation was determined to be insignificant.
Revision: 1 Page 11 of 24 RESULTS REPORT ISAP VII.c (Cont'6)
Appendix 14 (Cont'd) 2.0 DISCUSSION OF RESULTS (Cont'd)
If a similar deviation were to occur on the diesel fuel tanks it would have an insignificant effect on tank integrity because the diesel fuel tanks have shells and heads that are significantly thicker than those required by the ASME Code.
The remaining three cases of undercut occurred on structural and attachment welds. One of these was on a seismic bracket gusset plate, another was on a davit bracket, and the third was on the anchor bolt ring. In all of these cases, evaluations showed that even with the reduced area due to the observed undercut, the welds had adequate strength to withstand all loads specified. Therefore, it was determined that the O deviations were insignificant and did not have a detrimental impact on the structural integrity and fluid-retaining capability of the tank. If a similar deviation were to occur on the diesel fuel tanks it would have an insignificant effect on tank integrity.
Surfaces of Welds - Welds were inspected to determine whether the surfaces were sufficiently free of overlap, abrupt ridges, and ripples, so that proper interpretation of radiographic and other required nondestructive examinations (NDEs) could have been performed.
Out of approximately 1.100 inspection points, 48 deviations reported various surface irregularities. As all of the welds where these occurred were on support elements that have no Code or specification requirement for NDE except for visual examination, the deviations were determined to be insignificant.
Butt Weld Offset - Butt welds were inspected to determine whether the maximum offset of the finished weld was greater than that allowed by the ASME Code.
Of a total of 14 inspection points, three pressure-retaining veld deviations of excessive offset were reported.
ll
\
Revision: 1 Pags 12 of 24 j RESULTS REPORT ISAP VII.c (Cont'd) #
Appendix 14 (Cont'd) 2.0 DISCUSSION OF RESULTS (Cont'd) ;
The three offsets exceeded Code allowables by 3/32 j inch, and were observed on the lower plate walls of the recycle holdup tanks. An investigation of this condition revealed that the offset was caused by a plate that was thicker on the lowest wall section than.
on its adjoining plate. The intent of the offset standard is not to limit the offset of the plate surfcces, but to limit the offset of the center 11nes so that stresses due to offset overhang do not become excessive. The centerline offset for the as-built l installation wr0 calculated to be less than what the l Code-allowable offset would permit for a joint between two plates of equal thickness. It was therefore concluded that the installation is acceptable, and the O deviations were determined to be insignificant. As butt veld offset is not applicable to horizontal tanks, this condition could not occur in the diesel fuel oil tanks.
Fairing of Offsets - Offsets of all horizontal seam welds were inspected to determine whether they had been faired to at least a three-to-one taper over the width of the finished weld. In all 14 inspection points no deviations were reported.
Cracks, Lack of Fusion, and Crater Cra ks - Inspections were made to determine whether.the welds had visible cracks, lack of fusion, or crater cracks. Out of approximately 1,100 inspection points, two deviations were reported for lack of fusion. One case of lack of fusion was observed on a seismic support bracket veld, and the second on an anchor bolt chair veld.
- In the case of the seismic-support veld, lack of fusion was reported for 1-1/2 inches of a weld 20 inches long.
An evaluation showed that the capacity of the weld decreased 20.5 percent. The evaluation also showed, assuming the length of the weld with lack of fusion was missing, that the actual stress was 76 percent of the j code allowable. Because there was a 20.5 percent decrease in capacity this deviation has been determined to be notable. However, this deviation type will not have a significant impact on the structural integrity
Rsvision: 1 Pags 13 of 24 I
RESULTS REPORT ISAP VII.c (Cont'd)
Appendix 14 (Cont'd) 2.0 DISCUSSION OF RESULTS (Cont'd) or fluid-retaining capability of the tank. If a similar deviation were to occur on the diesel fuel tanks it would have an insignificant effect on tank integrity because all welds on these tanks have sufficient margin to accommodate this type of deviation. No adverse trend was identified.
In the case of the anchor bolt chair veld, lack of fusion was reported for 1/8 inch of a weld 28 inches long. Because the length of the portion of the weld where lack of fusion was reported was less than 0.5 percent of the total weld length, the effect on the stress level in the weld is negligible and within the accuracy of the original design calculation.
O Therefore, the deviation was determined to be insignificant and had no impact on the structural integrity and fluid-retaining capability of the tank.
Indications - Welds were inspected to determine whether there were indications of porosity that exceeded project acceptance standards. In approximately 1,100 inspet.M.on points no deviations were reported.
2.2.10 Attribute 10 - Base Material Defects Base material was inspected for surface defects within 2 feet on either side of weld joints. Of,approximately 2,200 inspection points, eight deviations were reported. The deepest gouge observed on the cutside surface of the tank was 3/16 inch deep and was located 1/2 inch above the elevation of the tank floor at a point where the plate is 1/2 inch thick. This deviation can be classified as a localized thin area.
Localized thin areas that are totally reinforced by surrounding metal are permitted by the Code. General membrane stress in the vicinity of the localized thin area is unchanged from the stress in the surrounding full thickness metal. Stress concentrations due to irregularities in the metal surface at the thin area O are not a concern because the operating conditions do not cause cyclical stresses. It was determined that !
the deviations were insignificant and had an insignificant impact on the structural integrity and
Rovision: 1 Pegs 14 of 24
(/-) RESULTS REPORT ISAP VII.c (Cont'd)-
Appendix 14 (Cont'd) 2.0 DISCUSSION OF RESULTS (Cont'd) fluid-retaining capability of the tanks. If a similar deviation were to occur on the diesel fuel tanks it would have an insignificant effect on tank integrity because the diesel fuel tanks have shells and heads that are significantly thicker than those required by the ASME Code.
2.2.11 Attribute 11 - Weld and Base Material Rust Each weld and base materials on both sides of the weld were inspected to determine whether there was rust-colored corrosion product build-up. In approximately 612 inspection points, no deviations were reported.
2.2.12 Attribute 12 - Grout An inspection was made to determine whether grout was installed as required. In all four inspection points, no deviations were reported.
2.3 Analysis of Documentation Review Results This section provides, by attribute, a discussion of the documentation review deviations and an analysis of the effect of the deviations on the ability of the dor, % .*ation to provide reasonable assurance of proper installat}on of the field-fabricated tanks.
2 . 3.1. Attribute 13 - Dimensional Verification A review of appropriate CBI inspection documentation was conducted to determine whether the following tank dimensions were recorded as being within the tolerances specified by CBI standards:
Tank diameter (for flat bottom tanks)
Shell diameter (for horizontal tanks)
Vessel head diameter (for horizontal tanks) l 1 i
q Rsvision: 1 Paga 15 of 24 m
(,) RESULTS REPORT ISAP VII.c (Cont'd)
Appendix 14 (Cont'd) 2.0 DISCUSSION OF RESULTS (Cont'd)
Temporary foundation levelness Tank bottom crown In all 20 review points, no deviations were reported.
2.3.2 Attribute 14 - Hydrostatic Testing Hydrostatic testing records were reviewed for presence of proper signatures. Test pressure recording charts for the horizontal tanks were reviewed to determine whether they were in accordance with CBI standards.
Test pressure recording is not required for the flat
,f s
-bottom tanks because they are not leak-tested with 9 pressurized water, but rather are tested by filling
' them with water. In all eight review points, no deviations were reported.
2.3.3 Attribute 15 - Record Drawings CBI record drawings identifying all tank seam welds were reviewed to determine whether the following information was included:
Joint identification Welders' identification ,
Welding procedure specification f
Record drawing table signoffs NDE procedure identification In 2,775 review points, 13 deviations were reported concerning welders' identification and NDE procedures identification.
CBI record drawings were reviewed to determine whether a welder's identification was given for each weld joint
_O shown on the record drawing. In two cases, weld seams were found for which there was no welders' identification. However, this is not significant
]
i
Revision: 1 Page 16 of 24 RESULTS REPORT ISAP VII.c (Cont'd)
Appendix 14 (Cont'd) 2.0 DISCUSSION OF RESULTS (Cont'd) because CBI has on file in the QA records center a list of all the welders who worked on these tanks showing what type of welding they were qualified to do. From this list it was determined that all of the welders that CBI had working on these tanks were qualified to make these welds.
The NDE procedure used was recorded on both a Leak Test Report form and the Record Drawing form. In the case of five welds the NDE test shown on the leak test report form did not agree with the NDE test indicated on the record drawing form. The leak test report form listed seven welds, two of which were tested to procedure SFT, and five of which were tested to procedure VB. Both of these tests are essentially the same in that they both are soap bubble tests that require the inspector to cover the weld with a film of soap, create an air presstre differential across the weld and then scrutinize the weld for bubbles which would indicate a flaw in the veld.
Procedure VB is known as a v.-e.uum box test. It is used to test butt welds where the two plates that have been joined form an even surface. In this test the veld is covered with a soap film, a box is placed over the weld, a vacuum is pulled in the box and the weld is inspected for bubbles. ,
Procedure SFT is known as a soap film test. It is used on velds where there is a step in the level of the two joined plates that would make it impossible to use the vacuum box described above. In this test a soap film l 1s applied to the weld, air pressure is applied behind the weld and the veld is examined for bubbles.
The five welds under discussion are butt welds, which were confirmed to be vacuum box tested. On the leak test report form the NDE procedure was correctly shown as VB for these welds. However, when the NDE procedure O was transferred to the record drawing form it was inadvertently shown as SFT and not VB as it should have been. These deviations were determined to be typographical errors of no significance.
i l'
Rsvision: 1 Page 17 of 24
/"
. k- N . RESULTS REPORT ISAP VII.c (Cont'd)
Appendix 14 (Con:'d) 2.0 DISCUSSION OF RESULTS (Cont'd)
In the case of six welds, the NDE procedure listed on the NDE report does not agree with the procedure shown on the record drawing table. Procedure VT-5 was listed on the NDE report whereas procedure VT-7 is listed on !
the Record Drawing table. The welds should have been inspected in accordance with procedure VT-7, which is a visual examination procedure for Class 2 and 3 atmospheric storage tanks. Procedure VT-5 is a visual examination procedure for nuclear component supports.
Finally, all welds that CBI made on the Field Fabricated Tanks were made in accordance with CBI's General' Welding Procedure Specification for the Shielded Metal Arc Process.
A review was made of the two examination procedures and the CBI's General Welding Procedure, and it was determined that all of the acceptance criteria listed in VT-7 are contained in either VT-5 or the CBI's General Welding Procedure. The only difference between the VT-5 and the VT-7 procedures is in the area of weld undercut, which was visually reinspected as described in Section 2.2.9. Therefore,'it is immaterial whether the welds were inspected to either VT-5 or VT-7. The six deviations were determined to be insignificant.
2.3.4 Attribute 16 - Welding Materials The CBI daily weld material distribution log was reviewed to determine whether the material issued to welders was correct in accordance with CBI's procedures. In all'555 review points, no deviations were reported.
2.3.5 Attribute 17 - Procedure Approval Each welding procedure specification, repair procedure specification, repair procedure, and NDE procedure was reviewed to determine whether it had been approved for use. In all 555 review points, no deviations were reported.
___a____ _ _ _ _ _ __. _ _ _ _ _ _ . - _ _ _ _ _
Rsvision: 1 Pags 18 of 24 RESULTS REPORT ISAP VII.c (Cont'd)
Appendix 14 (Cont'd) 2.0 DISCUSSION OF RESULTS (Cont'd) 2.3.6 Attribute 18 - Welder and Uelding Operator Qualifications 1
Appropriate drawings and documents were reviewed to determine whether a welder or welding operator was qualified to perform the welding operation indicated on the record drawings. In the 66 review points one deviation was reported.
1
! It was found that a welding report was dated after the Qualification Record indicates that the welder had terminated his employment. The weld completion date was 12/20/79 and the welder's termination date was 08/24/79, but the weld inspection date was 07/23/79
(
s
- when the welder was still on the job site. A conflict between the weld completion date and the welder's termination date is apparent. Since the weld inspection was completed when the welder was still on the job site it has been determined that a recording error was made by the inspection supervisor when transferring temporary records to the Record Drawing Table. Therefore, this deviation has been determined to be insignificant.
l 2.3.7 Attribute'19 - Nondestructive Examinations (NDE)
NDE reports were reviewed in order to determine whethee weld joints and NDE procedures were properly identified and whether the reviewer who signed the report was properly certified. In the approximately 1,665 review points one deviation was reported.
It was found that on one inspection report an improper NDE procedure revision number was recorded as being used. Because the correct revision, Rev. O, was correctly recorded in all other 1,664 review points and because there have not been any further revisions it was concluded that the proper test had been performed, and a recording error made by the person who recorded
,-7 the revision number. This deviation was determined to be insignificant.
)
l
Rsvision: 1 Pags 19 of 24 O
V RESULTS REPORT ISAP VII.c (Cont'd)
Appendix 14 (Cont'd) 2.0 DISCUSSION OF RESULTS (Cont'd) 2.3.8 Attribute 20 - NDE Personnel Certification There were 28 CBI NDE Performance Qualification and Certification forms checked to determine if they had been signed and dated by a certified individual and whether the. certification statement signature blanks i were signed, for a total of 84 review points.
Deviations were identified whenever evidence of certification to the established procedural requirements at the tima of inspection could not be found. However, as a recult of the ISAP I.d.1 evaluations of inspector certifications, it has been determined that such deviations do not necessarily p relate directly to actual inspector qualifications.
Consequently, the identified deviations were not relied -
upon (though they were considered) in determining !
inspector capability. Inspector qualifications were assersed for all inspectors (those with and without reported deviations) using the ISAP I.d.1 methodology.
It was discovered during the ISAP I.d.1 evaluation that the CBI QA procedures in effect at the time the tanks ,
were erected required the Welding and QA Supervisor to signoff for " final weld checked." This supervisor in turn had the authority to delegate this responsibility to a Welding Supervisor. Because the Welding I Supervisors were signing off for " final weld checked" i' instead of the persons who actually did the inspections, it is not now possible to determine from ,
this signature who actually did the inspections. j However, the ISAP I.d.1 evaluation of inspectors whose work affected this population concluded that all inspectors either were certifiable to applicable criteria at the time of inspections, were found to be capable of performing satisfactory inspections (including those with substantial positive evidence),
or were otherwise shown to be of no further concern.
Evaluation details are discussed further in the ISAP I.d.1 Results Report. It was therefore concluded that the inspections had been conducted by personnel who '
were capable of satisfactorily performing the work.
e Ravision: 1 Page 20 of 24 f%
-( ,) RESULTS REPORT ISAP VII.c (Cont'd)
Appendix 14 (Cont'd) 2.0 DISCUSSION OF RESULTS (Cont'd) 2.3.9 Attribute 21 - Seismic Restraint As-built Dimensions It was-the intent of the reinspection program to review as-built drawings to determine whether the elevation of the seismic brackets was shown within 1 inch of.the i location measurement recorded during reinspection. It was discovered, however, that no as-built drawings had been made. The seismic bracket elevations observed during reinspection were than compared with the elevations shown on the four design drawings (24 review pointe) and were found to be within the plus or minus 1-inch tolerance. It was therefore concluded that the lack of as-built information was of no consequence.
1
. () 2.3.10 Attribute 22 - Material Verification The material verification summary sheet was reviewed for proper signature and to determine whether the correct material was identified in accordance with CBI's procedures. In 548 review points, four deviations were reported regarding the material traceabil1Ly of a reinforcing pad and a shell insert on each of the two boric acid tanks. An investigation' determined that all four items were made from shell cut-outs from the recycle holdup tanks. All of the shell plates in the recycle holdup tanks ha/c material certificates on file, and therefore material traceability was verified. Therefore, these deviations were determined to be insignificant.
3.0 ROOT CAUSE AND GENERIC IMPLICATION No adverse trends or construction deficiencies were identified; a root cause and generic implication analysis is not required.
4.0 RECOPTINDATIONS
() No corrective action recommendations are required.
Revision: 1 Page 21 of 24 RESULTS REPORT ISAP VII.c (Cont'd)
Appendix 14 (Cont'd)
5.0 CONCLUSION
S Based on the findings of the reinspection and documentation reviews, there is reasonable assurance that the hardware in this construction work category is adequately installed in accordance with the design.
O t
Rsvision: 1 Page 22 of 24 RESULTS REPORT ISAP VII.c (Cont'd)
Appendix 14 (Cont'd)-
Table 14-1 Summary of Reinspection Results Field-Fabricated Tanks Deviation Classification Number of Inspection Number of Insigni- . Construction Attribute Points Deviations ficant Notable Deficiency
- 1. Nameplates 4 0 0. 0 0 Shell Top Angles 4 0 0 0 0 2.
- 3. Anchor Bolt Rings, O' Anchor Bolt Chairs, and Anchor Bolt Nut Tightness 180 0 0 0 0 Weld Seam Offsets 22 0 0 0 0 4.
- 5. Nozzle Type, Size, and Location 125 0 0 0 0-
- 6. Nozzle Insert 0 0 0 0 Reinforcement Plates 28
- 7. Nozzle and Manhole Reinforcing Pads 50 0 0 'O O 0 0 0 0
- 8. Seismic Restraints 72
- 9. Welding Configuration, Size, and Profile 1,100 338 261 77 0 Butt Weld 50 0 0 0 0 Reinforcement 2,200 4 4 0 0 Undercut
Prvisiot.: .1 Page 2 of 24
)
l .
RESULTS REPORT I
ISAP VII.c (Cont'd)
Appendix 14 (Cont'd)
Table 14-1 (Cont'd)
Deviation Classification Number of Inspection Number of Insigni- Construction Attribute Points Deviations ficant Notable Deficiency
- 9. Welding (Cont'd)
Surfaces of Welds 1,100 48 48 0 0 Butt Weld Offset 14 3 3 0 0 Fairing of Offsets 14 0 0 0 0 Cracks, Lack of Fusion, and Crater Cracks 1,100 2 1 1 0 Indications 1,100 0 0 0 0
- 10. Base Material Defects 2,200 8 8 0 0
- 11. Weld and Base Material Rust 612 0 0 0 0
- 12. Grout 4 0 0 0 0 TOTALS 9,979 403 325 78 0.
(Approximately 10,000) ,
o l
4
Rsvision: 1 Paga 24 of 24 RESULTS REPORT ISAP VII.c (Cont'd)
Appendix 14 (Cont'd)
Table 14-2 Summary of Documentation Review Results Field-Fabricated Tanks Deviation Classification Number of Review Number of Insigni-Attribute Points Deviations ficant Notable CD.
- 13. Dimensional Verification 20 0 0 0 0
- 14. Hydrostatic Testing 8 0 0 0 0
- 15. Record Drawings 2,775 13 13 0 0
- 16. Welding Materials 555 0 0 0 0
- 17. Procedure Approval 555 0 0 0 0
- 18. Welder and Welding Operator 0 0 Qualifications 66 1 1
- 19. Nondestructive Examinations (NDE) 1,665 1 1 0 0
- 20. NDE Personnel Certification * * * * *
- 21. Seismic Restrsint As-built Dimensions 24 4 4 0 0
- 22. Material Verification 548 4 4 0 0
- 0 TOTALS 6,216 23 23 0 (Approximately 6,200)
O
- QC Inspector Certification review points and deviations are excluded from totals and the deviations are not classified because the results of ISAP I.d.1 indicate that these deviations do not relate directly to actual inspector qualifications.
Rsvision: 1 Pags 1 of 58
/~'T RESULTS REPORT J ISAP VII.c (Cont'd)
Appendix 15 HVAC Ducts and Plenums 1.0 REVIEW PROGRAM IMPLEMENTATION 1.1 Construction Work Category Description The construction work category of heating, ventilating, and air conditioning (HVAC) ducts and plenums is comprised of seismic Category I sheet metal duct sections, with accessories, for all safety-related HVAC duct systems installed by the HVAC contractor, Bahnson Service Company (BSC).
1.2 Population Size and Sample Selection For this construction work category, a population of approximately 7,100 items was identified as QC-accepted as of the date of this reinspection effort. Reinspection and 7x documentation reviews were performed for a total of 112
( ,) accessible items. Of these, 72 first sample items were randomly selected to ensure that at least 60 inspections or documentation reviews of each attribute were performzd, with one exception as noted in Section 1.3. An additional 31 second sample items were randomly selected to ensure that at least 60 safe-shutdown hardware items were reinspected or reviewed. ,
1.3 Attributes Selected Sample items were reinspected for the following attributes:
Attribute 1 - Location 1
Attribute 2 - Configuration The configuration attribute is proportionally sampled in combination with the HVAC Equipment Installation population, however, at least 60 inspections of this i
attribute were perf ormed in this population.
1 l Attribute 3 - Longitudinal seam type Attribute 4 - Gasket installation Attribute 5 - Bolt location 7-m
Attribute 6 - Bolt installation Attribute 7 - Welding l . }
Revision: l'
- Page 2 of 58 RESULTS REPORT
< ISAP VII.c (Cont'd)
Appendix 15 (Cont'd) 1.0 REVIEW PROGRAM IMPLEMENTATION (Cont'd)
Sample item documentation was reviewed for the following.
attributes:
Attribute 8 - Pressure test-Attribute 9 - Duct section fabrication inspection l l
Attribute 10 - Weld procedure application Attribute 11 - Welder qualification Attribute 12 - QC inspector certification Attribute 13 - Material traceability
%.) Material traceability was required only through receipt inspection for items with a structural function; all receipt inspection reports for material were reviewed for seismic. duct or structural load carrying items in the population.
Attribute 14 - Welding inspection Attribute 15 - Touch-up galvanizing inspection 2.0 DISCUSSION 07 RESULTS 2.1 Summary of Results _
For reinspection, 199 Deviation Reports were issued describing 1,367 deviations. Approximately 100,000 inspection points were encountered in performing the reinspection. See Table 15-1 for results of the reinspection.
For documentation review, 72 Deviation Reports were issued describing 138 deviations. Approximately 1,100 review points were encountered in performing the documentation reviews. See Table 15-2 for results of the documentation reviews.
O
i Ravision: 1 Page 3 of 58 RESULTS REPORT j f,
l i
ISAP VII.c (Cont'd)
Appendix 15 (Cont'd) 2.0 DISCUSSION OF RESULTS (Cont'd)
Documentation Review deviations related to inspector certification have not been included in the aforementioned totals. As discussed in Section 2.3.5., it has been determined that these deviations do not necessarily relate directly to actual inspector qualifications. These deviations were considered in the ISAP I.d.1 evaluation together with supplemental information as necessary to determine actual inspector qualifications. The ISAP I.d.1 conclusions are summarized in Section 2.3.5.
In the seven reinspection attributes of this population, ther'e were no construction deficiencies, and no adverse trends were identified. An unclassified trend was identified related to hardware installed without approved design details and is discussed in Section 2.4.1. In addition, a trena was identified related to inadequate post installation
{
-(-- modifications end is discussed in Section 2.4.2.
In the seven documentation review attributes of this population (excluding QC inspector certification), 101 of the documentation review deviations were determined to be insignificant, and 37 were notable.
2.2 Analysis of Reinspection Results This section provides, by' attribute, a discussion of the reinspection deviations, an analysis of the effect of the deviations on the functional capability of the dpct section items and an analysis for the presence of trends. The function of each duct section is to provide flow capability and maintain the structural integrity of the HVAC duct system in which it is located.
Several attributes had deviations that could result in duct leakage, such as missing or loose bolts, missing or loose instrument test hole caps, and missing portions of gaskets.
This air leakage could be into or out of the duct system.
Ventilation systems that collect airborne radioactivity operate under negative pressure and any air leakage would be into the duct. Whenever radioactivity levels are of concern,
/~'\ this air is delivered under negative pressure to filtration k-- units. Ducts downstream of the filtration units operate under l
__..____________________.____w
Rsvision: 1 Pego 4 of 58
}
%J RESULTS REPORT ISAP VII.c (Cont'd)
Appendix 15 (Cont'd) 2.0 DISCUSSION OF RESULTS (Cont'd) positive pressure but any out-leakage would be filtered air.
Therefore, minor deviations leading to small leakage usually would be of no concern, and have negligible effect on the 1 function of the ventilation system. I 2.2.1 Attribute 1 - Location In 220 inspection points for location, one deviation was reported where the item inspected was not located-within the required 6 inches of the specified location on the duct installation drawings. In this deviation a horizontal section of duct, approximately 7 feet in length, was located 4 inches above its acceptable location in elevation and consequently increased the t'~s length of the vertical duct by 10 inches (the deviation
(,) plus the tolerance). The increased duct length was analyzed to increase the duct system pressure loss by 0.03 percent. The corresponding weight of the duct section increased 7.2 percent which reduced the duct support system design margin by 1.0 percent.
Therefore, this deviation was determined to be insignificant. No adverse trend was identified.
2.2.2 Attribute 2 - Configuration Proportional sampling was used in combination with the HVAC Equipment Installation population (HVIN) for this attribute. The following is a discussion of the deviations that relate to this population. In addition, an evaluation of the combined effects of the deviations for this attribute ir. Appendix 15 and 16 is also included herein.
In approximately 940 inspection points for configuration, 59 deviations were reported.
Fifty-eight were reported where the configuration, either inside or outside the duct section, did not comply with the fabrication drawings or installation details for accessories and one deviation was reported where the configuration could not be verified due to fs the lack of installation requirements.
()
Ravision: 1 !
Page 5 of 58
<-ms
( ) RESULTS REPORT w.)
ISAP VII.c (Cont'd)
Appendix 15 (Cont'd) 2.0 DISCUSSION OF RESULTS (Cont'd)
Configuration inspection focused on the ability of the HVAC contractor to fabricate a duct section to the requirements of a shop fabrication drawing, e.g., a Material Requisition Form, a design drawing or design change documentation. The important common characteristics checked under configuration were flow area and length which verifies the item's size and geometry. If a shop fabrication drawing required other special features, such as companion or reinforcing angles, those features (angle size and spacing) were also verified. Turning vane installation, sheet metal gage and duct gage were typically inaccessible and were verified by documentation review; they are discussed in Section 2.3, Analysis of Documentation Review Results.
p_
I k~ I Configuration also included the inspection of any accessories, e.g., access doors, instrument test hole caps, grilles, registers or diffusers, attached to the duct. This portion of the reinspection tested the HVAC contractor's ability to install accessories to the various dimensional or configuration requirements of the installation details for duct accessories. Only !
when an accessory appeared on a sample was it l inspected. The ability of the HVAC contractor to i construct duct sections in accordance with the overall configuration requirements has been assessed by analyzing the data collected under this attribute.
Deviations reported against configuration'are described in the following paragraphs.
Nineteen deviations were reported where 16 instrument test hole caps were loose and three were missing, the worst case being four loose on one item. Eighteen of these deviations were reported on negative pressure exhaust ducts and one on a positive pressure exhaust duct. An instrument' test hole has an area of less than 1 square inch, which is small when compared to the cross-sectional area of a duct. These test holes are occasionally located on a duct run and are used to L)
Rsvision: 1 Pegs 6 of 58
(~N RESULTS REPORT ISAP VII.c (Cont'd)
Appendix 15 (Cont'd) 2.0 DISCUSSION OF RESULTS (Cont'd) insert instruments into the duct to measure air flow races. A conservative analysis was made of the effects of leaking air on system performance due to these caps !
being missing. Each case showed that the deviations had a negligible effect on the air flow within the duct section where they occurred, with the worst case representing approximately 3 percent of the duct's ,
cross sectional area. Therefore, the deviations were determined to be insignificant. The largest number of test hole caps observed was 22 on a 68 by 32 inch duct section. If all of these caps were missing. . this would represent 1 percent of the duct's cross sectional area.
The worst case in the remainder of the population is estimated to be approximately 3 percent. Based upon 7- g this and the previous discussion on duct leakage at the
() start of Section 2.2., e.g., any in-leakage of air will be filtered and any out-leakage of air will be of air previously filtered, if similar deviations were to occur in the uninspected portion of the population, they too would have a negligible effect on air flow and would not be of concern regarding the release of airborne radioactivity. Therefore, these deviations )
also would be insignificant and no adverse trend was identified.
In addition to the discussion in Section 2.2 regarding concerns about duct leakage, battery room exhaust systems handle traces of hydrogen gas. The one deviation concerning test hole caps on a positive pressure exhaust duct occurred on a battery room exhaust system. This leakage would occur in areas outside of the battery room. An analysis concluded
- that the air leakage out of the duct has a negligible effect on the environment into which it is leaking. '
This is due to low concentrations of hydrogen being further diluted because the leakage occurred into welleventilated areas. Therefore, the deviation was determined to be insignificant. There are no other l l areas of the plant that are designed to allow production of concentrations of hydrogen gas like the battery rooms. Should similar deviations occur in the
[A-/) uninspected portion of the population, it is expected
Rsvision: 1 Page 7 of 58
( RESULTS REPORT ISAP VII.c (Cont'd)
Appendix 15 (Cont'd) 2.0 DISCUSSION OF RESULTS (Cont'd) that they also would be insignificant because the concentration of hydrogen would be negligible and leakage from other battery room exhaust systems would always be into a well-ventilated areas. No adverse trend was identified.
Six deviations were reported.where the duct section length was not within the 1/4 inch tolerance of its required length. Duct sections vary in length from approximately 10 to 48 inches. The range of these deviations is from 2-1/2 inches shorter to approximately 4 inches longer than. required. Three of these deviations occurred in small supply sections to diffusers and the remaining three were in straight 7- sections of duct. Evaluations showed that these deviations had a negligible effect on the system performance or structural integrity. Therefore, the deviations were determined to be insignificant. Based on the evaluation in Section 2.2.1 for a location deviation involving a duct section 10 inches too long, this deviation type would always be insignificant and' no adverse trend was identified.
- One deviation was reported concerning an internal tie rod that was installed where it was not required. The rod was installed in a 32-inch by 68-inch duct under negative pressure. An evaluation showed that the installed tie rod would not have any adverse effect on the strength or on the system performance of the item.
Tie rods are added inside large ducts (over 60 inches wide) that are subjected to internal positive pressure.
Therefore, the deviation was determined to be insignificant. There are no duct sections in this population that are large enough to require tie rods; therefore, dislocation of a tie rod in the uninspected portion of the population would not have any detrimental effect on the performance of an item. No adverse trend was identified.
Three deviations were reported against one air flow g
extractor because the extractor was mounted to the duct
Rsvision: 1 Page 8 of 58
/^),
( RESULTS REPORT x~/
ISAP VII.c (Cont'd)
Appendix 15 (Cont'd) 2.0 DISCUSSION OF RESULTS (Cont'd) j without reinforcing angles, spacer plates and associated bolting as required by its installation detail. An analysis of the three deviations was made that showed that without the reinforcing angles, the resultant stresses at the extractor increase approximately 455 percent. Therefore, the deviations were determined to be notable. The resulting stresses are still below the maximum allowable, by approximately 68 percent, and the functional capability of the item- i will not be impaired. A bounding analysis was performed that showed similar results should a similar deviation exist for any of the remaining extractors in the population. Since the extractors were the only special features utilizing a comparable angle configuration for installation, no adverse trend was
[}A/ identified.
One deviation was a collar that was missing from a plate flange used to install a flexible connection.
This particular arrangement occurred on a small, 8-inch diameter, molded flexible connection to a volume damper. An evaluation showed that the plate flange transmits all loads through the connection and performs the same function as the collar. Therefore, the deviation has no effect on the function of the item and it was determined to be insignificant. Should this alternate design have been used on similar connections in the uninspected portion of the population, the function of the connection would be fulfilled.
Therefore, no adverse trend was identified.
Two deviations were reported where the welding and the type of connection were not in accordance with the duct fabrication procedure because an angle-type flange was used for mating duct sections rather than the hem-type flange specified. These deviations occurred due to a modification that involved the shortening of the item in order to install a volume damper. The original hem flange was cut off and, for ease of fabrication, the
/ . angle-type flange was installed.- It is not likely that
( if an item inscalled with angle type flanges was shortened, a hem-type flange would be installed in its
I Revision: 1 Page 9 of 58 RESULTS REPORT ISAP VII.c (Cont'd) l i
Appendix 15 (Cont'd) 2.0 DISCUSSION OF RESULTS (Cont'd) place as a hem flunge is usually prepared in a shop and requires a bending machine that folds the sheet metal into the hem flange configuration. An evaluation showed that the angle flange was stronger than a hem flange for connecting duct sections and that the deviation had no effect on the functional performance of the item. Therefore, these deviations were determined to be insignificant. Should similar deviations occur in the uninspected portion of the population, it is expected that they also would be insignificant as an angle type flange is always stronger than a hem-type flange. Therefore, no adverse trend was identified.
Five angles were reported to be installed on a plenum section that were not detailed on the fabrication drawing. An evaluation showed that the added angles would enhance the strength of the plenum and have no detrimental effect on the structural integrity or system performance of the plenum. Therefore, the deviations were determined to be insignificant. Should similar deviations occur in the uninspected portion of the population, it is expected that they also would be insignificant as the added angles would not have any detrimental effect on the structural integrity of the item to which they are added. Therefore, no adverse trend was identified. .
Four deviations were reported where the configuration of four duct connections to plenums was not in accordance with the standard duct connection detail.
It was reported that the bolts connecting the ducts to the plenums were installed through the mating angles, gasket and the sheet metal exterior of the plenum. The standard duct connection detail does not include the sheet metal exterior of plenum. An evaluation of each case showed that the structural or leak-tightness aspects of the connection are not affected by extending the sheet metal exterior of the plenum so that the
Ravision: 1 Page 10 of 58
( '
( .
RESULTS REPORT ISAP VII.c (Cont'd)
Appendix 15 (Cont'd) 2.0 DISCUSSION OF RESULTS (Cont'd) bolts go through it, and the ability of the item to perform its function is maintained. Therefore, the deviations were determined to be insignificant. Should this type of deviation occur in the uninspected portion of the population, the item affected would still be capable of performing its intended function. No adverse trend was identified.
One deviation was reported where a plenum angle brace ended 1/4 inch away from the centerline of a field joint instead of the typical 2-inch dimension as shown on the fabrication drawing. Further review of the installation showed that all members and reinforcing
/~'g support steel were rigidly secured together with the
'Nj required welding and that the resulting joint was a totally integrated unit. The deviation has no effect on the ability of the plenum to perform its intended function and was determined to be insignificant.
Should this type of deviation occur in the uninspected portion of the population, the item affected would still be capable of performing its intended function.
No adverse trend was identified.
Three deviations were reported where the collar supporting a plenum exhaust register was formed as one piece folded out from the sheet metal exterior skin of the plenum rather than being a separate item attached to the plenum as detailed by the fabrication drawing.
An evaluation showed that the installed condition (collar prepared from the plenum skin) was equivalent in function and strength when compared to a separate piece attached to the plenum skin. Therefore, the reported deviations would not prevent the collar and register or the plenum from performing their intended function and the deviations were determined to be insignificant. Should similar deviations occur in the uninspected portion of the population, they also would be insignificant since a collar prepared from the sheet f-~
metal exterior skin is equivalent in function and
( strength to a separately attached collar. Therefore, no adverse trend was identified.
Rsvision: 1 Prgs 11 of 58 RESULTS REPORT
[G )/
ISAP VII.c (Cont'd)
Appendix 15 (Cont'd) 2.0 DISCUSSION OF RESULTS (Cont'd)
Related to the above three collar installations, three deviations were also reported because the arrangements of nuts, bolts and gasket were not in accordance with the typical collar installation used when the collar is a separate item. As the installed arrangement eliminates the need for the collar, there is no requirement for these items. The deviations will not affect the ability of the formed collar or plenum to perform their intended function. The deviations were-determined to be insignificant. Shoulds similar deviations occur in the uninspected portion of the populatiot., they also would be insignificant since having a collar arrangement that eliminates the need for nuts, bolts and gasket does not reduce the ability
("'}
\,/
of the item to perform its intended function.
Therefore, no adverse trend was identified.
One deviation was reported where an exhaust register was not connected to a duct by the required 5-inch collar as detailed in the accessories installation procedure. The register is horizontal and attached directly to the end of a horizontal duct. An evaluation was made that showed that the installation had the minimum 3 inch straight section necessary to ,
mount the register and that the deviation has no effect {'
on system performance or on the structural integrity of the connection. The deviation was determined to be insignificant. Should similar deviations' occur in the '
uninspected portion of the population, they also would be insignificant since a register, satisfactorily installed in the end of the duct without first being attached to a collar, will not reduce the ability of an ,
item to perform its intended function. Therefore, no l adverse trend was identified.
One deviation was reported where a transverse splice weld was included on a duct section but not detailed on the fabrication sheet for the item. An evaluation showed that the item had to be extended approximately g
Ravision: .1 Page 12 of 58 RESULTS REPORT ISAP VII.c (Cont'd)
Appendix 15 (Cont'd) 2.0 DISCUSSION OF RESULTS (Cont'd) 1-1/2_ inches to accommodate an approved change and-thus required a short filler piece welded into the system.
The evaluation also concluded that the properly performed weld will not reduce the structural strength of the sheet metal duct skin nor impair.the ability of the item to perform its intended function. The deviation was determined to be insignificant. Should similar deviations occur in the uninspected portion of the population, it is expected that they also would be insignificant since properly adding a short filler piece in the process of implementing an approved design change will not prevent the item from performing its intended function. Therefore, no-adverse-trend was identified.
Six deviations were cases where joints had companion-angle flanges fabricated from 3/16-inch rather than 1/4-inch thick angles. An analysis was made that concluded that under design loading conditions the 3/16-inch angle system was stronger than the 5/16-inch diameter bolting system installed in these angles and that the overall strength of the joint is not limited by the installed angle size. Therefore the deviations had no effect on the structural-integrity of the duct section and are insignificant. Since the 3/16-inch angle is the minimum size specified in the contractor's duct fabrication procedure, no deviations involving smaller sizes would be expected to exist in the uninspected portion of the population. Should similar deviations exist in the uninspected portion of the population it has been determined that a construction deficiency would be unlikely to result based on the results of the analysis performed. Therefore, no adverse trend was identified.
One deviation was a 12-by-8-inch duct section that was specified on the Material Requisition Form to be fabricated out of 18 rather than the 16-gage material (0.0516 versus 0.0635 inches thick, respectively) that O
4
- _ _ _ - - - _ - _ - _ _ _ - _ - _ _ _ _ _ _ _ . - _ _ - . _ _ __ 1
Rsvision: 1 Pags 13 of 58
() RESULTS REPORT ISAP VII.c (Cont'd)
Appendix 15 (Cont'd) 2.0 DISCUSSION OF RESULTS (Cont'd) was specified in the contractor's duct fabrication procedure for that size duct in that application. The reduction in material thickness used in this instance versus-that gc.nerically specified was approximately 20 percent, which resulted in the bending stress being 21 percent of the maximum allowable and the deviation was determined to be notable. However, an evaluation showed that this deviation had a negligible effect on the function of the duct section. The item inspected-is 12-3/4 inches long andHis located at the start of an exhaust system, which has an internal negative pressur'e of approximately 1/2-inch water gauge (0.02 psi) below the ambient pressure. Similar sized supply ducts in the same area are specified to be 18-gage material.
/~'i Therefore, under similar seismic loading conditions the
\s / item inspected of 18-gage material will maintain the structural integrity of the duct. The 12-inch duct is used only at the start of an exhaust system and will experience only low internal negative pressure. In fact, the 18-gage material was at one time permitted in exhaust systems up to 12 inches on the largest side.
The pressure loads are insignificant compared to seismic loads. Therefore, this deviation was determined not to have any effect on the functional performance of the item. Should similar deviations exist in the uninspected portion of the population, it has been determined that a construction deficiency would be unlikely to result based upon thy evaluation performed. Therefore, no adverse trend was identified.
! One deviation occurred where the flange leg perpendicular to the duct was 1-1/2 inches wide, rather than 2 ir.ches wide and the bolts were 3/8 inch rather than the specified minimum 1-3/16 inches from the duct skin. An evaluation showed that the narrower flange places the duct connecting bolts closer to the duct skin, improving the overall strength of the joint. A narrower flange could also have the effect of locating the bolts too close to the edge of the flange.
Deviations reporting this condition are discussed in
('
( Section 2.2.5 Bolt Location. Analyses of these deviations showed that bolts could be located 1-5/8 inches from the duct of a 2-inch flange or 3/8 inches
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ . _ _ _ _ _ _ _ _ _ __ _ m
Ravision: 1 Pegs 14 of 58 RESULTS REPORT ISAP VII.c (Cont'd)
Appendix 15 (Cont'd) 2.0 DISCUSSION OF RESULTS (Cont'd) from the edge of the flange. At this location the remaining metal from the bolt hole to edge of flange was stronger than the bolt. Applying the 3/8 inch dimension to this reported condition that the bolts were 3/8 inch rather than the specified minimum 1-3/16 inches from the duct skin,~would result in a minimum acceptable flange width-of 3/4 inch. As the minimum flange size reported is 1-1/2 inch, it is not likely that flanges less than 3/4 inch wide exist in the uninspected portion of the population. The narrower flange of 1-1/2 inch could also reduce the effective full face required gasket width of 2 inch to 1-1/2 inch. It has been evaluated that even a 3/4 inch wide gasket is sufficient to seal the duct. Therefore, this deviation is considered structurally insignificant and t the duct system will perform its intended function.
Should similar deviations exist in the uninspected portion of the population, it has been determined that a construction deficiency would be unlikely to result based upon the evaluation performed. Therefore, no adverse trend was identified.
In summary, the 58 previously reported deviations describe configurations that were not in accordance with the established fabrication details. In all cases, the resultant configuration was equal in function and performance when compared to the original configuration. Some cases provided added, quality to the installation by reducing the number of components in the design or by providing a structurally stronger installation. Additional configuration deviations may exist in the uninspected portion of the population; however, based upon the previous evaluations, it is expected that these conditions would not result in a construction deficiency. Therefore, no adverse trend was identified.
One deviation reported that configuration could not be verified as installation requirements for an item were not available. This involved the method for attaching g
a 90 degree branch circular duct to a rectangular main
' supply air duct. This circular duct was bolted directly to the sheet metal rectangular duct in a manner not detailed in the contractor's Duct
Revision: 1 l
Page 15 of 58
() RESULTS REPORT ISAP VII.c (Cont'd)
' Appendix 15 (Cont'd) 2.0 DISCUSSION OF RESULTS (Cont'd)
Fabrication Procedure and therefore this deviation remains unclassified. Deviations similarly resulting from the lack of applicable design information were identified for several attributes. This type of deviation led to the establishment of an unclassified trend that is further discussed in Section 2.4.1.
A review of the deviations discussed in Appendix 16 which relates to the aspects of configuration specifically associated with HVAC equipment ir_stallation was also performed. The deviations that relate to fire damper sleeves, fire damper blade catches, modulation dampers and fire damper sleeve retaining angles were all evaluated individually to be O insignificant and no trends were identified. A review of these deviations in combination with the deviations presented herein for this attribute did not identify any additional findings or trends associated with this attribute. In cddition, the unclassified trend resulting from a lack of applicable design information 13 act related to the types of configuration deviations sascussed in Appendix 16. Therefore, no additional trends relating to HVAC equipment installation were identified.
2.2.3 Attribute 3 - Longitudinal Seam Type In 100 inspection points, one deviation was reported where the type of longitudinal seam used was different from that designated in the duct fabrication procedure.
The longitudinal seam was a mechanical type (Pittsburgh Lock) rather than a welded-type seam. A mechanical-type seam is produced during the fabrication process of a duct section when the opposite edges of a sheet metal sheet are brought together and interlocked by overlaping and folding the edges together. The strength of the mechanical seam was 2quivalent to that of a welded seam based upon a previously performed test conducted by a consultant to the HVAC contractor and reviewed by CPRT personnel. The same test also O verified the pressure integrity of the mechanical seam.
The deviation does not affect leakage, had no reduction in the structural integrity of the item, and
Revision:- 1 Page :16 of 58 C~- RESULTS REPORT ISAP VII.c (Cont'd)
Appendix 15 (Cont'd) 2.0 DISCUSSION OF RESULTS (Cont'd) ,
i was determined to be insignificant. .Should this type of deviation occur in the uninspected portion of the _
population, they also would not result in a construction deficiency as the strength of the mechanical seam is' equivalent to that of a welded seam.
Therefore, no adverse trend was identified.
2.2.4 Attribute 4 - Gasher. Installation In 190 gasket inspection points, 43 deviations were reported where the gasket did not cover the entire 2-inch flange surface between mating sections of duct as required by the duct fabrication procedure.
Thirty-nine deviations were minor gasket depressions ranging from 1/8-inch to 1/2 inch deep from the edge of the flange up to 22 inches in length. There was also one deviation of 1/8 inch depression for 2 foot 9 inches and assorted deviations of depressions of indeterminate depth up to 22 inches long and various points of depression up to 1 inch deep. An analysis showed that in most cases sufficient portions of the 2-inch wide gasket material remained to prevent any air leakagc. In the' remaining cases, analyses showed that even if that deviating portion of the gasket was missing, an insignificant amount of air would leak (less than 1 percent of the' total air flow). These 39 deviations had no.effect'on system performance and therefore were determined to be insignificant. Should similar deviations occur in the uninspected portions of the population, it is expected that they also would be insignificant as they would not result in any air l 1eakage. Therefore, no adverse trend was identified.
Three deviations occurred where a 1/2-inch length of gasket in each of two exhaust ducts was missing for the l.
full width of the flange and a 1 inch leakage path
[, existed in a gasket of a supply duct. The 1 inch l leakage path was conservatively analyzed to allow 0.6
,~
percent of the total air flow to leak. This represents a negligible air flow and will have no effect on system performance. These deviations were determined to be insignificant. Should similar deviations occur in the uninspected portions of the population, it is expected
Ravision: 1 Pags 17 of 58 RESULTS REPORT ISAP VII.c (Cont'd)
Appendix 15 (Cont'd) 2.0 DISCUSSION OF RESULTS (Cont'd) that they also would be insignificant as they would result in a negligible air flow either leaving or entering the system, which would have no effect on system performance. Therefore, no adverse trend was identified.
One deviation was a 22-inch length of gasket that was missing from an air supply duct. This represented one full side of a 22-by-12-inch duct. The flow area of the missing gasket represents approximately 1 percent of the duct's cross sectional area.- The resulting air leakage out of the duct would have a negligible effect on the performance of the system. Therefore, the deviation was determined to be insignificant. If this leakage, representing 1 percent of the duct flow area, occurred in the uninspected portion of the population, this amount of leakage would also have a negligible effect on the performance of the system. Therefore, no adverse trend was identified.
A further check of documentation was made that showed that the above joint had been inspected and satisfactorily leak tested at one time in accordance with the HVAC contractor's inspection and pressure test procedures. The adjacent duct section was subsequently removed, shortened to accommodate the installation of a volume damper, reinstalled, and walked down prior to startup. Documentation verifying subsequent leak testing of this item is not available. See Section 2.4.2 for further discussion of this deviation, leading to a recommendation for improvement.
2.2.5 Attribute 5 - Bolt Location In approximately 4,500 inspection points, 206 deviations were reported for bolts that were not located in accordance with the duct fabrication procedure. The deviations were 118 cases of improper bolt spacing, 85 cases of improper bolt hole to duct distance and three cases in which bolt location could O* not be verified due to the lack of installation requirements.
I
1 i
Revision: 1 Page 18 of 58
[ '
RESULTS REPORT ISAP VII.c (Cont'd)
Appendix 15 (Cont'd) 2.0 DISCUSSION OF RESULTS (Cont'd)
An additional evaluation due to the number of deviations reported for this and other attributes is summarized in Section 2.4.3.
Bolt Spacing Of the 118 deviations for improper bolt spacing, 62 deviations were cases where the center-to-center distance between bolts on flanges was greater than the 4-1/16 inches allowed by the duct fabrication procedure. The largest reported spacing was 5-3/4 inches occurring on a circular duct flange having 8 bolts installed. The larger bolt spacing results in fewer bolts being installed than if the bolts were O installed on the required 4-inch centers. Analysis showed that at this distance, the bolt loading increases 26 percent and the 8 bolt spacing deviations were determined to be notable. The analysis also showed that the bolts were loaded to only 7 percent of their design capacity. Therefore, the deviations will not prevent the bolts from performing their safety function. Leakage.was also calculated not to occur.
Should similar deviations occur in the uninspected portion of the population, it is expected that they also would be insignificant as the increased spacing between bolts produces an insignificant increase in the bolt stress which would not affect the capability of the bolts to perform their intended function.
Therefore, no adverse trend was identified.
The remaining 56 deviations were for bolting on four identical plenum access doors, 18-by-24 inches, that were to have bolts installed on 4-inch centers. This spacing would have required 24 bolts, but only 17 were installed. Each door had three bolts a maximum distance of 9-1/2 inches apart on the side of the door with two hinges. The remaining 14 bolts had a reported maximum spacing of 5 inches. The function of the bolts is to hold the access door to the center portion of the
((
sheet metal plenum skin. The diameter of the bolts is 5/16 inches. The major plenum seismic loads are transferred through its angle framing and at the corners of the duct skin. Therefore, the loads at the access door bolts can be considered to be only due to
Revisions 1 Pags 19 of 58
( RESULTS REPORT ISAP VII.c (Cont'd)
Appendix 15 (Cont'd) 2.0 DISCUSSION OF RESULTS (Cont'd) the weight of the door. As the weight of the 18-by-24 inch door is small compared to structural loads normally resisted by identical bolts at structural connections, the fact that the 17 bolts are not on 4-inch centers will not alter the capability of these bolts to hold the door to the plenum during the design loading conditions. Based upon the typical 4-inch spacing requirement, 24 bolts should have been installed. Although only 17 of.the 24 were installed, '
the deviations were determined to be insignificant, as the bolts were evaluated to be so lightly loaded (less than 5 percent of allowable) when compared to their design capacity. Should similar deviations occur in the uninspected portion of the population, it is expected that they also would be insignificant as the A weight of any access door is small, resultiLg in negligible bolt loads when compared to the structural loads normally resisted by identical bolts at structural connections. Extrapolation of this deviation is limited to access doors as the 9 1/2-inch spacing deviations only occurred in the location of the door hinges. This situation is unique to doors and would not occur in the bolting of two duct sections.
Therefore, no adverse trend was identified.
Bolt Hole to Duct Distance The 85 deviations were reported for bolts,on a 2-inch flange that were not installed 1-3/16 to 1-5/16 inches from the duct as required by the duct fabrication procedure. The distances ranged from 3/8 to 1-13/16 inches. An evaluation showed that for bolts closer than 1-3/16 inches the structural integrity is improved. For the cases where bolts were installed farther than 1-5/16 and up to 1-5/8 inches from the duct, a negligible reduction in the structural integrity of the flange occurs. Sixty deviations reported bolts up to 1-5/8 inches from the duct. An evaluation of the condition where the bolt is 1-5/8 O
Revision: 1 Page 20 of 58 RESULTS REPORT
}
ISAP VII.c (Cont'd)
Appendix 15 (Cont'd) 2.0 DISCUSSION OF RESULTS (Cont'd) inches from the duct showed that the remaining metal from the bolt hole to edge of flange was stronger than the bolt. Similar deviations in the uninspected portions of the population are therefore not likely to affect the structural integrity of the flange. The 34 deviations with bolts up to 1-5/8 inches from the duct were determined to be insignificant and no adverse trend was identified.
The remaining 25 deviations reported bolts located further than 1-5/8 inches from the duct and occurred on one edge of a duct connection to a plenum. All 25 bolts were improperly located too close to the edge of the flange. Further investigation showed that the line G of tolts was not parallel to the edge of the flange and that five of these bolts were so close to the edge that they are considered ineffective. The remaining 20 bolts were located sufficiently far away from the edge of the flange so that they could develop their full load-carrying capacity. These bolts are part of a 76-bolt connection, and an evaluation that assumed that the five bolts were inef fective indicated that a stress increase in the remaining 71 bolts is less than ten percent. The evaluation also showed that the remaining 71 bolts are adequate to perform their function and the reported deviations will not affect the functional performance of the bolted connection.
The 25 deviations were determined to be insignificant.
It was also observed that two other lines of bolts of this same duct connection were parallel to the edge of the flange but 1-1/2 inches away from the duct. As these were not further than 1-5/8 inches from the duct as discussed in the previous paragraph, they have a negligible effect on the structural integrity of the duct flange and are insignificant. The plenums are rigidly supported from the building structure and each duct connection to the plenum is rigidly supported close to the connection. The spans between the connections are small, resulting in small seismic loads being transmitted to these bolted connections. This 9 results in overall connection loads that are less than ten percent of code allowables. Based upon this evaluation and an analysis of more severe situations
i Rsvision: 1 j Pags 21 of 58 RESULTS REPORT ISAP VII.c (Cont'd)
Appendix 15 (Cont'd) 2.0 DISCUSSION OF RESULTS (Cont'd) presented in Section 2.2.6 Attribute 6, Bolt Installation, under the heading Missing Bolts, should similar deviations occur in the uninspected portion of the population, it has been determined that a construction deficiency would be unlikely to occur.
Therefore, no adverse trend was identified.
Lack of Installation Requirements A total of three deviations were reported where the proper installation of bolts could not be verified as the bolting requirements were not available. Identical deviations were reported against the attachment of three control room air supply diffusers to their air
/] supply ducts. Each diffuser was secured to its
( /. circular supply duct with at least three, 1/4-inch bolts; however, the requirements for this bolting arrangement were not documented. Applicable drawings provide no indication of the type of connection intended and therefore, these deviations remain unclassified. Deviations similarly resulting from the lack of applicable design information were identified for several attributes. This type of deviation led to the establishment of an unclassified trend that is further discussed in Section 2.4.1.
2.2.6 Attribute 6 - Bolt Installation Of the approximately 10,500 inspection points for the presence of the nuts, volts and washers, and for bolt size and bolt tightness, 55 deviations were reported where the 1 installation was not in accordance with the requirements of the duct fabrication procedures. These deviations are discussed below and include 14 missing bolts, 17 loose bolts, and 24 missing washers. No deviations were reported for improper bolt size.
Missing Bolts Fourteen bolts were reported to be missing. One was missing from one duct connection flange, and 13 from a
'O f
Revision: 1 Page 22 of 58
- D .
\s RESULTS REPORT ISAP VII.c (Cont'd)
Appendix 15 (Cont'd) 2.0 DISCUSSION OF RESULTS (Cont'd) second duct connection flange. The evaluation of the one missing bolt on one connection is discussed under the heading of " Loose Bolts."
The 13 missing bolts were on the entire row of the long side of a 48-by-8-inch duct flange. As 13 out of 32 bolts that comprise the connection are missing, the deviations were determined to be notable. An analysis of this condition showed that the bolt tensile stress was approximately 50 percent of the maximum allowable stress. The top horizontal flange was also analyzed as supporting a bolt at each end. This analysis showed the bending stress to be at 34 percent the allowable stress. As the maximum allowable stresses in the bolting or duct flange were not exceeded, the x_-
structural integrity of the duct system was not impaired. Leakage was also determined to be insignificant for this duct section. Analyses were also made for the identical condition of a missing row of bolts on a flanged joint elsewhere in the uninspected portion of the population. The approach taken was to review the various duct locations of the plant and in each case to analyze the worst case of duct cross sectional height-to-width ratio against the design loading conditions for that location. There were no cases where a construction deficiency would occur. Therefore, no adverse trend was 1,dentified.
A further check of documentation was made which showed that the above joint had been inspected and satisfactorily leak-tested at one time. The adjacent duct section was subsequently removed, shortened to accommodate the installation of a volume damper, reinstalled, and walked down prior to turnover.
Documentation verifying subsequent leak testing of this item is not available. See Section 2.4.2 for further discussion of these deviations, leading to a recommendation for improvement.
O
(~ /
1 t
I
- - _ _ _ _ - - _ l
Rsvision: 1 Page 23 of 58 f%
(m- [ RESULTS REPORT ISAP Vll.c (Cont'd)
Appendix 15 (Cont'd) 2.0- DISCUSSION OF RESULTS (Cont'd)
Loose Bolts A total of 17 bolts were reported to be loose, with the worst case being eight of 34 bolts on one duct joint.
The remaining nine loose bolts occurred as follows:
three items had one bolt loose, one item had three of 28 bolts loose, and one item had three of 20 bolts loose.
Of the nine loose bolts, the item with three of 20 loose bolts was analyzed as the worst case. The analysis showed that stresses in some of the remaining (
bolts would be increased by 75 percent. Therefore, the ts three deviations were determined to be notable.
Although three bolts were loose, the remaining 17 bolts.
were installed properly. The evaluation of this condition demonstrated that the remaining stresses in the bolts are still below the allowable stress and the remaining tight bolts are adequate to maintain the structural' integrity of the joint. Should similar deviations occur in the uninspected portion of the population it was determined that a construction deficiency is unlikely to occur. This conclusion is based upon the bounding analysis performed under the previous section, Missing Bolts, which concluded that even under the limiting conditions of duct cross se:cional geometry and seismic loadings, an item with en entire row of bolts missing would not tesult in a construction deficiency. Therefore, no adverse trend was-identified.
The remaining six of the previous nine deviations were determined to be insignificant as the six loose bolts were distributed among four different items (three items had one loose bolt and one item had three of 28 loose bolts. The' item with three loose bolts had two loose bolts at one end of the duct and one loose bolt at the other end of the duct.) All the other bolts of the connections were properly installed and were more i
than adequate to support the design loading conditions, as there was only a negligible increase in stress due to the one or two loose bolts in each connection.
Therefore, no adverse trend was identified.
l 4
Rovision: 1 Page 24 of 58
<~') \
I V RESULTS REPORT ISAP VII.c (Cont'd)
Appendix 15 (Cont'd) 2.0 DISCUSSION OF RESULTS (Cont'd)
The duct connection with eight of 34 loose bolts also.
had one missing bolt as reported in the previous section, Missing Bolts. An analysis of this condition showed a significant increase in bolt stress but only to approximately 60 percent of the allowable stress.
The analysis also showed that 20 bolts would adequately support the design loading conditions. Twenty-five out of 34 bolts are properly installed. Although the reported deviations have a notable effect on the ~
remaining bolts, the duct and its flanges will perform their intended safety function. The nine deviations were classified as notable. Should similar deviations occur in the uninspected pr: tion of the population it was determined that a construction deficiency is
[s" unlikely to o: cur. This conclusion is based on the bounding analysis described in the previous section, Missing Bolts,-which concluded that even under the severest conditions of duct cross sectional geometry and seismic loadings, an item with an entire row of bolts missing would not result in a construction deficiency. Therefore, no adverse trend was identified.
A further check of documentation was made which showed that the above joint had been inspected and satisfactorily leak tested at one time. The adjacent duct section was subsequently removed, shprtened to j accommodate the installation of a volume damper, reinstalled, and walked down prior to turnover.
Documentation verifying subsequent testing is not available. See Section 2.4.2 for further discussion of the deviations associated with this joint, leading to a recommendation for improvement.
Missing Washers A total of 24 washers were reported missing, consisting of six lockwashers and 18 flat washers. One lockwasher was missing from each of four items, and two of 22 O lockwashers were missing from one item. Two case of
Revision: 1 Page 25 of 58 O RESULTS REPORT ISAP VII.c (Cont'd)
Appendix 15 (Cont'd) 2.0 DISCUSSION OF RESULTS (Cont'd) missing lockwashers were evaluated as though its associated bolt was missing. For the item with two lockwashers missing, the resulting bolt stress was only 33 percent of the allowable stress. In both cases the effect on the resultant stresses in the remaining bolts was insignificant. The remaining three items were evaluated by a generic analysis which concluded that the lockwasher missing from each installation would not result in the inability of the affected item to perform its intended safety function. The deviations were determined to be insufficient.
The missing 18 flat washers included eight missing from four instrument test holes and ten missing from one O flexible connection. The missing flat washers on the instrument test hole caps have negligible effect on the strength of the connection as the weight of the cap is negligible and the bolt heads all have adequate bearing based on the design hole size. The bolt holes on this particular flexible connection had to be elongated in the field to match-drill the connection between the flexible joint and the duct section. Field drilling in this instance is allowed per the HVAC contractor's Ductwork Installation Procedure. The elongated holes reduced the bearing area of the bolts, but an adequate bearing area remained to produce a tight joint as the gasket was observed to be compressed. As the flexible connection is made f rom flexible material, it does not function as a structural member. Thus the flexible connection flanges do not transfer seismic loads. Due to the light loading at this connection, the reduced bearing area of the bolts was evaluated to have negligible effect on the load-carrying capabilities of the bolts, and the deviations were determined to be insignificant. A bounding analysis was performed that extends similar deviations into the uninspected portion of the population. The result of this analysis showed that under the worst seismic loading conditions, the flanged joint O connection remains structurally adequate.
Based on the insignificance and low frequency of the deviations for missing washers, no adverse trend was identified.
Revision: 1 Page 26 of 58
,/
- RESULTS REPORT ISAP VII.c (Cont'd)
Appendix 15 (Cont'd) 2.0 DISCUSSION OF RESULTS (Cont'd) 2.2.7 Attribute 7 - Welding In approximately 80,400 inspection points, 1,002 deviations were reported where welds were not in accordance with the associated design documents. These included 117 location, 583 length, 134 size, 16 profile, 26 fusion, 79 undercut, 3 crack, and 44 touch-up galvanizing deviations. No deviations were reported against the categories of craters, porosity, overlap and surface slag. In addition, three deviations were reported for cases where welding could not be verified due to the lack of installation requirements.
f-~s k,-) An additional evaluation due to the number of deviations reported for this and other attributes is summarized in Section 2.4.3.
Location In approximately 6,700 inspection points, 117 welds were reported to be missing or not located in accordance with the generic weld detail requirements of the duct fabrication procedure.
Three duct sections were reported with 44 corner welds missing. The function of the corner welds is to prevent cracks from developing in the sheet metal duct due to the fabrication process for a hem-type flange.
During the fabrication of a hem-type flange, a corner piece is welded into the void created when the edges of a duct are folded up 90 degrees to form the hem flange.
j Corner welds are installed to prevent cracks from developing in the sheet metal duct during this fabrication process due to the excessive heat developed i when the corner pieces are welded in place or as a result of folding up of the sheet metal. The corner l
welds do not transmit any structural loads. An l rT examination of the installation showed no evidence of sheet metal cracks or cut propagation at the duct I
I ') corners that had the missing corner welds. The weld I
material used to connect the corner pieces to the l
flanges also functions as a cohesive joint for all components meeting at the corner joint. Therefore. l some weld support exists at the corner that would l
R2 vision: 1 Page 27 of 58
( p)-
- - RESULTS REPORT l
ISAP VII.c (Cont'd)
Appendix 15 (Cont'd) 2.0 DISCUSSION OF RESULTS (Cont'd) prevent any subsequent cuts or cracks from developing in the sheet metal even though the corner veld is missing. As these welds are not relied upon to transmit any load, the absence of these welds will not impair the functional performance of the item. The deviations were determined to be insignificant. Should similar deviations and conditions occur in the uninspected portion of the population, it is expected that they would also be insignificant based upon the above analysis. Therefore, no adverse trend was identified.
Forty-three welds were reported not to match the weld detail prescribed by fabrication procedures. This
/}
(_- includes 24 welds that were not in conformance with the weld detail for connecting intersecting reinforcing angles. However, these welds were in conformance with the weld detail for connecting intersecting companion angles. One case was reported where a welded joint was installed rather than a bolted joint. One case was i reported where one weld at the end of each of three plenum angles was not installed per a fabrication drawing note. Evaluations of these cases were performed, and in no case did the installed welding configuration result in reduced structural integrity or degraded leak-tightness of the installed item. These deviations were insignificant. .
Deviations were also reported for welds not matching weld details in nine locations on one plenum where the lip of a sheet metal skin was secured to the structural frame with intermittent welds rather than being continuously welded as detailed by the fabrication drawing. This same condition existed in six locations on another plenum section that was reinspected.
Evaluations of these plenum welds were performed that showed that a continuous seal weld was installed on the other side of the sheet metal plenum skin. In no case t
did the installed welding configuration result in
( ,g) reduced structural integrity or degraded leak-tightness of the assembled item. All the above deviations were therefore determined to be insignificant, and will not prevent the items from performing their intended functions.
1 Rsvision: 1 Pegs 28 of 58
- (g RESULTS REPORT ISAP VII.c (Cont'd)
Appendix 15
. (Cont'd) 2.0 DISCUSSION OF RESULTS (Cont'd)
If there are similar deviations (i.e., welds not matching weld details) in the uninspected portion of the population where welds are not precisely in accordance with the design details, they too would not affect the ability of the item to perform its intended function and would also be insignificant. This is based on the conclusions that the alternate welding arrangements did not produce any reduction in structural capacity or design margin, and there was no degradation in leak tightness. Therefore, no adverse trends were identified.
Fourteen stitch welds were located on a duct section.
However, the fabrication details for the item did not b show any requirement for those welds. An investigation showed that the welds were located along the common I
edges of two 90 degree elbows that combine air flows to form a' single duct. An evaluation was performed showing that the added welds would not reduce the structural strength of the items, and the deviations were determined to be insignificant. Should deviations of this type occur in the uninspected portion of the population, they too would be insignificant as the added welds do not reduce the structural strength or functional capab111ty of the item. Therefore, no i adverse trends were identified. l One deviation was reported where one of 10 required stitch welds was missing from a companion angle l connection for a circular duct. The stitch welds work !
in conjunction with a continuous internal circumferential weld. Inspection was made on !
individual welds. This missing weld reduces the calculated circumferential weld length by 2.8 percent.
An evaluation of the deviation was performed, guided by the maximum permissible tolerance in the acceptance criteria of the Visual Weld Acceptance Criteria procedure, which permits an 8.3 percent reduction in O
Rsvision: 1 Page 29 of 58 i i
RESULTS REPORT ISAP VII.c I (Cont'd)
Appendix 15 (Cont'd) 2.0 DISCUSSION OF RESULTS (Cont'd) weld length due to conservative design margins on allowable stresses. Since the missing weld length is well below this percentage, the missing weld will not prevent the angle from performing its intended function. This deviation was determined to be insignificant. Should this type of deviation occur in the uninspected portion of the population, the item affected would still be capable of performing its function as the total installed weld length would be within the acceptance tolerance as permitted by the Visual Weld Acceptance Criteria procedure. Therefore, no adverse trend was identified, f-'
Ten areas were reported where stitch welds were not
(,,)g spaced on the required 4-inch centers. Five of these were on a 14 by 10 inch duct with the maximum spacing between two welds being 5-1/8 inches and five of these were on a 20 by 26 inch duct with 5-1/4 inch spacing between welds. In both cases, the total installed length of weld was found to be greater than the required total length of weld. Based upon this, the '
reported deviations will not have any effect on the item's ability to perform its intended function and were determined to be insignificant. Should similar deviations occur in the uninspected portion of the population, they also would be insignificant as the total installed length of weld would exceed the required total length of weld. Therefore', no adverse trend was identified.
An item 48 inches long was reported to have two, 1-3/4 !
inch portions of its two longitudinal seams not welded at the intersection of the circumferential reinforcing angles. An evaluation was made that showed that the localized discontinuity of the longitudinal seam weld was not critical for the item's structural integrity.
An analysis concluded that should this lack of a seam weld result in air leakage, the maximum amount of O
+
Rsvision: 1 Page 30 of 58 '
V RESULTS REPORT ISAP VII.c (Cont'd) l Appendix 15 (Cont'd) 2.0 DISCUSSION OF RESULTS (Cont'd) leakage would be 0.037 percent of the total air flow, which is a negligible amount. Therefore, these deviations will not have any effect on the ability of the item to perform its function and were determined to be insignificant. Should similar deviations occur in the uninspected portion of the population, it has.been determined that a construction deficiency would be unlikely to occur, based upon the above analysis.
Therefore, no adverse trend was identified.
Length In approximately 6,700 inspection points, 583 deviations were reported for welds that did not meet the length requirements of the duct fabrication
.( -
procedure.
There were 449 reported underlength corner welds on 68 duct flanges. The shortest length reported was 1/4 inch. The minimum allowable length is 7/8 inches. The corner welds do not transmit any structural loads.
Their function is to prevent cracks from developing in the sheet metal duct due to the fabrication process as previously discussed for the characteristic " location" of this Attribute 7, welding. The weld material used to connect the corner pieces to the flange provides a cohesive joint for all components meeting,at the corner joint. Therefore, some weld support exists at the corner that would prevent any subsequent cracking of the sheet metal even though the corner weld is 1/4 inch in length. As these welds are not load-bearing, the deviations were determined to be insignificant. Should similar deviations occur in the uninspected portion of the population, it is expected that they also would be insignificant, i.e., without significant cracks, based upon the above evaluation. Therefore, no adverse trend was identified.
I u
1 l
Ravision: 1 Paga 31 of 58 f i
1
' '; RESULTS REPORT J
ISAP VII.c (Cont'd)
Appendix 15 (Cont'd) 2.0 DISCUSSION OF RESULTS (Cont'd)
There were 82 reported underlength stitch velds on connecting angles to the sheet metal duct skin on 18 items. The shortest length reported occurred in two cases where one of a series of 1-1/2 inch stitch welds was 1/2 inch long. All reported items with underlength stitch welds were evaluated. In all cases, even though j' some stitch welds were underlength, based on the requirement of 1-1/2 inch welds on 4 inch centers, the total length of installed weld material was greater than the total required weld length. Hence, there was no reduction in capacity as a result of these deviations, and the deviations were determined to be insignificant. Should similar deviations occur in the uninspected portion of the population, they also would fN be insignificant as the total length of installed weld
() ,
material would be greater than the total required weld length. Therefore, no adverse trend was identified.
On one plenum section, 52 intermittent sheet metal duct skin-to-structure welds were reported not to be 1-inch long each and not to be spaced on 9-inch centers. The welds varied in length down to 1/4 inch and varied in spacing down to 1-1/2 inches between welds. Analyses of these deviations were performed, and even though some stitch welds were underlength, based on the requirement of 1-inch long welds on 9-inch centers, the total length of installed weld material was greater than the total required weld material. The deviations will not cause any structural degradation or affect the functional integrity of the plenum and were determined to be insignificant. Should similar deviations occur in the uninspected portion of the population, they also would be insignificant as the total length of installed weld material would be greater than the total required weld length. Therefore, no adverse trend was identified.
Size fg In approximately 6,700 inspection points for welds
! ; checked for proper size, 134 deviations were reported. ,
%./
Revision: 1 j Page 32 of 58 O
(,,/ RESULTS REPORT ISAP VII.c (Cont'd)
Appendix 15 (Cont'd) 2.0 DISCUSSION OF RESULTS (Cont'd)
A total of 120 of the 134 deviations concerned angle-to-angle welds that were undersized a maximum of 1/16 inch from the 3/16 inch requirement of the duct fabrication procedure. Twelve of the 134 deviations concerned groove welds that were flush and two of the l I
134 were for groove welds underfilled for 3/8 inches.
An analysis was performed on each connection that showed that the cross-sectional area of the insta13ed welds was greater than the minimum weld size required to support the maximum loading of the angle. These analyses included undersized companion-angle-to-companion-angle welds, reinforcing-angle-to-reinforcing-angle welds and the circular duct companion f--
angle weld that joins the two ends of the companion
(,)g angle together after the angle is bent to its circular configuration. The highest loaded welds are those that join 2-by-2-by 3/16-inch companion angle to companion angle. In all these cases, the installed weld area was greater than the cross sectional area of the angle and the installed weld was larger than t'aa theoretical required weld size. The reinforcing-angle-to-reinforcing-angle welds only support the duct internal pressure loads while the previously mentioned companion angle welds support both internal pressure and seismic loads, e.g., the axial load from internal pressure is 47 pounds in one direction when compared to 2,410 pounds for seismic loads. For the circular d.uct angle weld, the forces are transferred through the additional circumferential welds that attach the angle to the duct section, and only a small load goes through the butt weld that joins r.he ends of the companion angle. In all cases, it was shown that the conservatism of the welding requirements are such that the deviations will not cause any significant structural degradation, or af f ect the functional integrity of the duct joint.
Therefore, these deviations were determined to be insignificant. Should similar deviations occur in the uninspected portion of the population, they also would be insignificant as the cross-sectional area of the installed welds would also be greater than the minimum g
's weld size required to support the design loading.
Therefore, no adverse trend was identified.
Rsvision: 1 ,
Page 33 of 58 I 7'~%
\s_) RESULTS REPORT ISAP VII.c (Cont'd)
Appendix 15 ,
(Cont'd) 2.0 DISCUSSION OF RESULTS (Cont'd)
Profile In approximately 6,700 inspection points, 16 deviations concerned weld profile.
The profile of two of a series of 32, 1-1/2 inch stitch welds was not in accordance with the required profile sketch included in the procedure for visual inspection of the welding. The installed weld profiles were described as being concave rather than straight. A conservative analysis, which considered no strength contribution by these welds, concluded that the reduction of the effective installed throat area was r~S 3.1 percent indicating a minimal decrease in strength.
(_,) When compared to a 10 percent reduction allowed by the Visual Weld Acceptance Criteria procedure for welds smaller than 1/8-inch, these deviations have a negligible effect on the structural capacity of the item and were determined to be insignificant.
The remaining deviations were for 14, 3/16-inch stitch welds 1-1/2 inch long attaching a reinforcing angle to the duct skin. These 3/16-inch welds did not meet the required profile sketch included in the procedures. An analysis was made assuming that, due to the reported concavity, only 1/8 inch of the installed weld size was effective. This resulted in a reduction in weld throat i
area of 7 percent. The Visual Weld Inspection Criteria I l procedure allows an 8.3 percent reduction in throat area for these weld lengths. Therefore, the reported deviations will not prevent the angle from performing its function and were determined to be insignificant.
Should similar deviations on profile exist in the uninspected portion of the population, it is expected that they also would be insignificant as analyses i similar to those described above would show that the installed welds would meet the acceptance criteria included in the Visual Weld Acceptance Criteria
( ,g procedure assuring significant margin for the typically l ('}
low expected loads. Therefore, no adverse trend was identified.
l l
l
L Rsvision: 1 Page 34 of 58
()
.m RESULTS REPORT ISAP VII.c (Cont'd)
Appendix 15 (Cont'd) 2.0 DISCUSSION OF RESULTS (Cont'd)
Fusion In approximately 6,700 inspection points, 26 deviations concerned lack of fusion on welds. In most cases the lack of fusion occurred on one of a series of welds that made up a connection. The worst situation reported six groove welds on one item with five of these located at one end of the item. The five groove welds are used to install three corner pieces of a hem flange. Corner pieces are used to fill the void !
created when the edges of a duct are folded up 90 degrees to form the hem flange. Corner pieces are secured to each corner by two groove welds. They are also drilled to accept the corner bolts. An evaluation considered that the three corner pieces secured by the
(\-} deviating welds were not effective and that the bolts in the corner pieces were also ineffective. As there are 16 bolts that make up the total flange and three bolts are assumed to be ineffective, the deviations are notable. The evaluation shows that even if all four corner bolts were missing, the resulting bolt stresses would be less than 18 percent of the allowable stress, and that the remaining welds could amply support the design loads. Although the deviations are notable, if these occurred in the uninspected portion of the population it has been determined that a construction deficiency would be unlikely to occur. This is based upon the low stress values in the remaining bolts even if all corner groove welds that support the corner bolts should fail. Therefore, no adverse trend was identified.
The remaining 20 deviations were evaluated by assuming that the portion of weld where the lack of fusion occurred was ineffective. In all cases the remaining welds were more than adequate to support the design loading conditions and the reported deviations had a negligible effect on the load-carrying capacity of the item. Therefore, the deviations were determined to be
,O insignificant. Should similar deviations occur in the
'N_, uninspected portion of the population, it is expected that they also would be insignificant because the unaffected lengths of welds would be more than adequate to support the design loading conditions. Therefore, no adverse trend was identified.
Revision: 1 Pcgs 35 of 58 7
() RESULTS REPORT ISAP VII.c (Cont'd)
Appendix 15 (Cont'd) 2.0 DISCUSSION OF RESULTS (Cont'd)
Undercut In approximately 13,400 inspection points, 79 deviations were identified for undercut on welds on 19 items. In most cases, undercut was reported in some portion of one or two welds on a particular item, with the worst cases being 18 of 165 welds on one plenum and 21 of 217 welds on a second plenum. Of these, the worst case of undercut appeared on a structural angle and was reported to be less than 5 percent, on an equivalent basis, of the total base metal. An evaluation shows that the Visual Weld Acceptance Criteria procedure permits up to 8.3 percent undercut.
Therefore, it was concluded that the reported deviations will not significantly reduce the
[)
\- ' load-carrying capacity of the joint, and the deviations were determined to be insignificant. Should similar deviations occur in the uninspected portion of the population, it is expected that they also would be insignificant as there would also be a negligible reduction in the capacity of the welded joint.
Therefore, no adverse trend was identified.
Cracks In approximately 6,700 inspection points, three welds were reported to have cracks. One of two, 2-inch groove welds that attaches the corner piece of a hem flange was reported to have a 1/4-inch long crack. The function of the weld is to attach one of two sides of the corner piece to the item through which a bolt is subsequently installed. An analysis was performed that showed that even if the entire weld should fail and the bolt became inef f ective, the remaining bolts would transfer all design loads to the adjacent duct supports. Therefore, the deviation was determined to be insignificant.
One deviation was reported where one, 1/2-inch long
/~S corner weld had a 1/4-inch long crack. As previously
(_-) discussed, the corner weld does not add to the structural strength of the connection, and the deviation was determined to be insignificant.
1 I
l Ravisions 1.
r Page 36 of 58-RESULTS REPORT 5__
ISAP VII.c (Cont'd)
Appendix 15 (Cont'd) 2.0 DISCUSSION OF RESULTS (Cont'd)
One deviation was reported where a 2-1/2 inch angle had either a 3/16-inch long crack or saw cut cnt one of two 2-inch welds that connect it to another angle. An examination and evaluation showed that the deviation was a saw cut in the wraparound portion of a fillet weld which is not required for strength. As such, the remaining welds will support all design loads. The deviation was determined to be insignificant.
The three weld crack deviations were determined'to be insignificant. Should similar deviations occur in the uninspected portion of the population, it is expected that they too would be insignificant based upon the evaluations that were performed. Therefore, no adverse i trends were identified.
Additionally, it was noted that cracks were reported on some plenum welds. An investigation and evaluation of this condition showed that the cracks occurred on tack welds located between structural welds. As tack welds precede the placement of structural welds they are not relied upon to support structural losds. .Therefore, cracks in tack welds can not affect the ability of the plenum to perform its intended function.
Touch-up Galvanizing In approximately 6,700 inspection points, 44 deviations l were reported in instances where galvanizing removed during the welding process was not replaced with touch-up galvanized paint. Each deviation was evaluated based upon the thickness of the sheet metal and the expected corrosion rates for the environment in which the item is located.
Twenty-one deviations involved the lack of touch-up galvanizing on sheet metal. Each case was evaluated, and the predicted amount of sheet metal reduction was notable, with the worst case being approximately 29
- percent of the original sheet metal thickness. The evaluations concluded that, nevertheless, these small, localized areas of sheet metal reduction would not affect the structural integrity of the duct. Leakage would not result as at least 71 percent of the material remains to contain the low differential pressure.
Revision: 1 Page 37 of 58
() RESULTS REPORT ISAP VII.c (Cont'd)
Appendix 15 (Cont'd) 2.0 DISCUSSION OF RESULTS (Cont'd)
Therefore, the deviations will not affect the functional performance of the item. Although the deviations are notable, if similar deviations were to occur in the uninspected portion of the population it is unlikely that a construction deficiency would occur.
This conclusion is based upon the small areas involved, IEEE?
the expected low rates of corrosion and the determination that no holes would develop.
The remaining 23 deviations involved the lack of touch-up galvanizing on welds joining angles. An analysis was performed that concluded that the amount of area reduction due to corrosion over the plant life would be less than 3 percent. The Visual Weld
() Acceptance Criteria procedure used in this analysis allows an 8 percent reduction in area; therefore, the deviations will not affect the items' capability of performing their function and were determined to be insignificant. Should similar deviations occur in the uninspected portion of the population, it is expected that they also would be insignificant as the amount of corrosion would not reduce the material thickness below the amount of reduction permitted by the Visual Weld Acceptance Criteria procedure.
Based upon the conclusions of the above analyses, no adverse trends were identified. .
Lack of Installation Requirements Three deviations, tabulated under " location", were reported for cases where the proper installation of welds could not be verified since applicable welding requirements were not available for plenums. The same deviation was reported against three plenum sections.
Welding details for plenums appear either on the field of the drawings, in the general notes of the plenum fabrication drawings or in the contractor's Duct Fabrication Procedure. Since there was no basis for evaluation, these deviations remained unclassified.
O However, deviations similarly resulting from the lack of applicable design information were identified for several attributes. This type of deviation led to the establishment of an unclassified trend that is further discussed in Section 2.4.1.
Rsvision: 1 Page 38 of 58 (m)
L.,/ RESULTS REPORT ISAP VII.c (Cont'd)
Appendix 15 (Cont'd) 2.0 DISCUSSION OF RESULTS (Cont'd) 2.3 Analysis of Documentation Review Results This section provides, by attribute, a discussion of the documentation review deviations and an analysis of the effect of the deviations on the ability of the evidence collected to provide reasonable assurance that the duct sections are properly fabricated and installed.
2.3.1 Attribute 8 - Pressure Test In 113 review points, 34 deviations were reported where pressure test documentation as required by the HVAC specification could not be located. Of these, 27
/~ occurred on HVAC systems outside the control room and
(,}f seven occurred on HVAC systems within the control room.
Given the frequency of these deviations, systems in the uninspected portions of the population may lack verification that pressure tests for leaks have been conducted on the latest installed configuration. These 34 similar deviations constitute a trend.
Since required documentation could not be located, these 27 deviations are notable. The lack of pressure test documentation for items outside the control room area alone is not sufficient to imply faulty hardware nor to indicate that inadequate quality of construction exists. The reinspection effort provided sufficient assurance, however, that even without a pressure test, the fabrication and erection of systems outside the control room are acceptable and that, as installed, they will be able to perform their required safety functions. Thus, the collected evidence demonstrates that the ducts and plenums outside the control room ;
will perform their intended safety function, and the missing documentation is of no consequence.
Seven deviations were reported on items located within the control room. Several reinspected duct sections
(,--) located in the control room had not been pressure-
\- tested in their latest installed configuration. These duct sections are described as follows: the duct section reported to have one full side of its gasket
_________-____ n
Revision: 1 Page -39 of 58
.r f RESULTS REPORT ISAP VII.c (Cont'd)
Appendix 15 (Cont'd) 2.0 DISCUSSION OF RESULTS (Cont'd) missing as discussed in Section 2.2.4; the duct section reported to have 13 missing bolts as discussedlin Section 2.2.6 and the duct section reported to have ,
one missing and eight loose bolts as discussed in Section 2.2.6. If a pressure test had been performed on these duct sections, these deviations of missing gasket, missing bolts and loose bolts would most likely have been detected and corrected at that time. These seven deviations reporting lack of pressure test documentation are determined to be notable as they provide supporting evidence identifying hardware that does not meet installation requirements. Based on the evaluations summarized in the referenced sections, however, none of the deviations were safety-significant nor were any adverse trends identified. Thus, the
[}
N- ' installed hardware has been determined to be acceptable. Further discussions on the lack of pressure test documentation are included in Section 2.4.2, leading to a recommendation for improvement.
2.3.2 Attribute 9 - Duct Section Fabrication Inspection In 127 review points to verify that duct section fabrication inspections were performed encompassing non-reinspectable configuration characteristics of duct gage (112 review points) and turning vane installation (15 review points), nine deviations were reported concerning lack of verification that the sage of the item inspected was as specified by the fabrication procedure. Five of these deviations were reported for cases where the required gage was not listed on the sheet signed by the HVAC contractor's QC inspector.
' Even though the required gage was not stated, however, each report was initialed by the inspector, indicating that the gage had been verified in accordance with the inspection procedure. For these five deviations, field measurements also verified that the item was fabricated out of the proper sheet metal thickness. These deviations were determined to be insignificant. Four of the nine deviations reported that documentation was I
l L - - - . - - . - - - . . - - - - - - - - - --_
Revision:- l' Page 40 of 58
() RESULTS REPORT ISAP VII.c (Cont'd)
'AppendLK 15 (Cont'd) 2.0 DISCUSSION OF RESULTS (Cont'd) not available to confirm the gage of collars used to attach accessories to the duct. Investigation in these cases revealed, however, that traceability for each collar was not required by procedure and that a single inspection report was permitted to address multiple collars without specific identification. In one of these four cases alternate documentation confirmed the proper material gage. It was determined that the lightest sheet metal gags-specified for ductwork was 18 gage.- For the remaining three cases,.an engineering'-
evaluation showed that even if the 18 gage' material was used, the strength of the collars was adequate to perform its function. These deviations were determined to be insignificant.- Therefore the quality of the hardware was confirmed either by further review of
'( ) inspection procedures, by field measurements or by evaluations, and it was determined that the items will function as required.
In all.15 review points,. installation of the turning vane was verified to have been inspected and accepted per the requirements of the contractor's approved installation procedures. No deviations were reported.
2.3.3 Attribute 10 - Weld Procedure Application In 405 review points, 52 deviations were reported where i documentation regarding weld procedure application could not be located. Of these, in 26 of 202 review points, deviations were reported where because of this, the weld procedure used to fabricate a duct section could not be verified. Also, in 26 of 199 review ,
points, deviations were reported where the weld filler I metal issued to the welder could not be verified to match the metal specified by the weld procedure used to fabricate a duct section. Reasons for these deviations are:
- Initials that appeared on the Material Requisition Form, or equivalent, do not exactly correspond to a welder's name as it appears on the welder's log.
l
R* vision: 1 Page 41 of 58 m RESULTS REPORT
( _-
)
ISAP VII.c (Cont'd)
Appendix 15 (Cont'd) 2.0 DISCUSSION OF RESULTS (Cont'd)
- Dates of welding were not correctly entered on the shop fabrication sheets, e.g., the i month and day is noted but the year is not or the date of welding was not entered or the date was not consistent with ot'her related dates on the form.
- Documentation was missing.
Each of the deviations have been analyzed. By investigating related records for the same date period or using other dates on the same record, by determining some nicknaues, and by comparing other logs with welder's identification, the date of welding and the I' identity of the welder could logically be established.
\~ On this basis, documentation was located in these instances that verified that the weld procedure employed and the weld material issued to the welder were correct; these 48 deviations are insignificant.
There were four cases where documentation could not be traced to the item being reviewed to establish weld procedure application, welder qualification and filler material. Two of these cases involved collars of similar function and construction. One of these cases involved a collar which attaches an 18 inch by 12 inch exhaust air register to the duct skin. Welding of the collar occurs at each corner, at the corner piece supporting the corner bolts. The bolting requirements for a register collar allows bolt spacing up to 12 inches maximum. The actual bolt spacing on the collar was 4 inches, which provides 18 bolts to secure this small register to the duct. In the unlikely event that these weld should fail and the corner bolts become ineffective, 14 bolts would remain and there would be little or no affect on the structural integrity of the collar due to the spacing and numbers of these bolts.
The second of these collars attaches on even lighter 10 gage sheet metal access door plate to a 14 inch by 16 inch duct. This collar has welding requirements and
(~}
(_,, bolts installed similar to the previously discussed register collar. Conclusions about any weld failure would also be similar. Also, investigation of collars
l Revision: 1:
L Page' 42 of 58 RESULTS REPORT ISAP VII.c (Cont'd)-
Appendix 15 (Cont'd) 2.0 DISCUSSION OF RESULTS (Cont'd) revealed that traceability of each collar was not required by procedure and that inspections were E permitted to address collars without specific identification. Based on the above, the lack of collar documentation will not affect the safety function of the systems and the deviations are considered insignificant. The remaining two cases are notable as documentation was missing. One involved a short section of duct, and the other one involved a flexible connection. In all four cases where documentation could not be traced to the duct section, a review was-made of the welding inspections that were conducted and the results discussed under Attribute 7, welding. In-all cases no deviations were reported against any.of
.O=
the 12 welding characteristics. These reinspection
- were judged to provide reasonable assurance that the .
welds will perform their required function and are acceptable for the two notable cases and to provide confirmatory evidence for the others. .It was thus determined that the installed hardware is acceptable.
2.3.4 Attribute 11 - Walder Qualification In 232 review points, 26 deviations were reported where documentation could not be located to substantiate that the welder was qualified to the appropriate weld procedure at the time of welding. The documentation used to identify the welder is the same documentation used to verify Attribute 10, weld procedure application. Therefore, reasons for the occurrence of these deviations are similar except for one deviation.
This deviation concerned a welder who welded to a procedure not listed on his Qualification Record. An evaluation showed that for the welding performed, the requirements of this unlisted procedure were equivalent to the requirements of a procedure listed on the Qualification Record, thereby providing reasonable assurance that he was qualified to perform the welding.
O
i Revision: 1 Page 43 of 58 3
k I RESULTS REPORT ISAP VII.c (Cont'd) i Appendix 15 l (Cont'd) J 2.0 DISCUSSION OF RESULTS (Cont'd)
The remaining deviations were evaluated by comparing ,
the identity of the welder with the veld procedure l employed and the Welder's Qualification Log or by reinspection. These evaluations provided assurance ,
that the welders were qualified at the time of welding l to perform the sheet metal and angle welding required to fabricate the duct sections. Therefore, the deviations were determined to be insignificant.
It was concluded that the deviations of the documentation regarding welder qualification are insignificant and that the welders were qualified for s
the welding of the duct sections.
\- 2.3.5 Attribute 12 - QC Inspector Certification A total of 419 review points were reviewed regarding inspector certification.
Deviations were identified whenever evidence of certification to the established procedural requirements at the time of inspection could not be found, However, as a result of the ISAP I.d.1 evaluations of inspector certifications, it has been determined that such deviations do not necessarily relate directly to actual inspector qualifications.
Consequently, the identified deviations were not relied upon (though they were considered) in determining inspector capability. Inspector qualifications were assessed for all inspectors (those with and without reported deviations) using the ISAP I.d.1 methodology.
The ISAP 1.d.1 evaluation of inspectors whose work affected this population concluded that all but nine inspectors either were certifiable to applicable criteria at the time of inspections, were found to be ,
capable of performing satisfactory inspections (including those with substantial positive evidence),
/)
(d or were otherwise shown to be of'no further concern.
Evaluation details are discussed further in the ISAP I.d.1 Results Report.
Revision: 1 Page 44 of 58 i )
RESULTS REPORT ISAP VII.c (Cont'd)
Appendix 15 (Cont'd) 2.0 DISCUSSION OF RESULTS (Cont'd)
For the nine inspectors not found to be either certifiable or capable of performing satisfactory inspections, it was determined that inspection results signed or initialed by the following inspectors may not be reliable evidence of the satisfactory performance of the indicated work processes. Corrective action, which will include the identification end evaluation or reinspection of inspections in this CWC previously performed by the inspectors discussed below, is recommended in the results report for ISAP I.d.l.
Nine Bahnson inspectors, responsible for conducting HVAC duct and duct support inspections, were determined 7-in ISAP I.d.1 to lack the education and/or experience
('- '
}
qualifications necessary for certification. BN-1 lacked adequate education and experience; BN-2, 3, 4, 5, 6, 7 and 9 lacked adequate experience; BN-8 lacked adequate education; and BN-4 had conducted questionable work while employed by Brown & Root as a mechanical QC inspector. Because the Bahnson record system did not allow ready identification of inspections conducted by individual inspectors and because a major corrective action program was being developed to correct other significant problems, no reinspection of these inspectors work were conducted. However, QA/QC-PDR-45 was issued during implementation of ISAP I.d.1 documenting the inadequacies of the certifications of these inspectors and requiring that these inadequacies be factored into the overall corrective action program for Bahnson-installed equirment.
2.3.6 Attribute 13 - Material Traceability The HVAC specification, 2323-MS-85, "HVAC Ducts, Louvers and Accessories", does not require material traceability after receipt inspection. Material traceability at receipt inspection is required for items that provide structural stability or integrity.
- A review of 100 percent of the purchase orders that
- l [' ' supplied these types of materials was performed.
(
R: vision: 1 Page 45 of 58 O
(_ ,1 RESULTS REPORT ISAP VII.c (Cont'd)
' Appendix 15 (Cont'd) 2.0 DISCUSSION OF RESULTS (Cont'd)
Receipt inspection was conducted for.the 49 structural material purchase orders. In 106 review points to verify structural load-carrying material traceability, three deviations were reported. One deviation was receipt of material to ASTM-A446 in lieu of the specified ASTM-A526 or A527 type sheet metal. An evaluation showed that the properties of the received material is equivalent to ASTM-A526 or A527, and the deviation was determined to be insignificant.
Two deviations were reported where nuts were received to ASTM-A307 in lieu of the originally specified ASTM-A563. A comparison of the two material specifications showed that the strength of material to f'"s ASTM-A307 was greater than that material to ASTM-A563.
( The deviations were determined to be insignificant.
2.3.7 Attribute 14 - Welding Inspection In 121 review points, seven deviations were reported where documentation could not be located to verify that welding had been shop-inspected on six items. Shop inspection includes internal and external welds.
Four of the deviations involved collars, each used to I attach an accessory to the duct. These deviations were judged insignificant for the same reasons discussed f or collars for Attribute 9. Nevertheless the welds were evaluated. The welds on a collar consist primarily of low load-carrying corner welds or externally applied welds. The four items were reinspected and no weld deviations were reported. Considering these results, the welds were evaluated to be sufficient to support the weight of the accessory, and the installed hardware was confirmed to be adequate.
One deviation involved the welding of a collar to a flange of a flexible connection. This deviation was determined to be notable based on the absence cf
() required documentation. An evaluation of the installed hardware was performed. The welding performed was one
___-----___________________.J
i' Revision: 1 Page 46 of 58
( . RESULTS REPORT ISAP VII.c (Cont'd) l Appendix 15 (Cont'd) 2.0 DISCUSSION OF RESULTS (Cont'd)
~
external circumferential weld and one groove weld. The welds were inspected, and no weld deviations were reported. An evaluation showed that the weld only supported the negligible weight of the collar.
Thus, the hardware fabrication was concluded to be acceptable.
One deviation was' reported where an inspector did not initial the Material Requisition Form for the fabrication of a 28-by-16-inch duct section. The inspector was identified via alternate documentation, the Inspector Tag Number Log, which provided assurance that the item was initially shop inspected. The deviation was determined to be insignificant.
O One deviation was reported where an inspector initialed a material requisition form but the date could not be distinguished. As the initials indicate that the item was inspected and the fact that the adjacent initials and dates of the tacker and welder are legible, the deviation was determined to be insignificant.
2.3.8 Attribute 15 - Touch-up Galvanizing Inspection In 121 review points, seven deviations were reported where documentation could not be located to verify that touch-up galvanizing had been inspected. These were reported on the same seven items discusseB in Section 2.3.7, Attribute 14, welding inspection. The lack of documentation to verify that touch-up galvanizing was inspected was determined to be insignificant for the same reasons given for the evaluation of touch-up galvanizing reinspection deviations in Section 2.2.7, Attribute 7, welding.
R vision: 1-Page 47 of 58 b
\/ RESULTS REPORT ISAP VII.c (Cont'd)
Appendix 15 (Cont'd) 2.0 DISCUSSION OF RESULTS (Cont'd) 2.4 Analysis of Combination of Deviations 2.4.1 Unclassified Trend - Hardware Installed without Approved Design Details A total of seven deviations reported under the attributes of configuration, bolt location, and welding reported that situations could not be inspected due to i the lack of design or installation requirements being.
clearly defined. None of these details were included.
in the HVAC contractor's documentation for duct systems fabrication or installation. An evaluation of this situation concluded that construction proceeded with the fabrication and installation of compcnents within
(~')
(, / the duct systems without having epproved design details.
An analysis of these deviations proceeded with the attempt to evaluate the as-constructed hardware in relationship to its intended or required performance as defined by the engineering requirements, e.g., design details. Since the engineering requirements were not clearly defined and since the deviations were judged potentially difficult to qualify for seismic conditions, this analysis was not performed and no conclusion precluding a construction deficiency could be reached. These seven deviations have gherefore been determined to constitute an unclassified trend. A root cause and generic implications analysis of this trend and recommendation for corrective action are included in Sections 3.0 and 4.0. In accordance with the procedures of this action plan, information regarding these deviations has been forwarded to the Design Adequacy Program for action.
! 2.4.2 Trend-Post Installation Modifications During reinspection, two cases of notable deviation i types were identified and discussed in Section 2.2.6,
( Attribute 6, bolt installation, under the headings of Missing Bolts and Loose Bolts. A thtrd case identified a related deviation and is discussed in Section 2.2.4, l
Ravision: 1 Page 48 of 58
/m
( ,) RESULTS REPORT ISAP VII.c (Cont'd)
Appendix 15 (Cont'd) 2.0 DISCUSSION OF RESULTS (Cont'd)
Attribute 4, gasket installation. The deviations involve a missing gasket section, missing bolts and loose bolts. Each deviation type occurred on a separate duct section in the Control Room Complex.
During the documentation review, one case of a notable deviation type has been identified and discussed in Section 2.3.1, Attribute 8, pressure test. This deviation type reported the lack of documentation that would confirm that a pressure leak test had been j conducted after an item was installed.
1 When the deviations are analyzed collectively, the
-~s following commonalities exist between each of the three duct sections inspected:
[L)
- Each duct section is part of the Control Room HVAC system and is located in the Control Room Complex.
- Each duct section was at one time installed, inspected, pressure-tested, and QC-accepted.
- As evidenced by the reported deviations and design change documentation, each duct section was removed from the system and modified by the HVAC contractor.
l
- No evidence of inspection after completion of l-the modification has been located, though an inspection walkdown was performed prior to turnover to TU Electric.
- Documentation verifying subsequent pressure i
testing is not available.
These deviations constitute a trend for post-installation modifications in this construction work category. The trend is determined not to be adverse as it is Enlikely that a construction
[}
A- / deficiency exists in the uninspected portion of the population, based on the reported evaluations of each of the deviations.
Rsvision: 1 Page 49 of 58 l)
- - RESULTS REPORT ISAP VII.c (Cont'd)
Appendix 15 (Cont'd) 2.0 DISCUSSION OF RESULTS (Cont'd)
The procedures used for fabrication, installation, inspection and pressure testing for duct section modifications are the same as those used for the original installation. The personnel and training of personnel are also the same for both the original installation and the modification of the duct sections.
The reason the pressure test was not repeated is that the contractor did not have this requirement documented until March 5, 1984. The modifications were completed prior to this date. The implication of the trend for inadequate post-installation modifications is that, most likely, other Control Room Complex HVAC systems contain similar deviations. To ensure that the Control 7._.
Room Complex HVAC systems have been constructed, and
's-)
installed in accordance with the construction requirements, a recommendation for improvement is provided in Section 4.2.
The recommendation is made for only the Control Room Complex area. The control room HVAC systems are important in that suitable ambient conditions must be maintained for sensitive control room equipment, and a habitable environment must also be provided for the personnel. The control room systems have also been the area of numerous modifications and the location of most of the notable deviations that were reported. Missing bolts, loose bolts and missing gasket all' occurred as a result of such modifications. Even though it has been concluded that those deviations will not result in a construction deficiency, it is desirable to ensure that similar deviations do not exist elsewhere in the Control Room Complex.
2.4.3 Summary Evaluation of Other Deviations The number of deviations identified during reinspection was judged to warrant additional evaluation.
O
R2 vision: 1 Page 50 of 58 RsSULTS REPORT ISAP VII.c (Cont'd)
Appendix 15 (Cont'd) 2.0 DISCUSSION OF RESULTS (Cont'd)
The fabrication and installation of HVAC ducts and plenums required a large number of welds in various configurations to fabricate the items and a large number of bolts and associated hardware to connect the flanges of the items. The tolerances on some of these requirements were unnecessarily restrictive,.given the relative insensitivity of the overall adequacy of the installation to variations in these parameters. For example, bolts were to be spaced every 4 inches along.a flange, with a plus-or-minus 1/16 inch tolerance. Many bolts were spaced closer than the minimum required distance, resulting in many deviations where the installed hardware was actually stronger than designed.
Even those deviations where requirements were exceeded O in an adverse direction were found to have minimal effects on the integrity of the ductwork. This is a result of the large design margin typical of HVAC systems, which in turn is due to the relatively low loading conditions. These bolting considerations also apply directly to the numerous stitch welding deviations reported.
The large number of deviations in the HVAC ducts and plenums for welds, bolts and associated hardware was not unexpected given the nature of the fabrication and installation procedures and the nature of the hardware.
The lack of any findings relating to the original construction of most items in this population was also not unexpected given the insensitivity of the overall functioning of the hardware to deviations on individual parts of an item, and given the typically large design margin for HVAC systems.
. l 3.0 ROOT CAUSE AND GENERIC IMPLICATIONS This section provides a root cause and generic implications analysis '
for the unclassified trend of hardware installed without approved design details and a generic implications analysis within this Oe construction work category for an unclassified trend identified for the entire population of HVAC Duct Supports, Appendix 31.
l
_ _ _ _ _ _ _ __-______________-_________________-______w
Rsvision: 1 Page 51 of 58
\' RESULTS REPORT ISAP VII.c (Cont'd)
Appendix 15 (Cont'd) 3.0 ROOT CAUSE AND GENERIC IMPLICATIONS (Cont'd) 3.1 Unclassified Trend - Hardware Installed without Approved Design Details Root Cause Deviations were reported where ducts were connected to ducts without approved connection details, diffusers were bolted to air supply ducts without bolting details and portions of plenuma contained welded configurations that did not have ,
defined welding requirements.. Construction was required to fabricate, install, and modify hardware in accordance with approved design documentation or procedures. Since construction proceeded with the installation of hardware without this documentation, the primary root cause for these O- deviations was concluded to be less-than-adequate construction supervision. Construction supervision should have recognized that some portions of the construction effort were proceeding without approved installation details. Activities in these areas should have been curtailed until approved design documentation had been made available to construction.
Another primary root cause was less-than-adequate procedures that lacked sufficient detail to control all necessary installations. Neither the HVAC contractor's Duct Fabrication, Duct Installation, Duct Accessories Installation Procedures or approved design documentation contained the proper details for construction to follow to complete the types of activities described by the deviations. HVAC duct systems include numerous possible combinations of fittings, accessories, duct section geometries and acceptable methods for fabricating and joining these items together. Evaluations of the reported deviations revealed that a complete package of design documents and procedures was not prepared that would clearly govern all facets of HVAC duct construction and erection. The fact that the current documents were approved by the HVAC contractor's engineering and QA personnel showed an underlying weakness in the ability to prepare appropriate craft instructions on the part of these organizations.
( The secondary root cause was concluded to be less-than-adequate QC inspections that accepted installation details without a documented design basis. The items against
Ravision: 1 Page 52 of 58 Y
! RESULTS REPORT ISAP VII.c (Cont'd) !
. Appendix 15 (Cont'd) 3.0 ROOT CAUSE AND GENERIC IMPLICATIONS (Cont'd) which the deviations were reported had been accepted by the HVAC contractor's QC organization. A review of the HVAC contractor's Duct Fabrication and Duct Installation Inspection Procedures shows that these inspection procedures were complete; however, the referenced Duct Fabrication and Duct Accessories Installation Procedures lacked the design details that were required to perform the inspection. ,QC acceptance-should not have been made if inspection criteria were not available in these areas.
Generic Implications The generic implications of these findings are limited to the vork perfctmed by the HVAC contractor (BSC) since they
(}
\,/ developed their own work processes that include the identified root causes.
Deviations similar to those reported have been observed in tha uninspected portion of the population. There are other uninspected 90 degree branch ducts and diffusers, for example, connected to supply ducts in a manner similar.co the ones teported by the deviations. The deviations regarding the lack of weld details on plenuus only relate to other uninspected plenums since plenum fabrication is to a unique set of plenum drawings not associated with the fabrication of sheet metal duct sections. The deviation type, as defined, applies to all reinspection attributes except longitudinal seam, type. Design details are availcble for this attribute. The deviation particularly applies to those with an element of configuration (specific deviations were identified f or three attributes).
Summary 6
The primary root cause for the deviations reporting that hardware was installed without approved design details was less-than-adequate construction supervision and less-than-adequate procedures lacking sufficient detail to r- '. control all necessary installations. Because the items had f been QC accepted, the secondary root cause was 4
less-than-adequate QC inspection that accepted installation details without a documented design basis.
l
Revision: 1 Page 53 of 58
/T RESULTS REPORT ISAP VII.c (Cont'd)
Appendix 15 (Cont'd) 3.0 ROOT CAUSE AND GENERIC IMPLICATIONS (Cont'd)
A recommendation for corrective action is presented in Section 4.1.
3.2 Unclassified Trend - Appendix 31 HVAC Duct Supports Generic Implications The generic implications identified in Appendix 31, HVAC Duct Supports, was limited to all work performed by BSC and their-duct support design subcontractor, Corporate Consulting and Development Company, LTD (CCL). This is because they developed their own unique approaches to the design and construction of
,-s supports, which introduced the root causes discussed in
( ) Section 3.0, Root Cause and Generic Implications, of Appendix
\~ 31. The contractor (BSC) who was responsible for HVAC duct supports was also responsible for the HVAC duct systems. In summary, the primary root cause of the unclassified trend was determined to be less-than-adequate performance of the HVAC contractor's (BSC) engineering personnel when preparing and approving "as-built" design documents and procedures.
The design of seismic duct supports by CCL and the BSC method of installing and inspecting seismic duct supports and then the subsequent method of developing and approving the as-built drawings is unique to Appendix 31. This method was not used for the installation of HVAC ducts or the development and approval of HVAC duct as-built drawings. Therefbre, this primary root cause and problems associated with this root cause do not apply to this Appendix 15, HVAC Ducts and Plenums.
A secondary root cause was determined to be less-than-adequate attention to detail by construction personnel when meeting the requirements of procedures, and a combination of less-than-adequate attention to detail by and training of QC personnel. Since personnel that f abricated, installed and inspected the ducts and their supports were the same as tnose who did the same activities on the duct systems, it is likely that the secondary root cause described 'in Appendix 31 would f'/}
\~_ apply to this appendix. By comparing the type of deviations
Revision: 1 Page 54 of 58 p-)
(
RESULTS REPORT ISAP VII.c (Cont'd)
Appendix 15 (Cont'd) 3.0 ROOT CAUSE AND GENERIC IMPLICATIONS (Cont'd) found in both appendices, this connection is confirmed. Of the types of deviations reported in Section 2.1, Summary of Results, of Appendix 31, the following same types were also reported in this appendix, namely: incorrect weld sizes, weld configuration, weld locations, weld lengths, missing welds, unacceptable weld quality, missing bolts and missing documentation. These identical deviations have been individually analyzed as they were reported in this appendix.
and in no case did they result in any adverse trend or construction deficiency. Therefore, it is concluded that the secondary root cause that contributed to establishing the unclassified trend for Appendix 31 confirms the findings of this Appendix 15, HVAC Ducts and Plenums, but does not require 7-~ further corrective action.
i j x_/
4.0 RECOMMENDATIONS 4.1 Recommendation for Corrective Action It is recommended that the as-built conditions described by the seven unclassified deviations discussed in Section 2.4.1 be analyzed to determine if any condition will prevent the items from performing their intended function. If not, then the current HVAC Duct Fabrication, Duct Installation or Duct Accessories Installation Procedures, or the design drawings should be revised to include these as-built conditions as new details. If any of the as-built conditions are not acceptable, then acceptable design details should be developed and included in the appropriate procedures; the related items should be reworked to the approved design. A walkdown inspection of the safety-related HVAC duct systems should follow to locate and rework similar, deviating installations.
A walkdown inspection of the safety-related HVAC duct systems should be performed to verify that fabrication and installation details exist for all installed work. For
[~ details that are missing, the as-built conditions should be
(~ analyzed to determina if they will prevent the items from
R0 vision: 1 Page 55 of 58
! i
' RESULTS REPORT b
l ISAP VII.c (Cont'd)
Appendix 15 l (Cont'd) l 4.0 RECOMMENDATIONS (Cont'd) performing their intended function. If not, then the current HVAC Duct Fabrication, Duct Installation or Duct Accessories i Installation Procedures, or the design drawings should be revised to include these as-built conditions. If any of the as-built conditions are not acceptable, then acceptable design details should be developed and included in the appropriate procedures; deviating items should be reworked to the approved l design.
l It is also recommended that the HVAC contractor's coast,ruction l
and QC inspection personnel undergo additional training to ensure that work and the inspection of completed work does not proceed unless approved drawings cre available or details of
[) the work are included in the fabrication, installation or inspection procedures.
\.__/
4.2 Recommendations for Improvement It is recommended that the HVAC duct systems within the Control Room Complex be reinspected to ensure that gaskets, nuts and bolts are properly installed. If an item is not properly instr.11ed, the work necessary to properly install the item should be performed.
5.0 CONCLUSION
S Based upon the findings of the reinspection and docu'entation m
reviews and satisfactory completion of the above recommendations and the applicable recommendations from ISAP I.d.1, there will be reasonable assurance that the hardware in this construction work category is adequately installed in conformance with the design.
p Y,
i Rsvision: 1 Page 56 of.58
)
-Q
( _j RESULTS REPORT ISAP VII.c ,
(Cont'd)
Appendix 15 (Cont'd)
Table 15-1 Summary of Reinspection Results HVAC Ducts and Plenums Deviation Classification Number of Inspection Number of Insigni- Construction Attribute Points Deviations ficant Notable Deficiency
- 1. Location 220 1 1 0 :0
- 2. Configuration #' 940 59* 54 4 0
- 3. Longitudinal Seam Type 100 1 1 0 0
. tO .4. Gasket Installation 190 43 43 0 0
- 5. Bolt Location 4,500 206** 195 8 0
- 6. Lolt Installation 10,500 55 30 25 0
- 7. Welding Location 6,700 117** 114 0 0 Length 6,700 583 583 0 0 Size .
6,700 134 134 'O O Profile 6,700 16 16 0 0 Fusion 6,700 26 21 5 0
.- Undercut 13,400 79 79 0 0 Cracks 6,700 3 3 0 0
- 1 deviation is Unclassified
Rsvision: 1 Page 57 of 58
- m.
I \
\- / RESULTS REPORT ISAP VII.c (Cont'd)
Appendix 15 (Cont'd)
Table 15-1 (Cont'd)
Deviation Classification Number of Inspection Number of Insigni- Construction Attribute Points Deviations ficant Notable Deficiency
- 7. Welding (Cont'd) ;
Touch-up .
j Galvanizing 6,700 44 23 21 0 i Craters 6,700 0 0 0 0 Porosity 6,700 0 0 0 0 Overlap 6,700 0 0 0 0 Surface Slag 6,700 0 0 0 0 TOTALS 103,550 1,367 1,297 63 0 (Approximately 100,000) l k l This number includes 7 Unclassified deviations indentified on the previous pages.
O
Revision: 'l Page 58 of 58 r
i RESULTS REPORT ISAP VII.c (Cont'd)
Appendix 15 (Cont'd)
Table 15-2 Summary of Documentation Review Results HVAC Ducts and Plenums Deviation Classification Number of Review Number of Insigni-Attribute Points Deviations ficant Notable 113 34- 0 34
- 8. Pressure Test
- 9. Duct Section Fabrication 127 9 9 0 Inspection 0 10. Weld Procedure Application 405 52' 50 2 232 26 26 0
- 11. Welder Qualification
- 12. QC Inspector (1) (1) (1) (1)
Certification Material Traceability 106 3 3 0 13.
Welding Inspection 121 7 6 1 14.
- 15. Touch-up Galvanizing 121 7 7 . O Inspection 1,225 138 101 37 Totals (
(Approximately 1,200)
( QC bsapector Certification review points and deviations are excluded from totals C and deviations are not classified because the results of ISAP I.d.1 indicate that these deviations do not relate directly to actual inspector qualifications.
__ _ _ _ _ - _ -