ML20212A145
| ML20212A145 | |
| Person / Time | |
|---|---|
| Site: | Oyster Creek |
| Issue date: | 12/18/1986 |
| From: | Huebsch P, Jerko D, Laggart M GENERAL PUBLIC UTILITIES CORP. |
| To: | |
| Shared Package | |
| ML20212A149 | List: |
| References | |
| 000243-002, 000243-002-R00, 243-2, 243-2-R, NUDOCS 8612220289 | |
| Download: ML20212A145 (96) | |
Text
- - -
" Nuclear Technic:1 Function 3 Safety / Environmental Determination and 50.59 Review UNIT Oyster Creek Nuclear Generatinr] Station PAGE 1 OF 19 SE No.000243-nn?
DOCUMENT NO.
(if ap licable) 0 NC gebtMkMR9 fMgion ACTIVM ITL e
Type of Activity Evaluation of Reduced Plate Thickness of the Drywell Steel Liner (Modification, procedure, test, experiment, or document) 1.
Is this activity / document listed in Section I or ll of the matnces in Corporate Eyes CNo Procedure 1000-ADM-1291.017 If the answer to question 1 is "no" stop here. (Section IV activities / documents should be reviewed on a case-by-case basis to determine if this procedure is applicable.) This procedure is not applicable and no documentation is required.
If the answer is "yes" proceed to question 2.
2.
Is this a new activity / document or a substantive revision to an activity /docu.
Eyes CNo ment? (See Exhibit 3, paragraph 3, this procedure for examples of non-substantive changes)
If the answer to question 2 is "no" stop here. This procedure is not applicable and no documentation is required. If the answer is "yes" proceed to answer all remaining questions. These answers become the Safety / Environmental Deter-mination and 5059 Review.
1 Does this activity / document have the potential to adversely affect nuclear safety E Yes CNo or safe plant operations?
l l
4.
Does the activity / document require revision of the system / component descrip-E Yes CNo tion in the FSAR or otherwise require revision of the Technical Specifications or any other Licensing Basis Document?
5.
Does the activity / document require revision of any procedural or operating Cyes $No description in the FSAR or otherwise require revision of the Technical Specifications or any other Licensing Basis Document?
6.
Are tests or experiments conducted which are not desenbed in the FSAR, the Cyes ZNo Technical Specifications or any other Licensing Basis Document?
7.
Does this document involve any potential Non-Nuclear environmental impact?
Cyes ENo 8.
Does the activityldocument require a review of criteria as outlined in SDD-Cyes ENo T1-000 TMI 1 Division i Plant Level Criteria?
If yes, identify TR/TFWR.
If any of the answers to questions 3,4,5, or 6 are yes, proceed to EXHIBIT 6 and prepare a written safety evaluation. If the answers to 3,4,5, or 6 are no, this precludes the occurrence of an Unreviewed Safety Question or Technical Specifications change. If the answer to question 7 is yes, either redesign or provide supporting documentation which will permit Environmental Licensing to determine if an adverse environmental impact exists and if regulatm approval is required (Ref. LP 010). If in doubt, consult the Radiological and Environmental Controls vivis' - or Environmental Licensing for assistance in com-pleting the evaluation.
Signatures - See attached sian-off sheet Date Engineer /Onginator Section Manager Responsible Technical Reviewer 1
Other Reviewer (s)
B612220289 861218 N55047 (1o 86)
PDR ADOCK 05000219 P
)
NWw 7 nni,ai gonctions Saf,ty Evaluation UNIT Oyster Creek Nuclear Generating Station PAGE 2 OF 1R SE N9. 000243-002 O
Rn No.
DrywellSteelShellPlateThicnbushonkntrenchment ss edu tion a the ACTIVITY / DOCUMENT TIT 1 p Base Sa Document No.
Region (if applicable)
Type of Activity /Documen, Evaluation of Reduced Thickness of the Drywell Steel ner.
(Modification, procedure, test, experiment, or document)
This Safety Evaluation provides the buis for determining whether this activity / document involves an Unreviewed Safety Question or impacts on nuclear safety.
Answer the following questions and provide reason (s) for each answer per Exhibit 7. A simple statement of conclusion in itself is not sufficient. The scope and depth of each reason should be commensurate with the safety significance and complexity of the proposed change.
1.
Is the margin of safety as defined in Licensing Basis Documents other than the Technical Specifications reduced?
Oyes CNo
- 2. Will implementation of the activity / document adversely affect nuclear safety or safe plant operations?
Oyes DNo The following questions comprise the 50.50 considerations and evaluation to determine if an Unreviewed Safety Question exists:
3.
Is the probability of occurre'nce or the consequences of an accident or malfunction of equipment important to safety previously evaluated in the Safety Analysis Report increased?
O Yes ENo 4.
Is the possibility for an accident or malfunction of a different type than any evaluated previously in the Safety Analysis Report created?
Oyes ENo 5.
Is the margin of safety as defined in the basis for any Technical Specification reduced?
Oyes ENo if any answer above is "yes" an impact on nuclear safety nr an Unreviewed Safety Question exists. If an adverse impact on nuclear safety exists revise or redesign. If cn unreviewed safe-ty question with no adverse impact on nuclear safety exists forward to Licensing with any ad.
ditional documentation to support a request for NRC approval prior to implementing approval.
6.
Specify whether or not any of the following are required, and if "yes" l
indicate how it was resolved Yes TR/TFWR/Other No a.
Does the activity / document require an update of the FSAR?
Explain: Yes; an analysis to support the new drywell shell thickness mo be included in the Final Safety Analysis Report Section 3.3.
b.
Does the activity / document require
(
a Technical Specification Amendment?
Explain: No. the minimum shell thickness found durinr1 the inspection noets the desian criteria specifled in the OCNGS Technical Soecificati ins N55046 (10 46)
SAFETY EVALUATION NO. 000243-002 Page 3 of 18 Preparers:
Signature Date M. Laqqart WA d W 2
/ 2 -/V 2(,
D. Jerko flaus#M fW
~/2 -/P-M P. Huebsch FMGL
/2 //P/ f6 L. Garibian
_o h 77 h
/ 2-/F-M S. Giacobbe T/
'/
g R. Greenwood p'E*4 Litu M / 2 _-_ 3
/ 2. - / a-s<.
Y. Nagai
_.2 hPN/-
v G
/
Responsible Technical Reviewers:
S. Leshnoff 2 d. bhb
/t////f6 G. VonNieda lW f
M. Sanford A4sslla(Wo /d n -/r-#4
/
Independe t afety Reviewer
/
/
11 18 36 17: 12 SFu A!:FCFT MG.
PC.0;5 00I
~~ ~
SAFETY EVALUATICM NO. 000243-002 Page 4 of 18 Preparaes:
Signatura Dato M. Laggart D. Jerko P. Huebsch L. Caribian j
S. Giacobbe 12-hffk R. Greenwood Y. Nagai Responsible Technical Reviewers:
S. Leshnoff C. YonNieda ht
!d//[/M M. $anford l
Independent Safety Reviewer J. R. Thorpe a
l
SENo.000243-dO2 Rev. O Page 5 of 18 TABLE OF CONTENTS Section Purpose 1.0 Systems Affected 2.0 Effects on Safety 3.0 Effects on the Environment 4.0 Conclusion
5.0 Attachments
1.
Description of Drywell Design 2.
Extent of Damage 3.
Causes of Corrosion and Corrosion Rate 4.
Structural Analysis e
SE No. 00243-002 Rev. O Page 6 of 18 1.0 PURPOSE The purpose of this safety evaluation is to assess the structural integrity of the Drywell steel pressure vessel in light of a recent (Inspection) finding that sections of the drywell shell near the base sand entrenchment region have a thickness which is below the thickness utilized in the original stress report prepared by Chicago Bridge & Iron Company (" Structural Design of the Pressure Suppression Containment Vessels", for JCPL/ Burns & Roe, Inc., Contract No. 9-0971, by CB&I Co.,
1965).
In addition, this evaluation provides a justification for operation up to the end of the lith operating cycle (18 months) for the Oyster Creek Nuclear Generating Station (OCNGS).
2.0 SYSTEMS AFFECTED 2.1 System No. 243, Drywell and Suppression System, particularly the drywell shell structure.
This structure is directly affected by the localized thinning.
2.2 Drawings showing original thickness - Chicago Bridge and Iron Co.,
Contract Drawings 9-0971, Drawings #1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11.
2.3 Documents that Describe the Drywell Structure are listed below.
2.3.1 Amendment #15 to OCNGS FDSAR, Primary Containment Design Report.
2.3.2 Updated FSAR, Paragraph 3.8.2.
2.3.3 OCNGS Technical Specification Section 5.2.
2.3.4 CB&I Stress Report, " Structural Design of the Pressure Suppression Containment Vessels" for JCPL/ Burns & Roe, Inc.,
CB&I Compar.y Contract No. 9-0971, 1965.
s SE No. 000243-002 Rev. 0 9
j Page 7 of 18 a
3.0 EFFECTS ON SAFETY 3.1 Identification of Documents 3.1.1 OCNGS Unit 1 Facility Description and Safety Analysis Report,
- Licensing Application, Amendment 3,Section V.
1
- LiceElsing Application, Amendment 11, Question III-18 t'
h, s.
- Licensing Application, Amendment 15
- Licensing' Application, Amendment 68
\\
\\
1 3.1.2' Technical Specification Documents 3.1. 2 '.1 Technical Specification and Bases - OCNGS Unit, Appendix A to Facility License DRP-16, JCP&L Docket
'o.
50-219, Sectior.s 3.5, 4.5, 5.2 N
y 3.1.3 Regulatory > Documents t
(
3.1.3.1 10CFR50, Appendix A, General Design Criteria for NuclearPobePlants
- Criterion 2 - Design Bases for Protection Against Natural Phenomena s
- Criterion 4 - Environmental and Missile Design Bases
- Criterion 16 - Containment Design
- Criterion 50 - Containment Design Basis
_ Y 3.1.4 Industry Codes and Standards 3.1.4.1 ASME Boller and Pressure Vessel Code,Section VIII, 1962 with Code cases 1270N-5, 1271N, and 1272N-5 Code cases,Section III, Div. 1, Subsection NE 3.1.4.2 See Attachment i for additional codes and standards.
N
^
'k I
i.
t-m
,-e
,e
,rw
-c
---,--m
r-
.,r,-,-.,---,-,-
r- - - - - - - - - - -,. ----
SE No. 000243-002 Rev. O Page 8 of 18 3.2 OryweII containment Structure 3.2.1 provides a description of the Oyster Creek Drywell Geometry, Design Bases, Materials, Shop and Field F,tbrication and Testing, and Concrete Interfaces.
3.2.2 Extent of Drywell Thinning Background information on the source of the sand cushion wetting, UT techniques, drywell thickness measurements, and core sample locations are included in Attachment 2.
Based on information contained in Attachmer)t 2, the following conclusions can be stated A.
The ultrasonic thickness probins of the drywell containment has been confirmed to give accurate but conservative results.
The physical measurements of the thicknesses of the plugs were approximately 0-47. greater than that determined by UT results.
B.
Destructive metallurgical examination of one of the containment plugs verified that the highly localized UT indication was an inclusion and that pitting did not exist.
SE No. 000243-002 Rev. O Page 9 of 18 C.
The general areas characterized as broad exterior corrosion have been vertfled to be general wastage.
D.
These broad areas of exterior corrosion are localized at an elevation corresponding to the exterior sand cushion.
Measurements of drywell thickness below the level of the interior concrete floor (which were made by removal of the interior concrete at two locations) show that wastage below the floor level is no greater than that measured just above the floor level. Measurements at the two locations show the drywell below floor level to be slightly thicker than the immediately adjacent area above the floor area.
E.
The drain line gasket was found to be leaking and was replaced.
Leak tests were performed on the bellows, and no leaks were detected. Observations of the areas where leakage had previously been found indicated that the leakage had been a'rrested.
F.
Based on the conservative methodology utilized in, the effective drywell thickness at the sand entrenchment region has a mean value of 0.87.
This value exceeds the minimum required shell thickness calculated for structural stability and integrity.
(See Attachment 4) t j
SE No. 000243-002 Rev. O Page 10 of 18 3.2.3 Drywell Corrosion Mechanism end Rate A teview of the potential causes of corrosian and a conservative prediction of a future corrosion rate is included in Attachment 3.
Based on information contained in Attachment 3 the following conclusions can be stated:
A.
In all ceses where general corrosion was present, the sand cushion appeared to be wet.
B.
No deep pitting was observed and no sulfide or substantial concentration of manganese was detected in the corrosion product.
This indicates that microbiological influenced corrosion is minimum..
C.
The corrosion observed can be explained by an aqueous corrosion mechanism assuming chloride contamination and oxygen depletion.
D.
A conservative corrosion allowance rate of 48 mils per year will account for any uncertainties in the assumptions i
of the corrosion mechanism.
l 3.2.4 Structural
, provides an assessment of the Drywell structural capability assuming a reduced shell thickness of.7 inches within the sand entrenchment area for two critical load combinations.
Conclusions which can be made from this assessment are:
SE No. 000243-002 Rev. O Page 11 of 18 A.
The original allowable stress criteria of ASME Soller and Pressure Vassel Code,Section VIII, 1962 with appropriate Code cases is met when credit is taken for the radially inward reaction due to the sand. Without the sand, (a beyond design basis condition) code allowable stresses are, exceeded by 2.7% with a reduced shell thickness of 0.7 inches at the sand entrenchment region. However, ASME Sect. III, Div. 1, Subsection NE allowable stress criteria are met without exception using stress intensities.
While peak local membrane stresses are less than the allowable, the meridional extent of these is more than allowed by Section III (but 2X).
The original Code placed no bounds on the extent of a local stress.
It is reasonable to neglect this departure from present Code guidance because the present situation is an in-service condition and not a design condition, and because the departure from present Code guidance is small.
B.
The load combinations selected for this analysis represent the design basis accident condition.
3.3 Effects of Thickness Reduction on the Safety Function of Drywell Containment Structure (DCS) i 3.3.1 Structural Performance The reduction in thickness of the drywell shell at the sand entrenchment region does not prevent the structure from performing its intended safety function.
SE No. 000243-002 Rev. O Page 12 of 18 3.3.2 Quality Standards Repair of the core samples taken were made in accordance with the quality standards of the plant.
3.3.3 Natural Phenomena Protection Since the DCS is protected from the cutside elements by a safety class.iructure capable of withstanding a tornado or hurricane, and since the plant elevation prevents natural flooding, these loadings do not contribute to the concerns posed by this activity.
However, in the evaluation of structural performance, seismic loads were included and found that this event does not affect the integrity of the DCS when the event occurs singly or in combination with other design loads.
3.3.4 Fire Protection The thinning of the drywell shell does not affect the fire protection program for the plant, since the drywell was not considered as one of the fire protection measures.
3.3.5 Environmental Qualifications The assumptions utilized in complying with 10CFR50.49
" Environmental Qualification of Electrical Equipment Important to Safety for Nuclear Power Plants" have not been altered, therefore there is no effect on Environmental Qualification.
SE No. 000243-002 Rev. O Page 13 of 18 3.3.6 Missile Protection The affected area is protected by a concrete shield wall as described in Section 3.2.1 and by the Reactor Building which provides protection from external missiles.
3.3.7 High Energy Line Break; Internal Flooding The maximum pressure inside the DCS after a high energy line break has been conservatively assumed to be 62 psig.
Subsequent evaluation of the affected area considering this pressure increase together with SSE and deadioad shows that DCS structural integrity is still maintained.
3.3.8 Electrical Separation The reduction in thickness of the affected area does not impact any electrical components.
3.3.9 Electrical Isolation The reduction in thickness of the affected area does not impact any electrical components.
3.3.10 Electrical Loading Impact on Emergency Diesel Generators and Safety Buses.
No effects per explanation 3.3.9.
3.3.11 Single Failure Criteria No effects on single failure criteria since the structural integrity and stability of DCS is assured.
SE No. 000243-002 Rev. O Page 14 of 18 3.4 Licensing Basis Documents Margin of Safety Review of the FDSAR requirements as to the structural integrity of the DCS during all modes of plant operation reveal that the minimum thickness of the affected regions still have ample margin of safety to satisfy Technical Specification 5.2 and the intended design as stated in the FDSAR.
This was ascertained after reanalysis (see ) of the structural response to the most severe load combinations considering the minimum thickness of the affected area.
3.5 Nuclear Safety / Safe Plant Operation Since the structural integrity and stability of the DCS have not been affected by the thinning of the affected regions of the shell, and the corrosion rate determined will not degrade the structural integrity and stability of the DCS during cycle 11, nuclear safety and safe plant operation will not be affected.
The thinning is limited to the area described in this evaluation; no evidence of damage to other drywell areas or other safety related equipment was found.
3.6 Probability of Occurrence or Consequences of an Accident Since the structural integrity and stability of the DCS is still maintained, the minimum thickness of the affected shell region of the DCS will not affect the probability of occurrence of any accident when the plant is in any mode of operation or plant condition.
Furthermore, since the containment isolation function of the DCS is intact, the consequences of any postulated accident at 0.C.N.G.S.
will not be affected.
SE No. 000243-002 Rev. O Page 15 of 18 3.7 Probability of Occurrence or Consequence of' Malfunction of Safety Equipment The fact that the structural integrity and stability of the DCS has not been affected by the condition, the probability of occurrence or consequence of a malfunction of safety equipment in the plant will not be affected.
3.8 Possibility for an Accident or Malfunction of a Different Type Than Any Previously Identified in FDSAR.
Since the DCS still meets design requirements no accident or malfunctions are different from what have been previously identified.
3.9 Margin of Safety on Basis of Technical Specification The thickness of the affected region of the shell has been ascertained to satisfy the original allowable stress criteria of ASME Boiler and Pressure Vessel Code,Section VIII, 1962 with appropriate Code cases when credit is taken for the radially inward reaction due to the sand.
Without the sand, (a beyond design basis condition) Code allowable stresses are exceeded by 2.7%.
However, ASME Sect. III, Div. 1, Subsection NE allowable j
stress criteria are met without exception.
While peak local l
membrane stresses are less than the allowable,
SE No. 000243-002 Rev. O Page 16 of 18 the meridional extent of these is more than allowed by Section III (but 2X). The original Code placed no bounds on the extent of'a local stress.
It is reasonable to neglect this departure from present Code guidance because the present situation is an in-service condition and not a design condition, and because the departure from present Code guidance is small.
~ 6 '(lolation of Plant Technical Specification The minimum thickness at the affected regions does not violate any section of the OCNGS Technical Specification. As stated in Section 3.9, the allowable stress criteria is satisfied.
3.11 Violation of Any Licensing Requirements or Regulations Review of OCNGS Licensing requirements and commitments reveal that the thinning of the drywell shell does not violate any of Licensing requirements or regulations.
This is primarily due to the fact that containrdnt isolation function and the structural integrity of the DCS have not been affected.
3.12 Radiological Safety Concerns The reduction in thickness of the drywell shell will not affect any radiological safety concerns because the containment isolation safety function of the DCS is still intact.
The i
drywell shell in the area of concern is within the biological shield, and adequate shielding of occupied plant areas will be maintained.
SE No. 000243-002 Rev. O Page 17 of 18 3.13 Change to FSAR This condition will require a change to the FSAR to reflect the change-in the plate thickness, and the results of the analysis which support this evaluation.
3.14 Change to Established Practice or Procedure This condition will not require any change to an established practice or procedure.
4.0 EFFECTS ON THE ENVIRONMENT 4.1 -Changes to Plant Environmental Interface The reduct'en in thickness of the affected shell area will impose no changes to the OCNGS plant environmental interfaces, because the structural integrity and stability of the DCS is still intact.
4.2 Potential Environmental Impact Since the activity does not affect the environment, it does not have any potential impact to the following:
A.
Environmental Technical Specification B.
Applicable Environmental Permit Requirements C.
Final Environmental Statement D.
Environmental Impact Statement Consequently, no additional evaluation is required.
5.0 CONCLUSION
Recent findings revealed that sections of the drywell shell near the base sand entrenchment region have a mean thickness of 0.87 inch.
This is less than the original thickness that was utilized
SE No. 000243-002 Page 18 of 18 in the evaluation of structural stability and integrity in support of Licensing tne OCNGS.
Extensive review of the original calculations, load combinations and different plant conditions, and new calculations generated to evaluate the structural stability and integrity of DCS show that:
1.
The structural performance of the DCS during the most severe plant condition (DBA) will not be affected.
The margins of safety found are more than enough to assure structural stability and integrity of the DCS.
2.
The containment isolation safety function of DCS is still intact, Consequently, no environmental or radiological concerns exist due p
to the reduced thickness.
3.
FSAR and Technical Specification Commitments have not been violated.
4.
Plant Procedures and Safe Practices are not affected.
5.
The corrosion rate determined for will not degrade the structural integrity and stability of the drywell during cycle 11.
6.
Based on Sections 3.6, 3.7, 3.8, and 3.9, there does not exist an unreviewed safety question as defined in 10CFR50.59.
SE No. 000243-002 Att. 1-1 DESCRIPTION OF DRYWELL DESIGN Primary Containment Geometry The primary containment consists of a pressure suppression system with two large chambers as shown in Figure 1.
The drywell houses the reactor vessel, the reactor coolant recirculating loops, and other components associated with the reactor system.
It is a 70 ft. diameter spherical steel shell with a 33 ft diameter by 23 ft high cylindrical steel shell extending from the top.
The pressure absorption chamber is a steel shell in the shape of a torus located below and around the base of the drywell.
The two chambers are interconnected through 10 vent pipes 6 ft. 6 in. in diameter equally spaced around the circumference of the pressure absorption chamber. The two chambers are structurally isolated by expansion bellows in the interconnecting piping and analysis for each unit may be censidered independently.
The drywell interior is filled with concrete to elevation 10 ft. 3 in. to provide a level floor. Concrete curbs follow the contour of the vessel up to elevation 12 ft. 3 in, with cutouts around the vent lines.
On the exterior, the drywell is encapsulated in concrete of varying thickness from the base elevation up to the elevation of the top head.
From there, the concrete continues vertically to the level of the top of the spent fuel pool.
SE No. 000243-002 Att. 1-2 The proximity of the concrete surface to the steel shell varies with elevation.
The concrete is in full contact with the shell over the bottom of the sphere at its invert elevation 2 ft. 3 in. up to elevation 8 ft. 11 1/4 in. At that point, the concrete is stepped back 15 inches radially to form a pocket which continues up to elevation 12 ft. 3 in.
That pocket is filled with sand which forms a cushion to smooth the transition of the shell plate from a condition of fully clamped between two concrete masses to a free standing condition.
The sand pocket is connected to drains provided to allow drainage of any water which might enter the sand.
Above elevation 12 ft. 3 in. the concrete is stepped back 3" measured radially from the steel shell.
Tnts gap was created during the construction by applying a compressible, inelastic material to the outside of the shell prior to concrete placement.
The material was later permanently compressed by controlled vessel expansion in order to create a gap between the vessel and the concrete.
Drywell Design Bases Design codes used for the original design are as follows with the effective dates at the time of design:
}
SE No. 000243-002 Att. 1-3
- ASME Boller and Pressure Vessel Code, Sections VIII and IX with all applicable addenda in effect at the time of design.
- Nuclear case interpretations 1270 N-5, 1271 N and 1272 N-5.
- ASME Boller and Pressure Vessel Code,Section II with all applicable addenda for the following material SA-212 High Tensile Strength Carbcn - Silicon Steel Plates for Bollers
.nd Other Pressure Vessels SA-300 Steel Plates for Pressure Vessels for Service at low Temperatures SA-333 Seamless and Welded Steel Pipe for Low Temperature Service SA-350 Forged or Rolled Carbon and Alloy Steel Flanges, Forged Fittings, and Valves and Parts for Low Temperature Service
- ASTM A-36 Structural Steel
- AISC Specification for the Design, Fabrication and Erection of Structural Steel for Buildings Pressure and temperature parameters in the original drywell design include:
- Drywell and connecting vent system tubes are designed for 62 psig internal pressure at 175'F and/or 35 psig at 281*F, and an external pressure of 2 psig at 205'F.
- In addition, the drywell is designed to withstand a local hotspot temperature of 300*F with a surrounding shell temperature of 150*F concurrent with the design pressure of 62 psig.
- The lowest temperature to which the primary containment vessel pressure containing parts are subject to while the plant is in service is 50*F.
To provide an additional factor of safety, 30*F was actually used for the design basis.
- During reactor operation, the vessel will be subjected to average temperatures up to 150*F at approximately atmospheric pressure.
SE No. 000243-002 l
Att. 1-4 Loadings considered in the design of the drywell include
- Loads caused by temperature and internal or external pressure conditions.
~* Gravity loads from the vessels, appurtenances and equipment supports.
- Horizontal and vertical seismic loads acting on the structures
- Live loads
\\
- Vent thrusts
- Jet forces on the downcomers
- Hater loadings under normal and flooded conditions
- Height of the contained gas in the vessels
- The effect of unrelieved deflection under temporary concrete loads during construction.
- Restraint due to compressible material
- Wind loads on the structures during erection
SE 000243-002 Att. 1-5 Load combinations used the design of the drywell and vent system for accident conditions include:
- Gravity load of vessel and appurtenances
- Gravity load from equipment supports
- Gravity-load of compressible material
- Gravity load on welding pads
- Seismic loads
- Design pressure: maximum positive pressure of 62 psig at 175*F decaying,
to 35 psig at maximum temperature of 281*F, to maximum negative pressure of 2 psig at 205'F.
- Restraint due to compressible material
- Vent thrusts
- Jet forces Allowable stress levels used in the design of the drywell are based on Code Case 1272 N-5
- General membrane (does not include thermal) - 19250 psi
- Local membrane (does not include thermal) - 28875 psi
- Surface stress - 52500 psi
~
SE 000243-002 Att. 1-6 Drywell Materials of Construction Steel plates are A-212-61T, Grade "B", made to ASTM A-300 requirements.
Minimum charpy vee notch impact test values of 20 ft.-lbs. at 0*F were used instead of 13 ft.-lbs. at 0*F as permitted by Code Case 1317.
Test specimens were taken both parallel to and transverse to the direction of final rolling of the plate.
Forgings are A-350 Grade LF1. Minimum charpy vee notch impact test values were 13 ft.-lbs. at O'F in addition to charpy keyhole impact test values required by the Burns and Roe specifications.
Pipe is A-333, Grade "0" seamless. Minimum charpy vee notch impact test values were 13 ft.-lbs. at O'F on full size test specimens in addition to charpy keyhole impact test values required by the Burns and Roe specifications.
Miscellaneous plate and structural steel (not within the scope of ASTM A-36):
All permanent structural attachments and lugs, welded to the shells, were made of impact tested material for a distance of not less than 16 times the plate thickness.
The erection skirt supporting the drywell was also made of impact q
tested material.
3 i
?
- +-
,e
-n
SE 000243-002 Att. 1-7 Drywell Shop Fabrication and Testing Components were shop welded, where possible, into large size shipping pieces, utilizing either submerged or metallic coated arc techniques.
In either case, low hydrogen electrodes were used, thus assuring the notch toughness requirements to meet the ASME Code Impact Tests.
All seam welds in the shell of the containment were of the double bevel butt type. All butt welds in any accessortes subject to the ASME Code were also of the double welded type or equivalent, and all the joints were full penetration welds. All welds subject to the Code were radiographed or otherwise examined in accordance with Code Case 1272 N-5.
All mandatory provisions of this code were followed and all recommended provisions were also followed where practical.
Heavy weldments and penetration weldments were furnace stress relieved as follows:
a.
Any plate segment wholly containing a penetration, nozzle, or column connection was furnace stress relieved at the shop after insertion of the penetration.
b.
All large penetrations intersecting more than one shell plate were stress relieved as follows. Any portion of a penetration containing seams joining metal over 1 1/2 in. thick at the joint was furnace stress relieved as a unit before welding into a penetration assembly or into the shell.
SE No. 000243-002 Att. 1-8 In keeping with the above, the vent line penetrations were shop assembled to the reinforcing collar and the completed assemblies were stress relieved.
The weld between the collar and the shell plate was made in the field and was not stress relieved.
All shop welds were radiographed in the shop. All welds in those parts of the work subject to the ASME Code were radiographed by methods complying with Paragraph UH-51 of the code.
Prior to shipment, all materials were cleaned and painted. Surface preparation and painting was in accordance with the paint manufacturer's recommendations. The interior of the drywell above the concrete floor, including jet deflectors and the exterior of the drywell above the water seal support bracket received one coat of Carboline Carbo-Zinc 11.
The interior of the drywell below elevation 8 f t.11 1/4 in. and the exterior surface of the drywell adjacent to concrete surfaces at completion of construction were not coated. All other surfaces of the drywell were given one coat of Carboline primer.
After erection and testing, all field welds and abraded places on the shop paint were cleaned by sandblasting and painted as noted above.
Drywell Field Fabrication and Testing During field fabrication the drywell steel was supported on a steel skirt of approximately 39 ft. diameter with its base plate at elevation-0 ft. 1 in. and invert of the sphere at elevation 2 ft. 3 in.
SE No. 000243-002 Att. 1-9 The 70 foot diameter spherical drywell and upper cylinder were field assembled and welded.
The transition knuckle and top head flanges were field stress relieved in accordance with the ASME Code.
The heavy plate flanges for the 33 foot diameter cover and neck flanges of the drywell were subassembled in segments, welded, x-rayed and stress relieved as complete units.
All completed shell plate assemblies, with penetrations installed, were stress relieved after fabrication. All butt welds were 100% x-rayed. Other welds which could not be 100% x-rayed were magnafluxed before ~d after stress relieving.
Upon completion of fabrication of the drywell and pressure absorption chamber, acceptance testing was initiated.
This included soapsuds testing at 5 psig, a holding period at 40.25 psig and a second soapsuds test at the design pressure of 35 psig.
This was followed by the overload test at a pressure of 71.3 psig which corresponds to a 115 percent overload.
The procedures for the overload test fulfilled the requirements of Section VIII of the ASME Code and Code Case 1272 N-5.
At the time of the tests, the downcomers, designed to pass the released steam and gases from the drywell into the suppression chamber were capped in order that a separate test could be conducted on each vessel.
The drywell was tested with no pressure in the suppression chamber.
The suppression chamber, however, was tested with a balancing pressure in the drywell to avoid an excessive external pressure on the verit lines and header inside the suppression chamber.
SE No. 000243-002 Att. 1-10 Drywell/ Concrete Interfaces The drywell shell is designed as a free standing structure and, with the exception of concentrated jet forces, will resist all required loads without interaction with the surrounding concrete.
The function of the concrete is to act as a radiation shield, provide a "back-up" to limit deformation due to concentrated jet forces and to form a support at the base of the sphere.
At the base of the sphere, subsequent to completloq of pneumatic testing, the volume inside the skirt was filled with concrete while simultaneously pouring the concrete floor inside the bottom of the shell.
The concrete pour outside the vessel' proceeded in full contact with the vessel up to elevation 8 ft, 11 1/4 in, where the concrete line was stepped back radially 15 inches.
This gap continues up to elevation 12 ft. 3 in. At points on the perimeter of the vessel where the vent lines penetrate the concrete, the forms were set back around the vent lines to provide clearance which would prevent contact between the vent lines and the concrete surface during any design condition.
The 15 inch radial gap was filled with sand to provide a cushion for the shell plate during the transition from clamped between two concrete surfaces to free standing.
At all elevations above the sand layer, the external concrete mass is set back from the surface of the steel vessel an amount calculated to allow unimpeded expansion of the steel shell during any design condition.
The gap was created by applying a compressible, inelastic material to the exterior surface of the
SE No. 000243-002 Att. 1-11 vessel prior to pouring concrete.
The material properties were chosen to provide resistance to crushing by the pressure induced by the head of concrete, but of low compressive strength to allow collapsing by induced vessel expansion.
The criteria for maximum gap was established to limit the deflection of the vessel wall due to local impact of jet forces.
The criteria used was that the space between the steel drywell vessel and the concrete shield outside must be sufficiently small that, although local yielding of the steel vessel may occur under concentrated forces, yielding to the extent causing rupture would be prevented. Using this criteria, the formed gap was 2 inches from elevation 12 ft. 3 in to elevation 23 ft. 6 in. Above 23 ft. 6 in the formed gap was increased to 3 inches.
This dimension allowed for inelastic compression due
(
to concrete pressure during the pour and residual thickness of gap material after compression by controlled vessel expansion.
The criteria used for selection of the gap material was as follows:
- It must adhere tightly to a curved, painted steel plate surface in flat, vertical and overhead positions.
- Could have relatively insignificant deformation under fluid pressure of wet concrete estimated at 3 psi.
- Would be reduced in thickness inelastically by about one inch from an initial thickness of 2 to 3 inches under a pressure of not more than 10 psi.
l
SE No. 000243-002 Att. 1-12
- WeJ1d remain dimensionally stable at the reduced thickness without significant flaking or powdering
- Would be unaffected by long term exposure to radiation and heat
- Should be susceptible to minimum damage which exposed on the vessel before concrete placement.
The 2 inch gap was formed using Owens-Corning Fiberglass SF Vapor Seal Duct Insulation.
The material was supplied with a factory applied laminated asphalt kraft paper waterproof exterior face, and was attached to the vessel with mastic and insulation pins. Joints between the boards, and edges and penetrations were sealed with glass fabric reinforced mastic.
The gap material used above elevation 23 ft. 6 in. was Firebar-D, a proprietary asbestos fiber - magnesite cement product applied as a spray coat. The solid materials, asbestos fibers, magnesite and magnesium sulphate (roughly 757. asbestos), were premixed and combined in a mortar mixing machine with water and, to control density, with foam to form a slurry suitable for spray application. After application and curing, the material surface was faced with polyethylene siteets with all edges sealed by tape and held in place by insulation pins.
The polyethylene sheets formed the bond-breaker for the concrete pour.
i f
l
SE No. 000243-002 Att. 1-13 Gap Formation and Results At the most critical location, drywell expansion at 281*F and 35 psig was expected to be approximately 0.7 inches. Considering an allowance for material rebound, it was calculated that the required vessel expansion could be achieved by raising its temperature 140*F above ambient. Concurrent with induced thermal loading, an internal pressure was created to balance the shell external compressive forces induced by the crushing of the gap material. An internal pressure of 40 psig was calculated as appropriate for this function, and considering the expansion induced by internal pressure, the temperature differential was reduced from 140*F to 130*F.
After placement of the gap material on the drywell shell, concrete placement continued in a staged schedule to complete encasement of the drywell.
The vessel was then expanded to create the required air gap required for thermal and pressure expansion.
Expansion of the vessel was monitored via use of pairs of extensometers at 7 points around the exterior of the vessel at locations of penetraticns.
The extensometers were read and recorded hourly and the readings compared with calculated theoretical values.
While the horizontal movements were in good agreement with calculated values, the upward accumulation of expansion expected due to the embedment of the lower region was at all points less than predicted.
Therefore, vessel discontinuity stresses at the embedment would have been less than calculated and the load on the concrete wall would have been more uniformly distributed and with a lower maximum.
= -
i
~
~
SE No. 000243-002 Att. 1-14 During the expansion, it was noted that the gap material had entrapped moisture due to incomplete curing and introduction of water from external sources.
This was evidenced by appearance of water at sleeves around several penetrations. This was deemed to be of no practical significance since the.
moisture's'effect on material compression characteristics would be a moderate improvement through a slight reduction in strength and a lesser rebound.
i i
4 l
l 1
2 6
T l
i
],
e==a====
- Mh*
m, e
./l
[BN=machy
' M
,,,,J
/
N n
y 1
a u-1 p
~_
^
j i
-~
l,
=:= '::.,- *
[m l
.~
==
! n
- ~
~
.r-l
.. - - -,.k 4, am 1
t-t-
'p' '.
s
-o-N M'
m
-t
.ui i:=ar.::s M
.dk i1
~."""
. f,.
- 3, i
W"\\.
~.r,r.w ~'
I L
~
f 3-,
H WM jA o
hL,i d - as==9 3.'
~' - f'l l
- gg r
w sif
- p..
!q F' --*
- _/
g *-
g.-
.e-I
- 2 s
~mt y
.)..k.
/ _g
%>=1 i
m.
.i 3
\\
/
Y 1
N.
/
r%l.
L r
E: h..- > <
,;X-eu
-a i
g
/x 1
w.,::::
tz::r-yl-[;. _, ' ~-#
f b-llElll 37
- !1
[#
~~'~l
,a
~. c..w,
g~
/"Q=
k&r) i s
~
y.
,daa.
.T
.w.e y
w
_, L..
mass 4<-
L.
- =
_a l
.Jl -
f
, F s,.
=_
u.
.c
- m..
1 r-fM
~
fN l
i l
1
\\
_ do l-2,
..u.e, - -- -
cs.
'l wy
[
'y.
6 i
e i
i i
Figure 1 Drywell Containment Structure
h'a If{,
9
/,
i 4f-n s
s s ld.* *c
=
49 i
p p *rn-$,
j h *
- A f +.P.; ' g c
n%,4 r.:
' 4
- 2
.,9
.h z
w w
e s>
m h em.u.82<(
4
.9
+E
.. a-
- s Po.
ec l ' ~;*i.p.
- 2.*> S
~
n w@-
j
%: m.. 4..3 4*: a$$
d f
.A g,
s'
/.f.S.
5
.a o ~ a n. xs'. :.. %.
~'
N, t-
' Y g,<
/gv0.4_
, y'
~
I
- 4.is4S
- -[0, 8
g
$f$$iLL
-h4.L>
h
~f
-l Bi o ul
/
hp in g
-d <
g z
$d>y '
4 h
@u 3,
N EE t
?
.a
= o.
)
as.
4 m"
.( <
v g 7 g 8. t
~
h
~
C..
E 5
C e
x o
4 g
E d
.,1 F
Y I
f.Yk-3 w
+. :
g Vf$
\\
l l
I l
Fiaure 2 Cross Section of Drywell & Torus
i
' b SE No. 000243-002 s
Att. 2-1 Attachment]
s Extent of Damage e
EXPECTED SOURCE OF SAND CUSHION HETTING t
During the 1980 Oyster Creek. plant outage, water was found leaking from various locations from the concrete surrounding the drywell. Containment penetration X-46 (Elev. 86'-0") on the south west, and penetration X-50 (Elev.
v l
47'-0") on the north east were reported to have water leaking from within the concr'ete biological shield. ~These identified areas correspond to Bays 7 and Bays'17 & 19, respectively.
In addition it was reported that water was coming from the sand cushion drain lines in Bays 3, 11, and 15 into the torus room.
a
'\\
Efforts were made to identify the Scurce of the water and its leak path.s The
^
leakage was found to have the same' range of radioactivity as that within,the
' (,
reactor. The 3eak path for the water was believed to have been from the reactor cavity located immediately above the drywell.
Thiscavity[isfilled e
with water during refueling operations.
It was believed that a leak from this cavity through the'6ellows seal at the bottom drained to the space between the
\\
drywell and the surr'ounding concrete (i.e., the spacestilled with insulat}on).
The Ydlume below th'e, bellows was pressurl2ed with service air
-dad the bellows checked for bubbles. Another leak test was' performed by l
rx Y
injecting helium benind the bellows and the bellows sniffed.
The results of thesel ests were negative.
The 2 inch reactor cavity drain line that includes t
t-a flexible pipe section was also tested with no,significant leakage detected.
. Plans were made during the following operating cycle;.to locpte and seal any potential leak path from the (eactpr refueling cavity.
s 1
w.
9 g
j
't
~
SE No. 000243-002 Att. 2-2 During the 1983 outage the welds of the refueling cavity were leak tested.
Some minor leaks were detected and repaired.
The bellows area between the containment and the refueling cavity was cleaned to remove contaminants.
The area was then inspected and attempts were made to apply various pressure tests to the bellows, however, no leaks were detected. Also, during the 1983 outage
~
the water level was dropped to the lowest reactor cavity shield plug step. At this time it was observed that leakage from penetrations X-46 and X-50 stopped.
Furthermore, leakage into the torus room had diminished.
Three of the four shield plug steps were inspected via liquid penetrant #or the full circumference; no indications were detected.
The single drain line used to detect leakage from the refueling cavity was suspected of being restricted. A tertriction in this line would cause any leakage to be directed into the area between the containment and biological shleid.
This drain line was purged
- with air and did not appear to have any flow restrictions.
When the refueling cavity was filled, similar leakage was found as previously described, however, it had been reduced appreciably.
During the cycle 11 outage outage the drain line from the refueling cavity was inspected. Drain line gasket (30"x7"F was found to have leaks, and it was replaced.
Leak tests were performed on the bellows, and no leaks were '
detected. Observations of the areas where leakage had previously been found indicated that the leakage had been arrested.
y
-n..
---, ~.,.
SE No. 000243-002 Att. 2-3 DRYWELL THICKNESS MEASUREMENTS Because of these wetting conditions, there was concern that repeated exposure of the drywell steel to water could result in degradation of the drywell.
Measurements of the drywell portion of the containment shell were made to verify its thickness during the llR outage.
These measurements were made using UT, a Non Destructive Examination (NDE) method, that is able to accurately determine the thickness of material or presence of abnormalities, i.e., nonmetallic inclusions.
UT plate thickness measurements were made on the Oyster Creek drywell.
Approximately 1,000 UT readings were eventually taken utilizing an ultrasonic thickness. gauge device (D-meter). Measurements were obtained by transmitting ultrasound through the plate and measuring the time it takes for the longitudinal wave mode to travel to a reflec' tor (front wall interface or mid-wall reflector or backwall) and back.
Since the electronic measurement.of time results in the digital thickness measurement of the first significant sound reflector, the probability of a mid-wall reflector being measured verses the backwall is dependent on the size of the reflector related to the surface area of the ultrasound transducer.
The larger the mid-wall reflector, the more likely the digital thickness reading will be the mid-wall number, and not the backwall value.
SE No. 000243-002 Att. 2-4 To further characterize the drywell and "A-Scan" UT technique was also employed.
"A-Scan" is important for the expanded analysis of the character, location and amplitude of various ultrasound reflectors.
The "A" scan is the ultrasonic indication displayed on a cathode ray tube (CRT).
The front surface pip or amplitude appears first, and the back surface pip or amplitude appears sometime later in the CRT sweep display.
The space between the pips is a measure of the distance between the surfaces.
Pips in between the front and back surfaces may be mid-wall reflectors such as laminations, inclusions or isolated holes and/or pits.
Other characteristics of the reflector can be observed by a qualified technician when using an "A" scan that are not available with a D-Meter.
Profile of the amplitude, break pattern at the baseline, number of doublets following t'he amplitude pip, multiples of original reflectors, and amplitude height on the screen and other characteristics all give information that may-be useful in analyzing origins of ultrasound reflectors.
MEASUREMENT LOCATION Initial UT measurements were made from the inside of the drywell containment at elevations 51 feet and 10 feet. A digital UT system was used.
The measurementsoppositethesandcushionatthe10ht.elevationintheBays corresponding to where water leaks were observed, indicated that the containment wall was thinner than expected. Measurements above these areas 'a the same plate indicated thicknesses within the original plate thickness variability.. Additional UT. readings in the same Bay quadrants at elevation 51
SE No. 000243-002 Att. 2-5 indicated no abnormal thickness variations. Although there are no specific requirements for surveillance of the containment wall thickness, it was considered prudent to make these measurements due to the wetted conditions that had occurred.
The initial measurements were made through the protective coatings on the inside of the containment. Since the effect of the protective coating on the UT measurements was questioned, special test blocks were made that included the coating material to quantify the effects of the coating on the UT readings.
The accuracy of the UT system was established for the coating thickness of the upper portions of the drywell. The effects of Carboline Carbo-Zinc 11 coating on the accuracy of UT measurements was verified through an experiment conducted by GPUN.
Two carbon steel plates approximately 1.15-inch thick and six by six-inch square were coated with carbon zinc. One plate had five mils coating and the other plate had 10 mils coating.
Both l
plates had a half inch wide strip on one edge left uncoated.
Both plates were laid out in a half inch grid pattern across the entire partially coated side including the uncoated strip. Similar equipment (D-meter of same make and model) transducers, and couplant as used in the field was utilized and measurements taken. Approximately 149 readings of thickness were taken for each plate. Additionally each grid (excluding the uncoated strip) was measured by Dry Film Technique (DFT) gauge to determine the coating thickness.
The uncoated strip for each was measured by micrometer.
The three readings:
- 1) ultrasonic (coated and uncoated); 2) dry film technique; and 3) micrometer (uncoated strip) were compiled,. averaged and final factors
SE No. 000243-002 Att. 2-6 developed.
The uncoated micrometer reading, plus the DFT reading was treated as the true reading of combined thickness.
The UT reading was found to overcall 0.3% for 5 mil coatings and 1.5% for 10 mil coatings after subtracting the DFT reading from the combined UT reading of steel and coating thickness.
It should be noted that the coating application on the test plates and the upper portion of the drywell were consistently uniform.
The coating along the basement wall, however, was found to be considerably thicker at places effecting the UT readings.
For this reason the coating was removed and a new set of UT measurements were made.
The new readings indicated that the containment wall was thinner than exp?cted in several areas along the basement floor.
The areas of indicated thinning was adjacent to the sand cushion.
EXTENDED UT MEASUREMENTS As a result of the initial UT readings adjacent to the sand cushion being considerably thinner than expected, a program was initiated to obtain detailed measurements to determine the extent and characterization of the thinning. UT neasurements were made in each Bay at the lowest accessible locations. Where thinning was detected, additional measurements were made in a cross pattern at the thinnest section to determine the extent and direction.
Measurements over a six by six inch grid were then made, mcving over the thinnest area to further quantify the wastage area.
To determine the verticai profile of the thinning, a trench was excavated into the floor in Bay 17 and Bay 5.
The concrete floor and rebar was removed to expose a portion of the drywell wall about 18 inches wide and sufficiently deep to allow measurement to the bottom of the sand cushion area.
Bay 17 was selected since the extent of thinning at the floor level was greatest in that area.
It was measured that the thinning below the initial measurements were s
9
SE No. 000243-002 Att. 2-7 no more severe and became less severe at the lower portions of the sand cushion.
Bay 5 was excavated to determine if the thinning line was lower than the floor level in areas where no thinning was detected although several inclusions were found, there were no significant indications of thinning.
Tre Safety Evaluation (SE No. 328227-001) for the excavation and its treatment for continued plant operation is separate from this evaluation.
Heat Affected Zones & Reinforcement Structure Other areas of concern requiring additional UT investigation were the plate to plate welds under the torus vents and the vent opening reinforcement plates.
These areas were given extra consideration on the basis that material sensitized by welding may have been attacked by a corrosion mechanism with greater damage or cracking occurring at those locations.
The extra UT investigation was conducted at three spots equal distance along side each toe of the vertical plate to plate weld and on either side of the bottom center gusset of the vent opening reinforcement plate.
D-meter thickness measurements were taken at all eight spots for Bay 5, 7 and
- 19. At these three Bay sites the six spot locations on each side of the plate to plate weld under the torus vent openings were also 45* shear tested to interrogate the weld Heat Affected Zone (HAZ).
The 45* shear wave test was especially done to detect HAZ cracking.
The top two spots were also the sites from which the plate to torus vent reinforcement plate weld was examined for v
SE No. 000243-002 Att. 2-8 HAZ cracking.
No crack indications were found and no wastage of the torus vent reinforcement plate was found.
The plate to plate weld HAZ as well as the weld when tested as part of a 8 or C location grid (6"x6") indicated wastage similar to the surrounding plate wastage.
Alternate UT Techniques and Verifications EPRI NDE Center UT personnel were invited to independently analyze the containment vessel plate.
Their objective was to independently analyze the condition of the drywell liner.
They scanned two areas using a "Zero Degree Longitudinal Wave Method". One area compared was just above the curb that we indicated had general wastage. Another area was where we had indications of mid-wall deflections or laminar inclusions.
Their observation and measurements independently verified GPUN's results.
Mapping of the' wall profile indicated a corrosion transition at seven to eight inches up from the concrete curb in Bay 19.
This detailed map was corroborated by the GE Ultra Image III "C" Scan topigraphical mapping system that will be used to obtain a baseline profile to track continued wastage.
GPUN experimentally utilized the I.D. Creeper or "30-70-70" technique (a UT integration method) to detect minor changes in back wall surface conditions.
This technique ccapared "A" scan presentations from one inch thick corroded samples the results from Bay 13 locations "A" and "E".
Reference standards were utilized representing light, moderate and heavy corrosion conditions.
This 30-70-70 technique defined surface roughness conditions by matching "A" Scan presentations from materials that have light, medium and heavy corrosion on their bzck surfaces.
It was able to verify the roughness condition of wastage and the light corrosion areas of the containment wall.
SE No. 000243-002 Att. 2-9 The "A" scan displays from the vessel plate were categorized by comparing them to the reference "A" scan displays.
Location A of Bay 13 (0"-6" up from concrete curb) showed typical "A" scan display of moderate corrosion on average and local sites of heavy corrosion. Bay 13 locations "A" and "E" indicated heavy corrosion between 0 to 6 inches above the curb, moderate corrosion 6 to 14 1/2 inches above the curb, and very low or no corrosion 14 1/2'to 17 inches above the curb.
LOGIC OF CORE SAMPLE LOCATION The selection of areas to obtain the core samples was made to evaluate if the UT measucements represented indicated material wastage or if there.vas localized " pitting".
Those measurement areas that indicated thickness readings of less than half of the thickness expected, i.e.,
.4 to.7 inches, and had adjacent measurements of the expected thicknesses (nominally 1.154"),
were designated as " pitted" areas. Areas that had indicated thinning at adjacent measurements were designed as wastage areas. A third area, above the wastage area, and within the sand cushion that appeared to have no thinning or
" pits", was also selected as a sample site.
The core sampling sequence and logic were to first obtain a sample of a suspected " pitted" area and two samples of a wastage areas but in different bays.
Should the " pitted" sample turn out to be an inclusion as suspected from the UT evaluation and the adjacent areas were actually the thickness as measured by UT, additional samples of areas that were suspected as being " pitted" would not be required.
SE No. 000243-002 Att. 2-10 Core Samples Core samples of the Drywell wall were taken at seven locations.
The samples were 2 inches in diameter. This was considered the minimum diameter to produce an adequate sample of the wastage area and provide an opening large enough to remove sand samples. The opening size also permitted insertion of a miniature video camera.
Larger. openings would have required a more complex plug design to restore the structure to its original condition.
The " pitted" sample #2 from bay 15 location "A" (GPUN 3E-SK-S-85) was found to be an inclusion in the plate with little~to no indication of corrosion on the outside of the sample.
Samples #1 and #3 were from bays 19 location "C" and bay 17 "0", respectively.
Both showed significant wastage with good correlation of actual micrometer measurement with the UT measurement (See Table 1).
The wastage samples (plug i & 3) were measured for thickness by' ultrasonic (D-meter) and dimensional (micrometer) in a four-point cross pattern and a center location.
The micrometer readings were taken with a ball micrometer to minimize the error observed when a flat bottom micrometer measures a locally irregular surface.
The micrometer measurements through the oxidized surface indicated the UT measurements to be between 0% and 4% less than the micrometer measurements.
- SE No.'000243-002~
Att. 2-11 Two additional wastage locations were selected below the severely thinned locations-(Samples 485) and two locations above the wastage areas (Samples 6&7) were selected to bound the conditions.
T;1e wastage samples 4&5 were similar to samples 1&3 confirming the UT measurement accuracy. Samples 6&7 did not have wastage and the sand behind them was found to be. dry, also confirming UT measurement accuracy.
/:
1 f
e 1
T e
~ --,
ene-
- e. -
.c-. -------.-
-.,e
,w.---,nm--
-, -+,
-,,-.w-m'-,.-
n-4-,--vgw g,m m
,,-,+-.,y..-
e,wg
.w--,,
,--,,-,-r-
-, -, -..,,,r,-w-,m,-r
-~
SE No. 000243-002 Att. 2-12 TABLE 1 CORE SAMPLE THICKNESS EVALUATION Sample No.
Location Type of Sample Pre-removal Thick.
Post-Removal Thick. (Ave.)
1 19C - 11'3 5/8" Wastage
.815" (avg.)
.825" 2
ISA - 11' 5 1/4" Pitting
.490" (min.)
1.170" center 1.17 (avg.)
only 3
170 - 11' 3 3/4" Hastage
.840" (avg.)
.860" 4
19A - 11' 3 3/8" Wastage
.830" (avg.)
.847"
.i 5
11A - 11' 3 Hastage
.860" (avg.)
.885" 6
11A - 12' 2 3/4" Above Wastage 1.170" (avg.)
1.19" center only 7
19A - 12' 1" Above Wastage 1.140" (avg.)
1.181" center only
SE No. 000243-002 Att. 2-13 The openings in the Drywell wall were repaired and sealed with a special designed and fabricated steel plug.
The final repair was accepted by the Authorized Nuclear Inspection (ANI) after successfully completing a magnetic partical examination of the welds on each p?ug. A final acceptance test for each plug was performed using a vacuum box bubble test.
In addition a local leak rate test was conducted on each plug and met the integrated leak rate requirements of the Code of Federal Regulations 10CFR50 Appendix J.
Actual leak rate measurement at each plug was 0.000 standard liters per minute at 35 psi.
The repair left the interior surface flush with the inside of the drywell wall. A separate Safety Evaluation (SE No. 328227-001) for the removal of the samples and for the repair of the Drywell openings has been conducted.
DATA
SUMMARY
The thickness measurements obtained adjacent to the sand cushion are tabulated on GPUN drawing number 3E-SK-S-85.
Initial measur emc-t3 were taken at four locations near the lower curb at each torus vent.
These locations, A-B-C-D, were selected to provide two thickness measurements of the left and right drywell plates that make up each Bay section.
Each tabulation heading defines the location of the tabulated matrix of measurements with respect to the top
SE No. 000243-002 Att. 2-14 of the curb and to the weld between the two plates at the center of the vent line. The matrix of measurements are at one inch increments both vertical and horizontal.
Those measurements around heat affected zones and on the vent line reinforcement were taken one inch on each side of the weld.
No degradation or wastage was indicated on the reinforcement plate or around the reinforcement plate to the containment plate weld. Wall thinning indications on the containment plate on each side of the containment plate weld was the same magnitude as surrounding areas indicating that the weld heat effected zone did not cause or accelerate wastage.
Data Reduction UT drywell thickness data was collected in each of the ten bays.
The UT data is presented on GPUN Drawing No. 3E-SK-5-85 Rev. 1.
The primary concentration of data was within a 6 inch wide circumferential band above the drywell ficar curb since data above this band indicated minimal wastage of the drywell wall materla1.
A new nominal wall thickness was sought for the affected lower portion, 6" wide band, of the drywell shell.
Two approaches were taken.
The first, was to establish the mean and standard deviation values of all the UT data in the affected region of the drywell.
The second approach was to establish the mean and standard deviation values of the UT data in the affected region which is contained within a 60 inen circumferential extent of the drywell.
The second approach using six measurement locations in each bay yielded nine (9) 60 inch combinations of mean and standard deviation values for each of the ten (10) drywell bays.
The significance of the 60 inch spans is that it represents a physical property of the sheII.
m
. - ~ -
....,,, ~ -.
SE No. 000243-002 Att. 2-15 This property is the deflection half wave length which defines the shell boundary relative to the location of applied primary and secondary loads beyond which the applied load does not cause shell deflection.
This property was calculated by Professor A. Kalnis of Lehigh University.
Although some of the low value UT indications were identified as inclusions in some of the areas measured, they were used as thickness measurements for the statistical reduction of data.
The first approach yleided a mean and standard deviation value of 0.96 inches for all of the UT data in the affected region.
The second approach yleided a value of 0.87 in. for the minimum mean wall thickness within the 60 inch arc length criteria.
This segment included 50 data points within a 26 x 6 inch segment of the drywell shell in Bay #19. A wall combination in the same bay within the 60 inch criteria included 148 data points with the data points extending over an area 57 x 6 inch and including the data within the former 26 inch segment.
This latter segment also yielded a value of 0.87 in. for the mean wall thickness.
For purposes of the engineering calculations regarding the structural integrity of the shell, based on the above minimum mean values, a nominal wall thickness of 0.87 in. should be utilized.
SE No. 000243-002 Att. 3-1 BACKGROUND Water Intrusion Detection The first documented evidence of the intrusion of water into the annular space between the drywell shell and concrete shield wall came to light during the 1980 refueling outage when water was visible around penetrations X-46 at elevation 86' 0" and running down the wall to floor elevation 75' 3".
Water was also observed at penetration X-50 at elevation 47' 0' and running down the wall to floor elevation 23' 6".
Water
('
collection was also observed on the torus room floor coming from the leak j
drains in bays 3, 11, and 15.
Informal, undocumented communications, however, also indicate water was observed on the torus room floor following construction.
Construction The primary containment pressure vessel is contained within a concrete j
l shield with a 3" annular space between the two structures.
The annulus is j
filled with sand specified as ASTM: C33 from elevation 8' 11 1/4" to elevation 12' 3" and from the bottom of this sand bed are 5 drain lines.
The sand appears to be a natural sand composed of silica with some alumina. An Owens-Corning Fiberglass SF vapor seal duct insulation
SE No. 000243-002 Att. 3-2 was applied to the vessel shell from elevation 12' 3" to 23' 6".
The insulation was supplied as individual boards 2" thick with a factory applied laminated asphalt kraft paper waterproof exterior face.
These boards were attached to the vessel shell with mastic and insulation pins.
The joints between the boards and edges and penetrations were then sealed with fabric reinforced mastic.
The remaining annular region above elevation 23' 6" is filled with a Firebar-D material.
It was applied as a spray coat (approximately 2.75" thick) over the vessel shell.
The material is composed of asbestos fiber (approximately 75%), magnesite, magnesium sulphate and a foaming agent (Aerosol PK) to control density. Over the top of the Firebar-D was placed a 4 mil thick polyethylene sheet.
The primary vessel is fabricated from ASTM-A 212 Grade B which is equivalent to SA-516 Grade 70.
The vessel was coated on the I.D. with Carbo-zinc 11 and on the 0.D. with " Red Lead" primer identified as TT-P-86C Type I.
Coating on the exterior of the vessel extends from elevation 8'-11 1/4".
+
O
A :.
SE. No. 000243-002 Att. 3-3 POTENTIAL SOURCES OF WATER INTRUSION Probable Sources Observations of leakage from the sand bed drains during the 1980 and 1983 refueling outages indicated that water had intruded into the annular region between the drywell shell and the concrete shield wall.
In addition, water samples withdrawn from the drains in 1980 were radiologically analyzed and showed activity similar to primary water.
From this information it was concluded that the probable sources of water were (1) the equipment storage pool, (2) the reactor cavity, or (3) the fuel pool.
It was further concluded that the leakage only occurred during refueling when the reactor cavity, the equipment storage pool, and the fuel pool are flooded. During the 1986 refueling outage, water samples were again taken from a drain line and analyzed.
In addition.to tritium, these samples were also analyzed for contaminants.
The results of these analyses are shown in Table 1-M.
e
=
wr
+s-c yw, 7----y--.
---r av-- - -., - - -- - -- -
SE No. 000243-002 Att. 3-4 TABLE I-M Drywell Drain Line Water Analy is Sample I Sample II Parameter (ppm)
(ppm)
Na 145 96 K
142 85 Ca 7.5 6.4 Mg 30 11 Al
.33
.02 Ni
<.01
<.02 l
Fe
<.01
.74 i
Cr
<.01
<.02 Mn
<.01
.02 Pb
.06
<.02 NH3(N) 3.6 C1 32.5 25 NO3 8.7 6
S0.
153 60 P0.
5 N.D.
F
<1 TOC 51 23.3 Organic Acid
.1 Total Sulfur 153 Conductivity 1100 us/cm 814 us/cm pH 8.9 8.7 Alkalinity (HCO3) 130 Samples taken 12/86 4
I
SE No. 000243-002 Att. 3-5 UT Data Interpretation Prior to core sample removal possible causes of the low UT thickness readings were attributed to external corrosion, laminations or a field of inclusions within the plate. Because the very low readings were localized it was expected that they would be a result of laminations.
The general wastage, however, extended from plate to plate and the affected areas of the shell were within the sand bed only.
Thus it was concluded that the plate thinning was most likely due to corrosion.
In addition, a qualitative assessment of the plate condition was made using an "A" scan presentation with a 5 mghz transducer.
This data was also indicative of corrosion on the outside.
Numerous ultrasonic thickness readings were taken in the drywell particularly at the elevation of 11' 3".
Review of this ultrasonic test data showed that potential corrosien damage appeared to be confined to regions in Bays 11, 13, 17 and 19.
Furthermore, the thinned parts of the drywell were limited to those areas which were in contact with the sand bed from elevation 10' to 11' 9".
Numerical analysis of this uata determined the minimum mean remaining wall thickness was.87".
UT thickness readings below the concrete floor elevation showed the thickness to be greater than.87" and at the bottom of the sand bed to be nearly nominal design thickness.
SE No. 000243-002 Att. 3-6 Sampling After the completion of the ultrasonic testing (UT) of each of the drywell bays above the concrete floor, the data was assembled and reviewed.
This data indicated that there were at least three regions which showed different characteristics. One set of data showed regions of overall general wall reduction which we characterized as wastage. Another set showed regions with little or no general wall reduction but localized areas with large wall reduction which we characterized as pitting /
inclusions.
The last set of data showed regions of little or no wall reduction and no random large reductions, which he characterized as minor wastage.
The characterization of each bay is summarized in Table 2-M.
SE No. 000243-002 Att 3-7 TABLE 2-M Bay No.
UT Characterization 1
Minor wastage 3
Minor wastage 5
Pitting / inclusion 7
Minor wastige 9
Pitting / inclusions 11 Wastage 13 Wastage 15 Pitting / inclusions 17 Wastage 19 Wastage In addition to the above general characterizations, it was also observed from the UT readings that above an elevation of approximately 11'9" the wall thickness would return to the nominal value.
This occurred even though the readings were still within the sand bed and there was wastage below this elevation.
Likewise, there were regions of the sand bed below the concrete which heretofore had not been ultrasonic tested and hence no characterization could be made.
d 4
..p.,
,,-_,,__,.,,y,-.
SE No. 000243-002 Att. 3-8 It was decided, therefore, that core samples should be removed from the drywell in each of these different regions in order to achieve the following goals:
a)
Verify UT thickness reading b)
Characterize the form of corrosion c)
Obtain sand samples and samples of other annulus materials d)
If corrosion existed, characterize corrosion products and environment e)
Provide access for visual examination of the outside surface of the drywell f)
Allow for sampling of sand and/or corrosion products for bacteria With these goals in mind, a first cut was made at selecting regions for sampilng of the drywell steel.
Twelve regions were selected:
four from wastage regions, four from " pitted" regions,.two from above the wastage region and two from below the concrete level.
These initial selections were, however, modified slightly as the program progressed and additional information became available from ultrasonic testing and initial core sample examinations, fl
SE No. 000243-002 Att. 3-9 Table 3-M identifies each of the seven core sample locations ultimately chosen and the types of samples obtained.
TABLE 3-M Core Se aples Sample Bay /
No.
Location Type Elevation Samples Obtained 1
19C Wastage 11'-3 5/8" Core, sand, bacteriological 2
15A Pitting /
11'-5 1/4" Core, sand, bacteriological Inclusion 3
170 Wastage 11'-3 3/4" Core, sand 4
19A Wastage 11'-3 3/8" Core, sand, bacteriological 5
11A Wastage 11'-3" Core, sand, bacteriological 6
- 11A, Minor wastage 12'-2 3/4" Core, sand 7
19A Minor wastage 12'-1" Core, sand
SE No. 000243-002 Att. 3-10 Evaluation of Pitting / Inclusion Sample Core sample #2 which was removed from bay 15 was taken to assess whether pitting or inclusions were responsible for the low ultrasonic thickness readings observed in random locations.
In region C where the sample was removed, the general area had thickness readings on the average of about 1.17" with random low readings of.48".
This particular plug had a region approximately 1/2" in diameter where the low readings resided.
Upon removal of this plug it was immediately evident visually that no serious corrosion or pitting had occurred.
The outside surface of the plug was covered with a reddish brown oxide and the actual measured thickness of the plug was 1.17" (avg.), Figure IM.
Elemental analysis of this oxide by EDAX indicated iron as the major constituent although in random location very high lev'els of lead were observed.
This lead is from reminants of the red lead primer originally applied to the shell. Other elements observed at trace levels were Al, Si, Mn, Ca, K, C1, S.
SE No. 000243-002 Att. 3-11 Metallographic specimens were prepared from the core plug both parallel to the rolling direction and perpendicular to it.
Examining the micro specimen at the outside surface of the core revealed some minor pitting.
These pits were filled with oxide which appeared normal for carbon steel corrosion. At the mid-plane of the spectren, however, a band of aluminide stringers was found in the region where the low UT readings existed, Figures 2M - 3M.
These stringers were sufficiently dense as to form a lamination which could easily reflect ultrasound.
This observation validates the conclusion drawn by the GPUN NDE people via their "A" scan UT analysis of this region that the low "0" meter readings were a result of laminations.
In addition, the examination of this plug also validated the accuracy of the thickness measurements.
It was concluded that UT could adequately define this type of condition and additional samples from pitting / inclusion regior.; were not required.
SE No. 000243-002 Att. 3-12 l
ii
?
e kW;y
~ ' -1.?g.
2X l
6 Figure 1-M Plug #2 outside surface of drywell.
Uniform red brown corrosion product.
l
SE No. 000143-002 Att. 3-13 et-t
- _ _ =:?T'..."%--- * - - ~ "l s
_4
~'
,~
50X S
e i
X' T
1 o
, 9 3.,.
-% "yvYN
'y*
9 1
'h I
500X l
1 l
1 Figure 2-M l
l Plug #2 Aluminide stringer at mid-wall.
Plane parallel to rolling direction.
l
._..m-.__.
l SE.No. 000263-002 i
Att. 3-14 l
.i
'^g l
t,
}
4 4:.
.........r:...
a. e. :. >. m., >...
,,........a-*.
~,.,;.4; u,..:.
c w,,,
1.
I I
e t
i
'l.
1.
l t
50X f
l 1
4 i
s 1
)
de
-E*ga *
.?. ',
i e
s*..
= ~
i s es -
o e
4 a
e e
a e
- e... 1... n n..-
se
- e
- e.
- P.. ' auP.e.
I e
e a en o.
i k
500X i
t 1
k Figure 3-M Plug #2 Aluminide Stringer at mid-wall.
l Plane perpendicular to rolling direction.
SE No. 000243-002 Att. 3-15 p
Examination of Wastage Samples As discussed previously, four samples were removed from wastage regions.
Three of these samples were sent to General Electric (Sample Nos. 3, 4 & 6) for analysis and one was analyzed by GPUN (Sample No. 1).
When core samples (numbers 1, 3, 4 and 5) were cut, it was noted that a hard black crust remained in the hole on top of the sand.
This crust was approximately 1/2" thick.
It was quickly realized that this crust was the corrosion product from the iron and as such was collected along with the sand beneath it for later analysis.
In general, all the wastage samples looked similar showing a relatively uniformly corroded surface with some hills and valleys (Figure 4-M).
- Overall, the surfaces were covered with a thin black adherant type deposit with some regions having a thicker more dense butidup of deposit (approx. 030" thick).
Elemental analysis of this deposit showed iron to be the major constituent with varying levels of chloride contamination. Minor traces of manganese, aluminum and silica were also noted and on occasion a trace of sulfur (Figure j
5-M).
On sample #1 a cross section was prepared through one of the valleys on the corroded surface (Figure 6-M).
This valley had a layer of corrosion product on it approximately 30 mils thick.
EDAX analysis of this deposit
SE No. 000243-002 Att. 3-16 revealed a high chloride concentration in a 2 mil thick layer of deposit adjacent to the steel, while further into the oxide but adjacent to this region the chloride levels were very low (Figure 7-M).
Although other samples did not show this dramatic variation in chloride, all did show that chlorine was a major contaminant.
In addition to EDAX analysis, x-ray diffraction was performed by GE on the black deposit. The results showed the material to be primarily Fe304 (magnetite).
This confirmed an initial observation that the deposit was magnetic; no other compounds were identified.
Metallography on the core samples showed that there was no deep pitting and no signs of any type of cracking or intergranular attack. Manganese sulfides were observed within the microstructure which were typical for this type material (Figure 8-M).
i I
l 1
SE No. 000243-002 Att. 3-17 e
e Plug #1 i
F^)
W...
b.-ly'=
)
t Figure 4-M Plug #2 outer wall uurface microplane is located
SE No. 000243-002 Att. 3-18 PLUGH 17C C H 1 J -gel 4ER AL DEFOFIi r :. - i.r.,. 3.. d
.y..
,, ' :, of,-
. r:
...n.li s ret r..I,ne,IAJ-L:,,7....,
Ucr uJ v; s7a
%h
- si 4(f;)
i u
e FE e
?
I f
f 9
s W
e l
l il 8
8 3M.'
y s
'n PI l
W 4
l F:
?.i L.4 I
u C
l El f
.i a
- )
c:
1
.s j,
6. *.y 6
r e
UL G!
b N i l.'
-. *,:....-....i...,.,
1 i.
.t
- 4. s'
- t.....,
, ' r ;s t 1 l Figure 5-M Typical elemental analysis of sample #1 corrosion product.
I SE No. 000243-002 Att. 3-19 i
e I
- f. ' :-
!!h.3 Ai.,'.i&yp i.lq(-
f$
l li, ;,.
l hkl i
6 iP'D41
.e 56X
[
Figure 6-M Thin layer of corrosion product remaining on sample #1 r
showing dif ferent layers and the presence of voids.
i i
i
SE No. 000243-002 Att. 3-20 F' LUG 1 GCALE SrECT.r? L4El.
"E PJ1 Fh
' LL z F "' t t ! ',;
yy3 e g
... ~.............
62.i t
i e
s lL.
i L
9i r
R ;a (
.i
=,5 pit h g.:l Ie b!!l C
- ti ! 2 M + !
- is {1 W 9 J t.'
ia, '4 613 J; % 1 a c.
- k b
[.r
(;W IH,. H u J2rm
,. d U a '.whtd b t,i ; n.a.1 h's /m; 2 M or m y..' s u w *J.I a A. g % ig'u - %d.i.e. :.
d a
t
.a 1 r
a*'
='~
(',2:m, h o u % w 5. m ei..c.a.~ *q ; dt ;.t:
d.* WJ.S&l. 3 t.:A=i' Aff Nl**;' "L M 6:M I.I 6 ' ! ' ' 'h. ~
?
<r'
.M ;
5 wir!/1Mhlw b J'; 2d LM s.*
, s r
?.:.,
1 m. > n :.<.w - 1 :n,. m L,t:.a.c e;n::b+t r: 4 e '
- .nm b m.s Wb t3 1 :e!M
' - - i a
.d a
trbs'n a.
.d..
.a.
6m-
.,t.+>
w i.s a
.v
?.
1
+ :. :
- y...- s
. p n h,..
A -
,u,,.,,:. s.w :.8.
- w
- :.+., : u.,. u.
..1 a t
's : 4.9 1. d. it Y, it M f?
- P. %'". W 3 ;b If :.$..a #.. A M A N 4.* '.',.* J '* S ' 2'. ; '.. wc a " sL.x,.x d. A E
.[.,.
- d p.*,. L #' i), a !.iM 3
a.a..;6. >.c. ' n;,..y.
s, 4., : M.
7a
~
,,,e
~ p.w ml:_meu
- .3 2. t...., ;. W M. e.a sa.wai.
u A '.
m.
-, p m. u, w b,l *,..m. a.;;,, p p q p s w >lm @,
.t n,b, 3 a>A >L*As M M-@
- .ukR f
- L iu. ali l :
2
. A u r.t h a d @ M. s
.a h a;1.1.w.e a. h w O.
o m.e.
y
. ; q.c.
a - u A si. e. mnacm pw wn n.ge t.r % m.
t.
_. 3:
n u.t a
a c'. c
'.2 'w>..+'-
'r...'.
i
- ri.e.P'
.it 1 t *ft /t% $. f 'JJ t
.,T.*r o,w S e m. ge..., &.
s-4 4(..,. -
. i s;,,
u
.s
-[t,.
3 S
'\\ I '. Sh
)
n, k ; k.
..,....,. h f'
s-
+..] t 1. e,. t t w w;a..
e 4
,, 4 f.,;../ 2.;..-,2-. '.:/ w
,2 ? +. --
' i, +
n,, J. '. 1 A -. - *.
L,?. Q. b ;,; :.... ;s.Q,,. W<r'-#
Ale. J 4.+;A 7.
h.- ;.n-U.:
...i.e. d f) W.;,;4 pf h:U.lw0 fl ";. d'4 H$.4*';a@.;if.I?; D,Z tf.E _ E /11. J.
'..r'?
ll - '
e' rY?,*f'lL*b $ f r *%.th:h I 4$$,* h*;*s '. 's.Y,.,J. N -' -- * '
*'/
~
1.$,,3 f,I '[ '! ) 'M $' k d A ki ils
'I T b I d.*In.. Ih'@.?bd,Mf/* u 7f.
! h.'...
8 L
q c.t-M t.W ;
t
,, 1 a[ '
..c.:.; : x,.e.
- _ a
- p. w.%a ",.
. 6 f.c.
s.
J. w
-.. : L.
-- htdi j L i ;.L
. t a 4;* L l
a'l u i..; ".d 4 t
. a s tin io mm
.... e, 4 J
....c
- Jtia,4,;;.
o,..,..
w l
um y. x -. s < '
end.wr
. n s.b N N hw,a r..:&e,R...e..& L lkkL'AQ Qg,-Qjf,y Q g Q Q &.
E j
...e.....m........n...e v.w.....e.. t.g u..n.....tW.r. - w.m m.
~ u a.v... a w
-+4
- 4. e
..i
.I 1
.c.
n.,
n..
I r 'il 6 Figure 7-M EDAX scan through deposit ahown in Figure 6M.
Deposit runs from 0% - 49% full screen.
Chloride peak is at steet/ oxide interface.
1 SE No. 000243-002
-21 i
i 1
1 i
l I
i.
l
.i I
100X l
1 0.D. surface; Plug #1 l
l I
.c 200X I
fins inclusions below surf ace.
l l
l Figure 8-Ft
SE No. 000243-002 Att. 3-22 Analysis of Sand and Firebar-0 As was shown in Table 3-M sand samples were removed from behind each core plug.
In addition, sand was removed from the Bay 11 drain line.
EDAX analysis as well as leachate analysis was performed on representative samples of the sand.
The results are shown in Figures 9-M and Table 4-M.
The sand appears to be a natural sand composed of silica with some alumina present. As noted on the EDAX spectrum, some chloride is present and this was confirmed by the leachate analysis which showed chloride in the range of 6.5 - 93 ug/gm.
Also noted in the leachage analysts was magnesium and sulfate which most probably came from the Firebar-D.
Some organic carbon was also detected.
These analysis indicate that a source of the chloride found in the corrosion product existed in the sand which was probably leached from the Firebar-0 and that organic material as well as a source of sulfur exist which could provide nutrients for bacterial growth.
A sample of the Firebar-0 was obtained through one of the drywell penetrations and subjected to a leachate analysis. As might be expected, this material was high in Na, K Ca, Mg and S04 as well as chlorine.
The results of these analyses are also shown in Table 4-M.
SE No. 000243-002 Att. 3-23 TABLE 4-M Sand Leachate Analysis Sand Imachate Sand Leachate Sand Imachate Sand Leachate Firebar-D* Leachate Bay 11 Drain Bay 11 Drain Plug #1 (19C)
Plug #2 (15A)
Analytical 1 Hr, 60* C 24 Hrs, Room Temp 1 He, 90* C 1 Hr, 60* C 1 Hr, 60* C Para:ne ters (ug/g)
(ug/g)
(ug/g)
(ug/g)
(ug/g)
Na 777 25 25 37 47 K
784 25 20 37 23 Ca 176 30 25 47
< 23 hg 1936 30 10 10 4 23 Al
< 0.3 40.5 1.5 39 2.3 Ni 4 0.3
< 0.5 0.5
<.33
< 2.3 Fe
<; 0.3 5.0 1.0 82 8.4 Cr 4 0.3
- 4. 0. 5
< 0.5 4.
33
<. 2.3 Mn
< 0.3 0.5
< 0.5 3.7
< 2.3 Pb 0.6 1.5
<. 0.5
<.33
< 2.3 Na3 (3)
C1 573 10.5 6.5 45 93 NO3 132 2.5 1.5 e 17 6
SO4 2850
< 25 32 28 79 PO4 N.D.
N.D.
N.D.
N.D.
N.D.
F 14 N.D.
N.D.
N.D.
N.D.
TOC 1056 39 37 46.6 N.D.
Organic Acids 4 20
< 5
<5 Total sulfur s 50 B
Conductivity 588 pH 8.46 7.43 7.58 7.02 5.99 Firebar-D is a composite of foam, fibers and concrete
l SE No. 000243-002 Att. 3-24 U A f4 D / G ri U U P OF G r1(4 L L P C '! B L L e i
gr E li.e L: +.
t
.I l
.: 7."O 7 ?>3LL FEE.:LE; t..
i31 6
i l
J k
I l
~
I L
r.
4 e.
s, L:
s s
l
}
.I',
v J
f 6
C,.
o i
t J
c r
to l
l 1
4 1
W
'eL I i.
l C l.
I
..'i, l
I.
i E' !* f, '(
. I K,li:
r I
Figure 9-M r
r Photo shows distribution arx! type of sand particles.
l Spectra shows htsic elemental composition.
I i
I I
I
SE No. 000243-002 Att. 3-25 Microbiological Assessment In order to assure a complete assessment of the corrosion damage it was decided early on that a microbiological analysis needed to be performed.
Because of limited in-house expertise in this area, an outside consultant, Dr. Carolyn Mathur from York College, was contracted to perform this analysis.
It was decided that four samples of material from the sand bed would be analyzed from the regions indicated in Table 3-M.
Two samples would be sand and two would be corrosion product.
These samples were secured immediately upon removal of the core plugs to assure minimal environmental effect on the bacteria.
Samples were treated for microscopic evaluation as weII as for future culturing. During core renoval close attention was also paid to metal temperatures to assure temperatures did not exceed 150* which would kill the bacterla.
Results of the cultures are not yet available; however, preliminary indications are that there is no strong presence of sulfate reducing bacteria (SRB).
The microscopic evaluation results are shown in Table 5-M.
Cell counts appear typical for levels of bacteria found in natural environments.
In addition, it was reported that the bactoria appear filamentous and in some cases bacteria was observed to be attached to the corrosion product.
SE No. 000243-002 Att. 3-26 Currently cultures are being grown aerobically and anaerobically to establish the type of bacteria present including the presence of sulfate reducing bacteria.
Ground Potential Measurements The possibility of stray currents influencing the corrosion rate was also considered.
In order to provide an assessment of this, external potential measurements were conducted to check for the presence of stray currents.
Potential measurements were taken between the ground and each of the five sand bed drain lines using a copper-copper sulfate reference cell.
The measurements revealed no evidence of stray currents while the reactor is shut down, however, these measurements will need to be repeated during power operation.
SE No. 000243-002 Att. 3-27 l
l t
TABLE 5-M i
l Bacteriological Studies Preliminary Results Sample No.
Type Cell Count
- SRB 2-15A Sand (dry) 1x10' cells /gm negative Adjacent to Drywell 71% viable 1-19C Corrosion Product 5x10' cells /gm weak pos.
Adjacent to Drywell 50% vlable 6-11A Sand (motst) 4x10' cells /gm weak pos.
Away from Orywell 74% viable
{
4-19A Corrosten Product 6x10' cells /gm negative Adjacent to Sand 40% viable Stained with fluoresceln isothlocyanate l
[
SE No. 000243-002 Att. 3-28 Corrosion Assessment As discussed in the background section, wetting of the sand bed may have occurred as early as initial construction.
The only other documented evidence of leakage was during the 1980, 1983 and 1986 outages. Although the exact source of leakage during construction is unknown, it was reported that during the application of the Firebar-D material that copious quantitles of water were observed coming from the Firebar and running down the drywell presumably into the sand bed. During outages water was most Ilkely coming from a leaking gasket in the seal plate region.
This gasket was replaced during the 1986 refueling outage and the leakage appears to be stopped. On the above basis and in view of the fact that there would not be other sources of water to enter the annular region behind the drywell during operation, it has been concluded that the introduction of water was an intermittent occurrence (i.e. during outages) which may have occurred during construction but definitely occurred in 1980, 1983 and 1986. Also, it can be concluded that the sand as a result of this water introduction is contaminated with chloride and sulfate along with numerous metal ions.
SE No. 000243-002 i
Att. 3-29 Sand is generally ascribed with good drainage properties which would allow for the bulk of the water which entered the sand bed to flow out of the drain lines; however, because this region is fairly enclosed with little air circulation, high humidity is believed to exist in the annular space which could result in the sand remaining moist for indefinite periods of time.
This is partially substantiated by the fact that high humidity and sweating is generally observed in the torus room where the sand bed drains exit. Above the sand bed, however, fiberglass boards and above that Firebar are applied to the exterior drywell steel which would help prevent moisture from coming directly in contact with it.
In addition, during operation the average drywell air temperature is approximately 140* F which again would prevent condensation from forming on any exposed steel surfaces.
The overall environment within the annular region can therefore best be described in the following manner:
Water was introduced into the sand bed possibly as early as in the late sixties and probably contained magnesite and magnesium sulfate from the Firebar.
The bulk of this water would have drained off leaving moist sand behind. We know that the exterior of the i
drywell was coated with red lead primer over which Firebar and fiberglass boards were applied which would afford general protection to these steel surfaces from corrosion. Coating damaged areas and with time all areas within the sand bed would be expected to expertence general corrotion as
{
long as the sand remained moist or until a protective oxido film built up on the steel surface as a result of the corrosion process.
It appears, I
i r
I I
w.....
SE No. 000243-002 Att. 3-30 however, that a completely protective film did not result most probably because of the presence of chloride.
The actual metal loss which may have occurred during the time frame from initial startup until the next time water was reintroduced as a result of leakage into the cavity is unknown.
The first documented incident of water intrusion following startup which would definitely initiate corrosion was in 1980. Water samples collected and analyzed at this time for radioactivity measurement, indicated that it was refueling water and hence adds credence to the assessment that the source of the water was the leaking bellows gasket. Corrosion rates would therefore be properly based on the assumption that the corrodent was refueling quality water contaminated with chloride from the Firebar and that the corrosion process was aqueous general corrosion.
Some shallow pltting 15 also occurring but it is considered only in view of its contribution to overall thinning.
The possibility of stress corrosion cracking and hydrogen embrittlement were also considered.
However, these forms of corrosion are generally associated with high strength steels or high temperatures and not considered a damage mechanism for the environment or material associated with the drywell. Ultrasonic examination of the welds and heat-affected zone in the wastage regions also showed no indication of cracking.
SE No. 000243-002 Att. 3-31 An upper bound general corrosion rate for carbon steel would be expected to be in the range of 10-20 mpy depending on the drywell plate temperature.
These corrosion rates, however, if applied genera 11y to the drywell region in contact with the sand bed are consistent with the average wall loss of.288" only if the corrosion is assumed to have occurred since 1969 which was the first possible time water could enter the sand bed drains.
In fact, however, close' scrutiny of the UT thickness data indicates that corrosion was extremely non-uniform as defined above in the section on UT measurements.
First, the region above the 11' 9" elevation shows little or no wall loss.
Then the region from 10' 3" to II' 9" shows the greatest wall loss followed by the region below 10' 3" which shows substantially less wall loss.
Lastly, only two regions of the drywell encompassing four bays show any significant wall loss. A possible explanation for this is that due to channeling only these regions became wetted.
This assumption is potentially confirmed by the observation that the sand in the minor wastage regions was dry. Also, the intimacy of the contact between the sand and the plate is a factor.
If the sand had been pushed away from the drywell in certain regions due to the preoperation pressure test causing the drywell to expand; this also can result in variations in corrosion rate.
The protectiveness of the red lead primer will be a function of its integrity in the various regions and again ma/ be leading to variable corrosion rates.
Lastly, differential aeration may be playing a rolo in where corrosion is occurring. Clearly the presence of magnetite, an oxygen defficient oxide, in some regions and hematite in other regions suggests this is occurring.
SE No. 000243-002 Att. 3-32 Conclusion Aqueous corrosion of the carbon steel drywell is estimated to have initiated in 1969 resulting from the first intrusion of water into the sand bed region.
This inventory of water may have been added to during subsequent outages but was definitely added during the 1980, 1983 and 1986 refueling outages.
This latter water is expected to flow down over the Insulation material in the annular space and pass through the sand and out through the drains. Depending on flow rates, an inventory of water may be accumulated in the sand bed or channeling of the water may also occur leading to wetting in specific locations.
Irrespective of the water flow rate some sand will become wetted with oxygen saturated water and corrosion will result.
This corrosion was most Ilkely influenced by the presence of chloride, leached from the Firebar-0, as it was found to be incorporated within the Fe 0. corrosion product.
Bacteria are not believed to have been a major influence on the corrosion.
This latter conclusion is based on the facts that no deep pitting was observed and no sulfide or substantial concentration of manganese was detected in the corrosion product, all of which are typical evidence of micreblological influenced corrosion.
In addition, the corrosion observed can be explained solely on the basis of chemical attack.
However, because there is viable bacteria present, any plans for Inhibiting future corrosion may also require destroying or rendering this bacteria harmless.
SE No. 000243-002 Att. 3-33 Review of the literature suggests corrosion rates can vary widely for carbon steel in aqueous environments.
Rates can be as low as 1-2 mpy in high pH aqueous environments or greater than 50 mpy in acid solutions.
Dr.
Uhllg in his " Corrosion Handbook" lists corrosion rates for ambient temperature, seawater at approximately 1-7 mpy which at 140*F would conservatively equate to approximately 17 mpy. Uhllg further states that with the formation of corrosion products the rate of corrosion will be less than it would be if the steel were in direct contact with seawater and that the rate will stabilize and not change with time.
In addition, he observed that, " specimens of steel have been exposed to seawater where sulfate reducing bacteria were known to be present, and in fact were found in tne corrosion products which contained appreciable percentages of Iron sulfide.
The observed rates of corrosion and pltting of such steel fell within the normal range previously defined."
If we then take the 17 mpy corrosion rate and project this over the 17 year Ilfe of the plant it correlates closely with the average corrosion loss of 288 mils.
However, in order to insure conservatism in the structural analysis a factor of safety should be applied to this rate.
To arrive at a defendable factor it has been assumed that all corrosion occurred over the past six years as a result of the water intrusion in 1980.
This would equate to a corrosion rate of 48 mpy and give a factor of safety of 2.8.
SE No. 000243-002 Att. 3-34 Conclusion Summary 1.
Hastage of the drywell plate is the result of an aqueous general corrosion process influenced by localized oxygen depletion, the degree to which moisture is present, temperature and chloride contamination.
2.
Although viable bacteria were identlfled in the sand and corrosion product, no substantive evidence exists as to its involvement in the corrosion process, at least in terms of currently pubilcized 1
mechanisms.
However, because of the variable nature of microbial induced corrosion any attempts at mitigating corrosion should consider this mechanism.
3.
0-Meter thickness readings, which initially were thought to be either i
pitting and later characterized by "A" scan UT as inclusions, were confirmed by metallography to be aluminide inclusions in the carbon steel.
l 4.
The combination of using a 0-Meter for ultrasonic thickness measurements and an "A" scan for qualitative assessment of the plate condition are adequate for engineering evaluations.
5.
Corroston is limited to the steel in contact with the sand bed and is present to a significant amount (i.e.,.25"
.35") only in bays 11, 13, 17 and 19 and only within elevations 10' 3" to 11' 9".
I t
i
~ _ _ _. _ -.
i 6
The areas of observed corrosion oppear to be those arcos in which the sand hos remained significantly wet,ted.
This wetting most likely occurred during initial construction and then periodically during refueling outages os o result of leokoge from the drywell bellows.
Documented evidence of such leokoge exists since 1980.
7.
Corrosion rotes have been conservatively set at 48 moy although more typically, through review of industry experience and corrosion literature, would be expected to be cooroximately 17 moy.
j k
)
1 1
l i
4 i
e s
i l
l I
I I
I i
t i
SE No. 000243-002 Att. 4-1 Structural Analysis Bases A reevaluation of the drywell containment structure has been performed to insure structural integrity for the combined effects of local shell thinning, operating basis earthquake, pressure and temperature due to a postulated Design Basis Accident (DBA) and the mechanical loads.
In performing this analysis the following design bases were used.
Appilcable Codes Establishing Allowable Stress Criteria (1) ASME Boller and Pressure Vessel Code,Section VIII, 1962 Edition.
(2) Nuclear Code Case 1270N-5, 1271N and 1272N-5 (3) ASME Boller and Pressure Vessel Code,Section III Division 1, appilcable portions of Subsection NE-3000, namely, NE-3213.10, NE-3221-2, NE-3221.4 and Table NE 3217-1.
Materials of Construction According to the Chicago Bridge and Iron drawing No. 9-0971 sheet No. 1, Rev.
2, the material used in the fabrication of the drywell shell is ASTM SA-212 grade B Firebox.
The examination of the original mill certificates reveals that all 1.154" plates used in fabrication of the drywell shell have a yleld strength of about 5 to 33% greater than the minimum specified in the ASTH.
SE No. 000243-002 Att. 4-2 Design Condition The drywell shell is analyzed for the maximum positive pressure 35 psig at 281*F and 62 psig at 175'F.
The former condition represents the double end breaks of a recirculation loop. This is the design basis accident.
Other Loads All other loads considered concomitant with accident conditions were taken from the Chicago Bridge and Iron original analysis.
Load Combination The load combination representing a DBA during normal operation, as specified in the original Chicago Bridge and Iron original report, was chosen for analysis.
This load combination includes the gravity load of vessel and appurtenances, gravity load from equipment supports, seismic loads (OBE), as well as accident conditions for temperature and pressure.
SE No. 000243-002 Methods Att. 4-3 Structural Model The mathematical model used to evaluate effects of the reduced shell thickness within the sand entrenchment area consists of a lower region of the spherical shell between elevations 23'-6 7/8" and the point of complete fixity against translational movement and rotation at the foundation level at elevation 8'-11 1/4".
This model is developed to calculate the membrane and bending stresses at the point of fixity due to the accident internal press"re and thermal loads as well as loads associated with normal operation.
The results of the structural analysis will allow the determination of the minimum allowable pressure boundary thickness using ASME Code allowable stress criteria.
Except for the sand pocket zone, all other shell thicknesses used in the analyses were those shown in the Chicago Bridge and Iron Drawing No. 9-0971 Sheet No. 4 Rev. 1.
The function of the sand pocket is to provide a proportional reaction so that the discontinuity stresses due to the embedment will be gradual and lower rather than abrupt and high.
In order to evaluate the sensitivity of the sand pocket, two separate structural models were considered.
In the first model, the sand is assumed to provide an inward reaction linear in proportion to shell displacement.
The secono model assumes the sand to offer no resistance against the drywell shell movement.
SE No. 000243-002 Att. 4-4 The thermal gradient in the sand entrenchment zone is assumed to be linear.
The attenuation of the thermal gradient in the meridional direction is assumed to be complete within the sand pocket, that is, the temperature distribution is 175*F/281*F at elevation 12'-3" and 60*F (ambient temperature) at elevation 8'-11 1/4".
'The drywell shell membrane loads from the original Chicago Bridge and Iron analysis are introduced at the top boundary of the structural models to simulate shell continuity.
The structural model and the loading are assumed to be symmetrical; the penetration and their effects are not considered.
This is reasonable since the reinforcement at the penetrations restores the shell to its original condition.
The fiberglass insulation material within the annulus between the drywell and the concrete shleid wall is assumed to have no structural stiffness.
SE No. 000243-002 Att. 4-5 Stress Analysis The drywell containment structure model described above is analyzed by the Chicago Bridge & Iron Corporation utilizing the Kalnins KSHEL Program for axisymmetric shells of revolution to evaluate the adequacy of the lower shell region within the sand entrenchment area.
Each of the two models, with and without sand entrenchment, is subjected to the mechanical loads, operating basis earthquake and the accident pressure and temperature conditions of 35 psig at 281*F and 62 psig at 175'F.
This analysis identifies meridional and circumferential membrane and meridional and clicumferential bending stresses for the dead weight, earthquake, pressure, and thermal loads.
SE No. 000243-002 Att. 4-6 The acceptance criteria used to establish structural adequacy of the drywell are taken from Section VIII, ASME Boller and Pressure Vessel Code, 1962 Edition, Nuclear Code Case 1272N-5, and Section III, ASME Boller and Pressure Vessel Code 1986 Edition, Division I, Subsection NE, paragraphs NE-3213-10, NE-3221-2, NE-3221-4 and Table NE 3217-1.
For purposes of analysis, the shell thickness in the sand entrenchment zone is taken to be equal to 0.700".
Mean of thickness readings as representing structural response Structural loads will follow paths through the affected region having the largest stiffness (thickness)'.
Less stiff (thinner) sections will follow the strain of the stiffer sections such that there will be a compatibility of strain through-out, as governed by the stiffer sections.
The condition of strain compatibility means that the stress in the thinner sections will be equal to the stress in the adjacent thicker sections.
It is reasonable to use the mean thickness, as opposed to the minimum thickness, because the mean represents the actual load reacting action of the shell.
i l
l
(
l
--a-r
-e m--e m
e m
-w-x.-
SE No. 000243-002 Att. 4-7 Potential for Buckling In addition, another analysis has been performed by Professor A. Kalnins of Lehigh University using the Kalnins shell of revolution computer program to evaluate the potential of buckling of the drywell shell in the sand entrenchment zone.
The mathematical model used to perform buckling analysis is basically similar to the model used for the stress analysis, except that credit was taken for the structural effect of the concrete that extends upward from the foundation around the inside of the drywell and for the sand which provides an inward reaction in direct proportion to shell expansion.
For purposes of analysis the shell thickness in the sand entrenchment zone is taken to be equal to 0.700".
SE No. 000243-002 Att. 4-6 STRESS ANALYSIS RESULTS FOR SHELL THICKNESS TAKEN TO BE EQUAL TO 0.700" The allowable stress criteria are:
1)
Local primary membrane stress (not including thermal) - P; -
1.5.Smc.- 28,875 psi (No change since 1962).
2)
Surface stresses (local membrane and secondary stresses, both thermal and mechanical axial and bending) - Q - 3 Sm - 52,500.
(Q-3 S..-57,900 from Sect. III, Div.1. Subsect NE)
Table 1 shows the results of the stress analysis at the point of embedment taking credit for the radially inward reaction because of the resistance of the sand and also for analytically removing it. Using stress intensity, the stresses satisfy the former Code allowable stress criteria except for the 4
condition of full sand removal when allowable stresses are exceeded by 2.7%.
(
This load combination considered the accident condition of 62 psig and 175*F, which is not the same as the design basis accident (DBA) representing a double ended break of a recirculation line.
Present Code Allowable stress criteria are satisfied in all cases.
l i
SE No. 000243-002 Att. 4-9 Except as mentioned above the stresses shown satisfy the allowable stress criteria of ASME Sect. VIII, 1962, with Nuclear Code Cases 1270 N-5, 1271 N and 1272 N-5, as well as those of ASME Sect. III, Div. 1, Subsection NE, 1986, using stress intensities as directed in the latter code. Meridional extent, but not the peak value, of local primary membrane stress slightly exceeds (but
< 2X) the guidance given in Sect. III.
It is reasonable to neglect this small departure from present code guidance because the present situation is an in-service condition and not a design condition, and because the departure is small.
i t
f SE No. 000243-002 Att. 4-10 Results of the analysis for buckling potential 4
Stability margin is identified in Table 2.
Margin is defined as the ratio of the calculated buckling load to the actual applied load.
The reference is the point of the embedment.
Normal and accident load combinations are considered with and without the radially inward resistance of the sand.
The stiffness of the concrete on the inside of the shell is included in both cases.
The shell is considered to be imperfect.
The minimum margin to safety is 3.80.
Conclusion Structural integrity of the primary pressure boundary is maintained with a local shell thickness reduction limited to the sand entrenchment region. Code allowable stress criteria are met using a thickness equal to 0.700".
A large margin to buckling exists such that buckling of a locally thinned shell is not a technical issue.
i I
l 1
l l
t
'ss 3
% r TABLE 1 STRESS IdTENSITIES AL0dG MERIDIAN (PSI)
Shell Thickneas t=0.7 in.
LOAD With Sand Pocket Without Sand Pocket COMBINATIONS MEMBRANE MEMBRAN8 & BEdDIdG MEMBRANE MEMBRANE & BERDIdG Calculated Allowabic Calculated Allowable Calculated isllowable Calculated Allowable P = 35 psig 25:251 32,147 8821 43,722 T = 281* F 1.5 S 38 l*0 8 38 ue u
mc m
23,875 52,500 28,875 52,500 i
I'/, edy 24,935 16,944 P = 62 psig 53,897 3S 3"
T = 175* F 57,900 57,900 l
TABLE 2 i
MARGIN AGAINST INSTABILITY t' hell thi ckrie m = 0.?"
Load i
Coubinations With Sand Pocket Without Sand Pocket i
- b Margin b
tiargin 17/ i n.
1/in.
r Safety
., f,,,
f,.
or Suroty l
Norual
-4031
+3312 6.12
-4031
+3312 6.12 Operation l
DBA P=35 psic 1
+7259
-10409 3.8 tension tencion H/A l
T=231 F
i l
P=62'paic 4
Lenalon tension d/A tension t,enaion 1;/4 f,
T.,=175*F
~
i l
)
OVERSIZE DOCUMENT PAGE PULLED i
l SEE APERTURE CARDS NUMBER OF OVERSIZE PAGES FILMED ON APERTURE CARDS i
l APERTURE CARD /HARD COPY AVAILABLE FROM RECORD SERVICES BRANCH,TIDC FTS 492-8989
-